analysis of published research articles

Write a Critical Review of a Scientific Journal Article

1. identify how and why the research was carried out, 2. establish the research context, 3. evaluate the research, 4. establish the significance of the research.

• Writing Your Critique

Ask Us: Chat, email, visit or call

Video: How to Integrate Critical Voice into Your Literature Review

Video: Note-taking and Writing Tips to Avoid Plagiarism

Get assistance

The library offers a range of helpful services.  All of our appointments are free of charge and confidential.

• Book an appointment

Read the article(s) carefully and use the questions below to help you identify how and why the research was carried out. Look at the following sections: 

Introduction

• What was the objective of the study?
• What methods were used to accomplish this purpose (e.g., systematic recording of observations, analysis and evaluation of published research, assessment of theory, etc.)?
• What techniques were used and how was each technique performed?
• What kind of data can be obtained using each technique?
• How are such data interpreted?
• What kind of information is produced by using the technique?
• What objective evidence was obtained from the authors’ efforts (observations, measurements, etc.)?
• What were the results of the study? 
• How was each technique used to obtain each result?
• What statistical tests were used to evaluate the significance of the conclusions based on numeric or graphic data?
• How did each result contribute to answering the question or testing the hypothesis raised in the introduction?
• How were the results interpreted? How were they related to the original problem (authors’ view of evidence rather than objective findings)? 
• Were the authors able to answer the question (test the hypothesis) raised?
• Did the research provide new factual information, a new understanding of a phenomenon in the field, or a new research technique?
• How was the significance of the work described?
• Do the authors relate the findings of the study to literature in the field?
• Did the reported observations or interpretations support or refute observations or interpretations made by other researchers?

These questions were adapted from the following sources:  Kuyper, B.J. (1991). Bringing up scientists in the art of critiquing research. Bioscience 41(4), 248-250. Wood, J.M. (2003). Research Lab Guide. MICR*3260 Microbial Adaptation and Development Web Site . Retrieved July 31, 2006.

Once you are familiar with the article, you can establish the research context by asking the following questions:

• Who conducted the research? What were/are their interests?
• When and where was the research conducted?
• Why did the authors do this research?
• Was this research pertinent only within the authors’ geographic locale, or did it have broader (even global) relevance?
• Were many other laboratories pursuing related research when the reported work was done? If so, why?
• For experimental research, what funding sources met the costs of the research?
• On what prior observations was the research based? What was and was not known at the time?
• How important was the research question posed by the researchers?

These questions were adapted from the following sources: Kuyper, B.J. (1991). Bringing up scientists in the art of critiquing research. Bioscience 41(4), 248-250. Wood, J.M. (2003). Research Lab Guide. MICR*3260 Microbial Adaptation and Development Web Site . Retrieved July 31, 2006.

Remember that simply disagreeing with the material is not considered to be a critical assessment of the material.  For example, stating that the sample size is insufficient is not a critical assessment.  Describing why the sample size is insufficient for the claims being made in the study would be a critical assessment.

Use the questions below to help you evaluate the quality of the authors’ research:

• Does the title precisely state the subject of the paper?
• Read the statement of purpose in the abstract. Does it match the one in the introduction?

Acknowledgments

• Could the source of the research funding have influenced the research topic or conclusions?
• Check the sequence of statements in the introduction. Does all the information lead coherently to the purpose of the study?
• Review all methods in relation to the objective(s) of the study. Are the methods valid for studying the problem?
• Check the methods for essential information. Could the study be duplicated from the methods and information given?
• Check the methods for flaws. Is the sample selection adequate? Is the experimental design sound?
• Check the sequence of statements in the methods. Does all the information belong there? Is the sequence of methods clear and pertinent?
• Was there mention of ethics? Which research ethics board approved the study?
• Carefully examine the data presented in the tables and diagrams. Does the title or legend accurately describe the content? 
• Are column headings and labels accurate? 
• Are the data organized for ready comparison and interpretation? (A table should be self-explanatory, with a title that accurately and concisely describes content and column headings that accurately describe information in the cells.)
• Review the results as presented in the text while referring to the data in the tables and diagrams. Does the text complement, and not simply repeat data? Are there discrepancies between the results in the text and those in the tables?
• Check all calculations and presentation of data.
• Review the results in light of the stated objectives. Does the study reveal what the researchers intended?
• Does the discussion clearly address the objectives and hypotheses?
• Check the interpretation against the results. Does the discussion merely repeat the results? 
• Does the interpretation arise logically from the data or is it too far-fetched? 
• Have the faults, flaws, or shortcomings of the research been addressed?
• Is the interpretation supported by other research cited in the study?
• Does the study consider key studies in the field?
• What is the significance of the research? Do the authors mention wider implications of the findings?
• Is there a section on recommendations for future research? Are there other research possibilities or directions suggested? 

Consider the article as a whole

• Reread the abstract. Does it accurately summarize the article?
• Check the structure of the article (first headings and then paragraphing). Is all the material organized under the appropriate headings? Are sections divided logically into subsections or paragraphs?
• Are stylistic concerns, logic, clarity, and economy of expression addressed?

These questions were adapted from the following sources:  Kuyper, B.J. (1991). Bringing up scientists in the art of critiquing research. Bioscience 41(4), 248-250. Wood, J.M. (2003). Research Lab Guide. MICR*3260 Microbial Adaptation and Development Web Site. Retrieved July 31, 2006.

After you have evaluated the research, consider whether the research has been successful. Has it led to new questions being asked, or new ways of using existing knowledge? Are other researchers citing this paper?

You should consider the following questions:

• How did other researchers view the significance of the research reported by your authors?
• Did the research reported in your article result in the formulation of new questions or hypotheses (by the authors or by other researchers)?
• Have other researchers subsequently supported or refuted the observations or interpretations of these authors?
• Did the research make a significant contribution to human knowledge?
• Did the research produce any practical applications?
• What are the social, political, technological, medical implications of this research?
• How do you evaluate the significance of the research?

To answer these questions, look at review articles to find out how reviewers view this piece of research. Look at research articles and databases like Web of Science to see how other people have used this work. What range of journals have cited this article?

These questions were adapted from the following sources:

Kuyper, B.J. (1991). Bringing up scientists in the art of critiquing research. Bioscience 41(4), 248-250. Wood, J.M. (2003). Research Lab Guide. MICR*3260 Microbial Adaptation and Development Web Site . Retrieved July 31, 2006.

• << Previous: Start Here
• Next: Writing Your Critique >>
• Last Updated: Jan 11, 2024 12:42 PM
• URL: https://guides.lib.uoguelph.ca/WriteCriticalReview

Suggest an edit to this guide

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Organizing Your Social Sciences Research Assignments

• Annotated Bibliography
• Analyzing a Scholarly Journal Article
• Group Presentations
• Dealing with Nervousness
• Using Visual Aids
• Grading Someone Else's Paper
• Types of Structured Group Activities
• Group Project Survival Skills
• Leading a Class Discussion
• Multiple Book Review Essay
• Reviewing Collected Works
• Writing a Case Analysis Paper
• Writing a Case Study
• About Informed Consent
• Writing Field Notes
• Writing a Policy Memo
• Writing a Reflective Paper
• Writing a Research Proposal
• Generative AI and Writing
• Acknowledgments

Definition and Introduction

Journal article analysis assignments require you to summarize and critically assess the quality of an empirical research study published in a scholarly [a.k.a., academic, peer-reviewed] journal. The article may be assigned by the professor, chosen from course readings listed in the syllabus, or you must locate an article on your own, usually with the requirement that you search using a reputable library database, such as, JSTOR or ProQuest . The article chosen is expected to relate to the overall discipline of the course, specific course content, or key concepts discussed in class. In some cases, the purpose of the assignment is to analyze an article that is part of the literature review for a future research project.

Analysis of an article can be assigned to students individually or as part of a small group project. The final product is usually in the form of a short paper [typically 1- 6 double-spaced pages] that addresses key questions the professor uses to guide your analysis or that assesses specific parts of a scholarly research study [e.g., the research problem, methodology, discussion, conclusions or findings]. The analysis paper may be shared on a digital course management platform and/or presented to the class for the purpose of promoting a wider discussion about the topic of the study. Although assigned in any level of undergraduate and graduate coursework in the social and behavioral sciences, professors frequently include this assignment in upper division courses to help students learn how to effectively identify, read, and analyze empirical research within their major.

Franco, Josue. “Introducing the Analysis of Journal Articles.” Prepared for presentation at the American Political Science Association’s 2020 Teaching and Learning Conference, February 7-9, 2020, Albuquerque, New Mexico; Sego, Sandra A. and Anne E. Stuart. "Learning to Read Empirical Articles in General Psychology." Teaching of Psychology 43 (2016): 38-42; Kershaw, Trina C., Jordan P. Lippman, and Jennifer Fugate. "Practice Makes Proficient: Teaching Undergraduate Students to Understand Published Research." Instructional Science 46 (2018): 921-946; Woodward-Kron, Robyn. "Critical Analysis and the Journal Article Review Assignment." Prospect 18 (August 2003): 20-36; MacMillan, Margy and Allison MacKenzie. "Strategies for Integrating Information Literacy and Academic Literacy: Helping Undergraduate Students make the most of Scholarly Articles." Library Management 33 (2012): 525-535.

Benefits of Journal Article Analysis Assignments

Analyzing and synthesizing a scholarly journal article is intended to help students obtain the reading and critical thinking skills needed to develop and write their own research papers. This assignment also supports workplace skills where you could be asked to summarize a report or other type of document and report it, for example, during a staff meeting or for a presentation.

There are two broadly defined ways that analyzing a scholarly journal article supports student learning:

Improve Reading Skills

Conducting research requires an ability to review, evaluate, and synthesize prior research studies. Reading prior research requires an understanding of the academic writing style , the type of epistemological beliefs or practices underpinning the research design, and the specific vocabulary and technical terminology [i.e., jargon] used within a discipline. Reading scholarly articles is important because academic writing is unfamiliar to most students; they have had limited exposure to using peer-reviewed journal articles prior to entering college or students have yet to gain exposure to the specific academic writing style of their disciplinary major. Learning how to read scholarly articles also requires careful and deliberate concentration on how authors use specific language and phrasing to convey their research, the problem it addresses, its relationship to prior research, its significance, its limitations, and how authors connect methods of data gathering to the results so as to develop recommended solutions derived from the overall research process.

Improve Comprehension Skills

In addition to knowing how to read scholarly journals articles, students must learn how to effectively interpret what the scholar(s) are trying to convey. Academic writing can be dense, multi-layered, and non-linear in how information is presented. In addition, scholarly articles contain footnotes or endnotes, references to sources, multiple appendices, and, in some cases, non-textual elements [e.g., graphs, charts] that can break-up the reader’s experience with the narrative flow of the study. Analyzing articles helps students practice comprehending these elements of writing, critiquing the arguments being made, reflecting upon the significance of the research, and how it relates to building new knowledge and understanding or applying new approaches to practice. Comprehending scholarly writing also involves thinking critically about where you fit within the overall dialogue among scholars concerning the research problem, finding possible gaps in the research that require further analysis, or identifying where the author(s) has failed to examine fully any specific elements of the study.

In addition, journal article analysis assignments are used by professors to strengthen discipline-specific information literacy skills, either alone or in relation to other tasks, such as, giving a class presentation or participating in a group project. These benefits can include the ability to:

• Effectively paraphrase text, which leads to a more thorough understanding of the overall study;
• Identify and describe strengths and weaknesses of the study and their implications;
• Relate the article to other course readings and in relation to particular research concepts or ideas discussed during class;
• Think critically about the research and summarize complex ideas contained within;
• Plan, organize, and write an effective inquiry-based paper that investigates a research study, evaluates evidence, expounds on the author’s main ideas, and presents an argument concerning the significance and impact of the research in a clear and concise manner;
• Model the type of source summary and critique you should do for any college-level research paper; and,
• Increase interest and engagement with the research problem of the study as well as with the discipline.

Kershaw, Trina C., Jennifer Fugate, and Aminda J. O'Hare. "Teaching Undergraduates to Understand Published Research through Structured Practice in Identifying Key Research Concepts." Scholarship of Teaching and Learning in Psychology . Advance online publication, 2020; Franco, Josue. “Introducing the Analysis of Journal Articles.” Prepared for presentation at the American Political Science Association’s 2020 Teaching and Learning Conference, February 7-9, 2020, Albuquerque, New Mexico; Sego, Sandra A. and Anne E. Stuart. "Learning to Read Empirical Articles in General Psychology." Teaching of Psychology 43 (2016): 38-42; Woodward-Kron, Robyn. "Critical Analysis and the Journal Article Review Assignment." Prospect 18 (August 2003): 20-36; MacMillan, Margy and Allison MacKenzie. "Strategies for Integrating Information Literacy and Academic Literacy: Helping Undergraduate Students make the most of Scholarly Articles." Library Management 33 (2012): 525-535; Kershaw, Trina C., Jordan P. Lippman, and Jennifer Fugate. "Practice Makes Proficient: Teaching Undergraduate Students to Understand Published Research." Instructional Science 46 (2018): 921-946.

Structure and Organization

A journal article analysis paper should be written in paragraph format and include an instruction to the study, your analysis of the research, and a conclusion that provides an overall assessment of the author's work, along with an explanation of what you believe is the study's overall impact and significance. Unless the purpose of the assignment is to examine foundational studies published many years ago, you should select articles that have been published relatively recently [e.g., within the past few years].

Since the research has been completed, reference to the study in your paper should be written in the past tense, with your analysis stated in the present tense [e.g., “The author portrayed access to health care services in rural areas as primarily a problem of having reliable transportation. However, I believe the author is overgeneralizing this issue because...”].

Introduction Section

The first section of a journal analysis paper should describe the topic of the article and highlight the author’s main points. This includes describing the research problem and theoretical framework, the rationale for the research, the methods of data gathering and analysis, the key findings, and the author’s final conclusions and recommendations. The narrative should focus on the act of describing rather than analyzing. Think of the introduction as a more comprehensive and detailed descriptive abstract of the study.

Possible questions to help guide your writing of the introduction section may include:

• Who are the authors and what credentials do they hold that contributes to the validity of the study?
• What was the research problem being investigated?
• What type of research design was used to investigate the research problem?
• What theoretical idea(s) and/or research questions were used to address the problem?
• What was the source of the data or information used as evidence for analysis?
• What methods were applied to investigate this evidence?
• What were the author's overall conclusions and key findings?

Critical Analysis Section

The second section of a journal analysis paper should describe the strengths and weaknesses of the study and analyze its significance and impact. This section is where you shift the narrative from describing to analyzing. Think critically about the research in relation to other course readings, what has been discussed in class, or based on your own life experiences. If you are struggling to identify any weaknesses, explain why you believe this to be true. However, no study is perfect, regardless of how laudable its design may be. Given this, think about the repercussions of the choices made by the author(s) and how you might have conducted the study differently. Examples can include contemplating the choice of what sources were included or excluded in support of examining the research problem, the choice of the method used to analyze the data, or the choice to highlight specific recommended courses of action and/or implications for practice over others. Another strategy is to place yourself within the research study itself by thinking reflectively about what may be missing if you had been a participant in the study or if the recommended courses of action specifically targeted you or your community.

Possible questions to help guide your writing of the analysis section may include:

Introduction

• Did the author clearly state the problem being investigated?
• What was your reaction to and perspective on the research problem?
• Was the study’s objective clearly stated? Did the author clearly explain why the study was necessary?
• How well did the introduction frame the scope of the study?
• Did the introduction conclude with a clear purpose statement?

Literature Review

• Did the literature review lay a foundation for understanding the significance of the research problem?
• Did the literature review provide enough background information to understand the problem in relation to relevant contexts [e.g., historical, economic, social, cultural, etc.].
• Did literature review effectively place the study within the domain of prior research? Is anything missing?
• Was the literature review organized by conceptual categories or did the author simply list and describe sources?
• Did the author accurately explain how the data or information were collected?
• Was the data used sufficient in supporting the study of the research problem?
• Was there another methodological approach that could have been more illuminating?
• Give your overall evaluation of the methods used in this article. How much trust would you put in generating relevant findings?

Results and Discussion

• Were the results clearly presented?
• Did you feel that the results support the theoretical and interpretive claims of the author? Why?
• What did the author(s) do especially well in describing or analyzing their results?
• Was the author's evaluation of the findings clearly stated?
• How well did the discussion of the results relate to what is already known about the research problem?
• Was the discussion of the results free of repetition and redundancies?
• What interpretations did the authors make that you think are in incomplete, unwarranted, or overstated?
• Did the conclusion effectively capture the main points of study?
• Did the conclusion address the research questions posed? Do they seem reasonable?
• Were the author’s conclusions consistent with the evidence and arguments presented?
• Has the author explained how the research added new knowledge or understanding?

Overall Writing Style

• If the article included tables, figures, or other non-textual elements, did they contribute to understanding the study?
• Were ideas developed and related in a logical sequence?
• Were transitions between sections of the article smooth and easy to follow?

Overall Evaluation Section

The final section of a journal analysis paper should bring your thoughts together into a coherent assessment of the value of the research study . This section is where the narrative flow transitions from analyzing specific elements of the article to critically evaluating the overall study. Explain what you view as the significance of the research in relation to the overall course content and any relevant discussions that occurred during class. Think about how the article contributes to understanding the overall research problem, how it fits within existing literature on the topic, how it relates to the course, and what it means to you as a student researcher. In some cases, your professor will also ask you to describe your experiences writing the journal article analysis paper as part of a reflective learning exercise.

Possible questions to help guide your writing of the conclusion and evaluation section may include:

• Was the structure of the article clear and well organized?
• Was the topic of current or enduring interest to you?
• What were the main weaknesses of the article? [this does not refer to limitations stated by the author, but what you believe are potential flaws]
• Was any of the information in the article unclear or ambiguous?
• What did you learn from the research? If nothing stood out to you, explain why.
• Assess the originality of the research. Did you believe it contributed new understanding of the research problem?
• Were you persuaded by the author’s arguments?
• If the author made any final recommendations, will they be impactful if applied to practice?
• In what ways could future research build off of this study?
• What implications does the study have for daily life?
• Was the use of non-textual elements, footnotes or endnotes, and/or appendices helpful in understanding the research?
• What lingering questions do you have after analyzing the article?

NOTE: Avoid using quotes. One of the main purposes of writing an article analysis paper is to learn how to effectively paraphrase and use your own words to summarize a scholarly research study and to explain what the research means to you. Using and citing a direct quote from the article should only be done to help emphasize a key point or to underscore an important concept or idea.

Business: The Article Analysis . Fred Meijer Center for Writing, Grand Valley State University; Bachiochi, Peter et al. "Using Empirical Article Analysis to Assess Research Methods Courses." Teaching of Psychology 38 (2011): 5-9; Brosowsky, Nicholaus P. et al. “Teaching Undergraduate Students to Read Empirical Articles: An Evaluation and Revision of the QALMRI Method.” PsyArXi Preprints , 2020; Holster, Kristin. “Article Evaluation Assignment”. TRAILS: Teaching Resources and Innovations Library for Sociology . Washington DC: American Sociological Association, 2016; Kershaw, Trina C., Jennifer Fugate, and Aminda J. O'Hare. "Teaching Undergraduates to Understand Published Research through Structured Practice in Identifying Key Research Concepts." Scholarship of Teaching and Learning in Psychology . Advance online publication, 2020; Franco, Josue. “Introducing the Analysis of Journal Articles.” Prepared for presentation at the American Political Science Association’s 2020 Teaching and Learning Conference, February 7-9, 2020, Albuquerque, New Mexico; Reviewer's Guide . SAGE Reviewer Gateway, SAGE Journals; Sego, Sandra A. and Anne E. Stuart. "Learning to Read Empirical Articles in General Psychology." Teaching of Psychology 43 (2016): 38-42; Kershaw, Trina C., Jordan P. Lippman, and Jennifer Fugate. "Practice Makes Proficient: Teaching Undergraduate Students to Understand Published Research." Instructional Science 46 (2018): 921-946; Gyuris, Emma, and Laura Castell. "To Tell Them or Show Them? How to Improve Science Students’ Skills of Critical Reading." International Journal of Innovation in Science and Mathematics Education 21 (2013): 70-80; Woodward-Kron, Robyn. "Critical Analysis and the Journal Article Review Assignment." Prospect 18 (August 2003): 20-36; MacMillan, Margy and Allison MacKenzie. "Strategies for Integrating Information Literacy and Academic Literacy: Helping Undergraduate Students Make the Most of Scholarly Articles." Library Management 33 (2012): 525-535.

Writing Tip

Not All Scholarly Journal Articles Can Be Critically Analyzed

There are a variety of articles published in scholarly journals that do not fit within the guidelines of an article analysis assignment. This is because the work cannot be empirically examined or it does not generate new knowledge in a way which can be critically analyzed.

If you are required to locate a research study on your own, avoid selecting these types of journal articles:

• Theoretical essays which discuss concepts, assumptions, and propositions, but report no empirical research;
• Statistical or methodological papers that may analyze data, but the bulk of the work is devoted to refining a new measurement, statistical technique, or modeling procedure;
• Articles that review, analyze, critique, and synthesize prior research, but do not report any original research;
• Brief essays devoted to research methods and findings;
• Articles written by scholars in popular magazines or industry trade journals;
• Academic commentary that discusses research trends or emerging concepts and ideas, but does not contain citations to sources; and
• Pre-print articles that have been posted online, but may undergo further editing and revision by the journal's editorial staff before final publication. An indication that an article is a pre-print is that it has no volume, issue, or page numbers assigned to it.

Journal Analysis Assignment - Myers . Writing@CSU, Colorado State University; Franco, Josue. “Introducing the Analysis of Journal Articles.” Prepared for presentation at the American Political Science Association’s 2020 Teaching and Learning Conference, February 7-9, 2020, Albuquerque, New Mexico; Woodward-Kron, Robyn. "Critical Analysis and the Journal Article Review Assignment." Prospect 18 (August 2003): 20-36.

• << Previous: Annotated Bibliography
• Next: Giving an Oral Presentation >>
• Last Updated: Jun 3, 2024 9:44 AM
• URL: https://libguides.usc.edu/writingguide/assignments

ON YOUR 1ST ORDER

How To Critically Analyse An Article – Become A Savvy Reader

By Laura Brown on 22nd September 2023

In the current academic scenario, knowing how to analyse an article critically is essential to attain stability and strength. It’s about reading between the lines, questioning what you encounter, and forming informed opinions based on evidence and sound reasoning.

• To critically analyse an article, read it thoroughly to grasp the author’s main points.
• Evaluate the evidence and arguments presented, checking for credibility and logical consistency.
• Consider the article’s structure, tone, and style while also assessing its sources.
• Formulate your critical response by synthesising your analysis and constructing a well-supported argument.

Have you ever wondered how to tell if an article is good or not? It’s important when it comes to your academic superiority. Critical analysis of an article is like being a detective. You check the article closely to see if it makes sense, if the facts are correct, and if the writer is trying to trick you.

But it’s not just something for school, college or university; it’s a superpower for everyday life. It helps you find the important stuff in an article, spot when someone is trying to persuade you and understand what the writer really thinks.

Think of it as a special skill that lets you dig deep into an article, like a treasure hunt. You uncover hidden biases, find the truth, and see how the writer tries to convince you. It’s a bit like being a detective and a wizard at the same time.

Get ready to become a smart reader. This guide will show you how to use this superpower to make sense of the information around us in just 8 simple steps.

Step 1: Read the Article

Before embarking on the journey to analyse an article critically, it is paramount to begin with the foundational step of reading the article itself. This step lays the groundwork for a comprehensive understanding of the material, enabling you to effectively evaluate its merits and demerits.

Reading an article critically starts with setting aside distractions and immersing yourself in the text. Instead of skimming through it hurriedly, take the time to read it meticulously.

To truly grasp the article’s essence, you must consider both its content and context. Content refers to the information and ideas presented within the article, while context encompasses the circumstances in which it was written.

• Why was this article written?
• Who is the intended audience?
• When was it published, and what was happening in the world at that time?
• What is the author’s background or expertise in the subject matter?

As you read, do not rely solely on your memory to retain key points and insights. Taking notes is an invaluable practice during this phase. Record significant ideas, quotes, and statistics that catch your attention.

Your initial impressions of the article can offer valuable insights into your subjective response. If a particular passage elicits a strong emotional reaction, make a note of it. Identifying your emotional responses can help you later in the analysis process when considering your own biases and reactions to the author’s arguments.

Step 2: Identify the Main Argument

While you are up to critically analyse an article, pinpointing the central argument is akin to finding the North Star guiding you through the article’s content. Every well-crafted article should possess a clear and concise main argument or thesis, which serves as the nucleus of the author’s message. Typically situated in the article’s introduction or abstract , this argument not only encapsulates the author’s viewpoint but also functions as a roadmap for the reader, outlining what to expect in the subsequent sections.

Identifying the main argument necessitates a discerning eye. Delve into the introductory paragraphs, abstract, or the initial sections of the article to locate this pivotal statement. This argument may be explicit, explicitly stated by the author, or implicit, inferred through careful examination of the content. Once you’ve grasped the main argument, keep it at the forefront of your mind as you proceed with your analysis, it will serve as the cornerstone against which all other elements are evaluated.

Step 3: Evaluate the Evidence

In order to solely understand how to analyse an article critically, it is imperative to know that an article’s persuasive power hinges on the quality of evidence presented to substantiate its main argument. In this critical step, it’s imperative to scrutinise the evidence with a discerning eye. Look beyond the surface to assess the data, statistics, examples, and citations provided by the author. You can run it through Turnitin for a plagiarism check. These elements serve as the pillars upon which the argument stands or crumbles.

Begin by evaluating the credibility and relevance of the sources used to support the argument. Are they authoritative and trustworthy? Are they current and pertinent to the subject matter? Assess the quality of evidence by considering the reliability of the data, the objectivity of the sources, and the breadth of examples. Moreover, consider the quantity of evidence; is there enough to convincingly underpin the thesis, or does it appear lacking or selective? A well-supported argument should be built upon a solid foundation of robust evidence.

Step 4: Examine the Reasoning

Critical analysis doesn’t stop at identifying the argument and assessing the evidence; it extends to examining the underlying reasoning that connects these elements. In this step, delve deeper into the author’s logic and the structure of the argument. The goal is to identify any logical fallacies or weak assumptions that might undermine the article’s credibility.

Scrutinise the coherence and consistency of the author’s reasoning. Are there any gaps in the argument, or does it flow logically from point to point? Identify any potential biases, emotional appeals, or rhetorical strategies employed by the author. Assess whether the argument is grounded in sound principles and reasoning.

Be on the lookout for flawed deductive or inductive reasoning, and question whether the evidence truly supports the conclusions drawn . Critical thinking is pivotal here, as it allows you to gauge the strength of the article’s argumentation and identify areas where it may be lacking or vulnerable to critique.

Step 5: Consider the Structure

The structure of an article is not merely a cosmetic feature but a fundamental aspect that can profoundly influence its overall effectiveness in conveying its message. A well-organised article possesses the power to captivate readers, enhance comprehension, and amplify its impact. To harness this power effectively, it’s crucial to pay close attention to various structural elements.

• Headings and Subheadings: Examine headings and subheadings to understand the article’s structure and main themes.
• Transitions Between Sections: Observe how transitions between sections maintain or disrupt the flow of ideas.
• Logical Progression: Assess if the article logically builds upon concepts or feels disjointed.
• Use of Visual Aids: Evaluate the integration and effectiveness of visual aids like graphs and charts.
• Paragraph Organisation: Analyse paragraph structure, including clear topic sentences.
• Conclusion and Summary: Review the conclusion for a strong reiteration of the main argument and key takeaways.

In essence, the structure of an article serves as the blueprint that shapes the reader’s journey. A thoughtfully organised article not only makes it easier for readers to navigate the content but also enhances their overall comprehension and retention. By paying attention to these structural elements, you can gain a deeper understanding of the author’s message and how it is effectively conveyed to the audience.

Step 6: Analyse Tone and Style

Exploring the tone and style of an article is like deciphering the author’s hidden intentions and underlying biases. It involves looking closely at how the author has crafted their words, examining their choice of language, tone, and use of rhetorical devices . Is the tone even-handed and impartial, or can you detect signs of favouritism or prejudice? Understanding the author’s perspective in this way allows you to place their argument within a broader context, helping you see beyond the surface of the text.

When you analyse tone, consider whether the author’s language carries any emotional weight. Are they using words that evoke strong feelings, or do they maintain an objective and rational tone throughout? Furthermore, observe how the author addresses counterarguments. Are they respectful and considerate, or do they employ ad hominem attacks? Evaluating tone and style can offer valuable insights into the author’s intentions and their ability to construct a persuasive argument.

Step 7: Assess Sources and References

A critical analysis wouldn’t be complete without examining the sources and references cited within the article. These citations form the foundation upon which the author’s arguments rest. To assess the credibility of the author’s research, it’s essential to scrutinise the origins of these sources. Are they drawn from reputable, well-established journals, books, or widely recognised and trusted websites? High-quality sources reflect positively on the author’s research and strengthen the overall validity of the argument.

While staying on the journey of how to critically analyse an article, be vigilant when encountering articles that heavily rely on sources that might be considered unreliable or biased. Investigate whether the author has balanced their sources and considered diverse perspectives. A well-researched article should draw upon a variety of reputable sources to provide a well-rounded view of the topic. By assessing the sources and references, you can gauge the robustness of the author’s supporting evidence.

Step 8: Formulate Your Critical Response

Having navigated through the previous steps, it’s now your turn to construct a critical response to the article. This step involves summarising your analysis by identifying the strengths and weaknesses within the article. Do you find yourself in agreement with the main argument, or do you have reservations? Highlight the evidence that you found compelling and areas where you believe the article falls short. Your critical response serves as a valuable contribution to the ongoing discourse surrounding the topic, adding your unique perspective to the conversation. Remember that constructive criticism can lead to deeper understanding and improved future discourse.

Now, let’s be specific on two of the most analysed articles, i.e. research articles and journal articles.

How To Critically Analyse A Research Article?

A research article is a scholarly document that presents the findings of original research conducted by the author(s) and is typically published in academic journals. It follows a structured format, including sections such as an abstract, introduction, methods, results, discussion, and references. To critically analyse a research article, you may go through the following six steps.

• Scrutinise the research question’s clarity and significance.
• Examine the appropriateness of research methods.
• Assess sample quality and data reliability.
• Evaluate the accuracy and significance of results.
• Review the discussion for supported conclusions.
• Check references for relevant and high-quality sources.

Never hesitate to ask our customer support for examples and relevant guides as you face any challenges while critically analysing a research paper .

How To Critically Analyse A Journal Article?

A journal article is a scholarly publication that presents research findings, analyses, or discussions within a specific academic or scientific field. These articles typically follow a structured format and are subject to peer review before publication. In order to critically analyse a journal article, take the following steps.

• Evaluate the article’s clarity and relevance.
• Examine the research methods and their suitability.
• Assess the credibility of data and sources.
• Scrutinise the presentation of results.
• Analyse the conclusions drawn.
• Consider the quality of references and citations.

If you have any difficulty conducting a good critical analysis, you can always ask our research paper service for help and relevant examples.

Concluding Upon How To Analyse An Article Critically

Mastering the art of analysing an article critically is a valuable skill that empowers you to navigate the vast sea of information with confidence. By following these eight steps, you can dissect articles effectively, separating reliable information from biased or poorly supported claims. Remember, critical analysis is not about tearing an article apart but understanding it deeply and thoughtfully. With practice, you’ll become a more discerning and informed reader, researcher, or student.

Laura Brown, a senior content writer who writes actionable blogs at Crowd Writer.

Free APA Journals ™ Articles

Recently published articles from subdisciplines of psychology covered by more than 90 APA Journals™ publications.

For additional free resources (such as article summaries, podcasts, and more), please visit the Highlights in Psychological Research page.

• Basic / Experimental Psychology
• Clinical Psychology
• Core of Psychology
• Developmental Psychology
• Educational Psychology, School Psychology, and Training
• Forensic Psychology
• Health Psychology & Medicine
• Industrial / Organizational Psychology & Management
• Neuroscience & Cognition
• Social Psychology & Social Processes
• Moving While Black: Intergroup Attitudes Influence Judgments of Speed (PDF, 71KB) Journal of Experimental Psychology: General February 2016 by Andreana C. Kenrick, Stacey Sinclair, Jennifer Richeson, Sara C. Verosky, and Janetta Lun
• Recognition Without Awareness: Encoding and Retrieval Factors (PDF, 116KB) Journal of Experimental Psychology: Learning, Memory, and Cognition September 2015 by Fergus I. M. Craik, Nathan S. Rose, and Nigel Gopie
• The Tip-of-the-Tongue Heuristic: How Tip-of-the-Tongue States Confer Perceptibility on Inaccessible Words (PDF, 91KB) Journal of Experimental Psychology: Learning, Memory, and Cognition September 2015 by Anne M. Cleary and Alexander B. Claxton
• Cognitive Processes in the Breakfast Task: Planning and Monitoring (PDF, 146KB) Canadian Journal of Experimental Psychology / Revue canadienne de psychologie expérimentale September 2015 by Nathan S. Rose, Lin Luo, Ellen Bialystok, Alexandra Hering, Karen Lau, and Fergus I. M. Craik
• Searching for Explanations: How the Internet Inflates Estimates of Internal Knowledge (PDF, 138KB) Journal of Experimental Psychology: General June 2015 by Matthew Fisher, Mariel K. Goddu, and Frank C. Keil
• Client Perceptions of Corrective Experiences in Cognitive Behavioral Therapy and Motivational Interviewing for Generalized Anxiety Disorder: An Exploratory Pilot Study (PDF, 62KB) Journal of Psychotherapy Integration March 2017 by Jasmine Khattra, Lynne Angus, Henny Westra, Christianne Macaulay, Kathrin Moertl, and Michael Constantino
• Attention-Deficit/Hyperactivity Disorder Developmental Trajectories Related to Parental Expressed Emotion (PDF, 160KB) Journal of Abnormal Psychology February 2016 by Erica D. Musser, Sarah L. Karalunas, Nathan Dieckmann, Tara S. Peris, and Joel T. Nigg
• The Integrated Scientist-Practitioner: A New Model for Combining Research and Clinical Practice in Fee-For-Service Settings (PDF, 58KB) Professional Psychology: Research and Practice December 2015 by Jenna T. LeJeune and Jason B. Luoma
• Psychotherapists as Gatekeepers: An Evidence-Based Case Study Highlighting the Role and Process of Letter Writing for Transgender Clients (PDF, 76KB) Psychotherapy September 2015 by Stephanie L. Budge
• Perspectives of Family and Veterans on Family Programs to Support Reintegration of Returning Veterans With Posttraumatic Stress Disorder (PDF, 70KB) Psychological Services August 2015 by Ellen P. Fischer, Michelle D. Sherman, Jean C. McSweeney, Jeffrey M. Pyne, Richard R. Owen, and Lisa B. Dixon
• "So What Are You?": Inappropriate Interview Questions for Psychology Doctoral and Internship Applicants (PDF, 79KB) Training and Education in Professional Psychology May 2015 by Mike C. Parent, Dana A. Weiser, and Andrea McCourt
• Cultural Competence as a Core Emphasis of Psychoanalytic Psychotherapy (PDF, 81KB) Psychoanalytic Psychology April 2015 by Pratyusha Tummala-Narra
• The Role of Gratitude in Spiritual Well-Being in Asymptomatic Heart Failure Patients (PDF, 123KB) Spirituality in Clinical Practice March 2015 by Paul J. Mills, Laura Redwine, Kathleen Wilson, Meredith A. Pung, Kelly Chinh, Barry H. Greenberg, Ottar Lunde, Alan Maisel, Ajit Raisinghani, Alex Wood, and Deepak Chopra
• Nepali Bhutanese Refugees Reap Support Through Community Gardening (PDF, 104KB) International Perspectives in Psychology: Research, Practice, Consultation January 2017 by Monica M. Gerber, Jennifer L. Callahan, Danielle N. Moyer, Melissa L. Connally, Pamela M. Holtz, and Beth M. Janis
• Does Monitoring Goal Progress Promote Goal Attainment? A Meta-Analysis of the Experimental Evidence (PDF, 384KB) Psychological Bulletin February 2016 by Benjamin Harkin, Thomas L. Webb, Betty P. I. Chang, Andrew Prestwich, Mark Conner, Ian Kellar, Yael Benn, and Paschal Sheeran
• Youth Violence: What We Know and What We Need to Know (PDF, 388KB) American Psychologist January 2016 by Brad J. Bushman, Katherine Newman, Sandra L. Calvert, Geraldine Downey, Mark Dredze, Michael Gottfredson, Nina G. Jablonski, Ann S. Masten, Calvin Morrill, Daniel B. Neill, Daniel Romer, and Daniel W. Webster
• Supervenience and Psychiatry: Are Mental Disorders Brain Disorders? (PDF, 113KB) Journal of Theoretical and Philosophical Psychology November 2015 by Charles M. Olbert and Gary J. Gala
• Constructing Psychological Objects: The Rhetoric of Constructs (PDF, 108KB) Journal of Theoretical and Philosophical Psychology November 2015 by Kathleen L. Slaney and Donald A. Garcia
• Expanding Opportunities for Diversity in Positive Psychology: An Examination of Gender, Race, and Ethnicity (PDF, 119KB) Canadian Psychology / Psychologie canadienne August 2015 by Meghana A. Rao and Stewart I. Donaldson
• Racial Microaggression Experiences and Coping Strategies of Black Women in Corporate Leadership (PDF, 132KB) Qualitative Psychology August 2015 by Aisha M. B. Holder, Margo A. Jackson, and Joseph G. Ponterotto
• An Appraisal Theory of Empathy and Other Vicarious Emotional Experiences (PDF, 151KB) Psychological Review July 2015 by Joshua D. Wondra and Phoebe C. Ellsworth
• An Attachment Theoretical Framework for Personality Disorders (PDF, 100KB) Canadian Psychology / Psychologie canadienne May 2015 by Kenneth N. Levy, Benjamin N. Johnson, Tracy L. Clouthier, J. Wesley Scala, and Christina M. Temes
• Emerging Approaches to the Conceptualization and Treatment of Personality Disorder (PDF, 111KB) Canadian Psychology / Psychologie canadienne May 2015 by John F. Clarkin, Kevin B. Meehan, and Mark F. Lenzenweger
• A Complementary Processes Account of the Development of Childhood Amnesia and a Personal Past (PDF, 585KB) Psychological Review April 2015 by Patricia J. Bauer
• Terminal Decline in Well-Being: The Role of Social Orientation (PDF, 238KB) Psychology and Aging March 2016 by Denis Gerstorf, Christiane A. Hoppmann, Corinna E. Löckenhoff, Frank J. Infurna, Jürgen Schupp, Gert G. Wagner, and Nilam Ram
• Student Threat Assessment as a Standard School Safety Practice: Results From a Statewide Implementation Study (PDF, 97KB) School Psychology Quarterly June 2018 by Dewey Cornell, Jennifer L. Maeng, Anna Grace Burnette, Yuane Jia, Francis Huang, Timothy Konold, Pooja Datta, Marisa Malone, and Patrick Meyer
• Can a Learner-Centered Syllabus Change Students’ Perceptions of Student–Professor Rapport and Master Teacher Behaviors? (PDF, 90KB) Scholarship of Teaching and Learning in Psychology September 2016 by Aaron S. Richmond, Jeanne M. Slattery, Nathanael Mitchell, Robin K. Morgan, and Jared Becknell
• Adolescents' Homework Performance in Mathematics and Science: Personal Factors and Teaching Practices (PDF, 170KB) Journal of Educational Psychology November 2015 by Rubén Fernández-Alonso, Javier Suárez-Álvarez, and José Muñiz
• Teacher-Ready Research Review: Clickers (PDF, 55KB) Scholarship of Teaching and Learning in Psychology September 2015 by R. Eric Landrum
• Enhancing Attention and Memory During Video-Recorded Lectures (PDF, 83KB) Scholarship of Teaching and Learning in Psychology March 2015 by Daniel L. Schacter and Karl K. Szpunar
• The Alleged "Ferguson Effect" and Police Willingness to Engage in Community Partnership (PDF, 70KB) Law and Human Behavior February 2016 by Scott E. Wolfe and Justin Nix
• Randomized Controlled Trial of an Internet Cognitive Behavioral Skills-Based Program for Auditory Hallucinations in Persons With Psychosis (PDF, 92KB) Psychiatric Rehabilitation Journal September 2017 by Jennifer D. Gottlieb, Vasudha Gidugu, Mihoko Maru, Miriam C. Tepper, Matthew J. Davis, Jennifer Greenwold, Ruth A. Barron, Brian P. Chiko, and Kim T. Mueser
• Preventing Unemployment and Disability Benefit Receipt Among People With Mental Illness: Evidence Review and Policy Significance (PDF, 134KB) Psychiatric Rehabilitation Journal June 2017 by Bonnie O'Day, Rebecca Kleinman, Benjamin Fischer, Eric Morris, and Crystal Blyler
• Sending Your Grandparents to University Increases Cognitive Reserve: The Tasmanian Healthy Brain Project (PDF, 88KB) Neuropsychology July 2016 by Megan E. Lenehan, Mathew J. Summers, Nichole L. Saunders, Jeffery J. Summers, David D. Ward, Karen Ritchie, and James C. Vickers
• The Foundational Principles as Psychological Lodestars: Theoretical Inspiration and Empirical Direction in Rehabilitation Psychology (PDF, 68KB) Rehabilitation Psychology February 2016 by Dana S. Dunn, Dawn M. Ehde, and Stephen T. Wegener
• Feeling Older and Risk of Hospitalization: Evidence From Three Longitudinal Cohorts (PDF, 55KB) Health Psychology Online First Publication — February 11, 2016 by Yannick Stephan, Angelina R. Sutin, and Antonio Terracciano
• Anger Intensification With Combat-Related PTSD and Depression Comorbidity (PDF, 81KB) Psychological Trauma: Theory, Research, Practice, and Policy January 2016 by Oscar I. Gonzalez, Raymond W. Novaco, Mark A. Reger, and Gregory A. Gahm
• Special Issue on eHealth and mHealth: Challenges and Future Directions for Assessment, Treatment, and Dissemination (PDF, 32KB) Health Psychology December 2015 by Belinda Borrelli and Lee M. Ritterband
• Posttraumatic Growth Among Combat Veterans: A Proposed Developmental Pathway (PDF, 110KB) Psychological Trauma: Theory, Research, Practice, and Policy July 2015 by Sylvia Marotta-Walters, Jaehwa Choi, and Megan Doughty Shaine
• Racial and Sexual Minority Women's Receipt of Medical Assistance to Become Pregnant (PDF, 111KB) Health Psychology June 2015 by Bernadette V. Blanchfield and Charlotte J. Patterson
• An Examination of Generational Stereotypes as a Path Towards Reverse Ageism (PDF, 205KB) The Psychologist-Manager Journal August 2017 By Michelle Raymer, Marissa Reed, Melissa Spiegel, and Radostina K. Purvanova
• Sexual Harassment: Have We Made Any Progress? (PDF, 121KB) Journal of Occupational Health Psychology July 2017 By James Campbell Quick and M. Ann McFadyen
• Multidimensional Suicide Inventory-28 (MSI-28) Within a Sample of Military Basic Trainees: An Examination of Psychometric Properties (PDF, 79KB) Military Psychology November 2015 By Serena Bezdjian, Danielle Burchett, Kristin G. Schneider, Monty T. Baker, and Howard N. Garb
• Cross-Cultural Competence: The Role of Emotion Regulation Ability and Optimism (PDF, 100KB) Military Psychology September 2015 By Bianca C. Trejo, Erin M. Richard, Marinus van Driel, and Daniel P. McDonald
• The Effects of Stress on Prospective Memory: A Systematic Review (PDF, 149KB) Psychology & Neuroscience September 2017 by Martina Piefke and Katharina Glienke
• Don't Aim Too High for Your Kids: Parental Overaspiration Undermines Students' Learning in Mathematics (PDF, 164KB) Journal of Personality and Social Psychology November 2016 by Kou Murayama, Reinhard Pekrun, Masayuki Suzuki, Herbert W. Marsh, and Stephanie Lichtenfeld
• Sex Differences in Sports Interest and Motivation: An Evolutionary Perspective (PDF, 155KB) Evolutionary Behavioral Sciences April 2016 by Robert O. Deaner, Shea M. Balish, and Michael P. Lombardo
• Asian Indian International Students' Trajectories of Depression, Acculturation, and Enculturation (PDF, 210KB) Asian American Journal of Psychology March 2016 By Dhara T. Meghani and Elizabeth A. Harvey
• Cynical Beliefs About Human Nature and Income: Longitudinal and Cross-Cultural Analyses (PDF, 163KB) January 2016 Journal of Personality and Social Psychology by Olga Stavrova and Daniel Ehlebracht
• Annual Review of Asian American Psychology, 2014 (PDF, 384KB) Asian American Journal of Psychology December 2015 By Su Yeong Kim, Yishan Shen, Yang Hou, Kelsey E. Tilton, Linda Juang, and Yijie Wang
• Resilience in the Study of Minority Stress and Health of Sexual and Gender Minorities (PDF, 40KB) Psychology of Sexual Orientation and Gender Diversity September 2015 by Ilan H. Meyer
• Self-Reported Psychopathy and Its Association With Criminal Cognition and Antisocial Behavior in a Sample of University Undergraduates (PDF, 91KB) Canadian Journal of Behavioural Science / Revue canadienne des sciences du comportement July 2015 by Samantha J. Riopka, Richard B. A. Coupland, and Mark E. Olver

Journals Publishing Resource Center

Find resources for writing, reviewing, and editing articles for publishing with APA Journals™.

Visit the resource center

Journals information

• Frequently Asked Questions
• Journal statistics and operations data
• Special Issues
• Email alerts
• Copyright and permissions

Contact APA Publications

Oxford University Press's Academic Insights for the Thinking World

Effective ways to communicate research in a journal article

Oxford Academic: Journals

We publish over 500 high-quality journals, with two-thirds in partnership with learned societies and prestigious institutions. Our diverse journal offerings ensure that your research finds a home alongside award-winning content, reaching a global audience and maximizing impact.

• By Megan Taphouse , Anne Foster , Eduardo Franco , Howard Browman , and Michael Schnoor
• August 12 th 2024

In this blog post, editors of OUP journals delve into the vital aspect of clear communication in a journal article. Anne Foster (Editor of Diplomatic History ), Eduardo Franco (Editor-in-Chief of JNCI: Journal of the National Cancer Institute and JNCI Monographs ), Howard Browman (Editor-in-Chief of ICES Journal of Marine Science ), and Michael Schnoor (Editor-in-Chief of Journal of Leukocyte Biology ) provide editorial recommendations on achieving clarity, avoiding common mistakes, and creating an effective structure.

Ensuring clear communication of research findings

AF : To ensure research findings are clearly communicated, you should be able to state the significance of those findings in one sentence—if you don’t have that simple, clear claim in your mind, you will not be able to communicate it.

MS : The most important thing is clear and concise language. It is also critical to have a logical flow of your story with clear transitions from one research question to the next.

EF : It is crucial to write with both experts and interested non-specialists in mind, valuing their diverse perspectives and insights.

Common mistakes that obscure authors’ arguments and data

AF : Many authors do a lovely job of contextualizing their work, acknowledging what other scholars have written about the topic, but then do not sufficiently distinguish what their work is adding to the conversation.

HB : Be succinct—eliminate repetition and superfluous material. Do not attempt to write a mini review. Do not overinterpret your results or extrapolate far beyond the limits of the study. Do not report the same data in the text, tables, and figures.

The importance of the introduction

AF : The introduction is absolutely critical. It needs to bring them straight into your argument and contribution, as quickly as possible.

EF : The introduction is where you make a promise to the reader. It is like you saying, “I identified this problem and will solve it.” What comes next in the paper is how you kept that promise.

Structural pitfalls

EF : Remember, editors are your first audience; make sure your writing is clear and compelling because if the editor cannot understand your writing, chances are that s/he will reject your paper without sending it out for external peer review.

HB : Authors often misplace content across sections, placing material in the introduction that belongs in methods, results, or discussion, and interpretive phrases in results instead of discussion. Additionally, they redundantly present information in multiple sections.

Creating an effective structure

AF : I have one tip which is more of a thinking and planning strategy. I write myself letters about what I think the argument is, what kinds of support it needs, how I will use the specific material I have to provide that support, how it fits together, etc.

EF : Effective writing comes from effective reading—try to appreciate good writing in the work of others as you read their papers. Do you like their writing? Do you like their strategy of advancing arguments? Are you suspicious of their methods, findings, or how they interpret them? Do you see yourself resisting? Examine your reactions. You should also write frequently. Effective writing is like a physical sport; you develop ‘muscle memory’ by hitting a golf ball or scoring a 3-pointer in basketball.

The importance of visualizing data and findings

MS : It is extremely important to present your data in clean and well-organized figures—they act as your business card. Also, understand and consider the page layout and page or column dimensions of your target journal and format your tables and figures accordingly.

EF : Be careful when cropping gels to assemble them in a figure. Make sure that image contrasts are preserved from the original blots. Image cleaning for the sake of readability can alter the meaning of results and eventually be flagged by readers as suspicious.

The power of editing

AF : Most of the time, our first draft is for ourselves. We write what we have been thinking about most, which means the article reflects our questions, our knowledge, and our interests. A round or two of editing and refining before submission to the journal is valuable.

HB : Editing does yourself a favour by minimizing distractions-annoyances-cosmetic points that a reviewer can criticize. Why give reviewers things to criticize when you can eliminate them by submitting a carefully prepared manuscript?

Editing mistakes to avoid

AF : Do not submit an article which is already at or above the word limit for articles in the journal. The review process rarely asks for cuts; usually, you will be asked to clarify or add material. If you are at the maximum word count in the initial submission, you then must cut something during the revision process.

EF : Wait 2-3 days and then reread your draft. You will be surprised to see how many passages in your great paper are too complicated and inscrutable even for you. And you wrote it!

Featured image by Charlotte May via Pexels .

Megan Taphouse , Marketing Executive

Anne Foster , (Editor of Diplomatic History)

Eduardo Franco , (Editor-in-Chief of JNCI: Journal of the National Cancer institute and JNCI Monographs)

Howard Browman , (Editor-in-Chief of ICES Journal of Marine Science)

Michael Schnoor , (Editor-in-Chief of Journal of Leukocyte Biology)

• Publishing 101
• Series & Columns

Our Privacy Policy sets out how Oxford University Press handles your personal information, and your rights to object to your personal information being used for marketing to you or being processed as part of our business activities.

We will only use your personal information to register you for OUPblog articles.

Or subscribe to articles in the subject area by email or RSS

Related posts:

Recent Comments

There are currently no comments.

Leave a Comment

Your email address will not be published. Required fields are marked *

Save my name, email, and website in this browser for the next time I comment.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 07 August 2024

Highest ocean heat in four centuries places Great Barrier Reef in danger

• Benjamin J. Henley   ORCID: orcid.org/0000-0003-3940-1963 1 , 2 , 3 ,
• Helen V. McGregor   ORCID: orcid.org/0000-0002-4031-2282 1 , 2 ,
• Andrew D. King   ORCID: orcid.org/0000-0001-9006-5745 4 , 5 ,
• Ove Hoegh-Guldberg   ORCID: orcid.org/0000-0001-7510-6713 6 ,
• Ariella K. Arzey 1 , 2 ,
• David J. Karoly 4 ,
• Janice M. Lough 7 ,
• Thomas M. DeCarlo   ORCID: orcid.org/0000-0003-3269-1320 8 , 9 &
• Braddock K. Linsley   ORCID: orcid.org/0000-0003-2085-0662 10  

Nature volume  632 ,  pages 320–326 ( 2024 ) Cite this article

16k Accesses

1 Citations

2973 Altmetric

Metrics details

• Climate change
• Environmental impact
• Palaeoclimate

Mass coral bleaching on the Great Barrier Reef (GBR) in Australia between 2016 and 2024 was driven by high sea surface temperatures (SST) 1 . The likelihood of temperature-induced bleaching is a key determinant for the future threat status of the GBR 2 , but the long-term context of recent temperatures in the region is unclear. Here we show that the January–March Coral Sea heat extremes in 2024, 2017 and 2020 (in order of descending mean SST anomalies) were the warmest in 400 years, exceeding the 95th-percentile uncertainty limit of our reconstructed pre-1900 maximum. The 2016, 2004 and 2022 events were the next warmest, exceeding the 90th-percentile limit. Climate model analysis confirms that human influence on the climate system is responsible for the rapid warming in recent decades. This attribution, together with the recent ocean temperature extremes, post-1900 warming trend and observed mass coral bleaching, shows that the existential threat to the GBR ecosystem from anthropogenic climate change is now realized. Without urgent intervention, the iconic GBR is at risk of experiencing temperatures conducive to near-annual coral bleaching 3 , with negative consequences for biodiversity and ecosystems services. A continuation on the current trajectory would further threaten the ecological function 4 and outstanding universal value 5 of one of Earth’s greatest natural wonders.

Similar content being viewed by others

Mesophotic coral bleaching associated with changes in thermocline depth

Atypical weather patterns cause coral bleaching on the Great Barrier Reef, Australia during the 2021–2022 La Niña

Internal tides can provide thermal refugia that will buffer some coral reefs from future global warming

Like many coral reefs globally, the World Heritage-listed GBR in Australia is under threat 4 , 6 . Mass coral bleaching, declining calcification rates 5 , 7 , outbreaks of crown-of-thorns starfish ( Acanthaster spp.) 8 , severe tropical cyclones 9 and overfishing 10 have placed compounding detrimental pressures on the reef ecosystem. Coral bleaching typically occurs when heat stress triggers the breakdown of the symbiosis between corals and their symbiotic dinoflagellates 11 . Although coral bleaching can occur locally as a result of low salinity, cold waters or pollution, regional and global mass bleaching events, in which the majority of corals in one or more regions bleach at once, are strongly associated with increasing SST linked to global warming 2 .

The first modern observations of mass coral bleaching on the GBR occurred in the 1980s, but these events were less widespread and generally less severe 3 than the bleaching events in the twenty-first century 4 . Stress bands in coral skeletal cores have provided potential evidence for pre-1980s bleaching in the GBR and Coral Sea, such as during the 1877–78 El Niño 12 . However, stress bands are evident in relatively few cores before 1980 (ref. 12 ),  suggesting that severe mass bleaching did not occur in the 1800s and most of the 1900s.

As the oceans have warmed, however, mass coral bleaching events have become increasingly lethal to corals 4 . Coral bleaching on the GBR 1 in 1998 coincided with a strong eastern-Pacific El Niño, and in 2002 with a weak El Niño. El Niño events can induce lower cloud cover and increased solar irradiance over the GBR 13 , increasing the risk of thermal stress and mass bleaching events 14 . In 2004, water temperatures were anomalously warm, and although bleaching occurred in the Coral Sea 15 , it was not widespread in the GBR, probably because there was reduced upwelling and an associated reduced influence of nutrients on symbiotic dinoflagellate expulsion 16 .

However, in the nine January–March periods from 2016 to 2024 (inclusive) there were five mass coral bleaching events on the GBR. Each was associated with high SSTs and affected large sections of the reef. GBR mass bleaching occurred in both 2016 and 2017, influenced by the presence of an El Niño event in 2016, and led to the death of at least 50% of shallow-water (depths of 5–10 m) reef-building corals 4 . Major bleaching events occurred again in quick succession in 2020 and 2022, with the accumulated heat stress for large sections of the GBR reaching levels conducive to widespread bleaching but lower levels of coral mortality 1 . The bleaching event in 2022 occurred, unusually, during a La Niña event, which is typically associated with cooler summer SSTs, higher than average rainfall and higher cloud cover on the GBR 1 . At the time of writing, researchers are assessing the impacts of the 2024 mass bleaching event.

The frequency of recent mass coral bleaching and mortality on the GBR is cause for concern. In 2021, the World Heritage Committee of the United Nations Educational, Scientific and Cultural Organization (UNESCO) drafted 17 a decision to inscribe the GBR on the List of World Heritage in Danger, stating that the reef is “facing ascertained danger”, citing recent mass coral bleaching events and insufficient progress by the State Party (Australia) in countering climate change, improving water quality and land management issues. The committee’s adopted decisions 18 have not included inscription of the ‘in danger’ status, but the draft inscription highlights the seriousness of the recent mass coral bleaching events. Authorities in Australia 5 have noted that climate change and coral bleaching have deteriorated the integrity of the outstanding universal value of the GBR, a defining feature of its World Heritage status.

Although rapidly rising SSTs are attributed to human activities with virtual certainty 19 , understanding the multi-century SST history of the GBR is critical to understanding the influence of SST on mass coral bleaching and mortality in recent decades. Putting aside a problematic attempt to do this 20 , which was discredited 21 , 22 , knowledge of the long-term context for GBR SSTs comes primarily from two multi-century reconstructions based on the geochemistry of coral cores collected from the inner shelf 23 and outer shelf 24 (Flinders Reef) in the central GBR. These reconstructions showed that SSTs in the early 2000s were not unusually high relative to levels in the past three centuries, with five-year mean SSTs (and salinities) estimated to be higher in the 1700s than in the 1900s. However, these records were limited by their relatively coarse five-year sampling resolution and their most recent data point being from the early 2000s. After these studies were published, SSTs in the GBR have continued to rise. Updated analysis of coral data from Flinders Reef provides valuable improved temporal resolution 25 , but interpretations of these records remain limited spatially.

Here, we investigate the recent high SST events in the GBR region in the context of the past four centuries. We combine a network of 22 coral Sr/Ca and δ 18 O palaeothermometer series (Supplementary Tables 1 and 2 ) located in and near to the Coral Sea region to infer spatial mean SST anomalies (SSTAs) for January–March, the months when maximum SST and thermal bleaching are most likely to occur in the Coral Sea 16 , 26 , each year from 1618 to 1995 ( Methods and Supplementary Information ). Anthropogenic climate change began and proceeded entirely within the multi-century lives of some of these massive coral colonies, offering a continuous multi-century record covering the industrial era. We use this 1618–1995 reconstruction and the available 1900–2024 instrumental data to contextualize the modern trend and rank four centuries of January–March SSTAs with greater precision than was previously possible. We then assess the degree of human influence on ocean temperatures in the region using climate model simulations run both with and without anthropogenic forcing.

The instrumental period (1900–present)

Mass coral bleaching on the GBR in 2016, 2017, 2020, 2022 and 2024 during January–March coincided with widespread warm SSTAs in the surrounding seas 1 , including the Coral Sea (Fig. 1a–e , using ERSSTv5 data 27 ). The Coral Sea and GBR have experienced a strong warming trend since 1900 (Fig. 1f ). January–March SSTAs averaged over the GBR are strongly correlated ( ρ  = 0.84, P   ≪  0.01) with those in the broader Coral Sea (Fig. 1f ), including when the long-term warming trend is removed from both time series ( ρ  = 0.69, P  < 0.01; Supplementary Fig. 4 ). Based on the strength of this correlation, we associate high January–March area-averaged Coral Sea SSTAs with increased thermal bleaching risk in the GBR.

a – e , SSTAs (using ERSSTv5 data) for January–March in the Australasian region relative to the 1961–90 average for the five recent GBR mass coral bleaching years: 2016, 2017, 2020, 2022 and 2024. The black box shows the Coral Sea region (4° S–26° S, 142° E–174° E). f , Coral Sea and GBR mean SSTAs for 1900–2024 in January–March relative to the 1961–90 average. The black vertical lines indicate the five recent GBR mass coral bleaching years.

Record temperatures were set in 2016 and 2017 in the Coral Sea, and in 2020 they peaked fractionally below the record high of 2017. The January–March of 2022 was another warm event, the fifth warmest on record at the time. Recent data (ERSSTv5) indicate that 2024 set a new record by a margin of more than 0.19 °C above the previous record for the region. The January–March mean SSTs averaged over the five mass bleaching years during the period 2016–2024 are 0.77 °C higher than the 1961–90 January–March averages in both the Coral Sea and the GBR. The multidecadal warming trend, extreme years and association between GBR and Coral Sea SSTs are similar for the HadISST 28 gridded SST dataset, with some notable differences in the 1900–40 period (Supplementary Fig. 3 ). Furthermore, analysis of modern temperature-sensitive Sr/Ca series from GBR corals for 1900–2017 provides coherent independent evidence of statistically significant multi-decadal warming trends in January–March SSTs in the central and southern GBR (Supplementary Information section  4.2 ).

A multi-century context (1618–present)

Reconstructing Coral Sea January–March SSTs from 1618 to 1995 extends the century-long instrumental record back in time by an additional three centuries (Fig. 2a and Methods ). The reconstruction (calibrated to ERSSTv5) shows that multi-decadal SST variability was a persistent feature in the past. At the centennial timescale, there is relative stability before 1900, with the exception that cooler temperatures prevailed in the 1600s. Warming during the industrial era has been evident since the early 1900s (Fig. 2a ). There is a warming trend for January–March of 0.09 °C per decade for 1900–2024 and 0.12 °C per decade for 1960–2024 (Fig. 1f ) using ERSSTv5 data. Calibrating our reconstruction to HadISST1.1 yields similar results, with some differences in the degree of pre-1900 variability at both multi-decadal and centennial timescales (Supplementary Information section  5.2.6 ).

a , Reconstructed and observed mean January–March SSTAs in the Coral Sea for 1618–2024 relative to 1961–90. Dark blue, highest skill (maximum coefficient of efficiency) reconstruction with the full proxy network; light blue, 5th–95th-percentile reconstruction uncertainty; black, observed (ERSSTv5) data. Red crosses indicate the five recent mass bleaching events. Dashed lines indicate the best estimate (highest skill, red) and 95th-percentile (pink) uncertainty bound for the maximum pre-1900 January–March SSTA. b , Central GBR SSTA for the inner shelf 23 in thick orange and outer shelf 25 (Flinders Reef) in thin orange lines; these series are aligned here (see Methods ) with modern observations of mean GBR SSTAs for January–March relative to 1961–90. Observed data are shown at annual (grey line) and five-year (black line with open circles, plotted at the centre of each five-year period and temporally aligned with the five-year coral series 23 ) resolution. Dashed lines indicate best-estimate pre-1900 January–March maxima for refs. 23 (red) and 25 (pink). Orange shading indicates 5th–95th-percentile uncertainty bounds. Red crosses indicate the five recent mass bleaching events. c , Evaluation metrics for the Coral Sea reconstruction (Supplementary Information section  3.1 ); RE, reduction of error; CE, coefficient of efficiency; Rsq-cal, R-squared in the calibration period; Rsq-ver, R-squared in the verification (evaluation) period. d , Coral data locations relative to source data region (orange box) and Coral Sea region (red box). Coral proxy metadata are given in Supplementary Tables 1 and 2 .

Our best-estimate (highest skill; Methods ) annual-resolution Coral Sea reconstruction (Fig. 2a ), using the full coral network calibrated to the ERSSTv5 instrumental data, indicates that the January–March mean SSTAs in 2016, 2017, 2020, 2022 and 2024 were, respectively, 1.50 °C, 1.54 °C, 1.53 °C, 1.46 °C and 1.73 °C above the 1618–1899 (hereafter ‘pre-1900’) reconstructed average. Using the same best-estimate reconstruction, Coral Sea January–March SSTs during these GBR mass bleaching years were five of the six warmest years the region has experienced in the past 400 years (Fig. 2a ).

By comparing the recent warm events to the reconstruction’s uncertainty range ( Methods ), we quantify, using likelihood terminology consistent with recent reports from the Intergovernmental Panel on Climate Change 19 , that the recent heat extremes in 2017, 2020 and 2024 are ‘extremely likely’ (>95th percentile; Fig. 2a ) to be higher than any January–March in the period 1618–1899. Furthermore, the heat extremes in 2016 and 2022 are (at least) ‘very likely’ (>90th percentile) to be above the pre-1900 maximum. We perform a series of tests that verify that our findings are not simply an artefact of the nature of the coral network itself (Supplementary Information section 5.2 ). In a network perturbation test, we generate 22 subsets of the reconstruction by adding proxy records incrementally in order from the highest to the lowest correlation with the target (Supplementary Information section  5.2.5 ). We confirm that 2017, 2020 and 2024 were ‘extremely likely’ (>95th percentile) to have been warmer than any year pre-1900 (using ERSSTv5 data) for all of these proxy subsets. Furthermore, in 20 of the 22 subsets, 2016 was also ‘extremely likely’ (>95th percentile), rather than ‘very likely’, to be warmer (2022 was ‘extremely likely’ in 14 of the 22 subsets). All our additional tests, including a reconstruction with only Sr/Ca coral data (thereby omitting the possibility of any non-temperature signal in δ 18 O coral on the reconstruction), achieve high reconstruction skill and confirm the extraordinary nature of recent extreme temperatures in the multi-century context (Supplementary Information section  5.2 ). Analyses using HadISST1.1 generally show lower correlations with the coral data and reconstructions with slightly warmer regional SSTs before 1900, along with more-muted centennial and multi-decadal variability in the pre-instrumental period. Nevertheless, the HadISST1.1-calibrated reconstructions show that the recent thermal extremes are well above the best estimate (highest skill) of the pre-1900 maximum of reconstructed January–March SSTAs (Supplementary Fig. 42 ). Furthermore, lower SSTAs (in the HadISST1.1 data) relative to the previous three centuries (as in our reconstructions calibrated to HadISST1.1), coupled with the recently observed mass coral bleaching events, could indicate that long-lived corals have a greater sensitivity to warming than is currently recognized.

Reconstructed regional GBR SSTAs based on a five-year-resolution, multi-century coral δ 18 O record from the central inshore GBR 23 (Fig. 2b ) show similarly strong warming since 1900 but more multi-decadal-to-centennial variability than the Coral Sea reconstruction. Recent five-year mean January–March GBR SSTAs narrowly exceed the best estimate of the maximum pre-1900 five-year mean since the early 1600s (Fig. 2b ). The averages for the five-year periods centred on 2018 and 2022 exceed the pre-1900 maximum by 0.11 °C and 0.06 °C, respectively. Results are similar using the five-year-resolution Flinders Reef (central outer shelf) 24 record (Supplementary Fig. 24 ), although its interpretation is limited by the lack of uncertainty estimates available for that record. Our Coral Sea reconstruction incorporates an updated (annual resolution) record from Flinders Reef 25 , which indicates similar centennial trends (thin orange line in Fig. 2b ) and shows that the recent high January–March SSTA events have approached the estimated local pre-1900 maximum SSTA. Although contiguous multi-century cores from within the GBR are limited in their spatial extent, twentieth-century warming is evident in these records.

The extraordinary nature of the recent Coral Sea January–March SSTs in the context of the past 400 years is further illustrated by comparing the ranked temperature anomalies (Fig. 3 ) for the combined reconstructed and instrumental period from 1618–2024, incorporating reconstruction uncertainty ( Methods ). The mass coral bleaching years of 2016, 2017, 2020, 2022 and 2024, and the heat event of 2004, stand out as the warmest events across the whole 407-year record. The warmest three years (2024, 2017 and 2020) exceed the upper uncertainty bound (95th percentile) of the warmest reconstructed January–March in the pre-1900 period (pink (upper) dashed line in Fig. 3 ); 2016, 2004 and 2022 exceed the 90th percentile bound (red (lower) dashed line in Fig. 3 ). The warming trend is clear in the association between the ascending rank of the temperature anomalies and the year (shown as the colour of the filled circles in Fig. 3 ). Despite high interannual variability, 78 of the warmest 100 January–March periods between 1618 and 2024 occurred after 1900, and the 23 warmest all occur after 1900. The warmest 20 January–March periods all occur after 1950, coinciding with accelerated global warming.

Ranked January–March SSTAs for 1618–2024 relative to 1961–90 (coloured circles) from the best-estimate (highest skill, full coral network) reconstruction (1618–1899) and instrumental (ERSSTv5) data (1900–2024). The year is indicated by the colour of the filled circles. The 5th–95th-percentile uncertainty bounds of the pre-1900 reconstructed SSTAs are shown by small grey dots. The year labels indicate the warmest six years on record, five of which were mass coral bleaching years on the GBR. The pink (upper) dashed line indicates the 95th-percentile uncertainty bound of the maximum pre-1900 reconstructed SSTA; the red (lower) dashed line indicates the 90th-percentile limit.

Assessing anthropogenic influence

Using climate model simulations from the most recent (sixth) phase of the Coupled Model Intercomparison Project 29 (CMIP6), we assess the human influence on January–March SSTAs in the Coral Sea. The model simulations are from two experiments in the Detection and Attribution Model Intercomparison Project (DAMIP) 30 . The first set of simulations represents historical climate conditions, including both the natural and human influences on the climate system over the 1850–2014 period (‘historical’; red in Fig. 4 ). The second experiment is a counterfactual climate that spans the same period and uses the same models but includes only natural influences on the climate, omitting all human influences (‘historical-natural’; blue in Fig. 4 ). The historical experiment includes anthropogenic emissions of greenhouse gases and aerosols, stratospheric ozone changes and anthropogenic land-use changes; the historical-natural experiment does not. Variations in natural climate forcings, such as from volcanic eruptions and solar variability, are incorporated in both experiments. We include models that have a transient climate response (the global mean surface-temperature anomaly at the time of a doubling of atmospheric CO 2 concentration) in the range 1.4–2.2 °C, which is deemed ‘likely’ by the science community 31 ( Methods and Supplementary Information ).

Climate-model simulations of Coral Sea January–March SSTAs relative to the 1850–1900 average for the period 1850–2014, for models within the ‘likely’ range for their transient climate response 31 . The blue line (median) and light blue shading (5th–95th-percentile limits) are from the ‘historical-natural’ climate model simulations (no anthropogenic climate forcing); the red line and light red shading are from the ‘historical’ simulations (anthropogenic influences on the climate included) using the same set of climate models. The climate-model-derived time of emergence of anthropogenic climate change, shown by the grey and black vertical lines (1976 and 1997), is when the ratio of the climate change signal to the standard deviation of noise/variability 32 across model ensemble members first rises above 1 and 2, respectively. All models are represented equally in the model ensemble.

It is only with the incorporation of anthropogenic influences on the climate that the model simulations capture the modern-era warming of the Coral Sea January–March SSTA (Fig. 4 ). The median of the historical simulations has statistically significant warming trends of 0.05 °C, 0.10 °C and 0.15 °C per decade for the periods from 1900, 1950 and 1970 to 2014, respectively; the equivalent historical-natural trends are smaller in magnitude than ±0.01 °C per decade. To further explore the centennial-scale trends, we use a bootstrap ensemble ( Methods ) of the two sets of 165-year simulations from 1850–2014. We found that 100% of the historical bootstrap ensemble has statistically significant positive trends ( Methods ) for 1900–2014, but this value is 0% for the historical-natural ensemble. The observed (ERSSTv5) mean SSTA for 2016–2024 of 0.60 °C relative to 1961–90 is warmer than any nine-year sequence in the 7,095 simulated years in the historical-natural experiments from models with transient climate responses in the ‘likely’ range 31 .

We also use the simulations to estimate the time of emergence of the anthropogenic influence on January–March Coral Sea SSTAs above the natural background variability. The anthropogenic warming signal 32 increases from near zero in 1900 to around 0.5 standard deviations of the variability (‘noise’) in 1960. The climate change signal-to-noise ratio then increases rapidly from 1960 to 2014, exceeding 1.0 in 1976, 2.0 in 1997 and around 2.8 by 2014, the end of these simulations (Fig. 4 , Methods and Supplementary Fig. 50 ). Anthropogenic impacts on the climate are virtually certain to be the primary driver of this long-term warming in the Coral Sea.

Previously, our knowledge of the SST history of the GBR and the Coral Sea region has been highly dependent on instrumental observations, with the exception of the five-year-resolution multi-century coral Sr/Ca and U/Ca SST reconstructions from the two point locations in the central GBR 23 , 24 , an update at one of these locations 25 , seasonal resolution ‘floating’ (in time) chronologies from the GBR in the Holocene 33 , 34 and point SST estimates further back in time 35 . Thus, the context of recent warming trends in the Coral Sea and GBR and their relation to natural variability on decadal to centennial timescales is largely unknown without reconstructions such as the one we developed here.

Our coral proxy network is located mostly beyond the GBR, in the Coral Sea, and some series are located outside the Coral Sea region (Fig. 2d ). The selection of the Coral Sea as a study region allowed for a larger sample of contributing coral proxy data than exists for the GBR. However, coral bleaching on the GBR can be influenced by factors other than large-scale SST, including local oceanic and atmospheric dynamics that can modulate the occurrence and severity of thermal bleaching and mortality events 13 . Nonetheless, warming of seasonal SSTs over the larger Coral Sea region is likely to prime the background state and increase the likelihood of smaller spatio-temporal-scale heat anomalies. Furthermore, where we use only the five-year resolution series directly from the GBR to reconstruct GBR SSTAs, we draw similar conclusions about the long-term trajectory of SSTAs as for our full coral network (Fig. 2b and Supplementary Fig. 24 ). Furthermore, short modern coral series from within the GBR, analysed in this study, document a multi-decadal warming signal that is coherent with instrumental data (Supplementary Figs. 29 and 30 ). Nonetheless, additional high-resolution, multi-century, temperature-sensitive coral geochemical series from within the GBR would help unravel the local and remote ocean–atmosphere contributions to past bleaching events and reduce uncertainties.

The focus on the larger Coral Sea study region also takes advantage of the global modelling efforts of CMIP6. The large number of ensemble members available for CMIP6 means that greater climate model diversity, and therefore greater certainty in our attribution analysis, is possible compared with most single model analyses. There is also a methodological benefit in having high replication of the same experiments run with multiple climate models. However, coarse-resolution global-scale models do not accurately simulate smaller-scale processes, such as inshore currents and mesoscale eddies in the Coral Sea or the Gulf of Carpentaria, which probably affect local surface temperatures and variations in nutrient upwelling in the GBR 36 , 37 . Upwelling on the GBR is linked to the strength of the East Australian Current 16 , the southward branch of the South Pacific subtropical gyre. The CMIP-scale models we use do capture these gyre dynamics. The models show that the East Australian Current is expected to increase in strength as the climate continues to warm through this century 38 , and this may lead to more nutrient inputs that can exacerbate coral sensitivity to rising heat stress 39 , 40 . As well as focusing our model analysis on the larger Coral Sea region, we use a three-month time step. In doing so, we minimize the impact of model spatio-temporal resolution on our inferences about the role of anthropogenic greenhouse-gas emissions on the SST conditions that give rise to GBR mass bleaching.

Remaining uncertainties

We present analyses and interpretations that are as robust as possible given currently available data and methods. However, several sources of remaining uncertainty mean that future reconstructions of past Coral Sea and GBR SSTs could differ from those presented here. Although bias corrections are applied to observational SST datasets such as ERSST and HadISST, these datasets probably retain biases, especially for the period during and before 1945 (ref. 41 ), and these may not be fully accounted for in the uncertainty estimates 42 . Because our reconstructions are calibrated directly to these datasets, future observational-bias corrections are likely to improve proxy-based reconstructions.

Reconstructions of SST that use coral δ 18 O records may be susceptible to the influence of changes in the coral δ 18 O–SST relationship on time periods longer than the instrumental training period, along with non-SST changes in the δ 18 O of seawater, which can covary with salinity. As such, new coral records of temperature-sensitive trace-element ratios such as Sr/Ca, Li/Mg or U/Ca may prove influential in future efforts to distinguish between changes in past temperature and hydroclimate. Owing to the limited availability of multi-century coral data from within the GBR itself, the reconstructed low-frequency variability of GBR SSTs in recent centuries is likely to change as more temperature proxy data become available. It is also likely that new sub-annual resolution records would aid in removing potential signal damping or bias from our use of some annual-resolution records to reconstruct seasonal SSTAs.

Ecological consequences

With global warming of 0.8–1.1 °C above pre-industrial levels 19 there has been a marked increase in mass coral bleaching globally 43 . Even limiting global warming to the Paris Agreement’s ambitious 1.5 °C level would be likely to lead to the loss of 70–90% of corals that are on reefs today 44 . If all current international mitigation commitments are implemented, global mean surface temperature is still estimated to increase in the coming decades, with estimates varying between 1.9 °C (ref. 45 ) and 3.2 °C (ref. 46 ) above pre-industrial levels by the end of this century. Global warming above 2 °C would have disastrous consequences for coral ecosystems 19 , 44 and the hundreds of millions of people who currently depend on them.

Coral reefs of the future, if they can persist, are likely to have a different community structure to those in the recent past, probably one with much less diversity in coral species 4 . This is because mass bleaching events have a differential impact on different coral species. For example, fast-growing branching and tabulate corals are affected more than slower-growing massive species because they have different thermal tolerance 4 . The simplification of reef structures will have adverse impacts on the many thousands of species that rely on the complex three-dimensional structure of reefs 4 . Therefore, even with an ambitious long-term international mitigation goal, the ecological function 4 of the GBR is likely to deteriorate further 5 before it stabilizes.

Coral adaptation and acclimatization may be the only realistic prospect for the conservation of some parts of the GBR this century. However, although adaptation opportunities may be plausible to some extent 47 , they are no panacea because evolutionary changes to fundamental variables such as temperature take decades, if not centuries, to occur, especially in long-lived species such as reef-building corals 48 . There is currently no clear evidence of the real-time evolution of thermally tolerant corals 48 . Most rapid changes depend on a history of exposure to key genetic types and extremes, and there are limitations to genetic adaptation that prevent species-level adaptation to environments outside of their ecological and evolutionary history 19 . Model projections also indicate that rates of coral adaptation are too slow to keep pace with global warming 49 . In a rapidly warming world, the temperature conditions that give rise to mass coral bleaching events are likely to soon become commonplace. So, although we may see some resilience of coral to future marine heat events through acclimatization, thermal refugia are likely to be overwhelmed 50 . Global warming of more than 1.5 °C above pre-industrial levels will probably be catastrophic for coral reefs 44 .

Our new multi-century reconstruction illustrates the exceptional nature of ocean surface warming in the Coral Sea today and the resulting existential risk for the reef-building corals that are the backbone of the GBR. The reconstruction shows that SSTs were relatively cool and stable for hundreds of years, and that recent January–March ocean surface heat in the Coral Sea is unprecedented in at least the past 400 years. The coral colonies and reefs that have lived through the past several centuries, and that yielded the valuable Sr/Ca and δ 18 O data on which our reconstruction is based, are themselves under serious threat. Our analysis of climate-model simulations confirms that human influence is the driver of recent January–March Coral Sea surface warming. Together, the evidence presented in our study indicates that the GBR is in danger. Given this, it is conceivable that UNESCO may in the future reconsider its determination that the iconic GBR is not in danger. In the absence of rapid, coordinated and ambitious global action to combat climate change, we will likely be witness to the demise of one of Earth’s great natural wonders.

Instrumental observations

The Coral Sea and GBR area-averaged monthly SSTAs relative to 1961–90 for January–March are obtained from version 5 of the Extended Reconstructed Sea Surface Temperature dataset (ERSSTv5) 27 . We compare our results using ERSSTv5 with those generated using the Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST1.1) 28 . We use only post-1900 instrumental SST observations here. Although gridded datasets have some coverage before 1900, ship-derived temperature data in the region for that period are too sparse to be reliable for calibrating our reconstruction (Supplementary Information section  1.2 ). The regional mean for the GBR is computed using the seven grid-cell locations used by the Australian Bureau of Meteorology (Supplementary Information section  1.1 ). We define the Coral Sea region as the ocean areas inside 4° S–26° S, 142° E–174° E.

Coral-derived temperature proxy data

We use a network of 22 published and publicly available sub-annual and annual resolution temperature-sensitive coral geochemical series (proxies; Fig. 2d , Supplementary Tables 1 and 2 , and Supplementary Fig. 5a–v ) from the western tropical Pacific in our source data region (4° N–27° S, 134° E–184° E) that cover at least the period from 1900 to 1995. Of these 22 series, 16 are δ 18 O, which are in per mil (‰) notation relative to Vienna PeeDee Belemnite (VPDB) 51 ; the remaining six are Sr/Ca series. The coral data are used as predictors in the reconstruction of January–March mean SSTAs in the Coral Sea region. We apply the inverse Rosenblatt transformation 52 , 53 to the coral data to ensure that our reconstruction predictors are normally distributed. Sub-annually resolved series are converted to the annual time step by averaging across the November–April window. This maximizes the detection of the summer peak values, allowing for some inaccuracy in sub-annual dating and the timing of coral skeleton deposition 54 , 55 . A small fraction (less than 0.8%) of missing data is infilled using the regularized expectation maximization (RegEM) algorithm 56 (Supplementary Information section  2.3 ), after which the proxy series are standardized such that each has a mean of zero and a standard deviation of one over their common 1900–1995 period.

Reconstruction method

To produce our Coral Sea reconstruction, we use nested principal component regression 57 (PCR), in which the principal components of the network of 22 coral proxies are used as regressors against the target-region January–March SSTA relative to the 1961–90 average. We perform the reconstructions separately for each nest of proxies, where a nest is a set of proxies that cover the same time period. The longest nest dates back to 1618, when at least two series are available. The nests allow for the use of all coral proxies over the full time period of their coverage. The 96-year portion of the instrumental period (1900–1995) that overlaps with the reconstruction period is used for calibration and evaluation (or equivalently, verification) against observations. We reconstruct regional SSTAs from the principal components of the coral network of δ 18 O and Sr/Ca data, rather than their local SST calibrations, to minimize the number of computational steps and to aid in representing the full reconstruction uncertainty.

Principal component analysis (PCA) is used to reduce the dimensionality of the proxy matrix, as follows. Let P ( t , r ) denote the palaeoclimate-data matrix during the time period t  = 1,..., n at an annual time step for proxy series r  = 1,..., p . PCA is undertaken on this matrix during the calibration period, P cal . We obtain the principal component coefficients matrix P coeff ( r , e ) for principal components e  = 1,..., n PC and principal component scores P score ( t , e ), which are representations of the input matrix P cal in the principal component space. P score is truncated to include n PC,use principal components to form \({P}_{{\rm{score}}}^{{\prime} }\) such that the variance of the proxy network explained by the n PC,use principal components is greater than \({\sigma }_{{\rm{expl}}}^{2}\) (which we set to 95%). Reconstruction tests in which \({\sigma }_{{\rm{expl}}}^{2}\) is varied from 70% to 95% show that our results are not strongly sensitive to this choice, and tests based on lag-one autoregressive noise for \({\sigma }_{{\rm{expl}}}^{2}\) from 50% to 99% further support this choice (Supplementary Information section  3.2 ). These principal components are used as predictors against which the Coral Sea January–March instrumental SSTAs are regressed. We regress the standardized SSTA target data during the calibration period, I cal , against the retained principal components of the predictor data, \({P}_{{\rm{score}}}^{{\prime} }\) :

Thus, we obtain n PC,use estimates of the regression coefficients γ e with gaussian error term ε t  ~  N (0, \({\sigma }_{N}^{2}\) ). The principal components are extended back into the pre-instrumental period by multiplying the entire proxy matrix P ( t , p ) with the truncated principal component coefficient matrix \({P}_{{\rm{coeff}}}^{{\prime} }\) ( t , e ) to obtain \({Q}_{{\rm{coeff}}}^{{\prime} }\) :

The reconstruction proceeds with the fitted regression coefficients γ e and extended coefficient matrix \({Q}_{{\rm{coeff}}}^{{\prime} }\) to obtain a reconstruction time series R m ( t ) for a given nest of proxy series

The standardized reconstruction R m ( t ) is then calibrated to the instrumental data such that the standard deviation and mean of the reconstruction and target during the calibration interval are equal. As well as obtaining reconstructions for each nest of available proxies, we compute stitched reconstructions S c ( t ) for each calibration period c , which include at each time step the reconstructed data for the proxy nest with maximum coefficient of efficiency 58 , 59 (Supplementary Information section  3.1 ). This procedure is performed for contiguous calibration intervals between 60 and 80 years duration between 1900 and 1995, with interval width and location increments of two years, reserving the remaining data in the overlapping period for independent evaluation, and for all proxy nests. The reconstruction error is modelled with a lag-one autoregressive process fitted to the residuals. We evaluate the capacity of our reconstruction method to achieve spurious skill from overfitting by performing a test in which we replace the coral data with synthetic noise (Supplementary Information section  3.2i ). We find that reconstructions based on synthetic noise achieve extremely low or zero skill and as more noise principal components are included in the regression, the evaluation metrics indicate declining skill. Our reconstruction and evaluation methods therefore guard against the potential for spurious skill.

Pseudo-proxy reconstructions

Our reconstruction method is further evaluated by using a pseudo-proxy modelling approach based on the Community Earth System Model (CESM) Last Millennium Experiment (LME) 60 , for which there are 13 full-forcing ensemble members covering the period 850–2005. We use the pseudo-proxy reconstructions to evaluate our reconstruction method and coral network in a fully coupled climate-model environment. We form pseudo-proxies by extracting from each LME ensemble member the SST and sea surface salinity (SSS) from the 1.5° × 1.5° grid cell located nearest to our coral data. We then apply proxy system models in the form of linear regression models, basing δ 18 O on both SST and SSS, and Sr/Ca on SST only (Supplementary Information section  3.3 ). We set the spatial and temporal availability of the pseudo-coral network to match that of the coral network. We then apply our PCR reconstruction and evaluation procedure to the pseudo-proxy network, taking advantage of the availability of the modelled Coral Sea SSTA data across the multi-century period of 1618–2005, which allows for the evaluation of the pseudo-proxy reconstruction over this entire time period. We first test our method using a ‘perfect proxy’ approach (with no proxy measurement error) before superimposing synthetic noise on the pseudo-proxy time series, evaluating our methodology at two separate levels of measurement error, quantified by signal-to-noise ratios of 1.0 and 4.0. The evaluation metrics for these tests indicate that our coral network and reconstruction method obtain skilful reconstructions of Coral Sea SSTAs in the climate-model environment (Supplementary Figs. 17b , 18 , 20b , 21 , 22b and 23 ).

Comparison with independent coral datasets

We use two multi-century five-year-resolution coral series from the central GBR 23 , 24 (Fig. 2b and Supplementary Fig. 24 ) and a network of sub-annual and annual resolution modern coral series (dated from 1900 onwards but not covering the full 1900–1995 period) from 44 sites in the GBR (Supplementary Information section  4.2 ) for independent evaluation of coral-derived evidence for warming in the region. We estimate five-year GBR SSTAs (Fig. 2b ) by aligning the post-1900 mean and variance of the proxy and instrumental (ERSSTv5) data.

Reconstruction sensitivity to non-SST influences

Of the 22 available coral series, 16 are records of δ 18 O, a widely used measure of the ratio of the stable isotopes 18 O and 16 O. In the tropical Pacific Ocean, δ 18 O is significantly correlated with SST 61 , 62 , 63 , 64 . Coral δ 18 O is also sensitive to the δ 18 O of seawater 65 , which can reflect advection of different water masses and/or changes in freshwater input, such as from riverine sources or precipitation, which in turn co-vary with SSS. Thus, it is generally considered that the main non-SST contributions to coral δ 18 O are processes that co-vary with SSS 62 , 66 . Our methodology minimizes the influence of non-temperature impacts on the reconstruction by exploiting the contrast in spatial heterogeneity between SST and SSS in January–March (Supplementary Information section  5.1 ). SSS is spatially inhomogeneous in the tropical Pacific 66 , 67 , leading to low coherence in SSS signals across our coral network. By contrast, the strong and coherent SST signal across our coral network locations and the Coral Sea region leads to principal components that are strongly representative of SST variations. This produces a skilful reconstruction of SST, as determined by evaluation against independent observations, and low correlations with SSS across the Coral Sea region (Supplementary Fig. 31 ).

Although the likelihood of non-SST influences on our SST reconstruction is low, we nonetheless test the sensitivity of our reconstruction and its associated interpretations to the possibility of these influences on the coral data. The tests compute the correlations between our best-estimate SSTA reconstruction (highest coefficient of efficiency) and observations of SSS, along with a series of additional reconstructions based on subsets of our coral network. The correlations between our highest coefficient of efficiency January–March Coral Sea SSTA reconstruction and January–March SSS are mapped for the Coral Sea and its neighbouring domain using three instrumental SSS datasets (Supplementary Fig. 31 ). Correlations are not statistically significant over most of the domain. Noting differing spatial correlation patterns between the instrumental SSS datasets 68 , which also cover different time periods (Supplementary Information section  5.1 ), we undertake six sensitivity tests using subsets of the coral network (Supplementary Information section  5.2 ). We use the following combinations of coral series: (1) the full network of 22 δ 18 O and Sr/Ca series (Figs. 2a and 3 ); (2) a subset of the six available Sr/Ca series (Supplementary Figs. 32 – 33 ), to test how the reconstruction is influenced by the inclusion of coral δ 18 O records; (3) a fixed nest subset of the five longest coral series, extending back to at least 1700 (Supplementary Figs. 34 – 35 ), to test for the potential influence of combining series of differing lengths (from our splicing of portions of the best reconstructions from each nest); (4) a subset of the ten coral series that are most strongly correlated with the target (Supplementary Figs. 36 and 37 ), to test how our reconstruction is influenced by the inclusion of coral series that are less strongly correlated with our target; (5) a subset of coral series that excludes the six records that are reported to potentially include biological mediation or non-climatic effects, or have low correlation with the target (Supplementary Figs. 38 and 39 ), to test their influence on the reconstruction; and (6) a network perturbation test comprising 22 separate subsets of proxies, in which proxy records are added incrementally in order of highest to lowest correlation with the target, starting with a single coral series and increasing the number of included proxies to all 22 series in our network (Supplementary Information section  5.2.5 ), to systematically quantify the influence of gradually including more coral datasets on our reconstruction and its interpretations.

The evaluation metrics (Fig. 2c and Supplementary Figs. 32b , 34b , 36b and 38b ) indicate a skilful reconstruction back to 1618 for the reconstructions based on the Full, Sr/Ca only, Long, Best-10 and OmitBioMed networks. These reconstructions explain 82.7%, 80.6%, 77.6%, 79.8% and 80.4% (R-squared values) of the variance in January–March SSTAs, respectively, in the independent evaluation periods (using ERSSTv5b). All coral subsets in the network perturbation test produce skilful reconstructions (Supplementary Fig. 40 ). The highest-skill reconstructions for all subsets in the network perturbation test align with our key interpretations (Supplementary Figs. 41 and 42 ). Together, our sensitivity tests show that the coral network, observational data and reconstruction methodology are a sound basis for reconstructing Coral Sea January–March SSTAs in past centuries and contextualizing recent high-SST events ( Supplementary Information ).

Climate-model attribution ensembles and experiments

The multi-model attribution analysis used here is based on simulations from CMIP6. We analyse simulations from the historical experiment (including natural and anthropogenic influences for 1850–2014) and the historical-natural experiment (natural-only forcings for 1850–2014). We select climate models for which monthly surface temperature is available in at least three historical and historical-natural simulations (Supplementary Table 5 ). All model simulations are interpolated to a common regular 1.5° × 1.5° latitude–longitude grid. January–March SSTAs relative to 1961–90 are calculated for each simulation. The full historical all-forcings ensemble is composed of 14 models with 268 simulations for 1850–2014. The natural-only ensemble is composed of the same 14 models with 95 individual simulations. A subset of climate models in the CMIP6 ensemble are considered by the science community to be ‘too hot’, simulating warming in response to increased atmospheric carbon dioxide concentrations that is larger than that supported by independent evidence 31 . We omit these models from our analysis by including only models with a transient climate response in the ‘likely’ range 31 of 1.4–2.2 °C. Our results are not strongly sensitive to this selection (Supplementary Information section  6.3 ). The ten remaining models yield a total of 25,410 years from 154 historical ensemble members and 7,095 years from 43 historical-natural ensemble members. We weight the models equally in our analysis using bootstrap sampling. We report linear trends based on simple linear regression models fitted with ordinary least squares. The statistical significance of linear trends is assessed using the Spearman’s rank correlation test 69 .

Time of emergence of the anthropogenic impact

We assess the anthropogenic influence on SSTAs in the Coral Sea region by starting with the assumption that any anthropogenic influence on SSTAs in the Coral Sea is indistinguishable from natural variability at the commencement of the model experiments. We measure the impact of anthropogenic influence on the climate in the region using a signal-to-noise approach 32 , 70 . We calculate the anthropogenic ‘signal’ as the mean of the difference between the smoothed (using a 41-year Lowess filter) modelled historical Coral Sea SSTA and the mean smoothed modelled historical-natural SSTA. Our ‘noise’ is the standard deviation of the difference between the modelled historical SSTA and its smoothed time series (Supplementary Information section  6 ).

Methods additionally rely on Supplementary Information and refs. 71 , 72 , 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 , 81 , 82 , 83 , 84 , 85 , 86 , 87 , 88 , 89 , 90 , 91 , 92 , 93 , 94 , 95 , 96 , 97 , 98 , 99 , 100 , 101 , 102 , 103 , 104 .

Data availability

The ERSSTv5 instrumental SST data are available from the US National Oceanic and Atmospheric Administration at https://psl.noaa.gov/data/gridded/data.noaa.ersst.v5.html . The HadISST1.1 data are available from the UK Met Office at https://www.metoffice.gov.uk/hadobs/hadisst/ . The original coral palaeoclimate data are available at the links provided in Supplementary Table 2 . Land areas for maps are obtained from the Mapping Toolbox v.23.2 in Matlab v.2023b and the Global Self-consistent, Hierarchical, High-resolution Geography (GSHHS) Database at https://www.soest.hawaii.edu/pwessel/gshhg/ through the m_map toolbox by R. Pawlowicz, available at https://www.eoas.ubc.ca/%7Erich/map.html . Prepared data from the coral geochemical series, reconstructions and climate models that support the findings of this study are available at: https://doi.org/10.24433/CO.4883292.v1 .

Code availability

The code that supports the findings of this study is available and can be run at : https://doi.org/10.24433/CO.4883292.v1 .

Australian Institute of Marine Science. Long-Term Monitoring Program. https://www.aims.gov.au/research-topics/monitoring-and-discovery/monitoring-great-barrier-reef/long-term-monitoring-program (2024).

Hughes, T. P. et al. Global warming and recurrent mass bleaching of corals. Nature 543 , 373–377 (2017).

Article   ADS   CAS   PubMed   Google Scholar  

Hoegh-Guldberg, O. Climate change, coral bleaching and the future of the world’s coral reefs. Mar. Freshw. Res. 50 , 839–866 (1999).

Google Scholar  

Hughes, T. P. et al. Global warming transforms coral reef assemblages. Nature 556 , 492–496 (2018).

Great Barrier Reef Marine Park Authority. Great Barrier Reef Outlook Report 2019 (Great Barrier Reef Marine Park Authority, 2019).

Hughes, T. P. et al. Coral reefs in the Anthropocene. Nature 546 , 82–90 (2017).

Davis, K. L., Colefax, A. P., Tucker, J. P., Kelaher, B. P. & Santos, I. R. Global coral reef ecosystems exhibit declining calcification and increasing primary productivity. Commun. Earth Environ. 2 , 105 (2021).

Article   ADS   Google Scholar  

Westcott, D. A. et al. Relative efficacy of three approaches to mitigate Crown-of-Thorns Starfish outbreaks on Australia’s Great Barrier Reef. Sci. Rep. 10 , 12594 (2020).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Mellin, C. et al. Spatial resilience of the Great Barrier Reef under cumulative disturbance impacts. Glob. Chang. Biol. 25 , 2431–2445 (2019).

Article   ADS   PubMed   Google Scholar  

Jackson, J. B. C. et al. Historical overfishing and the recent collapse of coastal ecosystems. Science 293 , 629–637 (2001).

Article   CAS   PubMed   Google Scholar  

Hoegh-Guldberg, O. & Smith, G. J. The effect of sudden changes in temperature, light and salinity on the population density and export of zooxanthellae from the reef corals Stylophora pistillata Esper and Seriatopora hystrix Dana. J. Exp. Mar. Biol. Ecol. 129 , 279–303 (1989).

Article   Google Scholar  

DeCarlo, T. M. et al. Acclimatization of massive reef-building corals to consecutive heatwaves. Proc. Biol. Sci. 286 , 20190235 (2019).

PubMed   PubMed Central   Google Scholar  

McGowan, H. & Theobald, A. ENSO weather and coral bleaching on the Great Barrier Reef, Australia. Geophys. Res. Lett. 44 , 10,601–10,607 (2017).

Zhao, W., Huang, Y., Siems, S. & Manton, M. The role of clouds in coral bleaching events over the Great Barrier Reef. Geophys. Res. Lett. 48 , e2021GL093936 (2021).

Oxley, W. G., Emslie, M., Muir, P. & Thompson, A. Marine Surveys Undertaken in the Lihou Reef National Nature Reserve (Australian Institute of Marine Science, 2004).

DeCarlo, T. M. & Harrison, H. B. An enigmatic decoupling between heat stress and coral bleaching on the Great Barrier Reef. PeerJ 7 , e7473 (2019).

Article   PubMed   PubMed Central   Google Scholar  

UNESCO World Heritage Committee. Extended 44th Session of the World Heritage Committee, Fuzhou (China) 16–31 July 2021 . Draft decision 44 COM 7B.90. https://whc.unesco.org/document/188005 (UNESCO, 2021).

UNESCO World Heritage Committee. Extended 45th Session of the World Heritage Committee, Riyadh (Saudi Arabia) 10–25 September 2023 . Decision 45 COM 7B.13. https://whc.unesco.org/document/199654 (UNESCO, 2023).

IPCC. Summary for Policymakers. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (eds Masson-Delmotte, V. et al.) (Cambridge Univ. Press, 2021).

Kamenos, N. A. & Hennige, S. J. Reconstructing four centuries of temperature-induced coral bleaching on the Great Barrier Reef. Front. Mar. Sci. 5 , 283 (2018).

Hoegh-Guldberg, O. et al. Commentary: reconstructing four centuries of temperature-induced coral bleaching on the Great Barrier Reef. Front. Mar. Sci. 6 , 86 (2019).

DeCarlo, T. M. Commentary: reconstructing four centuries of temperature-induced coral bleaching on the Great Barrier Reef. Front. Mar. Sci. 7 , 30 (2020).

Hendy, E. J. et al. Abrupt decrease in tropical Pacific sea surface salinity at end of Little Ice Age. Science 295 , 1511–1514 (2002).

Calvo, E. et al. Interdecadal climate variability in the Coral Sea since 1708 A.D. Palaeogeogr. Palaeoclimatol. Palaeoecol. 248 , 190–201 (2007).

Zinke, J. et al. North Flinders Reef (Coral Sea, Australia) Porites sp. corals as a candidate global boundary stratotype section and point for the Anthropocene series. Anthropocene Rev. 10 , 201–224 (2023).

Spady, B. L. et al. Global Coral Bleaching Database (NCEI Accession 0228498) (NOAA National Centers for Environmental Information, 2022); https://www.ncei.noaa.gov/archive/accession/0228498 .

Huang, B. et al. Extended reconstructed sea surface temperature, Version 5 (ERSSTv5): Upgrades, validations, and intercomparisons. J. Clim. 30 , 8179–8205 (2017).

Rayner, N. A. et al. Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res. Atmos. 108 , 4407 (2003).

Eyring, V. et al. Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9 , 1937–1958 (2016).

Gillett, N. P. et al. The Detection and Attribution Model Intercomparison Project (DAMIP v1.0) contribution to CMIP6. Geosci. Model Dev. 9 , 3685–3697 (2016).

Hausfather, Z., Marvel, K., Schmidt, G. A., Nielsen-Gammon, J. W. & Zelinka, M. Climate simulations: recognize the ‘hot model’ problem. Nature 605 , 26–29 (2022).

Hawkins, E. et al. Observed emergence of the climate change signal: from the familiar to the unknown. Geophys. Res. Lett. 47 , e2019GL086259 (2020).

Gagan, M. K. et al. Temperature and surface-ocean water balance of the mid-Holocene tropical western Pacific. Science 279 , 1014–1018 (1998).

Arzey, A. K. et al. Coral skeletal proxy records database for the Great Barrier Reef, Australia. Preprint at Earth Syst. Sci. Data https://doi.org/10.5194/essd-2024-159 (2024).

Brenner, L. D. et al. Coral record of Younger Dryas Chronozone warmth on the Great Barrier Reef. Paleoceanogr. Paleoclimatol. 35 , e2020PA003962 (2020).

Furnas, M. J. & Mitchell, A. W. Nutrient inputs into the central Great Barrier Reef (Australia) from subsurface intrusions of Coral Sea waters: a two-dimensional displacement model. Cont. Shelf Res. 16 , 1127–1148 (1996).

Wolanski, E., Andutta, F., Deleersnijder, E., Li, Y. & Thomas, C. J. The Gulf of Carpentaria heated Torres Strait and the Northern Great Barrier Reef during the 2016 mass coral bleaching event. Estuar. Coast. Shelf Sci. 194 , 172–181 (2017).

Oliver, E. C. J. & Holbrook, N. J. Extending our understanding of South Pacific gyre ‘spin-up’: modeling the East Australian Current in a future climate. J. Geophys. Res. Oceans 119 , 2788–2805 (2014).

DeCarlo, T. M. et al. Nutrient-supplying ocean currents modulate coral bleaching susceptibility. Sci. Adv. 6 , eabc5493 (2020).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Wiedenmann, J. et al. Nutrient enrichment can increase the susceptibility of reef corals to bleaching. Nat. Clim. Chang. 3 , 160–164 (2013).

Article   ADS   CAS   Google Scholar  

Chan, D. & Huybers, P. Correcting observational biases in sea surface temperature observations removes anomalous warmth during World War II. J. Clim. 34 , 4585–4602 (2021).

Chan, D., Kent, E. C., Berry, D. I. & Huybers, P. Correcting datasets leads to more homogeneous early-twentieth-century sea surface warming. Nature 571 , 393–397 (2019).

Hughes, T. P. et al. Spatial and temporal patterns of mass bleaching of corals in the Anthropocene. Science 359 , 80–83 (2018).

Hoegh-Guldberg, O. et al. Chapter 3: Impacts of 1.5°C global warming on natural and human systems. In Global Warming of 1.5°C (eds Masson-Delmotte, V. et al.) (IPCC, 2018).

Meinshausen, M. et al. Realization of Paris Agreement pledges may limit warming just below 2 °C. Nature 604 , 304–309 (2022).

Matthews, H. D. & Wynes, S. Current global efforts are insufficient to limit warming to 1.5 °C. Science 376 , 1404–1409 (2022).

Coles, S. L. et al. Evidence of acclimatization or adaptation in Hawaiian corals to higher ocean temperatures. PeerJ 6 , e5347 (2018).

Hughes, T. P., Baird, A. H., Morrison, T. H. & Torda, G. Principles for coral reef restoration in the anthropocene. One Earth 6 , 656–665 (2023).

Logan, C. A., Dunne, J. P., Ryan, J. S., Baskett, M. L. & Donner, S. D. Quantifying global potential for coral evolutionary response to climate change. Nat. Clim. Chang. 11 , 537–542 (2021).

Dixon, A. M., Forster, P. M., Heron, S. F., Stoner, A. M. K. & Beger, M. Future loss of local-scale thermal refugia in coral reef ecosystems. PLOS Clim. 1 , e0000004 (2022).

Coplen, T. B. Discontinuance of SMOW and PDB. Nature 375 , 285 (1995).

van Albada, S. J. & Robinson, P. A. Transformation of arbitrary distributions to the normal distribution with application to EEG test-retest reliability. J. Neurosci. Methods 161 , 205–211 (2007).

Article   PubMed   Google Scholar  

Emile-Geay, J. & Tingley, M. Inferring climate variability from nonlinear proxies: application to palaeo-ENSO studies. Clim. Past 12 , 31–50 (2016).

Barnes, D. J., Taylor, R. B. & Lough, J. M. On the inclusion of trace materials into massive coral skeletons. Part II: distortions in skeletal records of annual climate cycles due to growth processes. J. Exp. Mar. Biol. Ecol. 194 , 251–275 (1995).

Article   CAS   Google Scholar  

Gagan, M. K., Dunbar, G. B. & Suzuki, A. The effect of skeletal mass accumulation in Porites on coral Sr/Ca and δ 18 O paleothermometry. Paleoceanogr. Paleoclimatol. 27 , PA1203 (2012).

ADS   Google Scholar  

Schneider, T. Analysis of incomplete climate data: estimation of mean values and covariance matrices and imputation of missing values. J. Clim. 14 , 853–871 (2001).

PAGES 2k Consortium. Consistent multidecadal variability in global temperature reconstructions and simulations over the Common Era. Nat. Geosci. 12 , 643–649 (2019).

Cook, E. R., Briffa, K. R. & Jones, P. D. Spatial regression methods in dendroclimatology: a review and comparison of two techniques. Int. J. Climatol. 14 , 379–402 (1994).

Nash, J. E. & Sutcliffe, J. V. River flow forecasting through conceptual models part I − a discussion of principles. J. Hydrol. 10 , 282–290 (1970).

Otto-Bliesner, B. L. et al. Climate variability and change since 850 CE: an ensemble approach with the Community Earth System Model. Bull. Am. Meteorol. Soc. 97 , 735–754 (2016).

Evans, M. N., Kaplan, A. & Cane, M. A. Optimal sites for coral-based reconstruction of global sea surface temperature. Paleoceanogr. Paleoclimatol. 13 , 502–516 (1998).

Russon, T., Tudhope, A. W., Hegerl, G. C., Collins, M. & Tindall, J. Inter-annual tropical Pacific climate variability in an isotope-enabled CGCM: Implications for interpreting coral stable oxygen isotope records of ENSO. Clim. Past 9 , 1543–1557 (2013).

PAGES Hydro2k Consortium. Comparing proxy and model estimates of hydroclimate variability and change over the Common Era. Clim. Past 13 , 1851–1900 (2017).

Freund, M. B. et al. Higher frequency of Central Pacific El Niño events in recent decades relative to past centuries. Nat. Geosci. 12 , 450–455 (2019).

Gagan, M. K. et al. New views of tropical paleoclimates from corals. Quat. Sci. Rev. 19 , 45–64 (2000).

Thompson, D. M., Ault, T. R., Evans, M. N., Cole, J. E. & Emile-Geay, J. Comparison of observed and simulated tropical climate trends using a forward model of coral δ 18 O. Geophys. Res. Lett. 38 , L14706 (2011).

LeGrande, A. N. & Schmidt, G. A. Global gridded data set of the oxygen isotopic composition in seawater. Geophys. Res. Lett. 33 , L12604 (2006).

Reed, E. V., Thompson, D. M. & Anchukaitis, K. J. Coral-based sea surface salinity reconstructions and the role of observational uncertainties in inferred variability and trends. Paleoceanogr. Paleoclimatol. 37 , e2021PA004371 (2022).

Khaliq, M. N., Ouarda, T. B. M. J., Gachon, P., Sushama, L. & St-Hilaire, A. Identification of hydrological trends in the presence of serial and cross correlations: A review of selected methods and their application to annual flow regimes of Canadian rivers. J. Hydrol. 368 , 117–130 (2009).

Mahlstein, I., Hegerl, G. & Solomon, S. Emerging local warming signals in observational data. Geophys. Res. Lett. 39 , L21711 (2012).

Freeman, E. et al. ICOADS Release 3.0: a major update to the historical marine climate record. Int. J. Climatol. 37 , 2211–2232 (2017).

Huang, B. et al. Uncertainty estimates for sea surface temperature and land surface air temperature in NOAAGlobalTemp version 5. J. Clim. 33 , 1351–1379 (2020).

Druffel, E. R. M. & Griffin, S. Variability of surface ocean radiocarbon and stable isotopes in the southwestern Pacific. J. Geophys. Res. 104 , 23607–23613 (1999).

DeLong, K. L., Quinn, T. M., Taylor, F. W., Lin, K. & Shen, C.-C. Sea surface temperature variability in the southwest tropical Pacific since AD 1649. Nat. Clim. Change 2 , 799–804 (2012).

Quinn, T. et al. A multicentury stable isotope record from a New Caledonia coral: Interannual and decadal SST variability in the southwest Pacific since 1657. Paleoceanography 13 , 412–426 (1998).

Quinn, T. M., Crowley, T. J. & Taylor, F. W. New stable isotope results from a 173-year coral from Espiritu Santo, Vanuatu. Geophys. Res. Lett. 23 , 3413–3416 (1996).

Alibert, C. & Kinsley, L. A 170-year Sr/Ca and Ba/Ca coral record from the western Pacific warm pool: 1. What can we learn from an unusual coral record? J. Geophys. Res. Oceans 113 , C04008 (2008).

Tudhope, A. W. et al. Variability in the El Niño-Southern Oscillation through a glacial-interglacial cycle. Science 291 , 1511–1517 (2001).

Urban, F. E., Cole, J. E. & Overpeck, J. T. Influence of mean climate change on climate variability from a 155-year tropical Pacific coral record. Nature 407 , 989–993 (2000).

Guilderson, T. P. & Schrag, D. P. Reliability of coral isotope records from the western Pacific warm pool: A comparison using age-optimized records. Paleoceanography 14 , 457–464 (1999).

Quinn, T. M., Taylor, F. W. & Crowley, T. J. Coral-based climate variability in the Western Pacific Warm Pool since 1867. J. Geophys. Res. 111 , C11006 (2006).

Gorman, M. K. et al. A coral-based reconstruction of sea surface salinity at Sabine Bank, Vanuatu from 1842 to 2007 CE. Paleoceanography 27 , PA3226 (2012).

Bagnato, S., Linsley, B. K., Howe, S. S., Wellington, G. M. & Salinger, J. Coral oxygen isotope records of interdecadal climate variations in the South Pacific Convergence Zone region. Geochem. Geophys. Geosyst. 6 , Q06001 (2005).

Linsley, B. K. et al. Tracking the extent of the South Pacific Convergence Zone since the early 1600s. Geochem. Geophys. Geosyst. 7 , Q05003 (2006).

Cole, J. E., Fairbanks, R. G. & Shen, G. T. Recent variability in the Southern Oscillation: Isotopic results from a Tarawa Atoll coral. Science 260 , 1790–1793 (1993).

Dassié, E. P. et al. A Fiji multi-coral δ 18 O composite approach to obtaining a more accurate reconstruction of the last two-centuries of the ocean-climate variability in the South Pacific Convergence Zone region. Paleoceanography 29 , 1196–1213 (2014).

Carton, J. A., Chepurin, G. A. & Chen, L. SODA3: A new ocean climate reanalysis. J. Clim. 31 , 6967–6983 (2018).

Zuo, H., Balmaseda, M. A., Tietsche, S., Mogensen, K. & Mayer, M. The ECMWF operational ensemble reanalysis-analysis system for ocean and sea ice: A description of the system and assessment. Ocean Sci. 15 , 779–808 (2019).

Cheng, L. et al. Improved estimates of changes in upper ocean salinity and the hydrological cycle. J. Clim. 33 , 10357–10381 (2020).

Thompson, D. M. et al. Identifying hydro‐sensitive coral δ18O records for improved high‐resolution temperature and salinity reconstructions. Geophys. Res. Lett. 49 , e2021GL096153 (2022).

Wu, Y., Fallon, S. J., Cantin, N. E. & Lough, J. M. Assessing multiproxy approaches (Sr/Ca, U/Ca, Li/Mg, and B/Mg) to reconstruct sea surface temperature from coral skeletons throughout the Great Barrier Reef. Sci. Total Environ. 786 , 147393 (2021).

Sadler, J., Webb, G. E., Leonard, N. D., Nothdurft, L. D. & Clark, T. R. Reef core insights into mid-Holocene water temperatures of the southern Great Barrier Reef. Paleoceanography 31 , 1395–1408 (2016).

Roche, R. C. et al. Mid-Holocene sea surface conditions and riverine influence on the inshore Great Barrier Reef. Holocene 24 , 885–897 (2014).

Reed, E. V., Cole, J. E., Lough, J. M., Thompson, D. & Cantin, N. E. Linking climate variability and growth in coral skeletal records from the Great Barrier Reef. Coral Reefs 38 , 29–43 (2019).

Razak, T. B. et al. Use of skeletal Sr/Ca ratios to determine growth patterns in a branching coral Isopora palifera. Mar. Biol. 164 , 96 (2017).

Marshall, J. F. Decadal-scale, High Resolution Records of Sea Surface Temperature in the Eastern Indian and South Western Pacific Oceans from Proxy Records of the Strontium/calcium Ratio of Massive Porites Corals PhD thesis, Australian National Univ. (2000).

Marshall, J. F. & McCulloch, M. T. An assessment of the Sr/Ca ratio in shallow water hermatypic corals as a proxy for sea surface temperature. Geochim. Cosmochim. Acta 66 , 3263–3280 (2002).

Gagan, M. K. et al. Coral oxygen isotope evidence for recent groundwater fluxes to the Australian Great Barrier Reef. Geophys. Res. Lett. 29 , 43-1–43-4 (2002).

D’Olivo, J. P., Sinclair, D. J., Rankenburg, K. & McCulloch, M. T. A universal multi-trace element calibration for reconstructing sea surface temperatures from long-lived Porites corals: Removing ‘vital-effects’. Geochim. Cosmochim. Acta 239 , 109–135 (2018).

Fallon, S. J., McCulloch, M. T. & Alibert, C. Examining water temperature proxies in Porites corals from the Great Barrier Reef: a cross-shelf comparison. Coral Reefs 22 , 389–404 (2003).

Brenner, L. D., Linsley, B. K. & Potts, D. C. A modern Sr/Ca-δ 18 O-sea surface temperature calibration for Isopora corals on the Great Barrier Reef. Paleoceanography 32 , 182–194 (2017).

Alibert, C. et al. Source of trace element variability in Great Barrier Reef corals affected by the Burdekin flood plumes. Geochim. Cosmochim. Acta 67 , 231–246 (2003).

Murty, S. A. et al. Spatial and temporal robustness of Sr/Ca-SST calibrations in Red Sea corals: Evidence for influence of mean annual temperature on calibration slopes. Paleoceanogr. Paleoclimatol. 33 , 443–456 (2018).

Sayani, H. R., Cobb, K. M., DeLong, K., Hitt, N. T. & Druffel, E. R. M. Intercolony δ 18 O and Sr/Ca variability among Porites spp. corals at Palmyra Atoll: Toward more robust coral-based estimates of climate. Geochem. Geophys. Geosyst. 20 , 5270–5284 (2019).

Otto, F. E. L. Geert Jan van Oldenborgh 1961–2021. Nat. Clim. Chang. 11 , 1017 (2021).

Download references

Acknowledgements

We acknowledge the originators of the coral data cited in Supplementary Tables 1 and 2 ; S. E. Perkins-Kirkpatrick and the deceased G. J. van Oldenborgh 105 for contributions to an earlier version of this manuscript; E. P. Dassié and J. Zinke for discussions and data; R. Neukom for advice on an earlier version of the reconstruction code; and B. Trewin and K. Braganza for advice about the Bureau of Meteorology GBR SST time series. B.J.H. and H.V.M. acknowledge support from an Australian Research Council (ARC) SRIEAS grant, Securing Antarctica’s Environmental Future (SR200100005), and ARC Discovery Project DP200100206. A.D.K. acknowledges support from an ARC DECRA (DE180100638) and the Australian government’s National Environmental Science Program. B.J.H. and A.D.K. acknowledge an affiliation with the ARC Centre of Excellence for Climate Extremes (CE170100023). H.V.M. acknowledges support from an ARC Future Fellowship (FT140100286). A.K.A. acknowledges support from an Australian government research training program scholarship and an AINSE postgraduate research award. Funding was provided to B.K.L. by the Vetlesen Foundation through a gift to the Lamont-Doherty Earth Observatory. Grants to B.K.L. enabled the generation of coral oxygen isotope and Sr/Ca data from Fiji that were used in our reconstruction (US National Science Foundation OCE-0318296 and ATM-9901649 and US National Oceanic and Atmospheric Administration NA96GP0406). We acknowledge the support of the NCI facility in Australia and the World Climate Research Programme’s working group on coupled modelling, which is responsible for CMIP. We thank the climate-modelling groups for producing and making available their model output. For CMIP, the US Department of Energy’s Program for Climate Model Diagnosis and Intercomparison provided coordinating support and led the development of software infrastructure in partnership with the Global Organisation for Earth System Science Portals.

Author information

Authors and affiliations.

Environmental Futures, School of Earth, Atmospheric and Life Sciences, University of Wollongong, Wollongong, New South Wales, Australia

Benjamin J. Henley, Helen V. McGregor & Ariella K. Arzey

Securing Antarctica’s Environmental Future, University of Wollongong, Wollongong, New South Wales, Australia

School of Agriculture, Food and Ecosystem Sciences, University of Melbourne, Parkville, Victoria, Australia

Benjamin J. Henley

School of Geography, Earth and Atmospheric Sciences, University of Melbourne, Parkville, Victoria, Australia

Andrew D. King & David J. Karoly

ARC Centre of Excellence for Climate Extremes, University of Melbourne, Parkville, Victoria, Australia

Andrew D. King

School of the Environment, The University of Queensland, Brisbane, Queensland, Australia

Ove Hoegh-Guldberg

Australian Institute of Marine Science, Townsville, Queensland, Australia

Janice M. Lough

ARC Centre of Excellence for Coral Reef Studies and School of Earth Sciences, University of Western Australia, Crawley, Western Australia, Australia

Thomas M. DeCarlo

Department of Earth and Environmental Sciences, Tulane University, New Orleans, LA, USA

Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY, USA

Braddock K. Linsley

You can also search for this author in PubMed   Google Scholar

Contributions

B.J.H., H.V.M. and A.D.K. conceived the study and developed the methodology. B.J.H. did most of the analysis. A.K.A. contributed analysis of modern coral data (Supplementary Information section  4.2 ). T.M.D. contributed analysis of instrumental data coverage (Supplementary Information section  1.2 ). B.K.L. contributed sub-annual coral data. B.J.H. and H.V.M. led the preparation of the manuscript, with contributions from A.D.K., O.H.-G., A.K.A., D.J.K., J.M.L., T.M.D. and B.K.L. Generative artificial intelligence was not used in any aspect of this study or manuscript.

Corresponding author

Correspondence to Benjamin J. Henley .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature thanks Simon Michel, Miriam Pfeiffer, Claudia Tebaldi and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information ., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Henley, B.J., McGregor, H.V., King, A.D. et al. Highest ocean heat in four centuries places Great Barrier Reef in danger. Nature 632 , 320–326 (2024). https://doi.org/10.1038/s41586-024-07672-x

Download citation

Received : 02 November 2022

Accepted : 04 June 2024

Published : 07 August 2024

Issue Date : 08 August 2024

DOI : https://doi.org/10.1038/s41586-024-07672-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Great barrier reef’s temperature soars to 400-year high.

• Jeff Tollefson

Nature (2024)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Anthropocene newsletter — what matters in anthropocene research, free to your inbox weekly.

• Open access
• Published: 07 August 2024

Impact of neonatal sepsis on neurocognitive outcomes: a systematic review and meta-analysis

• Wei Jie Ong   ORCID: orcid.org/0000-0001-8244-2977 1   na1 ,
• Jun Jie Benjamin Seng   ORCID: orcid.org/0000-0002-3039-3816 1 , 2 , 3   na1 ,
• Beijun Yap 1 ,
• George He 4 ,
• Nooriyah Aliasgar Moochhala 4 ,
• Chen Lin Ng 1 ,
• Rehena Ganguly   ORCID: orcid.org/0000-0001-9347-5571 5 ,
• Jan Hau Lee   ORCID: orcid.org/0000-0002-8430-4217 6 &
• Shu-Ling Chong   ORCID: orcid.org/0000-0003-4647-0019 7  

BMC Pediatrics volume  24 , Article number:  505 ( 2024 ) Cite this article

311 Accesses

1 Altmetric

Metrics details

Introduction

Sepsis is associated with neurocognitive impairment among preterm neonates but less is known about term neonates with sepsis. This systematic review and meta-analysis aims to provide an update of neurocognitive outcomes including cognitive delay, visual impairment, auditory impairment, and cerebral palsy, among neonates with sepsis.

We performed a systematic review of PubMed, Embase, CENTRAL and Web of Science for eligible studies published between January 2011 and March 2023. We included case–control, cohort studies and cross-sectional studies. Case reports and articles not in English language were excluded. Using the adjusted estimates, we performed random effects model meta-analysis to evaluate the risk of developing neurocognitive impairment among neonates with sepsis.

Of 7,909 studies, 24 studies ( n  = 121,645) were included. Majority of studies were conducted in the United States ( n  = 7, 29.2%), and all studies were performed among neonates. 17 (70.8%) studies provided follow-up till 30 months. Sepsis was associated with increased risk of cognitive delay [adjusted odds ratio, aOR 1.14 (95% CI: 1.01—1.28)], visual impairment [aOR 2.57 (95%CI: 1.14- 5.82)], hearing impairment [aOR 1.70 (95% CI: 1.02–2.81)] and cerebral palsy [aOR 2.48 (95% CI: 1.03–5.99)].

Neonates surviving sepsis are at a higher risk of poorer neurodevelopment. Current evidence is limited by significant heterogeneity across studies, lack of data related to long-term neurodevelopmental outcomes and term infants.

Peer Review reports

Sepsis is a major cause of mortality and morbidity among neonates [ 1 , 2 , 3 , 4 ]. Young infants especially neonates, defined by age < 28 days old, have a relatively immature immune system and are susceptible to sepsis [ 5 , 6 ]. Annually, there are an estimated 1.3 to 3.9 million cases of infantile sepsis worldwide and up to 700,000 deaths [ 7 ]. Low-income and middle-income countries bear a disproportionate burden of neonatal sepsis cases and deaths [ 7 , 8 ]. While advances in medical care over the past decade have reduced mortality, neonates who survive sepsis are at risk of developing neurocognitive complications, which affect the quality of life for these children and their caregivers [ 9 ].

Previous reviews evaluating neurocognitive outcomes in neonates with infections or sepsis have focused on specific types of pathogens (e.g., Group B streptococcus or nosocomial infections [ 10 ]), or are limited to specific populations such as very low birth weight or very preterm neonates [ 11 ], and there remains paucity of data regarding neurocognitive outcomes among term and post-term neonates. There remains a gap for an updated comprehensive review which is not limited by type of pathogen or gestation. In this systematic review, we aim to provide a comprehensive update to the current literature on the association between sepsis and the following adverse neurocognitive outcomes (1) mental and psychomotor delay (cognitive delay (CD)), (2) visual impairment, (3) auditory impairment and (4) cerebral palsy (CP) among neonates [ 11 ].

We performed a systematic review using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines [ 12 ]. This study protocol was registered with Open Science Framework ( https://doi.org/10.17605/OSF.IO/B54SE ).

Eligibility criteria

We identified studies which evaluated neurocognitive outcomes in neonates less than 90 days old (regardless of gestational age) with sepsis. While the neonatal period is traditionally defined to be either the first 28 days postnatally for term and post-term infants, or 27 days after the expected date of delivery for preterm infants [ 13 ], serious late onset infections in the young infant population can present beyond the neonatal period [ 14 ], hence we defined the upper age limit as 90 days old to obtain a more complete picture of the burden of young infantile sepsis [ 15 ]. Post-term neonates was defined as a neonate delivered at >  = 42 weeks of gestational age in this study [ 16 ]. We included studies that either follow international sepsis definitions such as Surviving Sepsis Campaign guidelines definitions [ 17 ], or if they fulfilled clinical, microbiological and/or biochemical criteria for sepsis as defined by study authors. The primary outcome of interest was impaired neurocognitive outcome defined by the following domains of neurodevelopmental impairment (NDI) [ 11 ]: (1) CD, (2) visual impairment, (3) auditory impairment and (4) CP. We selected these domains because they were highlighted as key neurocognitive sequelae after intrauterine insults in a landmark review by Mwaniki et al. [ 18 ]. The authors’ definitions of these outcomes and their assessment tools were captured, including the use of common validated instruments (e.g., a common scale used for CD is the Bayley Scales of Infant Development (BSID) [ 19 ] while a common instrument used for CP was the Gross Motor Function Classification System (GMFCS) [ 20 ]. Specifically for BSID, its two summative indices score – Mental Development Index (MDI) and Psychomotor Development Index (PDI) were collected. The MDI assesses both the non-verbal cognitive and language skills, while PDI assess the combination of fine and gross motor skills. The cut-off points for mild, moderate and severe delay for MDI and PDI were < 85 or < 80, < 70 and < 55 respectively [ 21 ]. There were no restrictions on duration of follow-up or time of assessment of neurocognitive outcomes to allow capturing of both short- and long-term neurocognitive outcomes.

Case–control, cohort studies and cross-sectional studies published between January 2011 and March 2023 were included. Because the definition and management of sepsis has evolved over the years [ 22 ], we chose to include studies published from 2011 onwards. Case reports, animal studies, laboratory studies and publications that were not in English language were excluded. Hand-searching of previous systematic reviews were performed to ensure all relevant articles were included. To avoid small study effects, we also excluded studies with a sample size of less than 50 [ 23 ].

Information sources and search strategy

Four databases (PubMed, Cochrane Central, Embase and Web of Science) were used to identify eligible studies. The search strategy was developed in consultation with a research librarian. The first search was conducted on 4 December 2021 and an updated search was conducted on 3 April 2023. The detailed search strategy can be found in Supplementary Tables 1A and B.

Study selection process

Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia) [ 24 ] was utilized during this review. Five reviewers (WJO, BJY, NM, CLN and GH) independently conducted the database search and screened the title and abstracts for relevance. Following training on inclusion and exclusion eligibility, 4 reviewers (WJO, NM, CLN and GH) subsequently assessed the full text of shortlisted articles for eligibility. All full texts were independently assessed by at least 2 reviewers. Any conflict related to study eligibility were resolved in discussion with the senior author (S-LC). We recorded the reason(s) for exclusion of each non-eligible article.

Data collection process and data items

Four reviewers (WJO, NM, CLN and GH) independently carried out the data extraction using a standardized data collection form, and any conflict was resolved by discussion, or with input from the senior author (S-LC). A pilot search was performed for the first 200 citations to evaluate concordance among reviewers and showed good concordance among reviewers of 94%. For studies with missing data required for data collection or meta-analyses, we contacted the corresponding authors of articles to seek related information. If there was no reply from the authors, the data were labelled as missing.

Study risk of bias assessment

Three reviewers (BJY, GH and WJO) independently carried out the assessment of risk of bias using the Newcastle–Ottawa Scale (NOS) for all observational studies [ 25 ]. Studies were graded based on three domains namely, selection, comparability and outcomes. Studies were assigned as low, moderate and high risk of bias if they were rated 0–2 points, 3–5 points and 6–9 points respectively. Any conflict was resolved by discussion or with input from the senior author (S-LC).

Statistical analysis

All outcomes (i.e. CD, visual impairment, auditory impairment and CP) were analysed as categorical data. Analyses were done for each NDI domain separately. To ensure comparability across scales, results from different studies were only pooled if the same measurement tools were used to assess the outcomes and hence sub-group analyses were based on different scales and/or different definitions of neurocognitive outcomes used by authors. Both unadjusted and adjusted odds ratios (aOR) and/or relative risk (RR) for each NDI domain were recorded. Where source data were present, we calculated the unadjusted OR if the authors did not report one, together with the 95% confidence interval (CI). For adjusted odds ratio, these were extracted from individual studies and variables used for adjustment were determined at the individual study level.

Meta-analysis was conducted for all outcomes that were reported by at least 2 independent studies or cohorts. Studies were included in the meta-analysis only if they reported outcomes for individual NDI domains within 30 months from sepsis occurrence. For each domain, all selected studies were pooled using DerSimonian-Laird random effects model due to expected heterogeneity. Studies were pooled based on adjusted and unadjusted analyses. Case–control and cohort studies were pooled separately. The pooled results were expressed as unadjusted odds ratio (OR) or adjusted odds ratio (aOR) with corresponding 95% confidence interval (95% CI). If there was more than 1 study that utilized the same population, we only analysed data from the most recent publication or from the larger sample size, to avoid double counting. Standard error (SE) from studies with multiple arms with same control group were adjusted using SE = √(K/2), where K refers to number of treatment arms including control [ 26 ]. Heterogeneity across studies was evaluated using the I^2 statistic, for which ≥ 50% is indicative of significant heterogeneity. With regards to publication bias, this was performed using Egger’s test and funnel plots only if the number of studies pooled were 10 or more for each outcome.

For neurocognitive related outcomes, subgroup analyses were performed based on the severity of the NDI domain outcomes and distinct, non-overlapping populations of septic infants (such as late onset vs early onset sepsis, culture positive sepsis vs clinically diagnosed sepsis, term and post term patients).

All analyses were done using ‘meta’ library from R software (version 4.2.2) [ 27 ]. The statistical significance threshold was a two tailed P- value < 0.05.

Certainty of evidence

The certainty of evidence for outcomes in this review was performed during the GRADE criteria [ 28 ] which is centred on the study design, risk of bias, inconsistency, indirectness, imprecision, and other considerations.

Study selection

From 7,909 studies identified, a total of 24 articles were included (Fig.  1 ) [ 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 ]. A total of 101,657 and 19,988 preterm and term infants were included in this review.

PRISMA flowchart of the study selection process for search

Study characteristics

There were 2 case–control studies and 22 cohort studies, with a total of 121,645 infants (Table  1 ). Studies were conducted in 16 different countries (Fig.  2 ), with the most studies conducted in the United States of America (USA) (7 studies, n  = 92,358 patients) [ 30 , 33 , 37 , 41 , 42 , 47 , 52 ]. There were no studies that were conducted solely on term infants. 5 studies reported data specifically on ELBW infants (27,078 infants) and 6 studies on VLBW infants (3,322 infants). All studies were performed among neonates.

World map depicting distribution of studies that evaluate neurocognitive outcomes in infantile and neonatal sepsis

Risk of bias 

Overall, all 24 studies were classified as low risk (Supplementary Table 2). 5 papers scored high risk for outcome bias for having greater than 10% of initial population being lost to follow-up [ 29 , 32 , 40 , 41 , 42 ].

Outcome measures reported by domain

As the number of studies pooled for each outcome was less than 10, publication bias was not analysed in the meta-analyses.

Cognitive delay (CD)

Among 24 studies that assessed for CD, 16 studies reported either the incidence of CD among young infants with sepsis compared to those without, and/or the odds ratio (adjusted and/or unadjusted) comparing the two populations [ 29 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 40 , 41 , 42 , 45 , 46 , 48 , 49 ]. The scales used, authors’ definition of CD, incidence of CD among those with sepsis and those without are described in Table  2 . The most common tools used for assessment of CD were the Bayley Scales of Infant Development (BSID) ( n  = 13) and Denver Development Screening Test II ( n  = 2).

Infantile sepsis was associated with increased risk of overall CD delays [aOR 1.14 (95%CI: 1.01, 1.28)], overall PDI delay (aOR 1.73 (95%CI: 1.16, 2.58)) and moderate PDI delay [aOR 1.85 (95%CI: 1.01, 3.36)]. Conversely, infantile sepsis was not associated with increased risk for severe PDI delay nor overall MDI delay [aOR 1.30 (95%CI: 0.99, 1.71)] or its subgroups. There were no significant differences in outcomes between different subgroups of infections as well as culture-proven or clinically defined sepsis for either MDI or PDI (Table  8 , Fig.  3 A and B).

A Forest plot on adjusted odds ratios for neurocognitive outcomes related to MDI, PDI, visual impairment, hearing impairment and cerebral palsy. B Forest plot on unadjusted odds ratios for neurocognitive outcomes related to MDI, PDI, visual impairment, hearing impairment and cerebral palsy. Legend: MDI: Mental Developmental Index; PDI: Psychomotor Developmental Index. Foot note: Mild MDI or PDI: < 85 or < 80; Moderate MDI or PDI < 70; Severe MDI or PDI < 55

Visual impairment

Seven studies reported data on visual impairment (Table  3 ) [ 31 , 33 , 41 , 42 , 47 , 49 ]. The most common definition of visual impairment utilized was “visual acuity of < 20/200” ( n  = 4, 66.7%).

In the meta-analysis, infantile sepsis was associated with significantly increased risk of visual impairment [aOR 2.57 (95%CI: 1.14, 5.82)] but there were no statistically significant differences in visual impairment between subgroups of early or late onset sepsis, and blood culture negative conditions as compared to the non-septic population (Table  8 , Fig.  3 A and B).

Hearing impairment

Seven studies reported data on hearing impairment (Table  4 ) [ 31 , 33 , 41 , 42 , 47 , 49 ]. Two studies defined hearing impairment as permanent hearing loss affecting communication with or without amplification [ 42 , 47 ]. Other definitions included “sensorineural hearing loss requiring amplification” ( n  = 1), “bilateral hearing impairment with no functional hearing (with or without amplification)” ( n  = 1), “clinical hearing loss” ( n  = 1).

In the meta-analysis, sepsis was associated with increased risk of hearing impairment [aOR 1.70 (95% CI: 1.02–2.81)]. However, in the subgroup analyses, there were no differences in risk of hearing impairment between patients with late onset sepsis as compared to the non-septic population (Table  8 , Fig.  3 A and B).

Cerebral palsy

Nine studies [ 29 , 32 , 33 , 41 , 42 , 47 , 48 , 49 , 50 ] reported data on CP (Table  5 ), of which 5 studies [ 41 , 42 , 45 , 49 , 50 ] used the GMFCS scale. In the meta-analysis, infantile sepsis was associated with significantly increased risk of CP [aOR 2.48 (95%CI: 1.03; 5.99)]. There was no difference in rates of CP among patients with proven or suspected sepsis, as compared with infants with no sepsis (Table  8 , Fig.  3 A and B).

Differences in neurocognitive outcomes between neonates with culture-proven or clinically diagnosed sepsis as well as early or late onset sepsis

Tables 6 and 7 showed data related to differences in neurocognitive outcomes between neonates with culture-proven or clinically diagnosed sepsis as well as early or late onset sepsis. Meta-analyses were not be performed due to significant heterogeneity in definitions of sepsis, time of assessment of outcomes.

Differences in neurocognitive outcomes between term and post-term neonates

There were no studies which evaluated neurocognitive outcomes between term and post-term neonates and infants.

We found that the certainty of evidence to be very low to low for the four main neurocognitive outcomes selected. (Supplementary File 3).

In this review involving more than 121,000 infants, we provide an update to the literature regarding young infant sepsis and neurocognitive impairment. Current collective evidence demonstrate that young infant sepsis was associated with increased risk of developing neurocognitive impairment in all domains of CD, visual impairment, auditory impairment and cerebral palsy.

Cognitive delay

In this review, higher rates of cognitive delay were noted among infants with sepsis [ 29 , 31 , 33 , 34 , 35 , 36 , 37 , 38 , 40 , 41 , 42 , 45 , 46 , 48 , 49 , 52 ]. We found that infants with sepsis reported lower PDI scores (Table  8 ), which measures mainly neuromotor development. On the other hand, young infant sepsis was not associated with lower MDI scores (Table  8 ), which assesses cognitive and language development. The pathophysiological mechanism of young infant sepsis and its preferential impact on PDI remains unclear. Postulated mechanisms include development of white matter lesions which may arise from the susceptibility of oligodendrocyte precursors to inflammatory processes such as hypoxia and ischemia [ 53 ]. Future studies should look into evaluating the causes of the above findings. A majority of included studies focused on early CD outcomes while no studies evaluated long-term outcomes into adulthood. CD is known to involve complex genetic and experiential interactions [ 54 ] and may evolve overtime with brain maturation. Delays in speech and language, intellectual delay and borderline intellectual functioning are shown to be associated with poorer academic or employment outcomes in adulthood [ 55 , 56 ], and early assessment of CD may not fully reveal the extent of delays. The only study with follow-up to the adolescent phase showed a progressive increase in NDI rate as the participants aged, which provides evidence of incremental long-term negative outcomes associated with infantile sepsis [ 44 ]. Moving forward, studies with longer follow-up may allow for further examination of the long-term effects of neonatal sepsis on CD.

There were different versions of the BSID instrument (BSID-II and BSID-III) [ 19 , 57 , 58 ]. BSID-II lacked subscales in PDI and MDI scores, leading to the development of BSID-III with the segregation of PDI into fine and gross motor scales and MDI into cognitive, receptive language, and expressive language scales [ 59 ]. Although we pooled results of both BSID-II and BSID-III in our study, we recognize that comparisons between BSID-II and BSID-III are technically challenging due to differences in standardised scores [ 59 , 60 ]. In addition, the BSID-IV was created in 2019 which has fewer items, However, none of our studies utilized this instrument. Future studies should consider this instrument, as well as standardising the timepoints for assessment of CD.

Young infant sepsis was associated with increased risk of developing visual impairment. This was similar to results noted by a previous systematic review published in 2014 [ 61 ] and 2019 [ 62 ] which showed that neonatal sepsis was associated with twofold risk of developing retinopathy of prematurity in preterm infants. Specifically, meningitis was associated with a greater risk of visual impairment compared to just sepsis alone [ 47 ]. The mechanism of visual impairment has not been fully described although various theories have been suggested, including sepsis mediated vascular endothelial damage, increased body oxidative stress response as well as involvement of inflammatory cytokines and mediators [ 63 , 64 ].

Our meta-analysis showed an increased risk of hearing impairment for young infants with young infants with sepsis. This is consistent with a previous report that found an association between neonatal meningitis and sensorineural hearing loss [ 65 ]. One potential confounder which we were unable to account for may have been the use of ototoxic antimicrobial agents such as aminoglycosides. Additional confounders include very low birth weight, patient’s clinical states (e.g. hyperbilirubinemia requiring exchange transfusion) and use of mechanical ventilation or extracorporeal membrane support. To allow for meaningful comparisons of results across different study populations, it is imperative that a standardised definition of hearing impairment post neonatal sepsis be established for future studies.

Our meta-analysis found an association between neonatal sepsis and an increased risk of developing CP. This is also consistent with previous systematic reviews which had found a significant association of sepsis and CP in VLBW and early preterm infants [ 11 ]. One study found that infants born at full term and who experienced neonatal infections were at a higher risk of developing a spastic triplegia or quadriplegia phenotype of CP [ 66 ]. The pathophysiology and mechanism of injury to white matter resulting in increased motor dysfunction remains unclear and more research is required in this area.

Limitations and recommendations for future research

The main limitation of this review lies in the heterogeneity in the definitions of sepsis, exposures and assessment of outcomes across studies. This is likely attributed to the varying definition of sepsis used in different countries as well as lack of gold standard definitions or instruments for assessment of each component of NDI. A recent review of RCTs [ 67 ] also reported similar limitations where 128 different varying definitions of neonatal sepsis were used in literature. Notably, there is a critical need for developing international standardized guidelines for defining neonatal sepsis as well as assessment of NDI such as hearing and visual impairment. Another important limitation relates to the inability to assess quality of neonatal care delivered as well as temporal changes in medical practices which could have affected neurocognitive outcomes for neonates with sepsis. Improving quality of neonatal care has been shown to significantly reduce mortality risk among neonates with sepsis, especially in resource-poor countries [ 68 ]. We performed a comprehensive search strategy (PubMed, Embase, Web of Science and CENTRAL) coupled with hand searching of references within included systematic reviews, but did not evaluate grey literature. Future studies should include additional literature databases and grey literature. Another area of research gap lies in the paucity of data related to differences in neurocognitive outcomes between term and post-term neonates with sepsis and future research is required to bridge this area of research gap. Likewise, there are few studies which evaluated differences in neurocognitive outcomes between early or late onset sepsis and outcomes assessed were significantly heterogenous which limits meaningful meta-analyses. Similarly, there was significant heterogeneity in study outcomes, causative organisms and severity of disease.

We found a lack of long-term outcomes and recommend that future prospective cohorts include a longer follow-up duration as part of the study design. This is important given the implication of NDI on development into adulthood. Most data were reported for preterm infants with low birth weight, and there was a paucity of data for term infants in our literature review. Since prematurity itself is a significant cause of NDI [ 69 ], future studies should consider how gestational age and/or birth weight can be adequately adjusted for in the analysis.

Apart from the domains of NDI we chose to focus on in this review, there are other cognitive domains classified by the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V) [ 70 ] and/or recommended by the Common Data Elements (CDE) workgroup [ 71 ]. Future studies may wish to look into the implications of sepsis on other neuro-cognitive domains related to executive function, complex attention and societal cognition which are studied for other types of acquired brain injury [ 71 , 72 ].

Our systematic review and meta-analysis found that neonates surviving sepsis are at a higher risk of poorer neurodevelopment. However, the evidence is limited by significant heterogeneity and selection bias due to differing definitions used for NDI and for sepsis. There is also a lack of long-term follow-up data, as well as data specific for term and post-term infants. Future prospective studies should be conducted with long-term follow-up to assess the impact of neurodevelopmental impairment among all populations of neonates with sepsis.

Availability of data and materials

All data generated or analyzed in the study are found in the tables and supplementary materials.

Liu L, Johnson HL, Cousens S, Perin J, Scott S, Lawn JE, et al. Global, regional, and national causes of child mortality: an updated systematic analysis for 2010 with time trends since 2000. Lancet. 2012;379:2151–61.

Article   PubMed   Google Scholar  

WHO. Newborns: improving survival and well-being. Geneve: World Health Organisation; 2020.

Chiesa C, Panero A, Osborn JF, Simonetti AF, Pacifico L. Diagnosis of neonatal sepsis: a clinical and laboratory challenge. Clin Chem. 2004;50:279–87.

Article   CAS   PubMed   Google Scholar  

Ramaswamy VV, Abiramalatha T, Bandyopadhyay T, Shaik NB, Bandiya P, Nanda D, et al. ELBW and ELGAN outcomes in developing nations-Systematic review and meta-analysis. PLoS ONE. 2021;16:e0255352.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Zhang X, Zhivaki D, Lo-Man R. Unique aspects of the perinatal immune system. Nat Rev Immunol. 2017;17:495–507.

Prabhudas M, Adkins B, Gans H, King C, Levy O, Ramilo O, et al. Challenges in infant immunity: Implications for responses to infection and vaccines. Nat Immunol. 2011;12:189–94.

World Health Organization. Global report on the epidemiology and burden of sepsis. 2020. Available from: https://www.who.int/publications/i/item/9789240010789 .

Milton R, Gillespie D, Dyer C, Taiyari K, Carvalho MJ, Thomson K, et al. Neonatal sepsis and mortality in low-income and middle-income countries from a facility-based birth cohort: an international multisite prospective observational study. Lancet Glob Health. 2022May 1;10(5):e661-72. https://doi.org/10.1016/S2214-109X(22)00043-2 .

Li Y, Ji M, Yang J. Current understanding of long-term cognitive impairment after sepsis. Front Immunol. 2022;13:855006.

Haller S, Deindl P, Cassini A, Suetens C, Zingg W, Abu Sin M, et al. Neurological sequelae of healthcare-associated sepsis in very-low-birthweight infants: Umbrella review and evidence-based outcome tree. Euro Surveill. 2016;21:30143.

Alshaikh B, Yusuf K, Sauve R. Neurodevelopmental outcomes of very low birth weight infants with neonatal sepsis: Systematic review and meta-analysis. J Perinatol. 2013;33:558–64.

Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, Shamseer L, Tetzlaff JM, Akl EA, Brennan SE, Chou R, Glanville J, Grimshaw JM, Hróbjartsson A, Lalu MM, Li T, Loder EW, Mayo-Wilson E, McDonald S, McGuinness LA, Stewart LA, Thomas J, Tricco AC, Welch VA, Whiting P, Moher D. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. https://doi.org/10.1136/bmj.n71 .

U.S. Department of Health and Human Services F and D, Administration C for DE and R (CDER), (CBER) C for BE and R. General Clinical Pharmacology Considerations for Neonatal Studies for Drugs and Biological Products Guidance for Industry. 2019. Available from: https://www.fda.gov/media/129532/download . [cited 2022 Aug 9].

Bizzarro MJ, Shabanova V, Baltimore RS, Dembry LM, Ehrenkranz RA, Gallagher PG. Neonatal sepsis 2004–2013: the rise and fall of coagulase-negative staphylococci. J Pediatr. 2015;166:1193–9.

Article   PubMed   PubMed Central   Google Scholar  

Goddard B, Chang J, Sarkar IN. Using self organizing maps to compare sepsis patients from the neonatal and adult intensive care unit. AMIA Jt Summits Transl Sci Proc. 2019;2019:127–35.

PubMed   PubMed Central   Google Scholar  

Galal M, Symonds I, Murray H, Petraglia F, Smith R. Postterm pregnancy. Facts Views Vis Obgyn. 2012;4(3):175–87. Available from: https://www.ncbi.nlm.nih.gov/pubmed/24753906 .

Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47:1181–247.

Mwaniki MK, Atieno M, Lawn JE, Newton CRJC. Long-term neurodevelopmental outcomes after intrauterine and neonatal insults: a systematic review. Lancet. 2012;379:445–52.

Bayley N. Bayley scales of infant and toddler development, Third edition: screening test manual. San Antonio, Texas: Pearson Clinical Assessment PsychCorp; 2006.

Google Scholar  

Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol. 1997;39:214–23.

Spencer-Smith MM, Spittle AJ, Lee KJ, Doyle LW, Anderson PJ. Bayley-III cognitive and language scales in preterm children. Pediatrics. 2015;135(5):e1258-65.

Survival Sepsis Campaign. https://www.sccm.org/SurvivingSepsisCampaign/About-SSC/History . History of Surviving Sepsis Campaign | SCCM.

Tan B, Wong JJM, Sultana R, Koh JCJW, Jit M, Mok YH, et al. Global Case-Fatality Rates in Pediatric Severe Sepsis and Septic Shock: A Systematic Review and Meta-analysis. JAMA Pediatr. 2019;173:352–62.

Covidence systematic review software. Melbourne, Australia: Veritas Health Innovation; Available from: www.covidence.org .

GA Wells, B Shea, D O’Connell, J Peterson, V Welch, M Losos PT. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available from: https://www.ohri.ca/programs/clinical_epidemiology/oxford.asp . [cited 2022 Aug 9].

Rücker G, Cates CJ, Schwarzer G. Methods for including information from multi-arm trials in pairwise meta-analysis. Res Synth Methods. 2017;8:392–403.

R Core Team. R core team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org . 2021.

Schünemann H, Oxman JBGGA. GRADE Handbook. עלון הנוטע. 2013;66(1997).

Schlapbach LJ, Aebischer M, Adams M, Natalucci G, Bonhoeffer J, Latzin P, et al. Impact of sepsis on neurodevelopmental outcome in a swiss national cohort of extremely premature infants. Pediatrics. 2011;128:e348-57.

Adams-Chapman I, Bann CM, Das A, Goldberg RN, Stoll BJ, Walsh MC, et al. Neurodevelopmental outcome of extremely low birth weight infants with Candida infection. J Pediatr. 2013;163(4):961-7.e3.

de Haan TR, Beckers L, de Jonge RCJ, Spanjaard L, van Toledo L, Pajkrt D, et al. Neonatal gram negative and candida sepsis survival and neurodevelopmental outcome at the corrected age of 24 months. PLoS One. 2013;8:e59214.

Mitha A, Foix-L’Hélias L, Arnaud C, Marret S, Vieux R, Aujard Y, et al. Neonatal infection and 5-year neurodevelopmental outcome of very preterm infants. Pediatrics. 2013;132:e372-80.

Alshaikh B, Yee W, Lodha A, Henderson E, Yusuf K, Sauve R. Coagulase-negative staphylococcus sepsis in preterm infants and long-term neurodevelopmental outcome. J Perinatol. 2014;34:125–9.

Ferreira RC, Mello RR, Silva KS. Neonatal sepsis as a risk factor for neurodevelopmental changes in preterm infants with very low birth weight. J Pediatr (Rio J). 2014;90:293–9.

Hentges CR, Silveira RC, Procianoy RS, Carvalho CG, Filipouski GR, Fuentefria RN, et al. Association of late-onset neonatal sepsis with late neurodevelopment in the first two years of life of preterm infants with very low birth weight. J Pediatr (Rio J). 2014;90:50–7.

Dangor Z, Lala SG, Cutland CL, Koen A, Jose L, Nakwa F, et al. Burden of invasive group B Streptococcus disease and early neurological sequelae in South African infants. PLoS One. 2015;10:e0123014.

Savioli K, Rouse C, Susi A, Gorman G, Hisle-Gorman E. Suspected or known neonatal sepsis and neurodevelopmental delay by 5 years. J Perinatol. 2018;38:1573–80.

Singh L, Das S, Bhat VB, Plakkal N. Early neurodevelopmental outcome of very low birthweight neonates with culture-positive blood stream infection: a prospective cohort study. Cureus. 2018;10:e3492.

Zonnenberg IA, van Dijk-Lokkart EM, van den Dungen FAM, Vermeulen RJ, van Weissenbruch MM. Eur J Pediatr. 2019;178:673–80.

Nakwa FL, Lala SG, Madhi SA, Dangor Z. Neurodevelopmental impairment at 1 year of age in infants with previous invasive group B streptococcal sepsis and meningitis. Pediatric Infect Dis J. 2020;39:794–8.

Article   Google Scholar  

Mukhopadhyay S, Puopolo KM, Hansen NI, Lorch SA, Demauro SB, Greenberg RG, et al. Neurodevelopmental outcomes following neonatal late-onset sepsis and blood culture-negative conditions. Arch Dis Child Fetal Neonatal Ed. 2021;106:467–73.

Mukhopadhyay S, Puopolo KM, Hansen NI, Lorch SA, DeMauro SB, Greenberg RG, et al. Impact of early-onset sepsis and antibiotic use on death or survival with neurodevelopmental impairment at 2 years of age among extremely preterm infants. J Pediatr. 2020;221:39-46.e5.

Horváth-Puhó E, Snoek L, van Kassel MN, Gonçalves BP, Chandna J, Procter SR, et al. Prematurity modifies the risk of long-term neurodevelopmental impairments after invasive group B streptococcus infections during infancy in Denmark and the Netherlands. Clin Infect Dis. 2021;74:S44–53.

Article   PubMed Central   Google Scholar  

Horváth-Puhó E, van Kassel MN, Gonçalves BP, de Gier B, Procter SR, Paul P, et al. Mortality, neurodevelopmental impairments, and economic outcomes after invasive group B streptococcal disease in early infancy in Denmark and the Netherlands: a national matched cohort study. Lancet Child Adolesc Health. 2021;5:398–407.

Ortgies T, Rullmann M, Ziegelhöfer D, Bläser A, Thome UH. The role of early-onset-sepsis in the neurodevelopment of very low birth weight infants. BMC Pediatr. 2021;21:289.

Shim SY, Cho SJ, Park EA. Neurodevelopmental outcomes at 18–24 months of corrected age in very low birth weight infants with late-onset sepsis. J Korean Med Sci. 2021;36:e205.

Brumbaugh JE, Bell EF, Do BT, Greenberg RG, Stoll BJ, Demauro SB, et al. Incidence of and neurodevelopmental outcomes after late-onset meningitis among children born extremely preterm. JAMA Netw Open. 2022;5(12):e2245826.

Golin MO, Souza FIS, Paiva L da S, Sarni ROS. The value of clinical examination in preterm newborns after neonatal sepsis: a cross-sectional observational study. Dev Neurorehabil. 2022;25(2):80–6.

Humberg A, Fortmann MI, Spiegler J, Rausch TK, Siller B, Silwedel C, et al. Recurrent late-onset sepsis in extremely low birth weight infants is associated with motor deficits in early school age. Neonatology. 2022;119(6):695–702.

Kartam M, Embaireeg A, Albalool S, Almesafer A, Hammoud M, Al-Hathal M, et al. Late-onset sepsis in preterm neonates is associated with higher risks of cerebellar hemorrhage and lower motor scores at three years of age. Oman Med J. 2022;37(2):e368.

Paul P, Chandna J, Procter SR, Dangor Z, Leahy S, Santhanam S, et al. Neurodevelopmental and growth outcomes after invasive Group B Streptococcus in early infancy: a multi-country matched cohort study in South Africa, Mozambique, India, Kenya, and Argentina. EClinicalMedicine. 2022;47:101358.

Bright HR, Babata K, Allred EN, Erdei C, Kuban KCK, Joseph RM, et al. Neurocognitive outcomes at 10 years of age in extremely preterm newborns with late-onset bacteremia. Journal of Pediatrics. 2017;187:43-49. e1.

Romanelli RMC, Anchieta LM, Mourão MVA, Campos FA, Loyola FC, Mourão PHO, et al. Fatores de risco e letalidade de infecção da corrente sanguínea laboratorialmente confirmada, causada por patógenos não contaminantes da pele em recém-nascidos. J pediatr (Rio J). 2013;89(2):189–96.

Burgaleta M, Johnson W, Waber DP, Colom R, Karama S. Cognitive ability changes and dynamics of cortical thickness development in healthy children and adolescents. Neuroimage. 2014;84:810–9.

Peltopuro M, Ahonen T, Kaartinen J, Seppälä H, Närhi V. Borderline intellectual functioning: a systematic literature review. Intellect Dev Disabil. 2014;52:419–43.

Conti-Ramsden G, Durkin K, Toseeb U, Botting N, Pickles A. Education and employment outcomes of young adults with a history of developmental language disorder. Int J Lang Commun Disord. 2018;53:237–55.

Czeizel AE, Dudas I;, Murphy MM, Fernandez-Ballart JD, Arija V. Bayley scales of infant development-administration manual. Paediatr Perinat Epidemiol. 2019.

Bayley N. Manual for the Bayley Scales of Infant Development (2nd ed.). San Antonio: TX: The Psychological Corporation; 1993.

Bos AF. Bayley-Ii Or Bayley-Iii: what do the scores tell us? Dev Med Child Neurol. 2013;55:978–9.

Johnson S, Marlow N. Developmental screen or developmental testing? Early Hum Dev. 2006;82(3):173–83.

Bakhuizen SE, De Haan TR, Teune MJ, Van Wassenaer-Leemhuis AG, Van Der Heyden JL, Van Der Ham DP, et al. Meta-analysis shows that infants who have suffered neonatal sepsis face an increased risk of mortality and severe complications. Acta Paediatr Int J Paediatr. 2014;103:1211–8.

Cai S, Thompson DK, Yang JYM, Anderson PJ. Short-and long-term neurodevelopmental outcomes of very preterm infants with neonatal sepsis: a systematic review and meta-analysis. Children. 2019;6:131.

Joussen AM, Poulaki V, Le ML, Koizumi K, Esser C, Janicki H, Schraermeyer U, Kociok N, Fauser S, Kirchhof B, Kern TS, Adamis AP. A central role for inflammation in the pathogenesis of diabetic retinopathy. FASEB J. 2004;18(12):1450–2.

Ushio-Fukai M. VEGF signaling through NADPH oxidase-derived ROS. In: Antioxidants and Redox Signaling. 2007.

Sharma A, Leaf JM, Thomas S, Cane C, Stuart C, Tremlett C, et al. Sensorineural hearing loss after neonatal meningitis: a single-centre retrospective study. BMJ Paediatr Open. 2022;6(1):e001601.

Smilga AS, Garfinkle J, Ng P, Andersen J, Buckley D, Fehlings D, et al. Neonatal infection in children with cerebral palsy: a registry-based cohort study. Pediatr Neurol. 2018;80:77–83.

Hayes R, Hartnett J, Semova G, Murray C, Murphy K, Carroll L, et al. Neonatal sepsis definitions from randomised clinical trials. Pediatr Res. 2023;93:1141–8.

Rahman AE, Iqbal A, Hoque DME, Moinuddin M, Zaman SB, Rahman QSU, et al. Managing neonatal and early childhood syndromic sepsis in sub-district hospitals in resource poor settings: Improvement in quality of care through introduction of a package of interventions in rural Bangladesh. PLoS One. 2017;12(1):e0170267.

Singh M, Alsaleem M GCP. StatPearls. Treasure Island (FL): StatPearls Publishing. 2022. Neonatal Sepsis. [Updated 2022 Sep 29]. Available from: https://www-ncbi-nlm-nih-gov.libproxy1.nus.edu.sg/books/NBK531478/ .

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Diagnostic and Statistical Manual of Mental Disorders. 2022.

Book   Google Scholar  

McCauley SR, Wilde EA, Anderson VA, Bedell G, Beers SR, Campbell TF, et al. Recommendations for the Use of Common Outcome measures in pediatric traumatic brain injury research. J Neurotrauma. 2012;29:678–705.

Goh MSL, Looi DSH, Goh JL, Sultana R, Goh SSM, Lee JH, Chong SL. The Impact of Traumatic Brain Injury on Neurocognitive Outcomes in Children: a Systematic Review and Meta-Analysis. J Neurol Neurosurg Psychiatry. 2021:jnnp-2020-325066.

Download references

Acknowledgements

We would like to thank Ms. Wong Suei Nee, senior librarian from the National University of Singapore for helping us with the search strategy. We will also like to thank Dr Ming Ying Gan, Dr Shu Ting Tammie Seethor, Dr Jen Heng Pek, Dr Rachel Greenberg, Dr Christoph Hornik and Dr Bobby Tan, for their inputs in the initial design of this study.

Conflict of interest

No financial or non-financial benefits have been received or will be received from any party related directly or indirectly to the subject of this article.

Author information

Wei Jie Ong and Jun Jie Benjamin Seng are co-first authors.

Authors and Affiliations

MOH Holdings, Singapore, 1 Maritime Square, Singapore, 099253, Singapore

Wei Jie Ong, Jun Jie Benjamin Seng, Beijun Yap & Chen Lin Ng

SingHealth Regional Health System PULSES Centre, Singapore Health Services, Outram Rd, Singapore, 169608, Singapore

Jun Jie Benjamin Seng

SingHealth Duke-NUS Family Medicine Academic Clinical Programme, Singapore, Singapore

Yong Loo Lin School of Medicine, 10 Medical Dr, Yong Loo Lin School of Medicine, Singapore, Singapore

George He & Nooriyah Aliasgar Moochhala

Centre for Quantitative Medicine, Duke-NUS Medical School, Singapore, Singapore

Rehena Ganguly

Children’s Intensive Care Unit, KK Women’s and Children’s Hospital, SingHealth Paediatrics Academic Clinical Programme, 100 Bukit Timah Rd, Singapore, 229899, Singapore

Jan Hau Lee

Department of Emergency Medicine, KK Women’s and Children’s Hospital, SingHealth Paediatrics Academic Clinical Programme, SingHealth Emergency Medicine Academic Clinical Programme, 100 Bukit Timah Rd, Singapore, 229899, Singapore

Shu-Ling Chong

You can also search for this author in PubMed   Google Scholar

Contributions

SLC and JHL were the study’s principal investigators and were responsible for the conception and design of the study. WJO, JJBS, BY, GE, NAM and CLN were the co-investigators. WJO, JJBS, BY, GE, NAM and CLN were responsible for the screening and inclusion of articles and data extraction. All authors contributed to the data analyses and interpretation of data. WJO, JJBS, BY, GE, NAM and CLN prepared the initial draft of the manuscript. All authors revised the draft critically for important intellectual content and agreed to the final submission. All authors had access to all study data, revised the draft critically for important intellectual content and agreed to the final submission.

Corresponding author

Correspondence to Jun Jie Benjamin Seng .

Ethics declarations

Ethics approval and consent to participate.

As this was a systematic review with no access to patient data, ethical approval from the institutional review board was exempted.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary material 1., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ .

Reprints and permissions

About this article

Cite this article.

Ong, W.J., Seng, J.J.B., Yap, B. et al. Impact of neonatal sepsis on neurocognitive outcomes: a systematic review and meta-analysis. BMC Pediatr 24 , 505 (2024). https://doi.org/10.1186/s12887-024-04977-8

Download citation

Received : 11 May 2024

Accepted : 26 July 2024

Published : 07 August 2024

DOI : https://doi.org/10.1186/s12887-024-04977-8

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Neonatal sepsis
• Infantile sepsis
• Neurocognitive outcomes
• Systematic review

BMC Pediatrics

ISSN: 1471-2431

• Submission enquiries: [email protected]
• General enquiries: [email protected]

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Korean J Anesthesiol
• v.71(2); 2018 Apr

Introduction to systematic review and meta-analysis

1 Department of Anesthesiology and Pain Medicine, Inje University Seoul Paik Hospital, Seoul, Korea

2 Department of Anesthesiology and Pain Medicine, Chung-Ang University College of Medicine, Seoul, Korea

Systematic reviews and meta-analyses present results by combining and analyzing data from different studies conducted on similar research topics. In recent years, systematic reviews and meta-analyses have been actively performed in various fields including anesthesiology. These research methods are powerful tools that can overcome the difficulties in performing large-scale randomized controlled trials. However, the inclusion of studies with any biases or improperly assessed quality of evidence in systematic reviews and meta-analyses could yield misleading results. Therefore, various guidelines have been suggested for conducting systematic reviews and meta-analyses to help standardize them and improve their quality. Nonetheless, accepting the conclusions of many studies without understanding the meta-analysis can be dangerous. Therefore, this article provides an easy introduction to clinicians on performing and understanding meta-analyses.

Introduction

A systematic review collects all possible studies related to a given topic and design, and reviews and analyzes their results [ 1 ]. During the systematic review process, the quality of studies is evaluated, and a statistical meta-analysis of the study results is conducted on the basis of their quality. A meta-analysis is a valid, objective, and scientific method of analyzing and combining different results. Usually, in order to obtain more reliable results, a meta-analysis is mainly conducted on randomized controlled trials (RCTs), which have a high level of evidence [ 2 ] ( Fig. 1 ). Since 1999, various papers have presented guidelines for reporting meta-analyses of RCTs. Following the Quality of Reporting of Meta-analyses (QUORUM) statement [ 3 ], and the appearance of registers such as Cochrane Library’s Methodology Register, a large number of systematic literature reviews have been registered. In 2009, the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [ 4 ] was published, and it greatly helped standardize and improve the quality of systematic reviews and meta-analyses [ 5 ].

Levels of evidence.

In anesthesiology, the importance of systematic reviews and meta-analyses has been highlighted, and they provide diagnostic and therapeutic value to various areas, including not only perioperative management but also intensive care and outpatient anesthesia [6–13]. Systematic reviews and meta-analyses include various topics, such as comparing various treatments of postoperative nausea and vomiting [ 14 , 15 ], comparing general anesthesia and regional anesthesia [ 16 – 18 ], comparing airway maintenance devices [ 8 , 19 ], comparing various methods of postoperative pain control (e.g., patient-controlled analgesia pumps, nerve block, or analgesics) [ 20 – 23 ], comparing the precision of various monitoring instruments [ 7 ], and meta-analysis of dose-response in various drugs [ 12 ].

Thus, literature reviews and meta-analyses are being conducted in diverse medical fields, and the aim of highlighting their importance is to help better extract accurate, good quality data from the flood of data being produced. However, a lack of understanding about systematic reviews and meta-analyses can lead to incorrect outcomes being derived from the review and analysis processes. If readers indiscriminately accept the results of the many meta-analyses that are published, incorrect data may be obtained. Therefore, in this review, we aim to describe the contents and methods used in systematic reviews and meta-analyses in a way that is easy to understand for future authors and readers of systematic review and meta-analysis.

Study Planning

It is easy to confuse systematic reviews and meta-analyses. A systematic review is an objective, reproducible method to find answers to a certain research question, by collecting all available studies related to that question and reviewing and analyzing their results. A meta-analysis differs from a systematic review in that it uses statistical methods on estimates from two or more different studies to form a pooled estimate [ 1 ]. Following a systematic review, if it is not possible to form a pooled estimate, it can be published as is without progressing to a meta-analysis; however, if it is possible to form a pooled estimate from the extracted data, a meta-analysis can be attempted. Systematic reviews and meta-analyses usually proceed according to the flowchart presented in Fig. 2 . We explain each of the stages below.

Flowchart illustrating a systematic review.

Formulating research questions

A systematic review attempts to gather all available empirical research by using clearly defined, systematic methods to obtain answers to a specific question. A meta-analysis is the statistical process of analyzing and combining results from several similar studies. Here, the definition of the word “similar” is not made clear, but when selecting a topic for the meta-analysis, it is essential to ensure that the different studies present data that can be combined. If the studies contain data on the same topic that can be combined, a meta-analysis can even be performed using data from only two studies. However, study selection via a systematic review is a precondition for performing a meta-analysis, and it is important to clearly define the Population, Intervention, Comparison, Outcomes (PICO) parameters that are central to evidence-based research. In addition, selection of the research topic is based on logical evidence, and it is important to select a topic that is familiar to readers without clearly confirmed the evidence [ 24 ].

Protocols and registration

In systematic reviews, prior registration of a detailed research plan is very important. In order to make the research process transparent, primary/secondary outcomes and methods are set in advance, and in the event of changes to the method, other researchers and readers are informed when, how, and why. Many studies are registered with an organization like PROSPERO ( http://www.crd.york.ac.uk/PROSPERO/ ), and the registration number is recorded when reporting the study, in order to share the protocol at the time of planning.

Defining inclusion and exclusion criteria

Information is included on the study design, patient characteristics, publication status (published or unpublished), language used, and research period. If there is a discrepancy between the number of patients included in the study and the number of patients included in the analysis, this needs to be clearly explained while describing the patient characteristics, to avoid confusing the reader.

Literature search and study selection

In order to secure proper basis for evidence-based research, it is essential to perform a broad search that includes as many studies as possible that meet the inclusion and exclusion criteria. Typically, the three bibliographic databases Medline, Embase, and Cochrane Central Register of Controlled Trials (CENTRAL) are used. In domestic studies, the Korean databases KoreaMed, KMBASE, and RISS4U may be included. Effort is required to identify not only published studies but also abstracts, ongoing studies, and studies awaiting publication. Among the studies retrieved in the search, the researchers remove duplicate studies, select studies that meet the inclusion/exclusion criteria based on the abstracts, and then make the final selection of studies based on their full text. In order to maintain transparency and objectivity throughout this process, study selection is conducted independently by at least two investigators. When there is a inconsistency in opinions, intervention is required via debate or by a third reviewer. The methods for this process also need to be planned in advance. It is essential to ensure the reproducibility of the literature selection process [ 25 ].

Quality of evidence

However, well planned the systematic review or meta-analysis is, if the quality of evidence in the studies is low, the quality of the meta-analysis decreases and incorrect results can be obtained [ 26 ]. Even when using randomized studies with a high quality of evidence, evaluating the quality of evidence precisely helps determine the strength of recommendations in the meta-analysis. One method of evaluating the quality of evidence in non-randomized studies is the Newcastle-Ottawa Scale, provided by the Ottawa Hospital Research Institute 1) . However, we are mostly focusing on meta-analyses that use randomized studies.

If the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) system ( http://www.gradeworkinggroup.org/ ) is used, the quality of evidence is evaluated on the basis of the study limitations, inaccuracies, incompleteness of outcome data, indirectness of evidence, and risk of publication bias, and this is used to determine the strength of recommendations [ 27 ]. As shown in Table 1 , the study limitations are evaluated using the “risk of bias” method proposed by Cochrane 2) . This method classifies bias in randomized studies as “low,” “high,” or “unclear” on the basis of the presence or absence of six processes (random sequence generation, allocation concealment, blinding participants or investigators, incomplete outcome data, selective reporting, and other biases) [ 28 ].

The Cochrane Collaboration’s Tool for Assessing the Risk of Bias [ 28 ]

Data extraction

Two different investigators extract data based on the objectives and form of the study; thereafter, the extracted data are reviewed. Since the size and format of each variable are different, the size and format of the outcomes are also different, and slight changes may be required when combining the data [ 29 ]. If there are differences in the size and format of the outcome variables that cause difficulties combining the data, such as the use of different evaluation instruments or different evaluation timepoints, the analysis may be limited to a systematic review. The investigators resolve differences of opinion by debate, and if they fail to reach a consensus, a third-reviewer is consulted.

Data Analysis

The aim of a meta-analysis is to derive a conclusion with increased power and accuracy than what could not be able to achieve in individual studies. Therefore, before analysis, it is crucial to evaluate the direction of effect, size of effect, homogeneity of effects among studies, and strength of evidence [ 30 ]. Thereafter, the data are reviewed qualitatively and quantitatively. If it is determined that the different research outcomes cannot be combined, all the results and characteristics of the individual studies are displayed in a table or in a descriptive form; this is referred to as a qualitative review. A meta-analysis is a quantitative review, in which the clinical effectiveness is evaluated by calculating the weighted pooled estimate for the interventions in at least two separate studies.

The pooled estimate is the outcome of the meta-analysis, and is typically explained using a forest plot ( Figs. 3 and ​ and4). 4 ). The black squares in the forest plot are the odds ratios (ORs) and 95% confidence intervals in each study. The area of the squares represents the weight reflected in the meta-analysis. The black diamond represents the OR and 95% confidence interval calculated across all the included studies. The bold vertical line represents a lack of therapeutic effect (OR = 1); if the confidence interval includes OR = 1, it means no significant difference was found between the treatment and control groups.

Forest plot analyzed by two different models using the same data. (A) Fixed-effect model. (B) Random-effect model. The figure depicts individual trials as filled squares with the relative sample size and the solid line as the 95% confidence interval of the difference. The diamond shape indicates the pooled estimate and uncertainty for the combined effect. The vertical line indicates the treatment group shows no effect (OR = 1). Moreover, if the confidence interval includes 1, then the result shows no evidence of difference between the treatment and control groups.

Forest plot representing homogeneous data.

Dichotomous variables and continuous variables

In data analysis, outcome variables can be considered broadly in terms of dichotomous variables and continuous variables. When combining data from continuous variables, the mean difference (MD) and standardized mean difference (SMD) are used ( Table 2 ).

Summary of Meta-analysis Methods Available in RevMan [ 28 ]

The MD is the absolute difference in mean values between the groups, and the SMD is the mean difference between groups divided by the standard deviation. When results are presented in the same units, the MD can be used, but when results are presented in different units, the SMD should be used. When the MD is used, the combined units must be shown. A value of “0” for the MD or SMD indicates that the effects of the new treatment method and the existing treatment method are the same. A value lower than “0” means the new treatment method is less effective than the existing method, and a value greater than “0” means the new treatment is more effective than the existing method.

When combining data for dichotomous variables, the OR, risk ratio (RR), or risk difference (RD) can be used. The RR and RD can be used for RCTs, quasi-experimental studies, or cohort studies, and the OR can be used for other case-control studies or cross-sectional studies. However, because the OR is difficult to interpret, using the RR and RD, if possible, is recommended. If the outcome variable is a dichotomous variable, it can be presented as the number needed to treat (NNT), which is the minimum number of patients who need to be treated in the intervention group, compared to the control group, for a given event to occur in at least one patient. Based on Table 3 , in an RCT, if x is the probability of the event occurring in the control group and y is the probability of the event occurring in the intervention group, then x = c/(c + d), y = a/(a + b), and the absolute risk reduction (ARR) = x − y. NNT can be obtained as the reciprocal, 1/ARR.

Calculation of the Number Needed to Treat in the Dichotomous table

Fixed-effect models and random-effect models

In order to analyze effect size, two types of models can be used: a fixed-effect model or a random-effect model. A fixed-effect model assumes that the effect of treatment is the same, and that variation between results in different studies is due to random error. Thus, a fixed-effect model can be used when the studies are considered to have the same design and methodology, or when the variability in results within a study is small, and the variance is thought to be due to random error. Three common methods are used for weighted estimation in a fixed-effect model: 1) inverse variance-weighted estimation 3) , 2) Mantel-Haenszel estimation 4) , and 3) Peto estimation 5) .

A random-effect model assumes heterogeneity between the studies being combined, and these models are used when the studies are assumed different, even if a heterogeneity test does not show a significant result. Unlike a fixed-effect model, a random-effect model assumes that the size of the effect of treatment differs among studies. Thus, differences in variation among studies are thought to be due to not only random error but also between-study variability in results. Therefore, weight does not decrease greatly for studies with a small number of patients. Among methods for weighted estimation in a random-effect model, the DerSimonian and Laird method 6) is mostly used for dichotomous variables, as the simplest method, while inverse variance-weighted estimation is used for continuous variables, as with fixed-effect models. These four methods are all used in Review Manager software (The Cochrane Collaboration, UK), and are described in a study by Deeks et al. [ 31 ] ( Table 2 ). However, when the number of studies included in the analysis is less than 10, the Hartung-Knapp-Sidik-Jonkman method 7) can better reduce the risk of type 1 error than does the DerSimonian and Laird method [ 32 ].

Fig. 3 shows the results of analyzing outcome data using a fixed-effect model (A) and a random-effect model (B). As shown in Fig. 3 , while the results from large studies are weighted more heavily in the fixed-effect model, studies are given relatively similar weights irrespective of study size in the random-effect model. Although identical data were being analyzed, as shown in Fig. 3 , the significant result in the fixed-effect model was no longer significant in the random-effect model. One representative example of the small study effect in a random-effect model is the meta-analysis by Li et al. [ 33 ]. In a large-scale study, intravenous injection of magnesium was unrelated to acute myocardial infarction, but in the random-effect model, which included numerous small studies, the small study effect resulted in an association being found between intravenous injection of magnesium and myocardial infarction. This small study effect can be controlled for by using a sensitivity analysis, which is performed to examine the contribution of each of the included studies to the final meta-analysis result. In particular, when heterogeneity is suspected in the study methods or results, by changing certain data or analytical methods, this method makes it possible to verify whether the changes affect the robustness of the results, and to examine the causes of such effects [ 34 ].

Heterogeneity

Homogeneity test is a method whether the degree of heterogeneity is greater than would be expected to occur naturally when the effect size calculated from several studies is higher than the sampling error. This makes it possible to test whether the effect size calculated from several studies is the same. Three types of homogeneity tests can be used: 1) forest plot, 2) Cochrane’s Q test (chi-squared), and 3) Higgins I 2 statistics. In the forest plot, as shown in Fig. 4 , greater overlap between the confidence intervals indicates greater homogeneity. For the Q statistic, when the P value of the chi-squared test, calculated from the forest plot in Fig. 4 , is less than 0.1, it is considered to show statistical heterogeneity and a random-effect can be used. Finally, I 2 can be used [ 35 ].

I 2 , calculated as shown above, returns a value between 0 and 100%. A value less than 25% is considered to show strong homogeneity, a value of 50% is average, and a value greater than 75% indicates strong heterogeneity.

Even when the data cannot be shown to be homogeneous, a fixed-effect model can be used, ignoring the heterogeneity, and all the study results can be presented individually, without combining them. However, in many cases, a random-effect model is applied, as described above, and a subgroup analysis or meta-regression analysis is performed to explain the heterogeneity. In a subgroup analysis, the data are divided into subgroups that are expected to be homogeneous, and these subgroups are analyzed. This needs to be planned in the predetermined protocol before starting the meta-analysis. A meta-regression analysis is similar to a normal regression analysis, except that the heterogeneity between studies is modeled. This process involves performing a regression analysis of the pooled estimate for covariance at the study level, and so it is usually not considered when the number of studies is less than 10. Here, univariate and multivariate regression analyses can both be considered.

Publication bias

Publication bias is the most common type of reporting bias in meta-analyses. This refers to the distortion of meta-analysis outcomes due to the higher likelihood of publication of statistically significant studies rather than non-significant studies. In order to test the presence or absence of publication bias, first, a funnel plot can be used ( Fig. 5 ). Studies are plotted on a scatter plot with effect size on the x-axis and precision or total sample size on the y-axis. If the points form an upside-down funnel shape, with a broad base that narrows towards the top of the plot, this indicates the absence of a publication bias ( Fig. 5A ) [ 29 , 36 ]. On the other hand, if the plot shows an asymmetric shape, with no points on one side of the graph, then publication bias can be suspected ( Fig. 5B ). Second, to test publication bias statistically, Begg and Mazumdar’s rank correlation test 8) [ 37 ] or Egger’s test 9) [ 29 ] can be used. If publication bias is detected, the trim-and-fill method 10) can be used to correct the bias [ 38 ]. Fig. 6 displays results that show publication bias in Egger’s test, which has then been corrected using the trim-and-fill method using Comprehensive Meta-Analysis software (Biostat, USA).

Funnel plot showing the effect size on the x-axis and sample size on the y-axis as a scatter plot. (A) Funnel plot without publication bias. The individual plots are broader at the bottom and narrower at the top. (B) Funnel plot with publication bias. The individual plots are located asymmetrically.

Funnel plot adjusted using the trim-and-fill method. White circles: comparisons included. Black circles: inputted comparisons using the trim-and-fill method. White diamond: pooled observed log risk ratio. Black diamond: pooled inputted log risk ratio.

Result Presentation

When reporting the results of a systematic review or meta-analysis, the analytical content and methods should be described in detail. First, a flowchart is displayed with the literature search and selection process according to the inclusion/exclusion criteria. Second, a table is shown with the characteristics of the included studies. A table should also be included with information related to the quality of evidence, such as GRADE ( Table 4 ). Third, the results of data analysis are shown in a forest plot and funnel plot. Fourth, if the results use dichotomous data, the NNT values can be reported, as described above.

The GRADE Evidence Quality for Each Outcome

N: number of studies, ROB: risk of bias, PON: postoperative nausea, POV: postoperative vomiting, PONV: postoperative nausea and vomiting, CI: confidence interval, RR: risk ratio, AR: absolute risk.

When Review Manager software (The Cochrane Collaboration, UK) is used for the analysis, two types of P values are given. The first is the P value from the z-test, which tests the null hypothesis that the intervention has no effect. The second P value is from the chi-squared test, which tests the null hypothesis for a lack of heterogeneity. The statistical result for the intervention effect, which is generally considered the most important result in meta-analyses, is the z-test P value.

A common mistake when reporting results is, given a z-test P value greater than 0.05, to say there was “no statistical significance” or “no difference.” When evaluating statistical significance in a meta-analysis, a P value lower than 0.05 can be explained as “a significant difference in the effects of the two treatment methods.” However, the P value may appear non-significant whether or not there is a difference between the two treatment methods. In such a situation, it is better to announce “there was no strong evidence for an effect,” and to present the P value and confidence intervals. Another common mistake is to think that a smaller P value is indicative of a more significant effect. In meta-analyses of large-scale studies, the P value is more greatly affected by the number of studies and patients included, rather than by the significance of the results; therefore, care should be taken when interpreting the results of a meta-analysis.

When performing a systematic literature review or meta-analysis, if the quality of studies is not properly evaluated or if proper methodology is not strictly applied, the results can be biased and the outcomes can be incorrect. However, when systematic reviews and meta-analyses are properly implemented, they can yield powerful results that could usually only be achieved using large-scale RCTs, which are difficult to perform in individual studies. As our understanding of evidence-based medicine increases and its importance is better appreciated, the number of systematic reviews and meta-analyses will keep increasing. However, indiscriminate acceptance of the results of all these meta-analyses can be dangerous, and hence, we recommend that their results be received critically on the basis of a more accurate understanding.

1) http://www.ohri.ca .

2) http://methods.cochrane.org/bias/assessing-risk-bias-included-studies .

3) The inverse variance-weighted estimation method is useful if the number of studies is small with large sample sizes.

4) The Mantel-Haenszel estimation method is useful if the number of studies is large with small sample sizes.

5) The Peto estimation method is useful if the event rate is low or one of the two groups shows zero incidence.

6) The most popular and simplest statistical method used in Review Manager and Comprehensive Meta-analysis software.

7) Alternative random-effect model meta-analysis that has more adequate error rates than does the common DerSimonian and Laird method, especially when the number of studies is small. However, even with the Hartung-Knapp-Sidik-Jonkman method, when there are less than five studies with very unequal sizes, extra caution is needed.

8) The Begg and Mazumdar rank correlation test uses the correlation between the ranks of effect sizes and the ranks of their variances [ 37 ].

9) The degree of funnel plot asymmetry as measured by the intercept from the regression of standard normal deviates against precision [ 29 ].

10) If there are more small studies on one side, we expect the suppression of studies on the other side. Trimming yields the adjusted effect size and reduces the variance of the effects by adding the original studies back into the analysis as a mirror image of each study.

• Open access
• Published: 07 August 2024

Cord blood transfusions in extremely low gestational age neonates to reduce severe retinopathy of prematurity: results of a prespecified interim analysis of the randomized BORN trial

• Luciana Teofili   ORCID: orcid.org/0000-0002-7214-1561 1 , 2   na1 ,
• Patrizia Papacci 1 , 2   na1 ,
• Carlo Dani 3 ,
• Francesco Cresi 4 , 5 ,
• Giulia Remaschi 3 ,
• Claudio Pellegrino 1 , 2 ,
• Maria Bianchi 1 ,
• Giulia Ansaldi 4 ,
• Maria Francesca Campagnoli 5 ,
• Barbara Vania 4 ,
• Domenico Lepore 1 , 2 ,
• Fabrizio Gaetano Saverio Franco 3 ,
• Marco Fabbri 6 ,
• Roberta Penta de Vera d’ Aragona 7 ,
• Anna Molisso 8 ,
• Enrico Beccastrini 3 ,
• Antonella Dragonetti 4 ,
• Lorenzo Orazi 9 ,
• Tina Pasciuto 1 ,
• Iolanda Mozzetta 1 ,
• Antonio Baldascino 1 ,
• Emanuela Locatelli 4 ,
• Caterina Giovanna Valentini 1 ,
• Carmen Giannantonio 1 ,
• Brigida Carducci 1 ,
• Sabrina Gabbriellini 6 ,
• Roberto Albiani 4 ,
• Elena Ciabatti 6 ,
• Nicola Nicolotti 1 ,
• Silvia Baroni 1 , 2 ,
• Alessandro Mazzoni 6 ,
• Federico Genzano Besso 4 ,
• Francesca Serrao 1 ,
• Velia Purcaro 1 ,
• Alessandra Coscia 4 , 5 ,
• Roberta Pizzolo 4 ,
• Genny Raffaeli 10 , 11 ,
• Stefania Villa 10 ,
• Isabella Mondello 12 ,
• Alfonso Trimarchi 12 ,
• Flavia Beccia 1 , 2 ,
• Stefano Ghirardello 13 &
• Giovanni Vento 1 , 2  

Italian Journal of Pediatrics volume  50 , Article number:  142 ( 2024 ) Cite this article

198 Accesses

1 Altmetric

Metrics details

Preterm infants are at high risk for retinopathy of prematurity (ROP), with potential life-long visual impairment. Low fetal hemoglobin (HbF) levels predict ROP. It is unknown if preventing the HbF decrease also reduces ROP.

BORN is an ongoing multicenter double-blinded randomized controlled trial investigating whether transfusing HbF-enriched cord blood-red blood cells (CB-RBCs) instead of adult donor-RBC units (A-RBCs) reduces the incidence of severe ROP (NCT05100212). Neonates born between 24 and 27 + 6 weeks of gestation are enrolled and randomized 1:1 to receive adult donor-RBCs (A-RBCs, arm A) or allogeneic CB-RBCs (arm B) from birth to the postmenstrual age (PMA) of 31 + 6 weeks. Primary outcome is the rate of severe ROP at 40 weeks of PMA or discharge, with a sample size of 146 patients. A prespecified interim analysis was scheduled after the first 58 patients were enrolled, with the main purpose to evaluate the safety of CB-RBC transfusions.

Results in the intention-to-treat and per-protocol analysis are reported. Twenty-eight patients were in arm A and 30 in arm B. Overall, 104 A-RBC units and 49 CB-RBC units were transfused, with a high rate of protocol deviations. A total of 336 adverse events were recorded, with similar incidence and severity in the two arms. By per-protocol analysis, patients receiving A-RBCs or both RBC types experienced more adverse events than non-transfused patients or those transfused exclusively with CB-RBCs, and suffered from more severe forms of bradycardia, pulmonary hypertension, and hemodynamically significant patent ductus arteriosus. Serum potassium, lactate, and pH were similar after CB-RBCs or A-RBCs. Fourteen patients died and 44 were evaluated for ROP. Ten of them developed severe ROP, with no differences between arms. At per-protocol analysis each A-RBC transfusion carried a relative risk for severe ROP of 1.66 (95% CI 1.06–2.20) in comparison with CB-RBCs. The area under the curve of HbF suggested that HbF decrement before 30 weeks PMA is critical for severe ROP development. Subsequent CB-RBC transfusions do not lessen the ROP risk.

Conclusions

The interim analysis shows that CB-RBC transfusion strategy in preterm neonates is safe and, if early adopted, might protect them from severe ROP.

Trial registration

Prospectively registered at ClinicalTrials.gov on October 29, 2021. Identifier number NCT05100212.

Progressive improvements in obstetric and neonatal care have yielded significant decreases in mortality of preterm infants, particularly for those born at 23–25 weeks of gestational age [ 1 ]. Nonetheless, there are concerns regarding long-term morbidities faced by the surviving infants [ 2 ]. Among various complications, the incidence of retinopathy of prematurity (ROP) has recently remained stable or even increased [ 3 , 4 ]. ROP is a leading cause of childhood blindness and significantly influences neurodevelopmental outcomes of affected patients [ 5 , 6 ]. The pathogenesis of ROP involves abnormal angiogenesis in the immature retina, with low gestational age and birth weight as primary risk factors [ 7 ]. An association between fetal hemoglobin (HbF) reduction and ROP incidence was also reported, likely explaining the connection between red blood cell (RBC) transfusions and ROP [ 8 , 9 , 10 ].

RBC transfusions are acknowledged causes of oxidative stress in preterm neonates. Following RBC transfusions, non-transferrin bound iron has been reported to increase in preterm but not in term neonates [ 11 ]. Similarly, after transfusions, a raise of markers of hemolysis, inflammation and endothelial cell activation occurs [ 12 , 13 , 14 ]. In comparison with non-transfused similar-age neonates, preterm neonates receiving repeated RBC transfusions experience prolonged exposure to adult hemoglobin (HbA) [ 15 , 16 ]. The lower oxygen affinity of HbA compared to HbF likely perturbs physiological oxygen delivery to developing organs and tissues and was proposed to play a role in ROP development and progression [ 17 ]. Accordingly, preterm neonates with lower HbF show significantly higher oxidative stress biomarkers levels [ 18 ]. Allogenic cord blood (CB) collected from full-term deliveries is a suitable source of HbF-enriched RBCs (CB-RBCs) for transfusion into preterm neonates. In a pilot proof-of-concept study enrolling extremely low birth weight neonates, CB-RBC concentrates increased Hb levels as well as RBC concentrates from adult donors (A-RBCs), without decreasing HbF levels [ 19 ].

BORN is a randomized, multi-center, double-blinded, controlled trial to assess whether transfusing CB-RBCs, instead of A-RBCs, reduces the incidence of severe ROP in extremely low gestational age neonates (ELGANs, i.e., neonates born at less than 28 weeks of gestation) [ 19 ]. The current study reports the results of the prespecified interim analysis on the safety of this transfusion strategy and, secondarily, on the efficacy in preventing severe ROP.

Study aim, design and participants

BORN is an ongoing randomized, multi-center, double-blinded, controlled trial investigating CB-RBC instead of A-RBC transfusion to reduce the incidence of severe ROP. A detailed version of the study protocol has been previously published [ 20 ]. Inclusion criteria are birth at a gestational age (GA) between 24 + 0 and 27 + 6 weeks and signed informed consent of parents or legal representatives. Out-born infants are suitable for enrolment only if not previously transfused. Exclusion criteria include maternal-fetal immunization, hydrops fetalis, major congenital malformations associated, or not, with genetic syndromes [ 21 ], previous transfusions, hemorrhage at birth, congenital infections, and the health care team deeming it inappropriate to approach the infant’s family for informed consent. The study is approved by the Ethics Committee of Fondazione Policlinico A. Gemelli IRCCS (ID 4364, Prot. N. 003590/21) and of all participating centers, and is registered at https://clinicaltrials.gov (NCT05100212).

Enrollment, randomization and masking

Neonates are enrolled at NICU admission and randomized 1:1 to receive either standard A-RBCs (Arm A, control) or CB-RBCs (Arm B, intervention) until the post-menstrual age (PMA) of 31 + 6 weeks, with subsequent transfusions consisting of only standard A-RBCs. Twins are assigned to the same arm [ 22 ]. If compatible CB-RBC units are not available, A-RBC units are assigned. RBC types are given according to the arm allocation until the PMA of 31 + 6 weeks. The allocation sequence is generated using both stratifications for center and gestational age (< or ≥ 26 weeks), and permuted blocks with random block sizes and block order, using the NCSS 2020 Statistical Software 2020 (NCSS, LLC. Kaysville, UT, USA, ncss.com/software/ncss). The allocation table is not disclosed to ensure concealment, and randomization is provided through the Research Electronic Data Capture (RedCap) (RRID: SCR_003445) electronic data capture tool, hosted at Fondazione Policlinico Universitario A. Gemelli, IRCCS ( https://redcap-irccs.policlinicogemelli.it/ ). Randomization is performed by the blood bank medical staff; treating neonatologists are unaware to which arm neonates were assigned. To conceal the type of blood product, A-RBCs and CB-RBCs are distributed to neonatal intensive care units (NICUs) in the same type of bags (CompoFlex 4 F RCC storage pediatric bags).

Intervention and control description

Enrolled patients receive standard A-RBCs (Arm A, control) or allogenic CB-RBCs (Arm B, intervention). CB units are provided from public cord blood banks belonging to the Italian Cord Blood Bank Network, and processed into CB-RBC concentrates according to a standardized procedure detailed in the protocol study [ 20 ]. Both CB-RBC and A-RBC concentrates are pre-storage leukodepleted and prepared in compliance with quality standards required by Italian regulations and European guidelines. CB-RBC units are proved negative for bacterial and fungal contamination before distribution. All units are matched for ABO/RhD antigens, and γ-irradiated at the time of distribution. Transfusion therapy is managed according to Italian transfusion guidelines, and erythropoietin is administered as per center procedure [ 22 ].

Study outcome

The primary outcome of BORN is the incidence of severe ROP (stages 3 or higher according to International Classification criteria) [ 23 ] in the CB-RBC and A-RBC arms at discharge or at 40 weeks of PMA, whichever occurs first. ROP assessment is performed through indirect ophthalmoscopy, and RetCam imaging is used to confirm the diagnosis and to monitor disease status and treatment. According to the primary outcome, a sample size of 146 patients (73 per arm) was estimated. Due to the innovative type of the intervention, an interim analysis was planned after the first 58 patients were randomized, to confirm the safety of CB-RBC transfusion in ELGANs.

Data collection

Data are collected in REDCap in different record forms for patients and for CB unit processing. Baseline patient data include: demographics, obstetric pathology, gestational age, birth weight, antenatal prophylaxis for hyaline membrane disease, suspected (maternal fever, leukocytosis, uterine tenderness, malodorous discharge, tachycardia) or documented (histology) chorioamnionitis, post-natal steroid administration, hematological parameters [hemoglobin concentration (Hb), hematocrit (Htc), and HbF (assessed spectrophotometrically using point-of-care blood gas analyzers and expressed as percentage of total Hb), Apgar index at 1 and 5 min, Clinical Risk Index for Babies-II (CRIB-II) score, and mortality probability [ 24 ]. During hospitalization, the following clinical variables are recorded: hemodynamically-significant patent ductus arteriosus (hsPDA), maximal stage of ROP [ 23 ], necrotizing enterocolitis (NEC) [ 25 , 26 ], bronchopulmonary dysplasia (BPD) [ 27 ], intraventricular hemorrhage (IVH) [ 28 ], ROP treatment, erythropoietin administration, microbiologically-documented infections, duration of oxygen therapy, ventilation support (i.e., invasive ventilation: high frequency oscillatory ventilation and/or other conventional ventilation modalities; non-invasive ventilation: continuous positive airway pressure and/or high flow nasal cannula), surgery, and death. All complications during the clinical course were considered adverse events: the severity was reported according to the Common Terminology Criteria for Adverse Events (CTCAE) v4.0, ranging from 1 (no symptoms) to 5 (death), while the imputability to previous transfusions was scored according to the European regulations on transfusion surveillance from 0 to 3, where 0 indicates excluded/unlikely, 1 possible, 2 likely/probable, and 3 certain. Hb, Hct, and HbF values were recorded twice per week. At each RBC transfusion, the following were recorded: unit identifier number, date of transfusion, Hct value before and after transfusion, and post-transfusion (within 24 h) pH, lactate, and potassium levels. Data collected for each CB-RBC unit included parameters at: collection (collection date and CB volume before fractionation), processing (date, post-processing RBC recovery, platelet depletion, residual leukocyte count, hemolysis rate), and, for non-transfused units, the end-of-storage hemolysis rate (at 14 days).

Statistical analysis

A comprehensive descriptive analysis of the study population was performed. Demographic characteristics, clinical features, and relevant baseline variables were expressed as median and interquartile range (IQR), mean and standard deviation (SD), or proportions. To investigate differences between groups, we employed the Chi2 test, applying Fisher correction if needed. Relative Risk (RR) was calculated to assess the association between the RBC-type transfusion and the primary outcome (ROP). The impact of transfusion burden on the primary outcome was estimated by logistic regression analysis, using the primary outcome as a dependent variable. This method allowed us to assess the influence of the RBC type, adjusting for the number of transfusions and other covariates selected based on their clinical relevance and significance in univariate analyses. Adjusted odds ratios (ORs) and their corresponding confidence intervals were reported to quantify the association between the predictors and the outcome. To explore the impact of blood products on HbF levels, the area under the curve (AUC) of HbF and PMA was calculated, and results were compared among different groups of patients. Missing data, unless not otherwise specified, were below 5% and were not managed except for the AUC calculation of HbF and PMA. In this case, whenever possible, missing data were calculated based on the preceding HbF determination using previously reported algorithms [ 19 ].

Statistical analysis was performed in the “intention-to-treat” set (all enrolled patients, independently if they were or not transfused, including major protocol deviations, i.e., those patients in the CB-RBC arm who received A-RBC transfusions due to CB-RBC unit unavailability), and “per protocol” set (all patients categorized according to the transfusions received until the PMA of 32 weeks: no transfusion, only A-RBCs, only CB-RBCs, both A-RBCs and CB-RBCs).

The following software programs were used to perform statistical analysis and prepare illustrations: NCSS 10 Statistical Software 2015 (NCSS, LLC. Kaysville, Utah, USA, ncss.com/software/ncss), Stata Statistical Software, Release 18 (StataCorp. 2023. College Station, TX: StataCorp LLC) and GraphPad Prism version 10.0.0 (GraphPad Software, Boston, Massachusetts USA, www.graphpad.com ).

Study recruitment started on December 1, 2021, and the first patient was enrolled on December 17, 2021. Patients were recruited in NICUs of 4 tertiary Italian hospitals, and CB-RBC units were collected by 6 public CB banks. By August 31, 2023, 58 patients were enrolled and completed their follow-up: 14 died without reaching the primary endpoint (40 weeks of PMA or discharge, whichever occurred first), whereas 44 reached the primary endpoint and were evaluated for ROP (Fig.  1 A). Figure  1 B shows the patients’ distribution according to the types of blood products received and the two sets of analysis. Table  1 illustrates clinical and laboratory characteristics of the 58 patients: 28 were assigned to the arm A (A-RBCs) and 30 to the arm B (CB-RBCs). The arms were well balanced for prenatal and neonatal characteristics, hematological parameters at birth, and follow-up duration. Four patients in the Arm A and 6 in the arm B received erythropoietin during hospitalization, that was administered intravenously at a dose of 200 U/kg/die, followed by a maintenance dose of 400 U/Kg by subcutaneous injection three times per week. The overall mortality was 24.1% (95% CI 15.0-36.6) and the mortality rate was similar in the two arms (Table  1 ). Moreover, the mortality rate did not statistically differ among groups regardless of the type of transfusions received [31.3% (95% CI 14.2–55.6) in non-transfused patients, 29.2% (95% CI 14.9–49.2) in neonates receiving only A-RBCs, 25.0% (95% CI 4.4–59.1) in those receiving only CB-RBCs, and 0 (95% CI 0-27.8) in those receiving both A-RBCs and CB-RBCs, p  = 0.235)].

Adverse events and imputability of previous transfusions

During the study period, 336 adverse events were recorded: 165 occurred in the arm A and 171 in the arm B (Table  2 ). Individual patients could have multiple different types of events or multiple episodes of the same event. There was no difference between arms regarding the incidence of any of the events listed (Table  2 ). The rate of adverse events was also assessed in the “per protocol” setting (eTable  1 ), comparing patients enrolled according to the type of blood products received before the PMA of 32 weeks (Fig.  1 B). Comprehensively considered, patients receiving only A-RBCs or both types of RBCs experienced more adverse events than non-transfused neonates or those transfused with only CB-RBCs (p 0.014). A slightly higher incidence of surgery was observed in patients receiving both RBC types ( p  = 0.052), suggesting a particular complexity in the clinical course of these neonates. The severity of various adverse events was comparable between arms (Table  2 ). Nevertheless, when evaluated in the “per-protocol” setting, neonates receiving only A-RBCs or both types of RBCs tended to exhibit more severe forms of bradycardia ( p  = 0.002), pulmonary hypertension ( p  = 0.018), and hsPDA ( p  = 0.040) than non-transfused neonates or neonates receiving only CB-RBCs. In total, 25 adverse events were reported as fatal, with more events concurring to cause death: sepsis (6), bradycardia (6), acute renal failure (3), pulmonary hemorrhage (3), apnea (2), pneumonia (1), pneumothorax (1), NEC (1), coagulopathy (1), and tachyarrhythmia (1). There was no difference between arms A and B regarding the incidence of fatal adverse events.

Study population entering the interim analysis of safety of CB-RBC transfusions and of impact on the rate of severe ROP. ( A ) Flow-diagram of the study. ( B ) For the high rate of protocol deviations, a parallel analysis of safety and efficacy of CB-RBC transfusions was performed in the ”intention-to-treat” and “per protocol” settings

The association between adverse event and previous transfusion was evaluated in 285 out of the 336 transfusion episodes. No adverse events were certainly due to previous transfusions, whereas 8 cases were reported as possible and five as likely connected to transfusions (eFigure  1 ). The patient charts were individually reviewed to determine which types of RBC products were involved. The transfusions preceding the adverse events always consisted of A-RBCs, including an IVH in a patient in the arm B (CB-RBCs) who received also A-RBCs. Taken together, these results confirm that CB-RBC transfusions in ELGANs are at least as safe as standard RBC transfusions and do not increase the incidence or severity of clinical complications arising during the clinical course of these patients.

Transfusion needs, efficacy, and post-transfusion biochemical parameters

Forty-two of 58 patients (72.4%) required transfusions: 20 (71.4%) in the control arm and 22 (73.3%) in the experimental arm ( p  = 0.990); six of them also received erythropoietin ( p  = 0.334). A total of 153 RBC transfusions were administered, 72 in arm A and 81 in arm B, with similar transfusion needs in the two groups (3.0, IQR 1.0-5.7 and 3.0, IQR 1.0–7.0 RBC units in arm A and B, and respectively, p  = 0.867). Transfusion rate was similar in patients receiving or not erythropoietin (1.5, IQR 1.0–3.5 and 3.5, IQR 1.0–6.0, in patients receiving or not erythropoietin, respectively, p  = 0.163). Among 22 transfused patients in arm B, 8 received exclusively CB-RBCs, 10 both RBC types, and 4 exclusively A-RBCs (Fig.  1 B and eTable 2 ). PMA at first transfusion, indications to be transfused, and intervals between consecutive transfusions were also similar (eTable 2 ). In terms of storage duration, RBC content, transfused dose, and effects on hematological and biochemical parameters, CB-RBC units were significantly older than A-RBC units (3 days, IQR 2–5 and 8 days, IQR 6–12, in A-RBCs and CB-RBCs, respectively, p  < 0.001), had a lower Hct (61.6%, IQR 5.7–72.2 and 57.1%, IQR 53.0-62.3, in A-RBCs and CB-RBCs, respectively, p  = 0.009), and elicited a lower hematocrit increment (Δ Hct, 12.0, IQR 9.1–16.5, and 10.4 IQR 7.5–14.0, in A-RBCs and CB-RBCs, respectively, p  = 0.038). However, we found no correlation between ΔHct and storage duration (either with CB-RBCs or A-RBCs), so that the lower Hct increment after CB-RBCs should likely be ascribed to the lower RBC content of these units. Nonetheless, the interval between transfusions was similar after CB-RBCs and A-RBCs. Moreover, serum potassium and lactate levels, as well as pH values recorded within 24 h after CB-RBC or A-RBC transfusions, were comparable (eTable 3 ). In total, 230 CB units were processed into CB-RBC concentrates during the study period, but only 21.3% were transfused. eTable 4 provides characteristics of CB units at collection, after processing, and the hemolysis rate at the end of the storage.

RBC transfusions and ROP

The efficacy of CB-RBC transfusions in preventing severe ROP was preliminarily assessed in 44 evaluable patients (Fig.  1 B). Twenty-five patients (56.8%, 95%CI 42.2–70.3) had any stage of ROP, 10 (22.7%, 95% CI 12.8–36.9) had severe ROP, and 9 (20.4% 95%CI 11.1–34.5) required treatment. Therapy consisted of one or more administrations of anti-vascular endothelial growth factor (VEGF), combined with laser therapy (six cases) or vitrectomy (one case). All but one patient with severe ROP had received transfusions. The clinical course of the study population was typical for critically ill preterm neonates: 17 (38.6%) had IVH, 10 (22.7%) developed BPD, 4 (9.1%) surgical NEC, and 19 (43.2%) hsPDA. All but two patients experienced sepsis, 22 (50%) required inotropic agents, and all received oxygen therapy. Thirty-six patients (81.8%) needed invasive ventilation and 38 (86.3%) non-invasive ventilation. Erythropoietin was administered to 7 patients (15.9%) and 34 (77.3%) received RBC transfusions before discharge.

According to the arm allocation (intention-to-treat set), 14 (70.0%) out of 20 patients in the arm A and 19 (79.2%) out of 24 in the arm B received transfusions, with similar transfusion needs. Due to CB-RBC unavailability, ten neonates (52.6% of transfused patients) in arm B received both A-RBC and CB-RBC units and three patients (15.7%) exclusively received A-RBC units. The proportions of patients developing ROP (of any stage), severe ROP, or ROP requiring treatment in arm A and arm B were comparable (eTable 5 ). The two arms also did not differ for any other reported outcomes. Similar results were obtained excluding from the analysis non-transfused patients (data not shown).

Considering the high rate of protocol deviations in arm B, we further analyzed the data according to the per-protocol set. We categorized patients according to the types of RBC units received before the PMA of 32 weeks (treatment period). Thirteen neonates did not receive transfusions, 15 received only A-RBCs, 6 only CB-RBCs, and 10 both types of RBCs (Fig.  1 B). Severe ROP occurred in one out of 13 non-transfused patients (7.7%), in 5 of 15 receiving exclusively A-RBCs (30.0%), and in 4 of 10 transfused with both RBC types (40.0%). Severe ROP was not observed among infants receiving only CB-RBCs ( p  = 0.107).

The AUC of HbF at various weeks of PMA was compared for patients with and without severe ROP. The median AUC was 649 (IQR 462–778) for patients without severe ROP and 498 (IQR 259–597) for those with severe ROP ( p  = 0.035) (Fig.  2 A). The AUC profiles of patients who did or did not develop severe ROP, differed mainly in the earliest weeks of life. Therefore, we focused on to the type of RBC units received before the PMA of 30 weeks. In total, 16 patients were not transfused, 8 received only CB-RBCs, 15 received only A-RBCs, and 5 received both RBC types. Figure  2 B illustrates the AUC of HbF in these groups. Among infants transfused exclusively with CB-RBCs, none developed severe ROP, whereas 2 non-transfused patients (12.5%), 4 patients in the A-RBC group (26.7%), and 4 (80%) receiving both A-RBCs and CB-RBCs did develop this complication ( p  = 0.005). Considering only transfused patients, 8 of 20 (40.0%) receiving A-RBCs before the PMA of 30 weeks developed severe ROP, with a relative risk of 1.66 (95% CI 1.06–2.20) as compared to those receiving only CB-RBCs ( p  = 0.040). Table  3 shows that neonates receiving only CB-RBCs have GA, birth weight, Apgar, and CRIB scores, transfusion burden, rates of IVH, sepsis, NEC, and BPD as neonates receiving A-RBCs. Apart from a slightly shorter duration of oxygen therapy, they experienced similar periods of invasive or non-invasive ventilation, and more frequently required inotropic medications (Table  3 ). Collectively these data suggest that transfusing ELGANs with HbF-enriched RBCs in the first weeks of life may protect them from severe ROP.

To further corroborate these findings, we used logistic regression analysis to explore if A-RBC or CB-RBC transfusions exerted a diverging effect on severe ROP development. The model included severe ROP as the dependent variable and the number of A-RBC and CB-RBC transfusions before 30 week-PMA as independent variables. After adjusting for covariates with recognized effects on ROP (i.e., GA, BW, number of days on oxygen therapy), we found that the number of A-RBC, but not CB-RBC transfusions significantly predicted severe ROP, with an OR of 1.90 for each unit of A-RBCs transfused (95% CI 1.13–3.17, p  = 0.014).

Repeated standard transfusions of preterm neonates produce progressive non-physiological replacement of HbF by HbA [ 15 , 16 ]. Transfusing CB-RBC concentrates instead of RBC units obtained from adult donors limits HbF decrements [ 19 ]. The BORN trial is currently investigating if preserving HbF levels by transfusing CB-RBCs may limit the development of severe forms of ROP. This pre-specified interim analysis demonstrates that transfusing CB-RBCs in ELGANs is safe and can be used to treat anemia in these patients. Tissue establishments worldwide collect umbilical cord blood as a source of hematopoietic stem cells for transplanting patients with hematological diseases [ 29 ]. The use of cord blood for routine transfusion purposes has been largely reported in low-income countries, where health resources are limited and the blood supply does not meet the population’s needs [ 30 , 31 ]. More recently, umbilical cord blood has attracted attention as a source of HbF-containing RBCs, to prevent preterm neonates who need transfusions from untimely exposure to HbA [ 32 ]. Nevertheless, the safety of this transfusion approach was only assessed in initial pilot studies and remained to be explored in larger populations [ 19 ]. Apart from the predominant type of hemoglobin, adult and cord blood RBCs differ in other characteristics, ranging from shape and size, to membrane composition, cell metabolism, and rheological properties. The main concerns in using allogeneic CB-RBCs to transfuse preterm neonates are related to the higher rate of hemolysis during the storage in comparison to A-RBCs, placing neonates, particularly preterm neonates, at risk for hyperkaliemia, oxidative stress, and acidosis [ 33 , 34 ]. The CB-RBCs transfused in the BORN trial are collected according to criteria defined by the Italian public cord blood bank network. The CB-RBCs are filter leukodepleted and prepared using a standardized protocol designed to meet the same quality standards as A-RBC products. However, because CB-RBC units must be proven negative for fungal and bacterial contamination before distribution, the CB-RBCs transfused in this study had longer storage durations than A-RBC units. Nonetheless, the post-transfusion biochemical parameters were comparable using these two different products and the incidence and severity of adverse events were similar or even lower in patients exclusively receiving CB-RBCs. Similarly, no adverse events were definitively ascribed to transfusion and all adverse events with possible or likely connections to a previous blood product administration were seen only after A-RBC transfusions. Although this might be partly due to the higher number of A-RBC units transfused, the comprehensively collected findings allow us to conclude that CB-RBC units, as prepared and used in the BORN trial, are at least as safe as A-RBC units. Notably, the multicenter design of this study suggests that the fractionation method can be easily implemented to pursue a CB-RBC transfusion strategy. Nevertheless, the small volumes of the cord blood units that are typically collected might result in CB-RBC concentrates with lower amounts of hemoglobin as compared to A-RBCs, and this needs to be considered at transfusion. One disappointing issue emerging from this study is the difficulty of supporting the transfusion needs of patients in the experimental arm. In nearly 40% of cases, CB-RBC units were unavailable and were substituted with A-RBC units. This made comparing severe ROP rate in the two arms unfruitful. Likewise, only a minor portion of processed CB units led to CB-RBC transfusions; this suggests reorganizing CB unit procurement may be necessary to meet the ELGAN transfusion needs.

ROP is a multifactorial disease, whose pathogenesis relies on two consecutive phases affecting retinal vessel development in opposite ways [ 7 ]. The first phase occurs between 22 and 30 weeks of PMA, and is characterized by the attenuation of retinal vessels due to the hyperoxic post-natal environment. The second phase occurs between 31 and 36 weeks of PMA, and is characterized by rapid vessel growth and neovascularization stimulated by VEGF and other pro-angiogenic factors [ 7 ]. Modifying conditions that exacerbate hyperoxia in the first phase, or hypoxia in the second one, might reduce the risk for ROP development and progression [ 35 ]. Among modifiable risk factors, research has focused on the role of RBC transfusions and premature exposure to the adult form of hemoglobin, which has lower oxygen affinity than HbF. Using continuous near-infrared spectroscopy (NIRS) monitoring, we previously showed that CB-RBC and A-RBC transfusions are associated with different kinetics of cerebral regional oxygen saturation (crSO2) and cerebral fraction of tissue oxygen extraction (cFTOE) [ 36 ]. In particular, A-RBC transfusions resulted in higher crSO2 and lower cFTOE than CB-RBC transfusions, consistent with greater oxygen delivery to cerebral tissues [ 36 ]. These findings strongly support the rationale of transfusing HbF-enriched CB-RBCs, instead of RBC concentrates from adult donors to protect against ROP. The HbF trends in Fig.  2 A suggests that, when patients face prolonged exposure to low HbF levels before 30 weeks of PMA, the subsequent HbF increase fails to protect them against ROP progression. Hence, the greatest impact of hyperoxia on the immature retina likely occurs at lower PMA. In our patients, A-RBC, but not CB-RBC transfusions predicted severe ROP, and severe ROP did not occur in neonates receiving only CB-RBCs before 30 week-PMA, despite similar baseline characteristics and clinical course. Taken together, these findings support reserving the limited numbers of CB-RBC units for ELGANs from birth to the age of 30 weeks, during which time the hyperoxic load could inhibit retinal angiogenesis.

Area under the curve (AUC) of HbF and post-menstrual age (PMA). The lines represent median values of HbF. ( A ) HbF AUC of patients grouped according to the occurrence of severe ROP. ( B ) HbF AUC of patients grouped according to the types of RBC products received before 30 weeks of PMA

The data in this interim analysis comprise only a part of the planned study sample size of 146 patients; therefore, a protective effect of CB-RBC transfusions from ROP progression could not be conclusively assessed. The study has additional limitations. First, the evidence of the highest rate of severe ROP among patients receiving both A-RBCs and CB-RBCs, is in contrast with the presumed protective role of CB-RBCs. Nevertheless, the transfusion burden of these patients was higher than in other groups (on average 7 RBC units per patient, IQR 5–7), explaining why it could not be covered exclusively by CB-RBC units. Definitely, this finding emphasizes the connection between severe ROP and transfusions, and suggests that additional risk factors may have played a role in promoting ROP progression [ 35 ]. Second, the low number of patients evaluated prevented adjusting for any center-related variability, and possible confounders due to local practices could not be excluded. In addition, the evidence of protective effects of CB-RBC transfusions on severe ROP development emerged only at “per-protocol” set of analysis, which is more prone to bias than the “intention-to-treat” set, making the results less reliable. Finally, the impact of CB transfusions on other clinical outcomes such as BPD or NEC has not been extensively investigated at this early stage of the study.

Overall, this is the first randomized trial on allogeneic CB-RBC transfusions in preterm neonates and, albeit preliminary, these data may provide clues for the design of future randomized studies. In this regard, we decided to amend the protocol of the ongoing BORN trial, by reducing the treatment period in the experimental arm until a PMA of 30 weeks, prompted by the behavior of HbF levels in patients with severe ROP. Hopefully, this change will result in improved availability of HbF-enriched-RBCs for younger neonates, resulting in fewer protocol deviations.

BORN is the first randomized trial comparing CB-RBC and A-RBC transfusions to reduce severe ROP in ELGANs. The analysis of data collected in the first 58 patients proved the safety of this approach. Despite a high rate of protocol deviations, results at per-protocol analysis suggested the potential efficacy for this transfusion approach. The final analysis of the entire planned population, as well as further studies with larger sample size, are needed to confirm these promising results and support the use of new strategies to reduce ROP incidence.

Data availability

Data supporting the findings of this study are available in anonymized form upon reasonable request to the principal investigators.

Abbreviations

Retinopathy of prematurity

Fetal hemoglobin

Adult hemoglobin

Red blood cell

Cord blood-RBC

Adult donor-RBC

Postmenstrual age

Extremely low gestational age neonates

Gestational age

Clinical Risk Index for Babies-II score

Hemodynamically-significant patent ductus arteriosus

Necrotizing enterocolitis

Bronchopulmonary dysplasia

Intraventricular hemorrhage

Neonatal intensive care unit

Interquartile range

Confidence interval

Area under the curve

Stoll BJ, Hansen NI, Bell EF, et al. Trends in Care practices, Morbidity, and mortality of extremely Preterm neonates, 1993–2012. JAMA. 2015;314(10):1039–51.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Lee HC, Liu J, Profit J, Hintz SR, Gould JB. Survival without major morbidity among very low Birth Weight infants in California. Pediatrics. 2020;146(1):e20193865.

Article   PubMed   Google Scholar  

Painter SL, Wilkinson AR, Desai P, Goldacre MJ, Patel CK. Incidence and treatment of retinopathy of prematurity in England between 1990 and 2011: database study. Br J Ophthalmol. 2015;99(6):807–11.

Good WV. Retinopathy of Prematurity Incidence in Children. Ophthalmology. 2020;127(4S):S82–3.

Blencowe H, Lee AC, Cousens S, et al. Preterm birth-associated neurodevelopmental impairment estimates at regional and global levels for 2010. Pediatr Res. 2013;74(1):17–34.

Article   PubMed   PubMed Central   Google Scholar  

Solebo AL, Teoh L, Rahi J. Epidemiology of blindness in children [published correction appears in Arch Dis Child. 2017;102(10):995]. Arch Dis Child. 2017;102(9):853–8.

Hellström A, Smith LE, Dammann O. Retinopathy of prematurity. Lancet. 2013;382(9902):1445–57.

Stutchfield CJ, Jain A, Odd D, Williams C, Markham R. Foetal haemoglobin, blood transfusion, and retinopathy of prematurity in very preterm infants: a pilot prospective cohort study. Eye (Lond). 2017;31(10):1451–5.

Article   CAS   PubMed   Google Scholar  

Jiramongkolchai K, Repka MX, Tian J, et al. Lower foetal haemoglobin levels at 31- and 34-weeks post menstrual age is associated with the development of retinopathy of prematurity: PacIFiHER Report 1 PacIFiHER Study Group (Preterm Infants and fetal haemoglobin in ROP). Eye (Lond). 2021;35(2):659–64.

Hellström W, Martinsson T, Morsing E, Gränse L, Ley D, Hellström A. Low fraction of fetal haemoglobin is associated with retinopathy of prematurity in the very preterm infant. Br J Ophthalmol. 2022;106(7):970–4.

Hirano K, Morinobu T, Kim H, et al. Blood transfusion increases radical promoting non-transferrin bound iron in preterm infants. Arch Dis Child Fetal Neonatal Ed. 2001;84(3):F188–93.

Dani C, Martelli E, Bertini G, et al. Effect of blood transfusions on oxidative stress in preterm infants. Arch Dis Child Fetal Neonatal Ed. 2004;89(5):F408–11.

Keir AK, McPhee AJ, Andersen CC, Stark MJ. Plasma cytokines and markers of endothelial activation increase after packed red blood cell transfusion in the preterm infant. Pediatr Res. 2013;73(1):75–9.

Kalhan TG, Bateman DA, Bowker RM, et al. Effect of red blood cell storage time on markers of hemolysis and inflammation in transfused very low birth weight infants. Pediatr Res. 2017;82(6):964–9.

De Halleux V, Truttmann A, Gagnon C, Bard H. The effect of blood transfusion on the hemoglobin oxygen dissociation curve of very early preterm infants during the first week of life. Semin Perinatol. 2002;26(6):411–5.

Gavulic AE, Dougherty D, Li SH, et al. Fetal hemoglobin levels in premature newborns. Should we reconsider transfusion of adult donor blood? J Pediatr Surg. 2021;56(11):1944–8.

Podraza W. A new approach to neonatal medical management that could transform the prevention of retinopathy of prematurity: theoretical considerations. Med Hypotheses. 2020;137:109541.

Torrejón-Rodríguez L, Parra-Llorca A, Pinilla-González A, et al. Do lower levels of fetal hemoglobin in Preterm infants relate to oxidative stress? Antioxid Redox Signal. 2024;40(7–9):453–9.

Teofili L, Papacci P, Orlando N, et al. Allogeneic cord blood transfusions prevent fetal haemoglobin depletion in preterm neonates. Results of the CB-TrIP study. Br J Haematol. 2020;191(2):263–8.

Teofili L, Papacci P, Orlando N, et al. BORN study: a multicenter randomized trial investigating cord blood red blood cell transfusions to reduce the severity of retinopathy of prematurity in extremely low gestational age neonates. Trials. 2022;23(1):1010.

Piro E, Serra G, Schierz IAM, Giuffrè M, Corsello G. Neonatal ten-year retrospective study on neural tube defects in a second level University Hospital. Ital J Pediatr. 2020;46(1):72.

Girelli G, Antoncecchi S, Casadei AM, et al. Recommendations for transfusion therapy in neonatology. Blood Transfus. 2015;13(3):484–97.

PubMed   PubMed Central   Google Scholar  

International Committee for the Classification of Retinopathy of Prematurity. The International classification of retinopathy of Prematurity revisited. Arch Ophthalmol. 2005;123(7):991–9.

Article   Google Scholar  

Parry G, Tucker J, Tarnow-Mordi W, UK Neonatal Staffing Study Collaborative Group. CRIB II: an update of the clinical risk index for babies score. Lancet. 2003;361(9371):1789–91.

Bell MJ, Ternberg JL, Feigin RD, et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg. 1978;187(1):1–7.

Kliegman RM, Walsh MC. Neonatal necrotizing enterocolitis: pathogenesis, classification, and spectrum of illness. Curr Probl Pediatr. 1987;17(4):213–88.

CAS   PubMed   Google Scholar  

Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med. 2001;163(7):1723–9.

Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm. J Pediatr. 1978;92(4):529–34.

Mayani H, Wagner JE, Broxmeyer HE. Cord blood research, banking, and transplantation: achievements, challenges, and perspectives. Bone Marrow Transpl. 2020;55(1):48–6.

Bhattacharya N. Placental umbilical cord whole blood transfusion: a safe and genuine blood substitute for patients of the under-resourced world at emergency. J Am Coll Surg. 2005;200(4):557–63.

Hassall OW, Thitiri J, Fegan G, et al. Safety and efficacy of allogeneic umbilical cord red blood cell transfusion for children with severe anaemia in a Kenyan hospital: an open-label single-arm trial. Lancet Haematol. 2015;2(3):e101–7.

Rebulla P, Querol S, Pupella S, Prati D, Delgadillo J, De Angelis V. Recycling Apparent Waste Into biologicals: the case of umbilical cord blood in Italy and Spain. Front Cell Dev Biol. 2022;9:812038.

Khodabux CM, van Beckhoven JM, Scharenberg JG, El Barjiji F, Slot MC, Brand A. Processing cord blood from premature infants into autologous red-blood-cell products for transfusion. Vox Sang. 2011;100(4):367–73.

Bianchi M, Giannantonio C, Spartano S, et al. Allogeneic umbilical cord blood red cell concentrates: an innovative blood product for transfusion therapy of preterm infants. Neonatology. 2015;107(2):81–6.

Prasad M, Ingolfsland EC, Christiansen SP. Modifiable risk factors and preventative strategies for severe retinopathy of Prematurity. Life (Basel). 2023;13(5):1075.

Pellegrino C, Papacci P, Beccia F, et al. Differences in cerebral tissue oxygenation in Preterm neonates receiving adult or cord blood red blood cell transfusions. JAMA Netw Open. 2023;6(11):e2341643.

Download references

Acknowledgements

The authors are indebted to the personnel of the NICUs and public Cord Blood Banks, whose valuable work makes this study possible. The authors also emphasize the extraordinary generosity of the couples donating cord blood, and express deep empathy and gratitude to parents of all ‘born too soon’ babies enrolled in BORN. L.T. extends her deepest appreciation to Prof. Steven Spitalnik for his invaluable advice and the enriching exchange of ideas on this subject. The study is dedicated to the memory of Prof. Riccardo Saccardi.

The BORN trial is a no-profit study sponsored by Fondazione Policlinico A. Gemelli IRCCS and supported by Fresenius HemoCare Italia SRL (Prot. N 0038762/21 − 04/11/2021; grant number 5800134-FPG). Fondazione Policlinico A Gemelli IRCCS takes the responsibility of design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. .

Author information

Luciana Teofili and Patrizia Papacci equally contributed to the study.

Authors and Affiliations

Fondazione Policlinico A. Gemelli IRCCS, Largo Gemelli 8, 00168, Roma, Italy

Luciana Teofili, Patrizia Papacci, Claudio Pellegrino, Maria Bianchi, Domenico Lepore, Tina Pasciuto, Iolanda Mozzetta, Antonio Baldascino, Caterina Giovanna Valentini, Carmen Giannantonio, Brigida Carducci, Nicola Nicolotti, Silvia Baroni, Francesca Serrao, Velia Purcaro, Flavia Beccia & Giovanni Vento

Università Cattolica del Sacro Cuore, Roma, Italy

Luciana Teofili, Patrizia Papacci, Claudio Pellegrino, Domenico Lepore, Silvia Baroni, Flavia Beccia & Giovanni Vento

Azienda Ospedaliero Universitaria Careggi, Firenze, Italy

Carlo Dani, Giulia Remaschi, Fabrizio Gaetano Saverio Franco & Enrico Beccastrini

Città della Salute e della Scienza, Torino, Italy

Francesco Cresi, Giulia Ansaldi, Barbara Vania, Antonella Dragonetti, Emanuela Locatelli, Roberto Albiani, Federico Genzano Besso, Alessandra Coscia & Roberta Pizzolo

Department of Public Health and Pediatrics, Università di Torino, Torino, Italy

Francesco Cresi, Maria Francesca Campagnoli & Alessandra Coscia

Azienda Ospedaliero Universitaria Pisana, Pisa, Italy

Marco Fabbri, Sabrina Gabbriellini, Elena Ciabatti & Alessandro Mazzoni

Ospedale Santobono Pausilipon, Napoli, Italy

Roberta Penta de Vera d’ Aragona

Ospedale Evangelico Villa Betania, Napoli, Italy

Anna Molisso

Polo Nazionale Ipovisione IAPB Italia Onlus, Roma, Italy

Lorenzo Orazi

Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milano, Italy

Genny Raffaeli & Stefania Villa

Department of Clinical Sciences and Community Health, Università di Milano, Milano, Italy

Genny Raffaeli

Azienda Ospedaliera Bianchi Melacrino Morelli, Reggio Calabria, Italy

Isabella Mondello & Alfonso Trimarchi

Fondazione IRCCS Policlinico S. Matteo, Pavia, Italy

Stefano Ghirardello

You can also search for this author in PubMed   Google Scholar

Contributions

Conceptualization: LT, PP, SG, GV. Methodology: MB, TP, IM, FC, CG.

Software: TP, IM. Validation: MB, MF, RPdVdA, EB, AD, CGV, SaG, RA, EC, SB, AM, FGB, RP, SV, AT. Formal analysis: LT, CP, TP, IM, FB, NN. Investigation: PP, GR, CD, MB, GA, FC, BV, DL, FGSF, MF, RPdVdA, AM, EB, AD, LO, AB, EL, CGV, CG, SaG, RA, EC, SB, AM, FGB, FS, VP, AC, RP, GR, SV, IM, AT, SG, GV. Resources: PP, GR, CD, MB, GA, BV, DL, BC, FGSF, MF, RPdVdA, AM, EB, AD, LO, AB, EL, CGV, CG, SaG, RA, EC, SB, AM, FGB, FS, VP, AC, RP, GR, SV, IM, AT, SG, GV. Data Curation: TP, IM. Writing - Original Draft: LT, PP, CP. Writing - Review & Editing: LT, PP, CD, FC, CP, GR, FB, SG, GV. Reading and approval of the final manuscript: All authors. Visualization: LT, PP, CP. Supervision: LT, PP. Project administration: LT, PP. Funding acquisition: LT.

Corresponding author

Correspondence to Luciana Teofili .

Ethics declarations

Ethics approval and consent to participate.

All methods were performed in accordance with the ethical standards as laid down in the Declaration of Helsinki and its later amendments or comparable ethical standards The study was approved by the Ethics Committee of Fondazione Policlinico A. Gemelli IRCCS (ID 4364, Prot. N. 003590/21) and of all participating centers. Parents of enrolled patients provided a written informed consent to participate.

Consent for publication

Not applicable.

Conflict of interest

Carlo Dani is a member of the editorial board of Italian Journal of Pediatrics. The authors report no potential conflicts of interest, including relevant financial interests, activities, relationships, and affiliations.

Data access, responsibility, and analysis

LT, PP, TP and IM had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary material 2, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Teofili, L., Papacci, P., Dani, C. et al. Cord blood transfusions in extremely low gestational age neonates to reduce severe retinopathy of prematurity: results of a prespecified interim analysis of the randomized BORN trial. Ital J Pediatr 50 , 142 (2024). https://doi.org/10.1186/s13052-024-01714-w

Download citation

Received : 19 March 2024

Accepted : 25 July 2024

Published : 07 August 2024

DOI : https://doi.org/10.1186/s13052-024-01714-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Transfusions
• Umbilical blood
• Randomized controlled trial

Italian Journal of Pediatrics

ISSN: 1824-7288

• Submission enquiries: Access here and click Contact Us
• General enquiries: [email protected]

  Information Technologist (The) Journal / Information Technologist (The) / Vol. 21 No. 1 (2024) / Articles (function() { function async_load(){ var s = document.createElement('script'); s.type = 'text/javascript'; s.async = true; var theUrl = 'https://www.journalquality.info/journalquality/ratings/2408-www-ajol-info-ict'; s.src = theUrl + ( theUrl.indexOf("?") >= 0 ? "&" : "?") + 'ref=' + encodeURIComponent(window.location.href); var embedder = document.getElementById('jpps-embedder-ajol-ict'); embedder.parentNode.insertBefore(s, embedder); } if (window.attachEvent) window.attachEvent('onload', async_load); else window.addEventListener('load', async_load, false); })();

Article sidebar, article details, main article content, analysis of gender and use of social media for research: the case of social science lecturers in federal universities in south east nigeria, felicia n. ugwu, charles o. omekwu, chioma clarista onoh, harriet uche igbo.

The study investigated the correlation between the use of social media for research and scholarly research outputs of social sciences  lecturers in federal universities in South East Nigeria. The study was done with a view to finding the extent of use and possible differences  by gender on the use of social media and research outputs among the respondents, the effects of such differences and the  way forward. Three purposes of study, corresponding research questions and two hypotheses guided the study. Population of the study  consists of all the academic/teaching staff working in the faculty of the social sciences in all the federal universities in South East Nigeria.  Sample for the study include all the 459 academic staff in the faculty of the social sciences. Instrument for data collection was a structured questionnaire. Out of the 459 distributed, 354 useful copies ( 77% return rate). Research questions were analyzed using  Multiple Regression while the Hypotheses were tested using Ordinary Least Square (OLS) technique. Major findings include a significant  relationship between extent the use of social media and research outputs of lectures. Moreover, male lecturers use social media more than their female counterparts wchich increased their chances of having more research outputs. Many challenges on the use of social  media for research were reported based on which recommendations include provision of adequate power supply and internet  connectivity, regular training on the use of social media for research and increased research collaboration between male and female  lecturers. 

AJOL is a Non Profit Organisation that cannot function without donations. AJOL and the millions of African and international researchers who rely on our free services are deeply grateful for your contribution. AJOL is annually audited and was also independently assessed in 2019 by E&Y.

Your donation is guaranteed to directly contribute to Africans sharing their research output with a global readership.

• For annual AJOL Supporter contributions, please view our Supporters page.

Journal Identifiers

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

#ForYou? the impact of pro-ana TikTok content on body image dissatisfaction and internalisation of societal beauty standards

Roles Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing

Affiliation Faculty of Business, School of Psychology, Justice and Behavioural Science, Charles Sturt University, Wagga Wagga, New South Wales, Australia

Roles Conceptualization, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

• Madison R. Blackburn, 
• Rachel C. Hogg

• Published: August 7, 2024
• https://doi.org/10.1371/journal.pone.0307597
• Peer Review
• Reader Comments

Videos glamourising disordered eating practices and body image concerns readily circulate on TikTok. Minimal empirical research has investigated the impact of TikTok content on body image and eating behaviour. The present study aimed to fill this gap in current research by examining the influence of pro-anorexia TikTok content on young women’s body image and degree of internalisation of beauty standards, whilst also exploring the impact of daily time spent on TikTok and the development of disordered eating behaviours. An experimental and cross-sectional design was used to explore body image and internalisation of beauty standards in relation to pro-anorexia TikTok content. Time spent on TikTok was examined in relation to the risk of developing orthorexia nervosa. A sample of 273 female-identifying persons aged 18–28 years were exposed to either pro-anorexia or neutral TikTok content. Pre- and post-test measures of body image and internalisation of beauty standards were obtained. Participants were divided into four groups based on average daily time spent on TikTok. Women exposed to pro-anorexia content displayed the greatest decrease in body image satisfaction and an increase in internalisation of societal beauty standards. Women exposed to neutral content also reported a decrease in body image satisfaction. Participants categorised as high and extreme daily TikTok users reported greater average disordered eating behaviour on the EAT-26 than participants with low and moderate use, however this finding was not statistically significant in relation to orthorexic behaviours. This research has implications for the mental health of young female TikTok users, with exposure to pro-anorexia content having immediate consequences for internalisation and body image dissatisfaction, potentially increasing one’s risk of developing disordered eating beliefs and behaviours.

Citation: Blackburn MR, Hogg RC (2024) #ForYou? the impact of pro-ana TikTok content on body image dissatisfaction and internalisation of societal beauty standards. PLoS ONE 19(8): e0307597. https://doi.org/10.1371/journal.pone.0307597

Editor: Barbara Guidi, University of Pisa, ITALY

Received: November 2, 2023; Accepted: July 8, 2024; Published: August 7, 2024

Copyright: © 2024 Blackburn, Hogg. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The data for this study can be found on Figshare via the following link: https://doi.org/10.6084/m9.figshare.25756800.v1 .

Funding: We acknowledge the financial support provided by Charles Sturt University.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Social media is a self-presentation device, a mode of entertainment, and a means of connecting with others [ 1 ], allowing for performance and the performance of identity [ 2 ], with social rewards built into its systems. Five to six years of the average human lifespan are now spent on social media sites [ 3 ] and visual platforms such as Instagram and TikTok increasingly dominate the cultural landscape of social media. Such visually oriented platforms are associated with higher levels of dysfunction in body image [ 4 ], while the COVID-19 pandemic has seen a rise in disordered eating behaviour [ 5 ]. Despite this, the field lacks a clear theoretical framework for understanding how social media usage heightens body image issues [ 6 ] and little research has specifically examined the impacts of TikTok based content. In this research, we sought to explore the impact of pro-anorexia TikTok content on body image satisfaction and internalisation of beauty standards for young women. The forthcoming sections of this literature review will highlight the features of social media content that may be particularly pernicious for young female users and will explore disordered eating and orthorexia in a social media context, concluding with a theoretical analysis of the relationship between social media and body image and internalisation of beauty standards, respectively.

Social media offers instant, quantifiable feedback coupled with idealised online imagery that may intersect with the value adolescents attribute to peer relationships and the sociocultural gender socialisation processes germane to this period of development, creating the “perfect storm” for young social media users, especially females [ 6 ]. In a study of 85 young, largely female eating disorder patients, a rise in awareness of online sites emphasizing thinness as beauty was evident from 2017 to 2020, with 60% of participants indicating that they knew of pro-ana websites and 22% of participants admitting to visiting them [ 7 ]. Research suggests that social media may also trigger those with extant eating disorders while simultaneously influencing healthy individuals to engage in disordered eating behaviour [ 8 ].

“Pro” eating disorder communities, hereafter referred to as “pro-ana” (pro-anorexia) communities, are a particular concern in a social media context. These communities encourage disordered eating, normalise disordered behaviours, and provide a means of connection for individuals who endorse anti-recovery from eating disorders [ 8 ]. Weight-loss tips, excessive exercise routines, and images of emaciated figures are routinely shared in these online communities [ 9 ], with extant research highlighting the association between viewing eating disorder content online and offline eating disorder behaviour [ 8 ]. Women who view pro-ana websites display increased eating disturbances, lowered body satisfaction, an increased drive for thinness, and higher levels of perfectionism when compared to women who have not viewed pro-ana content [ 10 , 11 ]. In research on adolescent girls, Stice [ 12 ] investigated the influence of exposure to media portraying the “thin-ideal” and found that perceived pressure to be thin was a predictor of increased body image dissatisfaction, which in turn led to increases in disordered eating behaviour. In similar research, Green [ 10 ] found that individuals with diagnosed eating disorders reported worsening symptoms after just 10-minutes of exposure to pro-ana content on the online platform, Tumblr.

Disordered eating #ForYou

The most downloaded social application (app) of 2021, TikTok is a social media platform that allows short-form video creation and sharing within a social media context [ 13 ]. Since its launch in 2017, TikTok has had over two billion downloads and has an estimated one billion users, the vast majority of which are children and teenagers [ 14 ]. Unlike other social media platforms where users have greater autonomy over the content generated on their homepage newsfeed, TikTok’s algorithm records data from single users and proposes videos designed to catch a user’s attention specifically, by creating a personalised “For You” page [ 15 ]. This feed will suggest videos from any creator on the platform, not just followed accounts. As such, if a user ‘interacts’ with a video, such as liking, sharing, commenting, or searching for related content, the algorithm will continue to produce similar videos on their “For You” page. The speed with which TikTok content can be created and consumed online may also be key to its impact. Any given social media user could watch more than a thousand videos on TikTok in an hour, creating a reinforcing effect that may have more impact than longer form content from a single creator [ 2 ].

Whilst the popularity of TikTok’s “For You” page has prompted global leaders in social media to build their own recommended content features, this feature remains most pronounced on TikTok. The “For You” page is the homepage of TikTok where users spend the majority of their time, compared to other social media platforms where homepages consist of a curation of content from followed accounts. Instagram’s explore page continues to emphasise established influencer culture and promote accounts of public figures or influencers with large followings. Contrastingly, TikTok’s unique algorithm makes content discoverability an even playing field, as any user’s content has the potential to reach a vast audience regardless of follower count or celebrity status. TikTok users therefore have less control over their homepage newsfeed compared to other social media platforms where users elect who they follow.

Unlike other social media platforms that implicitly showcase body ideals, TikTok contains explicit eating disorder content [ 16 ], while the “For You” page means that simply interacting with health and fitness videos can lead to unintended exposure to disordered eating content. Even seemingly benign “fitspiration” content may have psychological consequences for viewers. Beyond explicit pro-ana content, #GymTok and #FoodTok are two popular areas of content that provide a forum for users to create and consume content around their and others’ daily eating routines, weight loss transformations, and workout routines [ 2 ]. TikTok also frequently features content promoting clean eating, detox cleanses, and limited ingredient diets reflective of the current “food as medicine” movement of western culture [ 17 ], otherwise known as orthorexia. Despite efforts to ban such pro-ana related content, some videos easily circumvent controls [ 18 ], in part because many TikTok creators are non-public figures who are not liable to the backlash or cancellation that a public figure might receive for circulating socially irresponsible content.

Orthorexia: The rise of ‘healthy’ eating pathologies

Psychological analyses of eating disorders have historically focused on restrictive eating and the binge-purge cycle, however, more recently “positive” interests in nutrition have been examined. Orthorexia nervosa is characterised by a restrictive diet, ritualized patterns of eating, and rigid avoidance of foods deemed unhealthy or impure that consumes an individual’s focus [ 19 ]. Despite frequent observation of this distinct behavioural pattern by clinicians, orthorexia has received limited empirical attention and is not formally recognised as a psychiatric disorder [ 19 ]. Orthorexia and anorexia nervosa share traits of perfectionism, high trait anxiety, a high need to exert control, plus the potential for significant weight loss [ 19 ]. Termed ‘the disorder that cannot be diagnosed’ due to limited consensus around its features and the line between healthy and pathological eating practices, orthorexia mirrors the narrative of neoliberal self-improvement culture, wherein the body is treated as a site of performance and transformation.

Orthorexic restrictions and obsessions are routinely interpreted as signs of morality, health consciousness, and wellness [ 20 , 21 ]. Social media wellness influencers have played a significant role in normalising “clean [disordered] eating”. As one example, Turner and Lefevre [ 22 ] conducted an online survey of social media users following health food accounts and found that higher Instagram use was associated with a greater tendency towards orthorexia, with the prevalence of orthorexia among the study population at 49%, substantially higher than the general population (<1%). Similar health and food-related content on TikTok may provoke orthorexic tendencies among TikTok users, however, limited research has investigated orthorexic eating behaviour in the context of TikTok. The current study aims to bridge this gap in the literature around TikTok use and orthorexic tendencies. Disordered eating behaviour in the present study was measured by two separate but related constructs. ‘Restrictive’ disordered eating relates to dieting, oral control, and bulimic symptoms, whilst ‘healthy’ disordered eating constitutes orthorexic-like preoccupation with health food.

Theoretical analysis of body image and social media

An established risk factor in the development and maintenance of disordered eating behaviour is negative body image. Body image is a multidimensional construct that represents an individual’s perceptions and attitudes about their physical-self and encompasses an evaluative function through which individuals compare perceptions of their actual “self” to “ideal” images [ 23 ]. This comparison may produce feelings of dissatisfaction about one’s own body image if a significant discrepancy exists between the actual and ideal self-image [ 23 ]. Body image is not necessarily congruent with actual physique, with research demonstrating that women categorised as having a healthy body mass index (BMI) nonetheless report dissatisfaction with their weight and engage in restrictive dietary behaviours to reduce their weight [ 24 ]. In addition, body image dissatisfaction is considered normative in Western society, particularly among adolescent women [ 25 ]. This may be attributable to the constant flow of media that exposes women to unrealistic images of thinness idealized within society [ 26 ].

One theoretical framework for understanding social media’s relationship with body image is the Social Comparison Theory, proposed by Festinger [ 27 ] who suggests that people naturally evaluate themselves in comparison to others via upward or downward social comparisons. Research supports the notion that women who frequently engage in maladaptive upward appearance-related social comparisons are more likely to experience body image dissatisfaction and disordered eating [ 25 , 28 ], while visual exposure to thin bodies may detrimentally modulate one’s level of body image satisfaction [ 29 – 31 ]. In their study of undergraduate females, Engeln-Maddox [ 29 ] found that participants made upward social comparisons to images of thin models which were strongly associated with decreases in body image satisfaction and internalisation of thinness. Similarly, Tiggemann [ 32 ] found that adolescents who spent more time watching television featuring attractive actors and actresses reported an increased desire for thinness, theorised to be a result of increased social comparison to attractive media personalities.

The Transactional Model [ 33 ] extends Social Comparison Theory by emphasising the multifaceted and complex nature of social media influences on body image. This model acknowledges that individual differences may predispose a person to utilise social media for gratification, and highlights that as time spent on social media increases, so too does body image dissatisfaction [ 33 ]. In line with this, a recent review of literature by Frieiro Padín and colleagues [ 34 ] indicated that time spent on social media was strongly correlated with eating disorder psychopathologies, as well as heightened body image concerns, internalisation of the thin ideal, and lower levels of self-esteem. Time on social media also correlated with heightened body image concerns to a far greater extent than general internet usage [ 35 , 36 ].

Body image ideals are not static. The traditional ideal of rib-protruding bodies from the 90s, known colloquially as “heroin chic”, have recently shifted to a celebration of the “slim-thicc” figure, consisting of a cinched, flat waist with curvy hips, ample breasts, and large behinds [ 37 ]. The “slim-thicc” aesthetic allows women to be bigger than previous body ideals, yet this figure is arguably more unattainable than the thin-ideal as surgical intervention is commonly needed to achieve it, depending on genetics and body type. The idealisation of the “slim-thicc” figure is highlighted by the “Brazilian butt lift” (BBL), a potentially life-threatening procedure that is nonetheless the fastest growing category of plastic surgery, doubling in growth over the past five years, despite the life-threatening potential of the procedure [ 38 ]. Research suggests that the slim-thicc ideal is no less damaging nor threatening of body image than the thin-ideal. Indeed, in experimental research on body ideals, McComb and Mills [ 39 ] found that the greatest body dissatisfaction levels in female undergraduate students were observed among those exposed to imagery of the slim-thicc physique, relative to that exhibited by those exposed to the thin-ideal and fit-ideal physique, as well as the control condition.

Recent body ideals have also favoured muscular thin presentations, considered to represent health and fitness as evident in the “#fitspiration” Instagram hashtag that features over 65 million images [ 40 ]. Fitspiration has the potential to positively influence women’s health and wellbeing by promoting exercise engagement and healthy eating, yet various content analyses of fitspiration images highlight aspects of fitspiration that warrant concern [see 40 , 41 ]. Notably, fitspiration typically showcases only one body type and women whose bodies do not meet this standard may experience body dissatisfaction [ 40 ], while the gamification of exercise, such as receiving likes for every ten sit-ups, segues with the intensive self-control and competitiveness that often underpins eating disorders and eating disorder communities [ 1 ].

In recent experimental research, Pryde and Prichard [ 42 ] examined the effect of exposure to fitspiration TikTok content on the body dissatisfaction, appearance comparison, and mood of young Australian women. Viewing fitspiration TikTok videos led to increased negative mood and increased appearance comparison but did not impact body dissatisfaction. This finding contradicts previous research and may be due to fitspiration content showcasing body functionality rather than aesthetic, which may lead to positive outcomes for viewers. The fitspiration content used by Pryde and Prichard [ 42 ] did not contain the harmful themes regularly found in other forms of fitspiration content. Appearance comparison was significant in the relationship between TikTok content and body dissatisfaction and mood, suggesting that this may be a key mechanism through which fitspiration content leads to negative body image outcomes and supporting the notion that fitspiration promotes a focus on appearance rather than health.

Body image dissatisfaction among women is associated with co-morbid psychological disturbances and the development of disordered eating behaviours [ 43 , 44 ]. A large body of research indicates that higher levels of both general and appearance-related social comparison are associated with disordered eating in undergraduate populations [ 10 , 28 , 45 – 48 ]. As one example, Lindner et al. [ 46 ] investigated the impact of the female-to-male ratio of college campuses on female students’ engagement in social comparison and eating pathology. Their findings lend support to the Social Comparison Theory, indicating that the highest levels of eating pathology and social comparison were found among women attending colleges with predominantly female undergraduate populations. A strong relationship was also found between eating pathology and engagement in appearance-related social comparisons independent of actual weight. Lindner et al. [ 46 ] surmised that these results suggest social comparison and eating pathology behaviours are due to students’ perceptual distortions of their own bodies, potentially fostered by pressures exerted from peers to be thin.

Similarly, Corning et al. [ 45 ] investigated the social comparison behaviours of women with eating disorder symptoms and their asymptomatic peers. Results illustrated that a greater tendency to engage in everyday social comparison predicted the presence of eating disorder symptoms, while women with eating disorder symptoms made significantly more social comparisons of their own bodies. Such findings are supported by subsequent research, with Hamel et al. [ 28 ] finding that adolescents with a diagnosed eating disorder engaged in significantly more body-related social comparison than adolescents diagnosed with a depressive disorder or no diagnosis. Body-related social comparison was also significantly positively correlated with disordered eating behaviours. While extant research has focused upon social comparison as it has occurred through traditional media outlets, less research has investigated the facilitation of social comparison through social media platforms, particularly contemporary platforms such as TikTok.

Theoretical analysis of internalisation processes and social media

The extent to which one’s body image is impacted by images and messages conveyed by the media is determined by the degree to which these images and messages are internalised. Some may argue that social media platforms are distinct from what occurs in “real” life, creating fewer opportunities for internalisation to occur. Yet as Pierce [ 2 ] argues, platforms such as TikTok create their own realities, allowing users to explore their identities, form relationships, engage with culture and world events, and even develop new patterns of speech and writing. TikTok trends commonly infiltrate society, underscoring the impact of social media on life beyond the online world and thus a sociocultural analysis of TikTok is warranted. Sociocultural theories suggest that society portrays thinness as the ideal body shape for women, resulting in an internationalisation of the “thin is good” assumption for women. This in turn results in lowered body image satisfaction and other negative outcomes [ 43 ]. The significance of social influences, including the role of family, peers, and the media, is emphasised by sociocultural theory, with individuals more likely to internalise the thin ideal when they encounter pressuring messages that they are not thin enough from social influences [ 48 ]. Within this theoretical approach, an individual’s degree of thin ideal internalisation is theorised to depend on their acceptance of socially defined ideals of attractiveness and is reflected in their engagement in behaviours that adhere to these socially defined ideals [ 49 ].

Building on this, the tripartite influence model suggests that disordered eating behaviours arise due to pressure from social agents, specifically media, family, and peers. This pressure centres on conforming to appearance ideals and may lead to engagement in social comparison and the internalisation of thin ideals [ 48 ]. This is relevant in a digital context given social media provides endless opportunities for individuals to practice social comparison and for many users, social comparison on TikTok is peer-based as well as media-based. According to the tripartite model, social comparisons have been consistently associated with a higher degree of thin ideal internalisation, self-objectification, drive for thinness, and weight dissatisfaction [ 50 ]. Furthermore, and in contrast to traditional media where social agents are mainly models, celebrities, and movie stars, social agents on social media can include peers, friends, family, and individuals who have a relationship with the individual. Social media content generated by “everyday” people, rather than super models or movie stars, may result in comparisons that are more horizontal in nature. This is particularly evident on TikTok where content creators are rarely famous before creating a TikTok account, often remain micro-influencers after achieving some notoriety, and are usually around the same age as those viewing their content.

Pressure to be thin from alike peers may have a particularly pronounced impact on one’s degree of internalisation of the thinness ideal. Indeed, Stice et al. [ 51 ] found that after listening to young thin women complain about “feeling fat”, their adolescent participant sample reported increased body image dissatisfaction, suggesting that pressure from peers perpetuates the thinness ideal, leading to internalisation of the ideal and subsequent body dissatisfaction. Similarly, it was found that adolescent females were more likely to engage in weight loss behaviour if a high portion of peers with a similar BMI were also engaging in these behaviours, illustrating that appearance pressure exerted by alike peers may result in thin-ideal internalisation and engagement in weight loss behaviours to control body weight and shape [ 52 ]. Such findings raise questions around whether those most similar to us have the greatest impact upon thin-ideal internalisation, body image dissatisfaction, and disordered eating behaviours.

In further support for the tripartite influence model, research by Thompson et al. [ 48 ] indicates that the ideals promoted through social media trends are internalized despite being unattainable, resulting in body image dissatisfaction and disordered eating behaviour. Similarly, Mingoia et al. [ 53 ] found a positive association between the use of social networking sites and thin ideal internalisation in women, indicating that greater use of social networking sites was linked to significantly higher internalisation of the thin ideal. Interestingly, the use of appearance-related features (e.g., posting or viewing photographs or videos) was more strongly related with internalisation than the broad use of social networking sites (e.g., writing status’, messaging features) [ 53 ]. Correlational and experimental research alike has demonstrated that thin ideal internalisation is related to body image dissatisfaction and leads to expressions of disordered eating such as restrictive dieting and binge-purge symptoms [ 31 , 48 , 54 , 55 ]. Subsequent expressions of disordered eating may be seen as an attempt to control weight and body shape to conform to societal beauty standards of thinness [ 51 ].

This sociocultural perspective is exemplified by Grabe et al’s. [ 54 ] meta-analysis of research on the associations between media exposure to women’s body dissatisfaction, internalisation of the thin ideal, and eating behaviours and beliefs, illustrating that exposure to media images propagating the thin ideal is related to and indeed, may lead to body image concerns and increased endorsement of disordered eating behaviours in women. Similarly, Groesz et al. [ 55 ] conducted a meta-analysis to examine experimental manipulations of the thin beauty ideal. They found that body image was significantly more negative after viewing thin media images than after viewing images of average size models, plus size models, or inanimate objects. This effect size was stronger for participants who were more vulnerable to activation of the thinness schema. Groesz et al. [ 55 ] conclude that their results align with the sociocultural theory perspective that media promulgates a thin ideal that in turn provokes body dissatisfaction.

Current research

Existing research has established the relationship between body image dissatisfaction and disordered eating behaviours and social media platforms such as Instagram and Twitter. The unique implications of the TikTok ‘For You Page’, as well as the dominance of peer-created and explicit disordered eating content on TikTok suggests that the influence of this platform warrants specific consideration. This study adds to extant literature by utilising an experimental design to examine the influence of exposure to pro-ana TikTok content on body image and internalisation of societal beauty standards. A cross-sectional design was used to investigate the effect of daily TikTok and the development of disordered eating behaviours. Although body image disturbance and eating disorders are not limited to women, varying sociocultural factors have been implicated in the development of disordered eating behaviour in men and women [ 45 ], while issues facing trans people warrant specific consideration beyond the scope of this study, therefore the present sample contains only female-identifying participants.

Aims and hypotheses.

The current study aimed to investigate the impact of pro-ana TikTok content on young women’s body image satisfaction and internalisation of beauty standards, as well as exploring daily TikTok use and the development of disordered eating behaviour. First, in line with the cross-sectional component of the study, it was hypothesized that women who spend greater time on TikTok per day would report significantly more disordered eating behaviour than women who spend low amounts of time on TikTok per day. Second, it was hypothesized that women in the pro-ana TikTok group would report a significant decrease in body image satisfaction state following exposure to the pro-ana content compared to women in the control group. Third, it was hypothesized that women in the pro-ana Tik Tok group would report increased internalisation of societal beauty standards following exposure to pro-ana TikTok content compared to women in the control group.

Participants

Participants in the current study included 273 women aged between 18 to 28 years sourced from the general population of TikTok users. The predominant country of residence of the sample was Australia, with 15 participants indicating they currently reside outside of Australia. Of the remaining data relating to the two conditions of the study, 126 participants were randomly allocated into the experimental condition, and 147 participants were randomly allocated into the control condition. Snowball sampling was used to recruit participants through social media, online survey sharing platforms, and word-of-mouth, with first-year University students targeted for recruitment by offering class credit in return for participation. Participants could withdraw their consent at any time by exiting the study prior to completion of the survey.

The current study employed a questionnaire set that included a demographic questionnaire, and five scales measuring disordered eating behaviour, body satisfaction, and internalisation of societal beauty standards, as well as perfectionism, the latter of which was not examined in the present study.

Demographic questionnaire.

The demographic questionnaire required participants to answer a series of questions relating to their gender, age, relationship status, ethnicity, country of residence, TikTok usage, and exercise routine. A screening question redirected non-female-identifying persons from the study. Responses to the TikTok usage items were examined cross-sectionally with responses on the EAT-26 and ORTO15 used to examine the influence of daily TikTok use and the presentation of disordered eating behaviours.

Eating attitudes test.

The Eating Attitudes Test (EAT-26, [ 56 ]) is a short form of the original 40-item EAT scale [ 57 ] which measures symptoms and concerns characteristic of eating disorders. The 26-item short-form version of the EAT was utilised in the present study due to its established reliability and validity, and strong correlation with the EAT-40 [ 56 ].

Responses to the 26-items are self-reported using a 6-point Likert scale ranging from Always (3) to Never (0) [ 56 ]. The EAT-26 consists of three subscales including dieting, bulimia and food preoccupation, and oral control. Five behavioural questions are included in Part C of the EAT-26 to determine the presence and frequency of extreme weight-control behaviours including binge eating, self-induced vomiting, laxative usage, and excessive exercise [ 56 ]. Higher scores indicate greater disordered eating behaviour, and those with a total score of 20 or greater are, in clinical contexts, typically highlighted as requiring further assessment and advice of a mental health professional [ 56 ].

Internal consistency of the EAT-26 was established in initial psychometric studies which reported a Cronbach’s alpha of.85 [ 58 ]. For the current study, the Cronbach’s alpha = .91. Previous research has also demonstrated that the EAT-26 has strong test-retest reliability (e.g., 0.84) [ 59 ], as well as acceptable criterion-related validity for differentiating between eating disorder populations and non-disordered populations [ 56 ]. In the current study, the EAT-26 was used to measure disordered eating behaviour, and the cut-off score of 20 and above was adopted to categorise increased disordered eating behaviour. Given how this construct is measured, from this point forward the present study will refer to EAT-26 responses as ‘restrictive’ type disordered eating.

The ORTO-15 is a 15-item screening measure that assesses orthorexia nervosa risk through questions regarding the perceived effects of eating healthy food (e.g. “Do you think that consuming healthy food may improve your appearance?”), eating habits (e.g. “At present, are you alone when having meals?”), and the extent to which concerns about food influence daily life (e.g. “Does the thought of food worry you for more than three hours a day?”) [ 19 ]. Responses are self-reported using a 4-point Likert scale ranging from always , often , sometimes , or never . Individual items are coded and summed to derive a total score. Donini et al. [ 60 ] established a cut off total score of 40; scores below 40 indicate orthorexia behaviours, whilst scores 40 or above reflect normal eating behaviour. This cut off score was determined by Donini et al. [ 60 ] as their results revealed the ORTO-15 demonstrated good predictive capability at the threshold of 40 compared to other potential threshold values.

Although the ORTO-15 is the most widely accepted screening tool to assess orthorexia risk, it is still only partially validated [ 61 ], and inconsistencies of the measures’ reliability and validity exist in current literature. For example, Roncero et al. [ 62 ] estimated that the reliability of the ORTO-15 using Cronbach’s alpha was between 0.20 and 0.23, however, after removing certain items, the reliability coefficients were between 0.74 and 0.83. Contrastingly, Costa and colleagues’ [ 63 ] review of current literature surrounding orthorexia suggested adequate internal consistency (Cronbach’s alpha = 0.83 to 0.91) with all 15-items.

In the present study, a reliability analysis revealed unacceptable reliability for the ORTO-15 (α = .24). Principal components factor analysis identified two factors within the ORTO-15, one relating to dieting and the other to preoccupation with health food. Separate reliability analyses were performed on the items that comprised these two factors and the diet-related items did not have acceptable reliability (α = -.40), whilst the health food-related items bordered on acceptable reliability at α = .63. Consequently, only the health food-related items were retained in the current study following consideration of Pallant’s [ 64 ] assertion that Cronbach alpha values are sensitive to the number of items on a scale and it is therefore common to obtain low values on scales with less than ten items. Pallant [ 64 ] notes that in cases such as this, it is appropriate to report the inter-item correlation of the items, while Briggs and Cheek [ 65 ] advise an optimal range for the inter-item correlation between.2 to.4, with the health food-related items in the current study obtaining an inter-item correlation of.25. Throughout this study, the construct measured by these ORTO-15 items will be referred to as ‘healthy’ type disordered eating to reflect this obsessive health food preoccupation and differentiate between the two disordered eating dependent variables measured in the current study.

Body image states scale.

The Body Image States Scale (BISS) by Cash and colleagues [ 66 ] is a six-item measure of momentary evaluative and affective experiences of one’s own physical appearance. The BISS evaluates the following aspects of current body experience: dissatisfaction-satisfaction with overall physical appearance; dissatisfaction-satisfaction with one’s body size and shape; dissatisfaction-satisfaction with one’s weight; feelings of physical attractiveness-unattractiveness; current feelings about one’s looks relative to how one usually feels; and evaluation of one’s appearance relative to how the average person looks [ 66 ]. Participants responded to these items using a 9-point Likert-type scale which is presented in a negative-to-positive direction for half of the items, and a positive-to-negative direction for the other half [ 66 ]. Respondents were instructed to select the statement that best captured how they felt “ right now at this very moment ”. A total BISS score was calculated by reverse-scoring the three positive-to-negative items, summing the six-items, and finding the mean, with higher total BISS scores indicating more favourable body image states.

During the development and implementation of the BISS, Cash and colleagues [ 66 ] report acceptable internal consistency and moderate stability over time, an anticipated outcome due to the nature of the BISS as a state assessment tool. The BISS was also appropriately correlated with a range of trait measures of body image, highlighting its convergent validity [ 66 ]. Cash and colleagues [ 66 ] also report that the BISS is sensitive to reactions in positive and negative situational contexts and has good construct validity. An acceptable Cronbach’s alpha coefficient of.88 was obtained in the current study.

Sociocultural Attitudes Towards Appearance Questionnaire—4.

The Sociocultural Attitudes Towards Appearance Questionnaire– 4 (SATAQ-4) [ 67 ] is a 22-item self-report questionnaire that assesses the influence of interpersonal and sociocultural appearance ideals on one’s body image, eating disturbance, and self-esteem. Ratings are captured on a 5-point Likert scale which asks participants to specify their level of agreement with each statement by choosing from 1 ( definitely disagree ) through to 5 ( definitely agree ), with higher scores indicative of greater pressure to conform to, or greater internalisation of, interpersonal and sociocultural appearance ideals [ 67 ]. The five subscales of the SATAQ-4 measure: internalisation of thin/low body fat ideals, internalisation of muscular/athletic ideals, influence of pressures from family, influence of pressure from peers, and influence of pressures from the media [ 67 ]. For the purposes of the present study, the questions from the media pressure subscale were modified to enquire specifically about social media rather than traditional forms of media.

Across all samples in Schaefer et al’s. [ 67 ] study, the internal consistency of the five SATAQ-4 subscales is considered acceptable to excellent, with Cronbach’s alpha scores between 0.75 and 0.95. These subscales also displayed good convergent validity with other measures of body satisfaction, eating disorder risk, and self-esteem [ 67 ]. Pearson product-moment correlations between the SATAQ-4 subscales and convergent measures revealed medium to large positive associations with eating disorder symptomology, medium negative associations with body satisfaction, and small negative associations with self-esteem [ 67 ]. A Cronbach’s alpha of.87 was obtained in the present study, demonstrating acceptable internal consistency.

Ethical approval for the present study was granted by the Charles Sturt University Human Research Ethics Committee (Approval number H21155) prior to data collection. Participants were directed to the study via an online link to QuestionPro where they were provided an explanation of the study, their rights, and contact details of relevant support services if they were to become distressed. Participants gave informed consent by clicking on a link that read, “I consent to participate” at the beginning of the survey and then again through the submission of their completed survey. Any incomplete responses were not included in the dataset. Data collection commenced on the 30 th of July 2021 and ceased on the 1st of October 2021. In line with the cross-sectional and descriptive aspects of the research, participants were asked demographic questions about their gender, age, relationship status, ethnicity, country of residence, TikTok usage, and exercise habits. Participants then completed the experimental set in the following order: BISS (pre-test), SATAQ-4 (pre-test), EAT-26, ORTO-15, Experimental intervention (control or experimental TikTok video condition), SATAQ-4 (post-test), BISS (post-test), and debrief. All questionnaires presented to each participant were identical. Measures were not randomised to ensure that body image and internalisation were assessed at both pre- and post-test to evaluate the experimental manipulation.

Participants were randomly allocated to one of two conditions: experimental (pro-ana TikTok video) or control (“normal” TikTok video). Participants allocated to the experimental condition watched a compilation of TikTok videos containing explicit disordered eating messages such as young women restricting their food, displaying gallows humour about their disordered eating behaviour, starving themselves, and providing weight loss tips such as eating ice cubes and chewing gum to curve hunger. Participants in the experimental condition were also exposed to more implicit body image ideals typical of fitspiration-style content. This included thin women displaying their abdomens, cinched waists, dancing in two-piece swimwear, along with workout and juice cleanse videos promising fast weight loss. Participants in the control condition viewed a compilation of TikTok videos containing scenes relating to nature, cooking and recipes, animals, and comedy. After viewing the 7- to 8-minute TikTok video, all participants completed measures of internalisation and body satisfaction again to assess the influence of either the pro-ana TikTok video or the normal TikTok video. The debrief statement made explicit to participants the rationale of the study and explained the non-normative content of the videos shown to the experimental group. A small financial incentive was offered via a prize draw of five vouchers.

Statistical analysis

The data from QuestionPro was collated and analysed using IBM SPSS Statistics software, Version 28. All measures and manipulations in the study have been disclosed, alongside the method of determining the final sample size. No data collection was conducted following analysis of the data. Data for this study is available via the Figshare data repository and can be accessed at https://doi.org/10.6084/m9.figshare.25756800.v1 . This study was not preregistered. Sample size was determined before any data analysis. A priori power analyses were conducted using G*Power to determine the minimum sample sizes required to test the study hypotheses. Results indicated the required sample sizes to achieve 90% power for detecting medium effects, with a significance criterion of α = 0.05, were: N = 108 for the mixed between-within subjects ANOVAs and N = 232 for the one-way between groups ANOVAs. According to these recommendations, adequate statistical power was achieved. All univariate and multivariate assumptions were checked and found to be met. All scales and independent variables were normally distributed.

The analysis of the current study including data screening processes, descriptive statistics, and hypothesis testing will be presented in this section. Hypothesis testing began with two separate mixed between-within subjects analysis of variance models (ANOVAs) to examine the impact of the experimental manipulation on the independent variables of body image and internalisation of appearance ideals and pressure. Finally, the effect of time spent using TikTok daily on restrictive and ‘healthy’ disordered eating behaviour was explored cross-sectionally using two separate one-way between-subjects ANOVAs.

Data screening

Prior to statistical analysis, data were screened for entry errors and missing data. Of the 838 participants who initially consented to participate in the survey, 555 responses were insufficiently complete for data analysis. As participants were permitted to withdraw their consent by exiting the online survey, these results were excluded from all subsequent analyses. Of those that did not complete the study, the majority withdrew during the BISS (pre-test) and the ORTO-15, suggesting that these participants potentially experienced discomfort or distress when asked to reflect on their appearance and their eating behaviours. Of the completed responses, nine were excluded due to not meeting the study’s stated age eligibility and another case was excluded due to disclosure of a previous eating disorder diagnosis. The remaining data set comprised of 273 participants.

Descriptive statistics

Demographic characteristics..

In the current sample, 50% of participants reported being currently single and most participants (83%) were Caucasian, with 71% of participants indicating that they spent up to two hours per day using TikTok. Further demographic information is provided in Table 1 .

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pone.0307597.t001

#ForYou: TikTok consumption demographics.

Participants in the current study reported that entertainment (75%), fashion (59%), beauty/skincare (54%), cooking/recipes (51%) and life hacks/advice (51%) content frequently occurred on their For You page. Largely in keeping with this, participants reported experiencing the most enjoyment from viewing entertainment (84%), life hacks/advice (57%), home renovation (56%), recipes/cooking (56%), and fashion (54%) content on their For You page.

In the current sample, 64% of participants reported being exposed to disordered eating content via their For You page. Only 15% of participants had not been exposed to any negative content themes. Further descriptive For You page content information is displayed below in Table 2 .

https://doi.org/10.1371/journal.pone.0307597.t002

Notably, 43% of the participant sample were frequently exposed to fitness and sports related content and the same percentage of the sample enjoyed seeing this content, suggesting that content broadly aligned with #fitspiration was consumed and appreciated by nearly half of participants. Concerningly, 40–60% of participants had been exposed to negative TikTok content via the For You Page, with content ranging from self-harm and suicidality to violence and illegal activity. No data was collected on the specifics of this content, however, and it is possible that some “negative” content may be framed from a proactive, preventative perspective, and this warrants further consideration.

Hypothesis testing: Cross-sectional analysis

Hypothesis 1: daily tiktok use and disordered eating behaviour..

To test the cross-sectional analysis of this study, two separate one-way between-groups ANOVAs were conducted to explore the impact of daily amount of TikTok use on ‘healthy’ disordered eating and restrictive disordered eating behaviour. This was necessary as time on TikTok was measured categorically. Participants were divided into four groups according to their average daily time spent using TikTok (Low use group: 1 hour or less; Moderate use group: 1–2 hours; High use group: 2–3 hours; Extreme use group: 3+ hours). Homogeneity of variance could be assumed for each ANOVA as indicated by non-significant Levene’s Test Statistics.

There was no statistically significant difference at the p < .05 level in ORTO15 scores for the four TikTok usage groups: F (3, 269) = .38, p = .78, indicating that ‘healthy’ disordered eating did not significantly differ across women who use TikTok for different periods of time per day. The effect size, calculated using eta squared, was.004, which is considered small in Cohen’s [ 68 ] terms. This small effect size is congruent with the non-significant finding.

The second ANOVA measuring differences among EAT-26 scores across the four TikTok usage groups also yielded a non-significant result: F (3, 269) = 1.21, p = .31. Eta squared was calculated as.01, representing a small effect size [ 68 ] consistent with this non-significant result. The means and standard deviations of the four TikTok usage groups across dependent variables of ‘healthy’ and restrictive disordered eating, as measured by the ORTO15 and the EAT-26 respectively, are displayed in Table 3 .

https://doi.org/10.1371/journal.pone.0307597.t003

Hypothesis testing: Experimental analyses

Hypothesis 2: body image satisfaction across groups from pre-test to post-test..

To evaluate the effect of the experimental intervention on body image, a 2 x 2 mixed between-within subjects ANOVA was conducted with condition (experimental vs control) as the between subjects factor and time (pre-manipulation vs post-manipulation) as the within subjects factor. All assumptions were upheld, including homogeneity of variance-covariance as indicated by Box’s M ( p >.001) and Levene’s ( p >.05) tests [ 64 ].

The interaction between condition and time was significant, Wilks’ Lambda = .98, F (1, 271) = 6.83, p = .009, partial eta squared = .03, demonstrating that the change in body image scores from pre-manipulation to post-manipulation was significantly different for the two groups. The body image satisfaction scores for women in both conditions decreased from pre-manipulation to post-manipulation. As anticipated, participants in the experimental condition reported a greater decrease in body image satisfaction than women in the control condition (see Table 4 ). This interaction effect is displayed in Fig 1 .

https://doi.org/10.1371/journal.pone.0307597.g001

https://doi.org/10.1371/journal.pone.0307597.t004

Although not consequential to the testing of the experimental manipulation, statistically significant main effects were also found for time, Wilks’ Lambda = .89, F (1, 271) = 32.99, p = < .001, partial eta squared = .109 and condition, F (1, 271) = 4.42, p = .036, partial eta squared = .016. The means and standard deviations of these main effects are displayed in Table 4 .

Hypothesis 3: Internalisation of societal beauty standards across groups from pre-test to post-test.

A second 2 x 2 mixed between-within subjects ANOVA was conducted to investigate the effect of the experimental manipulation on participants’ internalisation scores. All assumptions for the mixed model ANOVA were met with no violations.

A statistically significant interaction was found between group condition and time, Wilks’ Lambda = .97, F (1, 271) = 8.16, p = .005, partial eta squared = .029. This significant interaction highlights that the change in degree of internalisation at pre-manipulation and post-manipulation is not the same for the two conditions. Interestingly, the internalisation scores for women in the control group decreased from pre-manipulation to post-manipulation, whilst as anticipated, internalisation scores for women in the experimental group increased following exposure to the manipulation (see Table 5 ). This interaction is displayed in Fig 2 .

https://doi.org/10.1371/journal.pone.0307597.g002

https://doi.org/10.1371/journal.pone.0307597.t005

No statistically significant main effects were found for time, Wilks’ Lambda = .987, F (1, 271) = 3.59, p = .059, partial eta squared = .013 or condition, F (1, 271) = 2.65, p = .104, partial eta squared = .010. The means and standard deviations of internalisation scores for each condition at pre-manipulation and post-manipulation are displayed below in Table 5 .

The current study investigated the effect of TikTok content on women’s body image satisfaction and degree of internalisation of appearance ideals, and whether greater TikTok use contributed to increased disordered eating behaviour. In support of the hypotheses, exposure to pro-ana TikTok content significantly decreased participants’ body image satisfaction and increased participants’ degree of internalisation of appearance ideals. The hypothesis that greater daily TikTok use would contribute to increased disordered eating behaviour was not supported, as no statistically significant differences in restrictive disordered eating or ‘healthy’ disordered eating were found between the low, moderate, high, and extreme daily TikTok use groups.

Cross-sectional findings

Daily tiktok use and disordered eating behaviour..

Contrary to expectations, differences among groups on measures of restrictive disordered eating and ‘healthy’ disordered eating did not reach statistical significance. The proposed hypothesis that greater daily TikTok usage would be associated with disordered eating behaviour and attitudes was therefore unsupported. Despite lacking statistical support, participants categorised in the ‘high’ and ‘extreme’ daily TikTok use groups reported an average EAT-26 score of 18.16 and 19.09, respectively. Considering that an EAT-26 cut-off of ≥ 20 indicates potential clinical psychopathology, this mean score illustrates that exposure to TikTok content for two or more hours per day may contribute to a clinical degree of restrictive disordered eating.

The failure of the present study to detect any significant differences in disordered eating behaviours among participants with different TikTok daily usage does not align with the Transactional Model [ 33 ]. According to this model, risk factors such as low self-esteem and high thin ideal internalisation may predispose an individual to seek gratification via social media, resulting in body dissatisfaction and negative affect. The Transactional Model therefore proposes that a positive correlation exists between time spent on social media and body image dissatisfaction. Our findings also do not align with the conclusions Frieiro Padín et al. [ 34 ] drew from their review of the literature, in which a strong connection was identified between time on social media and heightened body image concerns and internalisation of the thin ideal, as well as eating disorder psychopathologies, though a distinction in outcome measures must be noted.

Based on the aforementioned sociocultural theory and previous research [see 28 , 43 , 48 ], it was assumed that increased body dissatisfaction as a result of increased time spent on social media (as stipulated by the Transactional Model), would lead to greater disordered eating behaviour. However, this was not supported statistically in the data. As postulated by Culbert et al. [ 69 ], disordered eating behaviour may instead only be a risk of media exposure if individuals are prone to endorse thin-ideals. Individuals in the present study that reported ‘high’ and ‘extreme’ daily TikTok use may have felt satisfied with their bodies and experienced lower thin-ideal internalisation. This could have potentially buffered the negative effect of greater TikTok content exposure and accounted for the lack of significant differences in disordered eating behaviour between groups. The quantity of TikTok consumption remains a pertinent question for disordered eating behaviour. As per the present study’s brief experimental manipulation, findings suggest that high frequency of daily TikTok use does not necessarily contribute to greater disordered eating behaviour than short exposures to this content.

Content presented to the pro-ana TikTok group included a mix of explicit and implicit pro- eating disorder messages as well as fitspiration content. Fitspiration content presented in the current study included workout videos to achieve a “smaller waist” and “toned abs” where female creators with slim, toned physiques sporting activewear took viewers through a series of exercises, advising viewers that they would “see results in a week”. In the present study, diet-related fitspiration content presented included the concoction of juices to “get rid of belly fat” and advice on the best “diet for a small waist” which requires avoidance of all meat, dairy, junk food, soda, and above all, to make “no excuses”. Fitspiration style content in the current study totalled one-minute, compared to disordered eating themes which totalled six minutes. The integration of these various types of content, although reflective of the For You function in TikTok, impeded our ability to determine the singular impact of fitspiration or disordered eating content, respectively, on body image and internalisation of societal beauty standards, but did reflect social media as it is consumed beyond experimental research settings.

Experimental findings

Tiktok and body image states..

The hypothesis that women exposed to pro-ana TikTok content would experience a significant decrease in body image compared to women who viewed the control TikTok content was supported. The present study found a significant interaction effect of body image between group condition (control vs experimental) and time period (pre-manipulation vs post-manipulation), as well as significant main effects. It is important to note that the statistic of interest in evaluating the success of the experimental manipulation is the interaction effect, thus main effects must be interpreted secondarily and with caution [ 64 ]. Women in the experimental group reported significantly lower body image satisfaction after exposure to the pro-ana TikTok content and compared to women who viewed the control content. This finding corroborates Festinger’s [ 27 ] Social Comparison Theory that posits people naturally evaluate themselves in comparison to others. Exposure to the pro-ana TikTok content, consisting of various thin bodies and messaging around weight loss, may have provided the opportunity for women to engage in maladaptive upward social comparisons, resulting in reduced body image satisfaction. The present study upholds previous findings of Engeln-Maddox, Tiggemann, McComb and Mills, and Gibson [ 29 , 32 , 39 , 70 ] who suggest that visual exposure to thin bodies may adversely affect one’s level of body image satisfaction and extends this research by replicating this finding in the context of a contemporary media platform, TikTok, and by utilising an experimental design.

Contradicting the present study and previous research, Pryde and Prichard [ 42 ] found no significant increase in young women’s body dissatisfaction following exposure to fitspiration TikTok content. A potential explanation for this finding is that the performance of physical movements captured in fitspiration videos may shift the focus of viewers from aesthetics to functionality, highlighting physical competencies and capabilities which has been shown to improve body image satisfaction in young women [ 71 ]. Pryde and Prichard’s [ 42 ] fitspiration content did not include typically occurring harmful themes as the present study did, potentially reducing the negative implications for body image satisfaction of exposure to such content in real world contexts.

Interestingly, women in the control group also reported a statistically significant decrease in body image satisfaction after viewing the neutral TikTok content, a finding that underscores the possible complexity of social media’s influence on body image, as identified in research by Huülsing [ 72 ]. This is an unexpected finding, as the TikTok content displayed to the control group was selected specifically to be unrelated to appearance ideals and pressures. One possible reason for this result is the repetition of administration of the BISS within a short time period. Completing the BISS twice may have caused participants to focus more attention on their body appearance than usual, resulting in more critical appraisals regardless of the experimental stimuli to which they were exposed. This notion aligns with previous research that found focusing on the appearance of body was associated with lower body image satisfaction, whereas focus on the function of the body was associated with more positive body image states [ 71 ].

One potential explanation for this finding is that the control group stimuli was contaminated and produced an unintentional effect on body image scores. Two-minutes of footage within the seven-minute control group TikTok compilation presented the human body including legs, arms, and hands. Although this body-related content was neutral in nature, it may be that even ‘harmless’ representations of the human body are sufficient to elicit a social comparison response in participants or in some capacity, reinforce the #fitspiration motifs commonly depicted on TikTok [ 1 ], therefore impacting body image scores at post-manipulation. This possible explanation has implications for TikTok use and women’s body image, as it suggests that viewing even benign content of human bodies for less than 10-minutes can have an immediate detrimental impact on body image states, even when this content is unrelated to body dissatisfaction, thinness, or weight loss. Furthermore, although a statistically significant body image decrease was detected in the control group, this finding must be interpreted with caution due to the significant interaction effect obtained.

TikTok and internalisation of societal beauty standards.

In accordance with the hypothesis, women in the experimental group reported a significant increase in their degree of internalisation of appearance ideals following exposure to pro-ana TikTok content. Women in the experimental group also reported significantly greater internalisation of appearance ideals than women in the control group. Conversely to the experimental group, internalisation scores of the control group decreased after viewing the neutral TikTok content. These findings are in line with the sociocultural theory, as women reported increased internalisation of societal beauty standards following exposure to media content explicitly and implicitly portraying the thinness ideal. The present study supports Mingoia et al’s. [ 53 ] meta-analysis, which yielded a positive association between social networking site use and the extent of internalisation of the thin ideal and furthers this notion by replicating the finding with TikTok specifically and utilising an experimental design.

In the current study, participants were subject to a single brief exposure of pro-ana TikTok content, whereas most of the sample indicated that their TikTok use was up to two hours per day. This suggests that the degree of internalisation of appearance ideals in participants lives outside of the experiment are likely to be much greater. Mingoia et al. [ 53 ] also found that the use of appearance-related features on social networking sites, such as posting and viewing photos and videos, demonstrated a stronger relationship with the internalisation of the thin ideal than the use of social networking features that were not appearance-related, such as messaging and writing status updates. As TikTok is a video sharing app and most of its content generally features full-body-length camera shots rather than a face or head shot, this finding suggests that TikTok users could potentially internalise body-related societal standards to a greater extent than users of other social media apps that typically feature head shots.

The finding that women internalised societal beauty standards to a greater degree after being exposed to pro-ana TikTok content corroborates the sociocultural theory’s emphasis of the significance of social influences in internalisation. TikTok users may be exposed to all three social influences (i.e., media, peers, and family) simultaneously on a single platform which may encourage internalisation of appearance-ideals in a more profound manner than any of these three influences in isolation. One point of difference between TikTok and other social media apps is that much content on the app is generated by “ordinary” individuals, rather than supermodels or celebrities. This enables blatantly insidious and diet-related content to circulate the app with less policing and scrutiny compared to content produced by an influencer or celebrity who may be more likely to be criticised or cancelled for socially irresponsible messaging and also provides the opportunity for more horizontal social comparisons and peer-to-peer style interactions rather than upward social comparisons.

Indeed, in their study of American teens, Mueller et al. [ 52 ] identified that girls were especially likely to engage in weight loss behaviour if a high proportion of girls with a similar BMI were also engaging in weight loss behaviours. This indicates that internalisation was strongest when appearance-ideals were promoted by alike peers. Due to the fact that much pro-ana TikTok content is created by young women, Mueller et al’s. [ 52 ] finding has problematic implications for the young female users of TikTok, in that harmful diet-related messages could be internalised to a greater extent on TikTok than on other platforms and potentially lead to body image disturbances, disordered eating behaviour, and other negative outcomes among young women.

General discussion

The findings of the current study are important but must also be understood within the broader context of participant’s daily lives beyond their participation in this study. Everyday female-identifying individuals are exposed to a multitude of different sources of information from which body image related stimuli can be drawn. The present study’s experiment was not conducted in a controlled environment due to its online nature, therefore researchers did not have the ability to assess and control for other pieces of body image-related information that participants might have consumed prior to participation that may have been salient for their body image. Further research is required to identify how sustained a change in body image states as measured by the BISS may be over time.

The findings of this study provide some insights into how social media influences disordered eating behaviour and mental health; a theoretical gap in the literature that Choukas-Bradley et al. [ 6 ] highlight as holding back research in this domain. In particular, the findings of the current study indicate that short periods of exposure to disordered TikTok content have an effect, while the high-range EAT-26 scores observed for those who engaged with TikTok for two or more hours a day also raise questions about the duration of exposure. Nonetheless, our findings demonstrate that short exposure periods are sufficient to have a negative effect on body image and internalisation of the thin ideal.

One point that may be readily overlooked in developing a theoretical framework around social media’s influence is that the narrative arc of TikTok videos is such that users are exposed to many short stories in quick succession, which may have a different effect to longer form content from a single content creator. As Pierce [ 2 ] notes, the speed of exposure to overlapping, but separate narratives depicted in successive videos, is an important feature of TikTok content and may contribute to the influence of such platforms on disordered eating and body attitudes. Each piece of content serves as a standalone narrative but may also overlap and interact with the viewer’s experience of the next video they watch to build a cumulative, normalised narrative of disordered body- and eating-practices.

In the current study, participants who engaged with TikTok for two-three hours a day were classified as high users, and those who used TikTok for three or more hours were classed as extreme. These rates of usage may, however, be quite normative, with Santarossa and Woodruff [ 73 ] citing three-four hours a day on social media as normative for their sample of young adults, though notably participants in the current study were only questioned about their TikTok usage, not their general use of social media.

While we examined the effect of pro-ana content in this study, that some changes were observed in the control group as well as the experimental group indicates that the social media environment, characterised as it is by idealisation, instant feedback, and readily available social comparison [ 6 ], may play a general role in diminishing positive body image attitudes and healthy aspirations. This is supported by Tiggemann and Slater’s [ 35 , 36 ] research in which social media usage was found to correlate positively with higher levels of body image concerns, in contrast to time spent on the internet more generally, and this may be particularly true for visually oriented platforms that sensitize viewers to their own appearance and that of others. As noted previously, of the visually-oriented social media platforms, predominantly TikTok and Instagram, videos are commonly framed on TikTok so that the subject’s whole body is visible, particularly in dance videos and in #GymTok content, where on Instagram, cameo style head-shot videos appear more likely to feature, which further suggests that TikTok may provide more body-related stimuli than other platforms, even when the intention of the content does not relate to body-image or #fitspiration.

Importantly, the algorithm on TikTok functions in such a way that those who actively seek out body positivity content may also be exposed to nefarious body-related content such as body checking, a competitive, self-surveillance type of content where users are encouraged to test out their weight by attempting to drink from a glass of water while their arm encircles another’s waist. As McGuigan [ 74 ] reports, watching just one body checking video may result in hundreds more filtering through a user’s For You page, with those actively attempting to seek out positive body image content likely to be inadvertently exposed to disordered content due to the configuration of the algorithm. This function of the For You page is demonstrated in the current study, with 64% of participants reporting having seen disordered eating content on their For You page, higher than any other kind of harmful content, including suicide and bullying. The current study did not assess participants’ consumption of #FoodTok, #GymTok, and #Fitspiration. Engagement with these dimensions of TikTok and the type of content that participants seek out via the search function warrant consideration in future research.

The TikTok algorithm underscores Logrieco et al’s. [ 18 ] findings that even anti-anorexia content can be problematic, especially given complexities in determining and controlling what is performatively problematic, including videos discussing recovery and positive body attitudes that may somewhat paradoxically further body policing and competition among users and consumers of social media content. Furthermore, as Logrieco et al. [ 18 ] highlight, TikTok is replete in both pro-ana and much more implicit body-related content that may be harmful to viewers, not to mention those creating the content, whose experiences also warrant consideration.

Theoretical and practical implications

The present study bridged an important gap in the literature by utilising both experimental and cross-sectional designs to examine the influence of pro-ana TikTok content on users’ body image satisfaction, internalisation of body ideals, and disordered eating behaviours. While the negative impact of social media on body image and eating behaviours has been established in relation to platforms such as Instagram and Twitter, TikTok’s rapid emergence and unique algorithm warrant independent analysis.

The present findings have important theoretical implications for the understanding of sociocultural influences of orthorexia nervosa development. Notably, this study is one of the first to highlight the association between orthorexia nervosa and the tripartite model of disordered eating using an experimental design. The results illustrate that the internalisation of sociocultural appearance ideals predicts the development of ‘healthy’ disordered eating, as suggested by the tripartite theory. Western culture ideals do seem to influence the expression of orthorexic tendencies, thus caution should be exercised by women when interacting with appearance-related TikTok content.

Unlike explicit pro-ana content, which is open to condemnation, the moral and health-related discourses underpinning much body-related content in which thinness and health are espoused as goodness, reflects a new trend in diet culture masquerading as wellness culture [ 20 , 21 ]. Questions are raised around the ethics of social media algorithms when the technologically fostered link between recovery-focused content and disordered-content on TikTok is laid bare, particularly considering that extant research has found individuals with experience of eating disorders often seek out support, safety, and connection online [ 49 ] and in doing so on a platform like TikTok, may be exposed to more disordered eating content than the average user. Given visual social media platforms are associated with higher levels of dysfunction in relation to body image [ 4 ], the policy and ethics of such platforms warrant scrutiny from a variety of stakeholders in management, marketing, technology regulation, with psychology playing an important role in the marketing of these platforms. As traditional journalistic platforms have been subjected to scrutiny and reform, so too must a climate of accountability be established within the social media nexus.

The widespread growth of social media may warrant greater concern than traditional forms of mass media, not only because of the full-time accessibility and diverse range of platforms, but also due to the prevalence of peer-to-peer interactions. According to the social comparison theory, comparison of oneself to others has traditionally considered more removed, higher status influences (e.g., celebrities, actors/actresses, supermodels) as a greater source of pressure than those in the individuals’ natural environment (e.g., family and peers). Re-examination of this theoretical perspective is warranted considering the contemporary challenges of social media and the perpetuation of body image messages from alike peers. Furthermore, a diverse range of “content” may trigger disordered body- and eating-related attitudes, including #fitspiration and #GymTok, which poses challenges for social media platforms in regulating content. The inclusion of orthorexia in the milieu highlights the disordered nature of seemingly benign health practices and social media content.

That TikTok content containing explicit and implicit pro-ana themes may readily remain on the app uncensored exemplifies the importance of protective strategies to build resilience at the individual level. One such protective strategy is shifting focus from body appearance to functionality. Alleva and colleagues [ 71 ] investigated the Expand Your Horizon programme, designed to improve body image by training women to focus on body functionality. They report that women who engaged with the Expand Your Horizon programme experienced greater satisfaction with body image and functionality, body appreciation, and reduced self-objectification compared to women who did not engage with the program. Health professionals involved in the care of women with eating disorders and other mental health issues should also be educated to ensure they are knowledgeable about the social media content their clients may be exposed to, equipping them with skills to engage in conversations about the potential detrimental impacts of viewing pro-ana and other harmful TikTok content [ 53 ].

The administration of such programs in schools, universities, community groups, and clinical settings could prove effectual in the prevention of disordered eating and body image disturbance development and may reduce symptom severity of a pre-established disorder. Such programs must be developed with great care, however, given the propensity for even anti-anorexia content to have a negative effect on those consuming it [ 18 ]. The development of self-compassion may also build resilience in women, with research confirming that self-compassion can be effectively taught [ 75 ]. Subsequently, programs have been developed such as Compassion Focused Therapy (CFT) in which clients are trained to develop more compassionate self-talk during negative thought processes and to foster more constructive thought patterns [ 76 ]. The value of CFT has been established in the literature with both clinical and non-clinical samples and has promising outcomes particularly for those high in self-criticism [ 77 ].

Young women should be provided with media literacy tools that can assist in advancing critical evaluations of the online world. Digital manipulation of advertising and celebrity images is well known to many people, however, this awareness may be lacking regarding social media images, as they are generally disseminated within one’s peer network rather than outside of it [ 33 ]. Media literacy interventions may educate women about how social media perpetuates appearance-ideals that are often unrealistic and unattainable [ 53 ]. As an example, Posavac et al. [ 78 ] revealed that a single media literacy intervention resulted in a reduction in women’s social comparison to body ideals portrayed in the media.

Such interventions might be extended to female-identifying TikTok users to educate them on the manipulation of videos to produce idealised portrayals of the self. Media literacy should be commenced from an early age by teaching children, adolescents, and adults to understand the influence of implicit messages conveyed through social media and to create media content that is responsible and psychologically safe for others [ 79 ]. Increased understanding of messages portrayed by social media content may prevent thin-ideal endorsement and internet misuse. Notably, however, the most effective approach would be to address the problem at its source and increase the regulation of social media companies, rather than upskill users in how to respond to harmful online environments, which creates further labour for the individual while allowing organisations to continue to produce harmful but easily monetizable content.

Limitations and future directions

To meet the requirements to run multivariate analyses, the continuous data of body image and internalisation scores were dichotomised using a median split to create ‘low’ and ‘high’ groups for each variable. Although dichotomisation was necessary to perform appropriate analyses and power analyses deemed the sample size as adequate following performance of the median split, dichotomising these variables may have contributed to a loss of statistical power to detect true effects.

Limitations are implicated in the use of the ORTO-15 in the present study. The ORTO-15 does not account for different lifestyle factors that may alter a participants’ response, such as dietary restrictions, food intolerances, or medical dietary guidelines. The discrepancies in literature surrounding the psychometric properties of the ORTO-15 may be attributable to the lack of established diagnostic criteria of orthorexia nervosa, cultural differences in expressions of eating disorders, and difficulty comparing research results in determining orthorexia nervosa diagnoses due to inconsistencies in testing questions and cut-off values [ 61 ]. Due to unacceptable reliability in the present study, a factor analysis was performed which identified a factor relating to health food preoccupation. This identified factor was used as the ORTO-15 measure and data from these 5-items were used in analyses and referred to throughout the present study as ‘healthy’ disordered eating. Using the 5-items related to ‘healthy’ disordered eating rather than the complete 15-item scale may not have accurately assessed participants’ degree of orthorexic tendencies. Despite these limitations, the ORTO-15 is the only accepted measure of orthorexic tendencies available [ 63 ]. Additionally, more limitations would likely have been encountered by using the full 15-item measure lacking reliability, compared to utilising the 5-item factor with acceptable reliability.

Future studies of TikTok and disordered eating behaviour should incorporate a measure of social comparison to verify whether social comparison is the vehicle through which women experience decreased body image satisfaction after viewing TikTok content. Future research should also examine the influence of TikTok content creation on body image, internalisation of thinness, and disordered eating behaviour and explore the association between what individuals consume on TikTok and the social media content that they produce. This research should be conducted using more diverse samples of women, including transgender women, to determine whether the findings of the present study are relevant for this population given the unique challenges regarding body image and societal beauty standards that they may experience.

Longitudinal studies are also warranted to examine the effect of exposure to pro-ana TikTok content over time, and to assess the effects of pro-ana TikTok content on body image satisfaction and eating disorder symptomology over time. Further research on orthorexia nervosa is needed to establish a more reliable measure of orthorexic tendencies and this would enable future investigation of the impact of pro-ana TikTok content on the development of orthorexia nervosa, as well as individual differences as predisposing factors in the development of orthorexic tendencies. Finally, future research should examine the efficacy of media literacy and self-compassion intervention programs as a protective factor specifically in the TikTok context, where disordered eating messages are more explicit in nature than traditional media and other social media platforms.

The findings of the current study support the notion that pro-ana TikTok content decreases body image satisfaction and increases internalisation of societal beauty standards in young women. This research is timely given reliance on social media for social interaction, particularly for young adults. Our findings indicate that female-identifying TikTok users may experience psychological harm even when explicit pro-ana content is not sought out and even when their TikTok use is time-limited in nature. The findings of this study suggest cultural and organisation change is needed. There is a need for more stringent controls and regulations from TikTok in relation to pro-ana content as well as more subtle forms of disordered eating- and body-related content. Prohibiting or restricting access to pro-ana content on TikTok may reduce the development of disordered eating and the longevity and severity of established eating disorder symptomatology among young women in the TikTok community. There are current steps being taken to delete dangerous content, including blocking searches such as “#anorexia”, however, there are various ways users circumvent these controls and further regulation is required. Unless effective controls are implemented within the platform to prevent the circulation of pro-ana content, female-identifying TikTok users may continue to experience immediate detrimental consequences for body image satisfaction, thin-ideal internalisation, and may experience an increased risk of developing disordered eating behaviours.

• 1. Burger MR. Correlation Between Social Media Use and Eating Disorder Symptoms: A Literature Review. California Polytechnic State University. 2022. https://digitalcommons.calpoly.edu/kinesp/20/ .
• 2. Pierce S. Alimentary Politics and Algorithms: The Spread of Information about “Healthy” Eating and Diet on TikTok. Washington University. 2022. https://openscholarship.wustl.edu/undergrad_etd/40\ .
• 3. Asano E. How Much Time Do People Spend on Social Media? 2017 Jan 1 [cited 15 December 2022]. In: Social Media Today [Internet]. Industry Dive 2023. https://www.socialmediatoday.com/marketing/how-much-time-do-people-spend-social-media-infographic .
• View Article
• PubMed/NCBI
• Google Scholar
• 10. Green S. An Experimental Study on the Effects of Pro-Anorexia Content on Eating Disorder Development. M.A. Thesis, Western Kentucky University. 2019. https://digitalcommons.wku.edu/theses/3138 .
• 13. Shepherd J. 20 Essential TikTok statistics you need to know in 2022. 2022 Oct 1 [cited 1 December 2022]. In: The Social Shephard [Internet]. 2023. https://thesocialshepherd.com/blog/tiktok-statistics .
• 14. Kemp S. Digital 2020: Global Digital Overview. Datareportal. 2020 Jan 30 [Cited 2022 November 29]. https://datareportal.com/reports/digital-2020-global-digital-overview .
• 15. TikTok. Privacy Policy. Updated 2023 Aug 4 [cited 5 May 2021]. https://www.tiktok.com/legal/page/row/privacy-policy/en .
• 17. Muno DA. Orthorexia nervosa as a distinct eating disorder category: Similarities in alexithymia, attachment, perfectionism, body dissatisfaction, & eating attitudes. Psy.D. Dissertation, Alliant International University. 2020. https://www.proquest.com/openview/e8979ba30dd9eeefd80214e6556c1960/1?pq-origsite=gscholar&cbl=18750&diss=y .
• 23. Cash T. Cognitive-behavioural perspectives on body image. In: Cash T, Pruzinsky T, editors. Body Image: A handbook of theory, research, and clinical practice. New York: Guilford Press; 2002. pp. 38–46.
• 37. DeMuynck JP. The Femininity Diet: A Rhetorical Analysis of the Discursive Formation of Femininity and Weight Loss in Contemporary Social Media Promotions. M.A. Thesis, Texas State University. 2020. https://digital.library.txst.edu/items/a15097cd-b076-4ccd-8f80-2b5d373550eb .
• 48. Thompson JK, Heinberg LJ, Altabe M, Tantleff-Dunn S. Exacting beauty: Theory, assessment, and treatment of body image disturbance. American Psychological Association; 1999. https://psycnet.apa.org/record/1999-02140-000 .
• 64. Pallant J. SPSS survival manual: A step by step guide to data analysis using IBM SPSS. 6 th ed. McGraw-hill education (UK); 2016.
• 68. Cohen JW. Statistical power analysis for the behavioural sciences. 2 nd ed. Hillsdale NJ: Lawrence Erlbaum Associates; 1988.
• 70. Gibson KS. Thinspiration and Fatspiration on Body Dissatisfaction: The Roles of Social Comparisons and Anti-Fat Attitudes. M.A. Dissertation, Texas State University. 2021.
• 72. Hülsing GM. # Triggerwarning: Body Image: A qualitative study on the influences of TikTok consumption on the Body Image of adolescents. BSc Thesis, University of Twente. 2021.
• 74. McGuigan S. Body checking is dangerous and it’s all over TikTok. 2022 June 14 [cited 1 December 2022]. In Refinery29 [Internet]. 2022. https://www.refinery29.com/en-au/body-checking-tiktok-trend .
• 76. Gilbert P. Compassion: Conceptualisations, research, and use in psychotherapy. London, UK: Routledge; 2005. https://books.google.com.au/books?id=-I6NAgAAQBAJ&dq=Gilbert,+P.+(2005).+Compassion:+Conceptualisations,+research,+and+use+in+psychotherapy&lr=&source=gbs_navlinks_s .

• Open access
• Published: 05 August 2024

Benefits of budesonide/glycopyrronium/formoterol fumarate dihydrate on lung function and exacerbations of COPD: a post-hoc analysis of the KRONOS study by blood eosinophil level and exacerbation history

• Shigeo Muro   ORCID: orcid.org/0000-0001-7452-9191 1 ,
• Tomotaka Kawayama 2 ,
• Hisatoshi Sugiura 3 ,
• Munehiro Seki 4 ,
• Elizabeth A. Duncan 5 ,
• Karin Bowen 6 ,
• Jonathan Marshall 7 ,
• Ayman Megally 8 &
• Mehul Patel   ORCID: orcid.org/0000-0003-0435-858X 9  

Respiratory Research volume  25 , Article number:  297 ( 2024 ) Cite this article

810 Accesses

Metrics details

Japanese guidelines recommend triple inhaled corticosteroid (ICS)/long-acting muscarinic antagonist (LAMA)/long-acting β 2 -agonist (LABA) therapy in patients with chronic obstructive pulmonary disease (COPD) and no concurrent asthma diagnosis who experience frequent exacerbations and have blood eosinophil (EOS) count ≥ 300 cells/mm 3 , and in patients with COPD and asthma with continuing/worsening symptoms despite receiving dual ICS/LABA therapy. These post-hoc analyses of the KRONOS study in patients with COPD and without an asthma diagnosis, examine the effects of fixed-dose triple therapy with budesonide/glycopyrronium/formoterol fumarate dihydrate (BGF) versus dual therapies on lung function and exacerbations based on blood EOS count – focusing on blood EOS count 100 to < 300 cells/mm 3 – as a function of exacerbation history and COPD severity.

In KRONOS, patients were randomized to receive treatments that included BGF 320/14.4/10 µg, glycopyrronium/formoterol fumarate dihydrate (GFF) 14.4/10 µg, or budesonide/formoterol fumarate dihydrate (BFF) 320/10 µg via metered dose inhaler (two inhalations twice-daily for 24 weeks). These post-hoc analyses assessed changes from baseline in morning pre-dose trough forced expiratory volume in 1 s (FEV 1 ) over 12–24 weeks and moderate or severe COPD exacerbations rates over 24 weeks. The KRONOS study was not prospectively powered for these subgroup analyses.

Among patients with blood EOS count 100 to < 300 cells/mm 3 , least squares mean treatment differences for lung function improvement favored BGF over BFF in patients without an exacerbation history in the past year and in patients with moderate and severe COPD, with observed differences ranging from 62 ml to 73 ml across populations. In this same blood EOS population, moderate or severe exacerbation rates were reduced for BGF relative to GFF by 56% in patients without an exacerbation history in the past year, by 47% in patients with moderate COPD, and by 50% in patients with severe COPD.

Conclusions

These post-hoc analyses of patients with moderate-to-very severe COPD from the KRONOS study seem to indicate clinicians may want to consider a step-up to triple therapy in patients with persistent/worsening symptoms with blood EOS count > 100 cells/mm 3 , even if disease severity is moderate and there is no recent history of exacerbations.

Trial registration

ClinicalTrials.gov registry number NCT02497001 (registration date, 13 July 2015).

Chronic obstructive pulmonary disease (COPD) is a leading cause of morbidity and mortality worldwide, with economic and social burdens that are both substantial and increasing [ 1 , 2 ]. Three fixed-dose triple therapies with an inhaled corticosteroid (ICS), a long-acting muscarinic antagonist (LAMA), and a long-acting β 2 -agonist (LABA) are approved for the maintenance treatment of COPD [ 3 , 4 , 5 ].

The Global Initiative For Chronic Obstructive Lung Disease (GOLD) 2023 report recommends triple therapy with an ICS/LAMA/LABA be considered as an initial treatment option in patients with blood eosinophil (EOS) count ≥ 300 cells/mm 3 with frequent (≥ 2/year) moderate exacerbations or (≥ 1) exacerbation leading to hospitalization [ 2 ]. A step up to ICS/LAMA/LABA triple therapy is also recommended in patients with blood EOS count ≥ 100 cells/mm 3 who experience exacerbations despite receiving LAMA/LABA dual therapy [ 2 ]. According to COPD treatment guidelines in Japan [ 6 ], ICS-containing treatment is recommended for patients with COPD and a clinical asthma diagnosis when dual therapy is not sufficient; however, in patients with COPD and no asthma diagnosis, ICS/LAMA/LABA triple therapy is only recommended for those who experience frequent exacerbations (≥ 2 moderate or ≥ 1 severe per year) and have blood EOS count ≥ 300 cells/mm 3 .

In ETHOS (NCT02465567), a study of patients with moderate-to-very severe COPD with exacerbations and receiving at least two inhaled maintenance therapies at screening, the fixed-dose triple combination therapy budesonide/glycopyrronium/formoterol fumarate dihydrate (BGF) 320/14.4/10 µg significantly reduced the annual rate of moderate or severe exacerbations (the primary study end point) [ 7 ] and significantly improved lung function (pulmonary function test sub-study primary endpoint) versus glycopyrronium/formoterol fumarate dihydrate (GFF) and budesonide/formoterol fumarate dihydrate (BFF) [ 8 ]. Similarly, in KRONOS (NCT02497001), a study of patients with moderate-to-very severe COPD and no requirement for prior exacerbations, BGF 320/14.4/10 significantly improved lung function versus GFF, BFF, and open-label budesonide/formoterol fumarate dry-powder inhaler (BUD/FORM), and significantly reduced the rate of moderate or severe exacerbations versus GFF [ 9 ].

Importantly, benefits of BGF over dual therapy were observed across a range of blood EOS counts in post-hoc analyses of ETHOS and KRONOS [ 10 , 11 , 12 ]. Given recommendations in the GOLD 2023 report [ 2 ], current Japanese treatment guidelines [ 6 ], and evidence for benefits of BGF over dual therapy across a range of blood EOS counts (including below 300 cells/mm 3 ) in patients with COPD [ 10 , 11 ], post-hoc analyses of the KRONOS study were conducted to further examine the effects of BGF versus dual LAMA/LABA and ICS/LABA therapies on lung function and exacerbation rates in patients with COPD based on blood EOS count (100 to < 300 and ≥ 100 cells/mm³) as a function of exacerbation history (exacerbations in the past year and no exacerbations in the past year) and COPD severity (moderate, severe, very severe).

Study design

A detailed description of the study design and patient population in KRONOS (ClinicalTrials.gov registry number NCT02497001; registration date, 13 July 2015), including inclusion and exclusion criteria, has been previously published [ 9 ]. In brief, KRONOS was a 24-week, double-blind, parallel-group, phase III randomized controlled study conducted at 215 sites across four countries (Canada, China, Japan, and the United States).

At screening, eligible patients discontinued current COPD medications (i.e., LAMA, LABA, or both) for the study duration and received open-label ipratropium bromide four times daily as COPD maintenance therapy. ICS use was permitted during screening, provided patients were on a stable dose for at least 4 weeks before screening; however, both ipratropium and ICS were stopped before randomization. Rescue use of salbutamol was permitted throughout the study.

After screening, patients were randomized 2:2:1:1 to receive BGF 320/14.4/10 µg, GFF 14.4/10 µg, or BFF 320/10 µg via a single Aerosphere ™ metered dose inhaler, or open-label BUD/FORM 400/12 µg via a dry powder inhaler (Symbicort ® Turbuhaler ® ), as two inhalations twice-daily for 24 weeks. As BFF was not an approved COPD therapy at the time KRONOS was conducted, BUD/FORM (which was already approved for COPD treatment) was included as an active comparator to support BFF as a comparator for BGF. However, for the purposes of this post-hoc analysis, only data for BFF and GFF are reported.

The study was conducted in accordance with Good Clinical Practice, including the Declaration of Helsinki. The protocol and informed consent form were approved by appropriate institutional review boards or independent ethics committees prior to the start of the study (a full listing of appropriate institutional review boards or independent ethics committees has been published [ 9 ]). All patients provided written informed consent before screening.

Key inclusion criteria for the KRONOS study have been described in detail previously [ 9 ]. Eligible patients were aged 40–80 years; were current or former smokers (smoking history of ≥ 10 pack-years); had an established COPD clinical history, as defined by the American Thoracic Society/European Respiratory Society [ 13 ] or Japanese local guidelines [ 14 ]; had moderate-to-very severe COPD, defined as post-bronchodilator FEV 1 of 25–80% of predicted normal values based on National Health and Nutrition Examination Survey III reference equations [ 15 ] or applicable local reference norms [ 14 , 15 , 16 , 17 ]; and were symptomatic (as defined by a COPD Assessment Test score ≥ 10) despite treatment with ≥ 2 inhaled maintenance therapies for ≥ 6 weeks before screening. Patients were not required to have a history of COPD exacerbations in the previous 12 months and were excluded if they had a current diagnosis of asthma or any respiratory disease other than COPD, evaluated by the investigator, that could affect study results.

In the KRONOS study, the primary lung function endpoint, according to the Japanese/Chinese regulatory approach, was change from baseline in morning pre-dose trough FEV 1 over 12–24 weeks; the rate of moderate or severe COPD exacerbations over 24 weeks was a secondary efficacy endpoint [ 12 ].

A COPD exacerbation was defined as a change in the patient’s usual COPD symptoms lasting for ≥ 2 days that was beyond normal day-to-day variation, acute in onset, and may have warranted a change in regular medication. An exacerbation was considered moderate if it resulted in systemic corticosteroid and/or antibiotic use for at least 3 days, and as severe if it resulted in an inpatient COPD-related hospitalization or death.

Data presentation and statistical analyses

For the current post-hoc analyses, change from baseline in morning pre-dose trough FEV 1 over 12–24 weeks and the rate of moderate or severe COPD exacerbations over 24 weeks were analyzed in patients with blood EOS counts of 100 to < 300 cells/mm 3 and ≥ 100 cells/mm 3 as a function of exacerbation history (any moderate or severe exacerbations in the past year; no exacerbations in the past year) and COPD severity (moderate [FEV 1 50–<80% predicted], severe [FEV 1 30–<50% predicted], very severe [FEV 1  < 30% predicted]). Analyses were conducted in the modified intention-to-treat (mITT) population, which included all patients with post-randomization data obtained before treatment discontinuation.

The primary baseline EOS subgroup of interest included those with blood EOS count 100 to < 300 cells/mm 3 , as assessment of this subgroup will provide insight into the benefits of BGF among patients with blood EOS count < 300 cells/mm 3 . The blood EOS count ≥ 100 cells/mm 3 subgroup was included to provide supportive evidence that inclusion of patients with blood EOS count > 300 cells/mm 3 in the analysis did not result in substantively different findings. Patients with blood EOS count < 100 cells/mm 3 were not included in the post-hoc analyses because the population size would be small and the published literature supports greater ICS benefits with higher EOS count [ 7 , 9 , 18 , 19 , 20 , 21 ] and lesser ICS efficacy with low blood EOS count [ 2 , 8 , 20 ].

Demographic and clinical characteristics are reported descriptively across treatment arms for each subgroup. Change from baseline in morning pre-dose trough FEV 1 over 12–24 weeks in each EOS subgroup by exacerbation history in the preceding 12 months or COPD severity was assessed using a linear repeated measures model that included baseline FEV 1 , percent reversibility to salbutamol, and baseline blood EOS count as continuous covariates and visit, treatment, treatment-by-visit interaction, and ICS use at screening (yes or no), as categorical covariates. Data reported includes the least squares (LS) mean change from baseline with 95% confidence intervals (CIs) for each treatment and LS mean differences with 95% CIs in the change from baseline for each treatment versus BGF.

The rate of moderate or severe exacerbations over 24 weeks in each EOS subgroup by exacerbation history in the preceding 12 months or COPD severity was assessed using negative binomial regression; treatments were compared with adjustment for baseline post-bronchodilator percent predicted FEV 1 , baseline COPD exacerbation history (0, 1, or ≥ 2) in the preceding 12 months, log baseline blood EOS count, region, and ICS use at screening (yes or no). The logarithm of the time at risk of experiencing an exacerbation was used as an offset variable in the model. The data reported includes the number (%) of patients with exacerbations, the total time at risk for an exacerbation, and the adjusted (standard error [SE]) rate of moderate or severe exacerbations; treatment differences between BGF and the other treatment arms are reported using rate ratios (RR) with 95% CIs. As the KRONOS study was not prospectively powered for any of the reported post-hoc analyses, reported P -values are nominal, unadjusted for multiplicity, and provided for descriptive purposes only.

Patient disposition and characteristics

The disposition and demographic/clinical characteristics of patients in the KRONOS study has been described in detail previously [ 9 ]. In brief, of 1902 randomized patients, 1896 were included in the mITT population (BGF, n  = 639; GFF, n  = 625; BFF, n  = 314). Across treatment groups in the overall mITT population, the average age was approximately 65 years, and the median blood EOS count was approximately 150 cells/mm 3 ; approximately 74% of patients did not report having an exacerbation in the preceding 12 months.

Demographic and clinical characteristics in patients with blood EOS count 100 to < 300 cells/mm 3 with and without exacerbations in the preceding 12 months are summarized in Table  1 and in patients categorized based on COPD severity in Additional file 1 supplementary Table S1 . Across treatment groups, demographic and clinical characteristics within each exacerbation history subgroup and COPD severity subgroup were well balanced, with the exception of those variables associated with categorization (i.e., exacerbation history or FEV 1 % predicted). Similarly, among patients with blood EOS count ≥ 100 cells/mm 3 , patient characteristics in each exacerbation history subgroup (Additional file 1 supplementary Table S2 ) or COPD severity subgroup (Additional file 1 supplementary Table S3 ) were also well balanced across treatment groups.

• Lung function

Across treatment groups, increases from baseline in morning pre-dose trough FEV 1 were observed over 12–24 weeks for all blood EOS counts by exacerbation history and COPD severity subgroups (Additional file 1 supplementary Table S4 ). Among patients with blood EOS count 100 to < 300 cells/mm 3 , improvement in lung function with BGF versus BFF was observed among those without an exacerbation history in the preceding 12 months (nominal P  < 0.0001; Fig.  1 A); treatment differences in the changes from baseline in morning pre-dose trough FEV 1 were not suggestive of differences between BGF and GFF (Fig.  1 A). Improvements in lung function with BGF versus BFF were observed among those with moderate and severe COPD (both nominal P  < 0.05; Fig.  1 B), with a similar trend among those with very severe COPD; treatment differences in the changes from baseline in morning pre-dose trough FEV 1 were not suggestive of differences between BGF and GFF (Fig.  1 B).

Lung function difference versus BGF a, b : EOS subgroups by exacerbation history or COPD severity, mITT population. a Change from baseline in morning pre-dose trough FEV 1 over 12–24 weeks. b From a linear repeated measures model which included the following covariates: baseline FEV 1 , percent reversibility to salbutamol, and baseline EOS count as continuous covariates and visit, treatment, treatment-by-visit interaction, and ICS use at screening (yes/no) as categorical covariates. Abbreviations: BFF: budesonide/formoterol fumarate dihydrate; BGF: budesonide/glycopyrronium/formoterol fumarate dihydrate; CI: confidence interval; COPD: chronic obstructive pulmonary disease; EOS: eosinophil; FEV 1 : forced expiratory volume in 1 s; GFF: glycopyrronium/formoterol fumarate dihydrate; ICS: inhaled corticosteroid; LS: least squares; mITT: modified intention-to-treat

Similarly, among patients with blood EOS count ≥ 100 cells/mm 3 , improvements in lung function with BGF versus BFF were observed among those without an exacerbation history in the preceding 12 months (nominal P  < 0.0001; Fig.  1 C); treatment differences in the change from baseline in morning pre-dose trough FEV 1 were not suggestive of differences between BGF and GFF (Fig.  1 C). Improvement in lung function with BGF versus BFF was observed regardless of COPD severity (all nominal P  < 0.05; Fig.  1 D). Treatment differences in the change from baseline in morning pre-dose trough FEV 1 were not suggestive of differences between BGF and GFF in any COPD severity subgroup (Fig.  1 D).

• Exacerbation rates

Across blood EOS counts by exacerbation history in the preceding 12 months or COPD severity, the adjusted rate of moderate or severe exacerbations was greater with GFF than any other treatment (Table  2 ). Among patients with blood EOS count 100 to < 300 cells/mm 3 , the risk of moderate or severe exacerbations was 56% lower for BGF versus GFF in patients without exacerbation history in the preceding 12 months (nominal P  < 0.0001; Fig.  2 A), with a similar trend observed in those with exacerbation history in the preceding 12 months. Risk of moderate or severe exacerbations were 47% and 50% lower, respectively, for BGF versus GFF in patients with moderate and severe COPD (both nominal P  < 0.05; Fig.  2 B), with a similar trend observed for very severe COPD. Examination of RRs for moderate or severe exacerbations between BGF versus BFF was not suggestive of treatment differences for either exacerbation history subgroup (Fig.  2 A) or COPD severity group (Fig.  2 B).

Moderate/severe exacerbation risk versus BGF a : EOS subgroups by exacerbation history or COPD severity, mITT population. a Treatments compared adjusting for baseline post-bronchodilator percent predicted FEV 1 , baseline COPD exacerbation history (0, 1, or ≥ 2) in the preceding 12 months, log baseline blood EOS count, region, and ICS use at screening (yes/no) using negative binomial regression; the logarithm of the time at risk of experiencing an exacerbation was used as an offset variable in the model. Abbreviations: BFF: budesonide/formoterol fumarate dihydrate; BGF: budesonide/glycopyrronium/formoterol fumarate dihydrate; CI: confidence interval; COPD: chronic obstructive pulmonary disease; EOS: eosinophil; FEV 1 : forced expiratory volume in 1 s; GFF: glycopyrronium/formoterol fumarate dihydrate; ICS: inhaled corticosteroid; mITT: modified intention-to-treat; RR: rate ratio

Among patients with blood EOS count ≥ 100 cells/mm 3 , similar trends were observed in patients with blood EOS count 100 to < 300 cells/mm 3 (Fig.  2 C-D). However, this is not surprising as those with blood EOS count 100 to < 300 cells/mm 3 constitute the majority of the sample; only 12.4% of patients in the KRONOS mITT had blood EOS count > 300 cells/mm 3 .

In this post-hoc analysis of the KRONOS study, lung function and exacerbation rates with BGF versus dual LAMA/LABA and ICS/LABA therapies were evaluated in patients with moderate-to-very severe COPD in blood EOS count subgroups, as a function of exacerbation history in the preceding 12 months and COPD severity. To the best of our knowledge, these are the first analyses to suggest that triple therapy is effective even in patients with no history of exacerbations and low levels of peripheral eosinophilia.

Triple therapy with BGF improved lung function, as measured by greater increases from baseline in morning pre-dose trough FEV 1 , versus dual ICS/LABA therapy with BFF, in patients with blood EOS count 100 to < 300 cells/mm 3 without an exacerbation history in the preceding 12 months and among patients with moderate and severe COPD. Similar findings were observed among patients with blood EOS count ≥ 100 cells/mm 3 , which included a relatively small number of patients with blood EOS count ≥ 300 cells/mm 3 . Additionally, triple therapy with BGF reduced the annual moderate or severe exacerbations rate versus LAMA/LABA dual therapy with GFF in patients with blood EOS count 100 to < 300 cells/mm 3 without an exacerbation history in the preceding 12 months and among those with moderate and severe COPD severity, with a similar trend observed for very severe COPD. Overall, these findings seem to indicate that benefits of triple BGF therapy versus dual LAMA/LABA and ICS/LABA therapy are observed across a range of blood EOS counts (even when blood EOS counts are 100 to < 300 cells/mm 3 ) and exacerbation histories (including in the absence of exacerbations in the past year), and COPD severity (including those with moderate COPD). These findings may suggest that triple therapy with BGF is more effective than treatment without ICS, i.e., LAMA/LABA, in terms of exacerbations, and more effective than treatment without LAMA, i.e., ICS/LABA, in terms of lung function in some patients.

The observation that BGF conveys benefits over dual ICS/LABA and LAMA/LABA therapy in patients with blood EOS count 100 to < 300 cells/mm 3 is consistent with previously published reports [ 9 , 10 , 20 ]. In post-hoc analyses of the 52-week ETHOS study, BGF improved morning pre-dose trough FEV 1 versus BFF and GFF as well as reduced moderate or severe exacerbation rates versus GFF across a range of blood EOS counts (≥ 100, ≥ 100−<300, and ≥ 300 cells/mm³) [ 10 ]. In the KRONOS study, change from morning pre-dose trough FEV 1 with BGF versus BFF and BUD/FORM, as well as reductions in the rate of moderate or severe exacerbations for BGF versus GFF, were observed in patients with blood EOS count < 150 cells/mm 3 [ 9 ]. Similarly, results of the triple therapy studied in the 52-week IMPACT trial indicated that moderate or severe exacerbation rates with fluticasone furoate/umeclidinium/vilanterol triple therapy were lower compared with dual LAMA/LABA therapy with umeclidinium/vilanterol across a range of blood EOS levels, including at blood EOS count of approximately 100 to 300 cells/mm 3 [ 20 ]. Although the duration of the intervention was not long enough, the reduction in exacerbation rate with BGF triple therapy may be considered clinically meaningful. The clinical significance of the improvement in respiratory function needs to be clarified in future studies.

In the KRONOS study, exacerbation history reported in the year before study entry was lower than the model-estimated rates observed during the study [ 9 ]. This suggests that there are other factors that lead to the risk of exacerbations, and not only exacerbation history in the preceding 12 months. Although, not having an exacerbation history in the preceding 12 months is not synonymous with reduced risk, it is widely accepted that those with a history of exacerbations are more likely to experience a future exacerbation [ 22 ]. This is supported by observations in the current analyses, as patients with an exacerbation history in the preceding 12 months before entering the study had numerically higher exacerbations rates during the study, irrespective of treatment arm or blood EOS level, compared with those without an exacerbation history in the preceding 12 months.

Current guidance in Japan recommends ICS/LAMA/LABA triple therapy in patients with COPD and no diagnosis of asthma who experience frequent exacerbations and have blood EOS count ≥ 300 cells/mm 3 , and in patients with COPD and features of asthma with continuing/worsening symptoms despite receiving dual ICS/LABA therapy [ 6 ]. Our analyses suggest BGF has beneficial effects on lung function versus dual ICS/LABA therapy and on moderate or severe exacerbation rates versus dual LAMA/LABA therapy in patients with and without recent exacerbation histories and among those with moderate and severe COPD who have blood EOS count 100 to < 300 cells/mm 3 . Similar results were generally observed for both exacerbation history and COPD severity in supportive analyses of patients with blood EOS count ≥ 100 cells/mm 3 (i.e., when patients with blood EOS count > 300 cells/mm 3 were included; BGF, n  = 55; GFF, n  = 56; BFF, n  = 32). However, treatment differences on exacerbation rate reductions for BGF versus GFF did appear more robust in this subgroup in some instances, with beneficial effects observed in those with and without exacerbation histories. This is expected since a threshold of blood EOS count > 300 cells/mm 3 identifies patients most likely to benefit from ICS [ 2 ].

ICS withdrawal has been raised as a concern in triple therapy studies among participants previously treated with an ICS who discontinued ICS following randomization to a non-ICS containing treatment arm [ 23 ]. In this regard, it is possible that those patients randomized to LAMA/LABA with GFF might have exhibited increased exacerbation rates due to removal of the ICS treatment component. However, a previously published post-hoc analysis of the ETHOS study, which examined the relationship between prior ICS use and benefits of BGF on exacerbations, symptoms, health-related quality of life, and lung function in patients with COPD, indicated there are benefits of BGF versus GFF regardless of ICS use within the 30 days before screening [ 24 ], suggesting ICS withdrawal may not account for the current findings.

Though the current findings seem to suggest benefits of ICS-containing triple therapy versus dual therapy on lung function and exacerbations, observations from a real-world observational study of triple therapy in COPD among ICS-naive patients highlight that triple therapy may have potential negative impacts, including increased incidence of severe pneumonia [ 25 ]. Other studies have also reported increased risk of other respiratory infections and pneumonia associated with ICS [ 26 , 27 , 28 ]. This emphasizes the importance of tailoring treatment plans to individual patient needs.

A few limitations of these analyses should be considered when interpreting these results. As the KRONOS study was not prospectively powered for any of the reported post-hoc analyses, reported P -values are nominal, unadjusted for multiplicity, and provided for descriptive purposes only. In addition, 74% of patients had no exacerbations in the last 12 months in the KRONOS study [ 9 ]. As such, sample sizes for post-hoc analyses of patients with an exacerbation history in the preceding 12 months were relatively small and subject to greater levels of variability. However, as the most compelling and clinically relevant findings from the perspective of current treatment guidelines relate to triple therapy use in patients without exacerbation history in the preceding 12 months, this limitation is not considered to be critical. It should be acknowledged that exacerbations are not a stable phenotype. Even though previous reports suggest the most important determinant and the singular predictive tool of frequent exacerbations is a history of exacerbations [ 29 ], there also patients who experience exacerbations in the previous year who do not experience exacerbations in the following year [ 29 ]. Therefore, when considering the exacerbation-suppressing effects of drug interventions, it is essential to consider the possibility some patients might not have experienced exacerbations even without drug intervention.

In post-hoc analyses of patients with moderate-to-very severe COPD from the KRONOS study, benefits of ICS/LAMA/LABA triple therapy with BGF were observed for lung function versus dual ICS/LABA therapy, and for exacerbation rates versus dual LAMA/LABA therapy in patients with blood EOS count 100 to < 300 cells/mm 3 who had less severe disease and no history of exacerbations in the last 12 months. Taken together, these data may suggest patients with blood EOS count > 100 cells/mm 3 without a recent history of exacerbations and those with moderate disease could benefit from ICS/LAMA/LABA triple therapy with BGF relative to dual therapy with ICS/LABA or LAMA/LABA. Therefore, clinicians should consider a step-up to triple therapy in patients with persistent/worsening symptoms whose blood EOS count is ≥ 100 cells/mm 3 , even if overall disease severity is moderate and there is no recent history of exacerbations. However, these findings require confirmation in adequately controlled studies that are statistically powered to assess these endpoints.

Data availability

Data underlying the findings described in this manuscript may be obtained in accordance with AstraZeneca’s data sharing policy described at https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure. Data for studies directly listed on Vivli can be requested through Vivli at www.vivli.org. Data for studies not listed on Vivli could be requested through Vivli at https://vivli.org/members/enquiries-about-studies-not-listed-on-the-vivli-platform/. The AstraZeneca Vivli member page is also available outlining further details: https://vivli.org/ourmember/astrazeneca/.

Abbreviations

Budesonide/formoterol fumarate dihydrate

Budesonide/glycopyrronium/formoterol fumarate dihydrate

Budesonide/formoterol fumarate dihydrate (via dry-powder inhaler)

Confidence interval

Chronic obstructive pulmonary disease

Forced expiratory volume in 1 s

Global initiative for Chronic Obstructive Lung disease

Glycopyrronium/formoterol fumarate dihydrate

Inhaled corticosteroid

Long-acting β 2 -agonist

Long-acting muscarinic antagonist

Least squares

Modified intention-to-treat

Standard error

Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, Abraham J, Adair T, Aggarwal R, Ahn SY, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the global burden of Disease Study 2010. Lancet. 2012;380:2095–128.

Article   PubMed   PubMed Central   Google Scholar  

Global Initiative for Chronic Obstructive Lung Disease. 2023 GOLD Report. Global strategy for prevention, diagnosis and management of COPD. https://goldcopd.org/2023-gold-report-2/ . Accessed 20 Nov 2023.

AstraZeneca Pharmaceuticals LP. Breztri Aerosphere™ Prescribing Information. https://www.azpicentral.com/breztri/breztri.pdf . Access 13 June 2023. 2020.

Chiesi Limited. Trimbow 87 µg/5 µg/9 µg pressurised inhalation solution SmPC. https://www.medicines.org.uk/emc/product/761/smpc . Access 1 March 2024. 2022.

US Food and Drug Administration. Trelegy Ellipta. Highlights of prescribing information. https://www.accessdata.fda.gov/drugsatfda_docs/label/2017/209482s000lbl.pdf . Access 1 March 2024. 2017.

Shibata Y, Kawayama T, Muro S, Sugiura H, members of Japanese Respiratory Society COPD Guideline 6th Edition Editing Committee. Management goals and stable phase management of patients with chronic obstructive pulmonary disease in the Japanese Respiratory Society guideline for the management of chronic obstructive pulmonary disease 2022 (6th edition). Respir Investig, : 2023, 61:773–780.

Rabe KF, Martinez FJ, Ferguson GT, Wang C, Singh D, Wedzicha JA, Trivedi R, St Rose E, Ballal S, McLaren J, et al. Triple inhaled therapy at two glucocorticoid doses in moderate-to-very-severe COPD. N Engl J Med. 2020;383:35–48.

Article   CAS   PubMed   Google Scholar  

Rabe KF, Martinez FJ, Singh D, Trivedi R, Jenkins M, Darken P, Aurivillius M, Dorinsky P. Improvements in lung function with budesonide/glycopyrrolate/formoterol fumarate metered dose inhaler versus dual therapies in patients with COPD: a sub-study of the ETHOS trial. Ther Adv Respir Dis. 2021;15:17534666211034329.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Ferguson GT, Rabe KF, Martinez FJ, Fabbri LM, Wang C, Ichinose M, Bourne E, Ballal S, Darken P, DeAngelis K, et al. Triple therapy with budesonide/glycopyrrolate/formoterol fumarate with co-suspension delivery technology versus dual therapies in chronic obstructive pulmonary disease (KRONOS): a double-blind, parallel-group, multicentre, phase 3 randomised controlled trial. Lancet Respir Med. 2018;6:747–58.

Bafadhel M, Rabe KF, Martinez FJ, Singh D, Darken P, Jenkins M, Aurivillius M, Patel M, Dorinsky P. Benefits of budesonide/glycopyrronium/formoterol fumarate dihydrate on COPD exacerbations, lung function, symptoms, and quality of life across blood eosinophil ranges: a post-hoc analysis of data from ETHOS. Int J Chron Obstruct Pulmon Dis. 2022;17:3061–73.

Muro S, Sugiura H, Darken P, Dorinsky P. Efficacy of budesonide/glycopyrronium/formoterol metered dose inhaler in patients with COPD: post-hoc analysis from the KRONOS study excluding patients with airway reversibility and high eosinophil counts. Respir Res. 2021;22:187.

Ichinose M, Fukushima Y, Inoue Y, Hataji O, Ferguson GT, Rabe KF, Hayashi N, Okada H, Takikawa M, Bourne E, et al. Efficacy and safety of budesonide/glycopyrrolate/formoterol fumarate metered dose inhaler formulated using co-suspension delivery technology in Japanese patients with COPD: a subgroup analysis of the KRONOS study. Int J Chron Obstruct Pulmon Dis. 2019;14:2979–91.

Celli BR, MacNee W, ATS/ERS Task Force. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J. 2004;23:932–46.

The Japanese Respiratory Society. Guidelines for the diagnosis and treatment of COPD, 4th Edition (in Japanese). http://www.jrs.or.jp/ . Access 9 April 2019. 2013.

Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general U.S. population. Am J Respir Crit Care Med. 1999;159:179–87.

Sasaki H, Nakamura M, Kida K, Kambe M, Takahashi K, Fujimura M, Sakakibara H, Horie T, Nishimura M, Takagi K, et al. Reference values for spirogram and blood gas analysis in Japanese adults. J Jpn Respir Soc. 2001;39:S1–17.

Google Scholar  

Hankinson JL, Kawut SM, Shahar E, Smith LJ, Stukovsky KH, Barr RG. Performance of American thoracic society-recommended spirometry reference values in a multiethnic sample of adults: the multi-ethnic study of atherosclerosis (MESA) lung study. Chest. 2010;137:138–45.

Article   PubMed   Google Scholar  

Bafadhel M, Peterson S, De Blas MA, Calverley PM, Rennard SI, Richter K, Fagerås M. Predictors of exacerbation risk and response to budesonide in patients with chronic obstructive pulmonary disease: a post-hoc analysis of three randomised trials. Lancet Respir Med. 2018;6:117–26.

Siddiqui SH, Guasconi A, Vestbo J, Jones P, Agusti A, Paggiaro P, Wedzicha JA, Singh D. Blood eosinophils: a biomarker of response to extrafine beclomethasone/formoterol in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2015;192:523–5.

Pascoe S, Barnes N, Brusselle G, Compton C, Criner GJ, Dransfield MT, Halpin DMG, Han MK, Hartley B, Lange P, et al. Blood eosinophils and treatment response with triple and dual combination therapy in chronic obstructive pulmonary disease: analysis of the IMPACT trial. Lancet Respir Med. 2019;7:745–56.

Martinez FJ, Ferguson GT, Bourne E, Ballal S, Darken P, Aurivillius M, Dorinsky P, Reisner C. Budesonide/glycopyrrolate/formoterol fumarate metered dose inhaler improves exacerbation outcomes in patients with COPD without a recent exacerbation history: a subgroup analysis of KRONOS. Int J Chron Obstruct Pulmon Dis. 2021;16:179–89.

Whittaker H, Rubino A, Müllerová H, Morris T, Varghese P, Xu Y, De Nigris E, Quint JK. Frequency and severity of exacerbations of COPD associated with future risk of exacerbations and mortality: a UK routine health care data study. Int J Chron Obstruct Pulmon Dis. 2022;17:427–37.

Suissa S, Dell’Aniello S, Ernst P. Triple inhaler versus dual bronchodilator therapy in COPD: real-world effectiveness on mortality. COPD. 2022;19:1–9.

Singh D, Rabe KF, Martinez FJ, Krüll M, Jenkins M, Patel M, Dorinsky P. Relationship between prior inhaled corticosteroid use and benefits of budesonide/glycopyrronium/formoterol fumarate dihydrate on exacerbations, symptoms, health-related quality of life, and lung function in patients with chronic obstructive pulmonary disease: analyses from the ETHOS study. Respir Med. 2022;197:106857.

Suissa S, Dell’Aniello S, Ernst P. Single-inhaler triple versus dual bronchodilator therapy in COPD: real-world comparative effectiveness and safety. Int J Chron Obstruct Pulmon Dis. 2022;17:1975–86.

Oishi K, Matsunaga K, Yamamoto T, Matsuda K, Murata Y, Hirano T. Practical recommendations for a selection of inhaled corticosteroids in COPD: a composite ICO chart. Biomolecules 2023, 13.

Quint JK, Ariel A, Barnes PJ. Rational use of inhaled corticosteroids for the treatment of COPD. NPJ Prim Care Respir Med. 2023;33:27.

Johnsen RH, Heerfordt CK, Boel JB, Dessau RB, Ostergaard C, Sivapalan P, Eklof J, Jensen JS. Inhaled corticosteroids and risk of lower respiratory tract infection with Moraxella catarrhalis in patients with chronic obstructive pulmonary disease. BMJ Open Respir Res 2023, 10.

Hurst JR, Vestbo J, Anzueto A, Locantore N, Müllerová H, Tal-Singer R, Miller B, Lomas DA, Agusti A, Macnee W, et al. Susceptibility to exacerbation in chronic obstructive pulmonary disease. N Engl J Med. 2010;363:1128–38.

DeTora LM, Toroser D, Sykes A, Vanderlinden C, Plunkett FJ, Lane T, Hanekamp E, Dormer L, DiBiasi F, Bridges D, et al. Good publication practice (GPP) guidelines for company-sponsored biomedical research: 2022 update. Ann Intern Med. 2022;175:1298–304.

Download references

Acknowledgements

The authors would like to thank all the patients and the investigators of the KRONOS study. Medical writing support, under the direction of the authors, was provided by Stephanie Lee, MSc, CMC Connect, a division of IPG Health Medical Communications, funded by AstraZeneca, in accordance with Good Publication Practice (GPP 2022) guidelines [ 30 ]. The sponsor was involved in the study design; the collection, analysis, and interpretation of data; the writing of the report; and in the decision to submit the article for publication.

The KRONOS study was sponsored by AstraZeneca.

Author information

Authors and affiliations.

Department of Respiratory Medicine, Nara Medical University, 840 Shijo-cho, Nara, 634-8522, Japan

Shigeo Muro

Division of Respirology, Neurology and Rheumatology, Department of Medicine, Kurume University School of Medicine, Kurume, Japan

Tomotaka Kawayama

Department of Respiratory Medicine, Tohoku University Graduate School of Medicine, Sendai, Japan

Hisatoshi Sugiura

Respiratory Inhalation, Medical Department, AstraZeneca K.K. Kita-ku, Osaka, Japan

Munehiro Seki

Formerly of AstraZeneca, Durham, NC, USA

Elizabeth A. Duncan

Late RIA Biometrics, AstraZeneca, Gaithersburg, MD, USA

Karin Bowen

BioPharmaceuticals, R&I Medical, AstraZeneca, Cambridge, UK

Jonathan Marshall

Late RIA, BioPharmaceuticals R&D, AstraZeneca, Gaithersburg, MD, USA

Ayman Megally

Late RIA, BioPharmaceuticals R&D, AstraZeneca, Cambridge, UK

Mehul Patel

You can also search for this author in PubMed   Google Scholar

Contributions

SM, TK, HS, MS, JM, and MP are responsible for the conception of the analysis. MP, EAD, and AM contributed to formal analysis. SM and HS contributed to the investigation. SM, HS, and KB contributed to the study methodology. KB acquired resources for this analysis. SM, TK, HS, MS, MP, and KB supervised the analysis. SM, TK, HS, MS, MP, and AM validated the analysis. MP, EAD, and AM contributed to data visualization. SM, TK, HS, MS, MP, EAD, KB, JM, and AM critically reviewed and edited the manuscript.

Corresponding author

Correspondence to Shigeo Muro .

Ethics declarations

Ethics approval and consent to participate, consent for publication.

Not applicable.

Competing interests

SM has received lecture fees from AstraZeneca, GlaxoSmithKline, Nippon Boehringer Ingelheim, and Novartis Pharma. TK has received grants from Helios co. Ltd. and lecture fees from AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, Kyorin, Novartis, Sanofi, and Teijin healthcare. HS has received lecture fees from AstraZeneca, GlaxoSmithKline, Nippon Boehringer Ingelheim, Novartis Pharma, and Sanofi. MS is an employee of AstraZeneca K.K. Kita-ku and owns stock and/or stock options in the company. EAD is a former employee of AstraZeneca and owns stock and/or stock options in the company. KB, JM, AM, and MP are employees of AstraZeneca and own stock and/or stock options in the company.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Muro, S., Kawayama, T., Sugiura, H. et al. Benefits of budesonide/glycopyrronium/formoterol fumarate dihydrate on lung function and exacerbations of COPD: a post-hoc analysis of the KRONOS study by blood eosinophil level and exacerbation history. Respir Res 25 , 297 (2024). https://doi.org/10.1186/s12931-024-02918-8

Download citation

Received : 28 March 2024

Accepted : 17 July 2024

Published : 05 August 2024

DOI : https://doi.org/10.1186/s12931-024-02918-8

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Blood eosinophils
• Budesonide/glycopyrronium/formoterol fumarate dihydrate (BGF)
• Chronic obstructive pulmonary disease (COPD)
• Disease severity
• Fixed-dose triple therapy

Respiratory Research

ISSN: 1465-993X

• General enquiries: [email protected]

• Download PDF
• Share X Facebook Email LinkedIn
• Permissions

Rape-Related Pregnancies in the 14 US States With Total Abortion Bans

• 1 Planned Parenthood of Montana, Billings, Montana
• 2 Resound Research for Reproductive Health, Austin, Texas
• 3 Hunter College, City University of New York, New York
• 4 Department of Medicine, Cambridge Health Alliance, Cambridge, Massachusetts
• 5 Department of Medicine, University of California, San Francisco
• Editor's Note Access to Safe Abortion for Survivors of Rape Deborah Grady, MD, MPH; Sharon K. Inouye, MD, MPH; Mitchell H. Katz, MD JAMA Internal Medicine
• Medical News in Brief 65 000 Rape-Related Pregnancies Took Place in US States With Abortion Bans Emily Harris JAMA
• Correction Error in Methods, Results, and Table 2 JAMA Internal Medicine

Many US women report experiencing sexual violence, and many seek abortion for rape-related pregnancies. 1 Following the US Supreme Court’s 2022 Dobbs v Jackson Women’s Health Organization ( Dobbs ) decision overturning Roe v Wade , 14 states have outlawed abortion at any gestational duration. 2 Although 5 of these states allow exceptions for rape-related pregnancies, stringent gestational duration limits apply, and survivors must report the rape to law enforcement, a requirement likely to disqualify most survivors of rape, of whom only 21% report their rape to police. 3

• Editor's Note Access to Safe Abortion for Survivors of Rape JAMA Internal Medicine

Read More About

Dickman SL , White K , Himmelstein DU , Lupez E , Schrier E , Woolhandler S. Rape-Related Pregnancies in the 14 US States With Total Abortion Bans. JAMA Intern Med. 2024;184(3):330–332. doi:10.1001/jamainternmed.2024.0014

Manage citations:

© 2024

Artificial Intelligence Resource Center

Best of JAMA Network 2022

Browse and subscribe to JAMA Network podcasts!

Others Also Liked

Select your interests.

Customize your JAMA Network experience by selecting one or more topics from the list below.

• Academic Medicine
• Acid Base, Electrolytes, Fluids
• Allergy and Clinical Immunology
• American Indian or Alaska Natives
• Anesthesiology
• Anticoagulation
• Art and Images in Psychiatry
• Artificial Intelligence
• Assisted Reproduction
• Bleeding and Transfusion
• Caring for the Critically Ill Patient
• Challenges in Clinical Electrocardiography
• Climate and Health
• Climate Change
• Clinical Challenge
• Clinical Decision Support
• Clinical Implications of Basic Neuroscience
• Clinical Pharmacy and Pharmacology
• Complementary and Alternative Medicine
• Consensus Statements
• Coronavirus (COVID-19)
• Critical Care Medicine
• Cultural Competency
• Dental Medicine
• Dermatology
• Diabetes and Endocrinology
• Diagnostic Test Interpretation
• Drug Development
• Electronic Health Records
• Emergency Medicine
• End of Life, Hospice, Palliative Care
• Environmental Health
• Equity, Diversity, and Inclusion
• Facial Plastic Surgery
• Gastroenterology and Hepatology
• Genetics and Genomics
• Genomics and Precision Health
• Global Health
• Guide to Statistics and Methods
• Hair Disorders
• Health Care Delivery Models
• Health Care Economics, Insurance, Payment
• Health Care Quality
• Health Care Reform
• Health Care Safety
• Health Care Workforce
• Health Disparities
• Health Inequities
• Health Policy
• Health Systems Science
• History of Medicine
• Hypertension
• Images in Neurology
• Implementation Science
• Infectious Diseases
• Innovations in Health Care Delivery
• JAMA Infographic
• Law and Medicine
• Leading Change
• Less is More
• LGBTQIA Medicine
• Lifestyle Behaviors
• Medical Coding
• Medical Devices and Equipment
• Medical Education
• Medical Education and Training
• Medical Journals and Publishing
• Mobile Health and Telemedicine
• Narrative Medicine
• Neuroscience and Psychiatry
• Notable Notes
• Nutrition, Obesity, Exercise
• Obstetrics and Gynecology
• Occupational Health
• Ophthalmology
• Orthopedics
• Otolaryngology
• Pain Medicine
• Palliative Care
• Pathology and Laboratory Medicine
• Patient Care
• Patient Information
• Performance Improvement
• Performance Measures
• Perioperative Care and Consultation
• Pharmacoeconomics
• Pharmacoepidemiology
• Pharmacogenetics
• Pharmacy and Clinical Pharmacology
• Physical Medicine and Rehabilitation
• Physical Therapy
• Physician Leadership
• Population Health
• Primary Care
• Professional Well-being
• Professionalism
• Psychiatry and Behavioral Health
• Public Health
• Pulmonary Medicine
• Regulatory Agencies
• Reproductive Health
• Research, Methods, Statistics
• Resuscitation
• Rheumatology
• Risk Management
• Scientific Discovery and the Future of Medicine
• Shared Decision Making and Communication
• Sleep Medicine
• Sports Medicine
• Stem Cell Transplantation
• Substance Use and Addiction Medicine
• Surgical Innovation
• Surgical Pearls
• Teachable Moment
• Technology and Finance
• The Art of JAMA
• The Arts and Medicine
• The Rational Clinical Examination
• Tobacco and e-Cigarettes
• Translational Medicine
• Trauma and Injury
• Treatment Adherence
• Ultrasonography
• Users' Guide to the Medical Literature
• Vaccination
• Venous Thromboembolism
• Veterans Health
• Women's Health
• Workflow and Process
• Wound Care, Infection, Healing
• Register for email alerts with links to free full-text articles
• Access PDFs of free articles
• Manage your interests
• Save searches and receive search alerts

Chuck Todd: Republicans are stuck with Trump's obsessions

To say the current political environment is a moving target is an understatement. A steady state through the first six months of 2024 has turned into anything but since that fateful debate night in June. And if you think trying to steer a presidential campaign in this unsteady environment is hard, try running a House or Senate race.

A month ago, it was panicking Democrats down the ballot who were pushing hardest to persuade President Joe Biden to drop out. Biden said so himself in an interview with CBS News over the weekend.

“Polls we had showed that it was a neck-and-neck race; it would have been down to wire,” Biden said. “But what happened was a number of my Democratic colleagues in the House and Senate thought that I was going to hurt them in their races. And I was concerned if I stayed in the race that would be the topic.”

This was about as transparent as anyone has been about the process to persuade Biden to step down. But let’s explain more about what Biden meant by “that would be the topic.” What was made clear to him, during the three-week campaign to persuade him to step aside, was that every battleground House and Senate Democrat was likely to be forced into the uncomfortable position of tossing their presidential standard-bearer overboard. That’s because Republicans were telegraphing in the early aftermath of that June debate debacle that every Democrat was going to have to answer for “hiding” Biden’s diminished abilities.

This digital ad posted by Republican Senate candidate Dave McCormick in Pennsylvania, about a week before Biden eventually dropped his bid, was the blueprint. It was a compilation of Democratic Sen. Bob Casey’s defending Biden’s abilities to serve another four years, as well as establishing Casey’s close ties to the then-very unpopular Biden.

Incumbent Democrats were bracing for more ads just like the one above all over the House and Senate battlegrounds.

Given his lack of a large, devoted, personal political base, Biden realized he was a man on an island. Despite his belief that there was enough anti-Trump sentiment to power him to victory, if the rest of those on the Democratic ticket were running away from him and styling themselves as a “check on Trump” ahead of GOP victory, it would have become self-fulfilling for voters to decide to go against him.

So Biden chose the most rational path out of the conundrum he put his party in: He stepped off the stage.

Almost immediately, Democrats saw a surge of support — mostly from disillusioned Democratic voters. And that renewed relief among the rank and file has expressed itself in polls of not just the presidential race but down-ballot campaigns, as well.

So far, Vice President Kamala Harris has turned that expression of relief into real momentum when it comes to money and resources, which has also trickled down the ballot.

Nearly a month into this new Democratic reality, the question of who has the weaker presidential nominee for down-ballot candidates is back up for debate. As of this writing, it’s now Republicans in battleground states and districts who are begging their nominee to change course. And so far, former President Donald Trump isn’t listening.

A good example of this gentle public nudging of Trump to change his approach comes from Vivek Ramaswamy, the former presidential candidate who has been making the media rounds calling for a more focused and even subdued Trump campaign. Here’s his most recent public appearance, via NPR , responding to the question about how Trump should pivot:

“Who’s going to secure the border, who’s going to grow the economy, who’s going to stay out of World War III? And more intangibly ... who’s going to restore national pride in this country? I think Donald Trump has a strong case on all of those counts, and I think he and the Republican Party would be well served to focus on the policy contrasts.”

And yet, in just about every public forum Trump has been involved in since Harris’ ascension, he has been incapable of pivoting to a message that’s anywhere close to the advice from Ramaswamy.

This wouldn’t be hard for a normal candidate. And yet for Trump, it has been quite hard. In talking with folks who know him well, it’s clear he can’t get over losing to Biden in 2020. His only personal hope of moving past the Biden defeat in 2020 was to defeat him in 2024. Now, Biden has denied him that opportunity.

He also doesn’t seem to respect Harris as an opponent, and voters are picking up on that. Disdain from one candidate can turn swing voters against that campaign and toward the target. The best example of that? Hillary Clinton, who clearly never viewed Trump as a worthy opponent in 2016. The “deplorables” comment only reinforced that view.

Well, Trump is falling into the same trap. It’s possible the less respect he shows for Harris, the less scared the public will be to give her a chance.

Now, some might argue that if he can’t get over that and can’t pivot, then maybe he should walk away from the nomination like Biden. Putting aside the fact that Republicans have already had their convention for a moment, it’s fairly obvious that if the GOP’s sole focus was to deny the Democrats the White House, the best course of action would be to persuade Trump to step aside and nominate the primary runner-up, Nikki Haley.

But as obvious as that might be to a political scientist, the reality of the GOP is quite different. Since Trump’s shock election in 2016, the Democratic Party has been organized around, basically, one principle: stopping Trump. It has allowed the party to paper over plenty of ideological differences for nearly a decade.

As for the GOP, Trump and his allies have done everything they can to mold the GOP in Trump’s image, and anyone who doesn’t share that view has essentially been purged. Even if down-ballot GOP candidates wanted to push Trump out of the party, the rank-and-file Republican voters wouldn’t acquiesce the way Democrats just did. In fact, it was more than acquiescing: Rank-and-file Democrats joined with Democratic elites in a primal scream to get Biden to step aside. (See the polls that showed less than half of Democrats being “satisfied” with or enthusiastic about Biden as the nominee long before the debate.)

All of this explains why Republicans, especially those invested in trying to flip the Senate or hold the House, are begging Trump in public to pivot from a grievance-focused campaign to a policy-focused one. These Republicans know they’d be in a better position if they had a nominee who was future-oriented rather than past-obsessed.

But what happens if Trump never pivots? Then what do these down-ballot Republicans do? 

In 2016, many down-ballot Republicans were comfortable running against Trump or simply ignoring him, both because they (and many Republican voters) were skeptical of him and because most Republicans — let alone Democrats and the media — thought he wouldn’t win. All of that created a permission structure that allowed for Republicans like then-Sen. Pat Toomey of Pennsylvania to run his own race. He won re-election without having to endorse Trump or appear with him.

No battleground-state Republican could do what Toomey did in 2016 and win in 2024. Toomey felt no pressure to appear with Trump at any Pennsylvania rallies in 2016, but the same can’t be said for McCormick, this year’s GOP Senate nominee in the state. If you doubt me, let me remind you what Trump did in Georgia just a few weeks ago, when he attacked Republican Gov. Brian Kemp for refusing to use (or abuse) his powers as governor to assist in his efforts to question the results in 2020. 

The last thing Republican nominees in key Senate races want in 2024 is Trump dogging them in public (which he would do) for not showing up at his rallies. Perhaps the only Republicans who might get a pass for that are former Gov. Larry Hogan in Maryland or maybe Nella Domenici, the GOP nominee for the Senate in New Mexico, where her name is synonymous with Republican politics. 

The point is that while Democratic base voters wouldn’t punish a Democrat from running away from their presidential nominee (have you noticed how well Sherrod Brown is holding up in Ohio while he makes a show of skipping the Democratic convention?), there’s plenty of evidence GOP base voters would do that. 

This all brings me back to the question: What do Republican down-ballot candidates do if Trump never fully pivots? At some point, the obvious answer would be to run a campaign based on being a “check” on a Harris presidency, similar to what Republicans ran in 2016 against Clinton. But executing that campaign wouldn’t be easy. Just one stray public comment from an anonymous staffer of some campaign that starts to pivot away from Trump, and he could rain you-know-what down on said campaign.

Among all the competitive battleground Senate races, I’d argue McCormick in Pennsylvania and former Rep. Mike Rogers of Michigan are in the most precarious positions when it comes to the need for Trump to pivot. So keep an eye on the messages those two campaigns develop. The swing voters who don’t like Trump personally are the voters Rogers and McCormick need. Can they find a way to message to them without alienating Trump — especially if Trump, himself, can’t seem to make the obvious pivot any generic candidate in his position would be making?

Of course, the real challenge for both Republicans will be figuring out how to navigate Trump’s appearances in their respective states. If Trump’s numbers continue to languish, it’s going to be tempting for those candidates to find some distance from him. Ultimately, it might be a fool’s errand, even if the reality is staring the campaign in the face. After all, there’s only so much a down-ballot candidate can do if the top of the ticket is unpopular or undisciplined.

Chuck Todd is NBC News' chief political analyst and the former moderator of " Meet The Press ."