review of literature for ratio analysis

Have a language expert improve your writing

Run a free plagiarism check in 10 minutes, generate accurate citations for free.

• Knowledge Base

Methodology

• How to Write a Literature Review | Guide, Examples, & Templates

How to Write a Literature Review | Guide, Examples, & Templates

Published on January 2, 2023 by Shona McCombes . Revised on September 11, 2023.

What is a literature review? A literature review is a survey of scholarly sources on a specific topic. It provides an overview of current knowledge, allowing you to identify relevant theories, methods, and gaps in the existing research that you can later apply to your paper, thesis, or dissertation topic .

There are five key steps to writing a literature review:

• Search for relevant literature
• Evaluate sources
• Identify themes, debates, and gaps
• Outline the structure
• Write your literature review

A good literature review doesn’t just summarize sources—it analyzes, synthesizes , and critically evaluates to give a clear picture of the state of knowledge on the subject.

Table of contents

What is the purpose of a literature review, examples of literature reviews, step 1 – search for relevant literature, step 2 – evaluate and select sources, step 3 – identify themes, debates, and gaps, step 4 – outline your literature review’s structure, step 5 – write your literature review, free lecture slides, other interesting articles, frequently asked questions, introduction.

• Quick Run-through
• Step 1 & 2

When you write a thesis , dissertation , or research paper , you will likely have to conduct a literature review to situate your research within existing knowledge. The literature review gives you a chance to:

• Demonstrate your familiarity with the topic and its scholarly context
• Develop a theoretical framework and methodology for your research
• Position your work in relation to other researchers and theorists
• Show how your research addresses a gap or contributes to a debate
• Evaluate the current state of research and demonstrate your knowledge of the scholarly debates around your topic.

Writing literature reviews is a particularly important skill if you want to apply for graduate school or pursue a career in research. We’ve written a step-by-step guide that you can follow below.

A faster, more affordable way to improve your paper

Scribbr’s new AI Proofreader checks your document and corrects spelling, grammar, and punctuation mistakes with near-human accuracy and the efficiency of AI!

Proofread my paper

Writing literature reviews can be quite challenging! A good starting point could be to look at some examples, depending on what kind of literature review you’d like to write.

• Example literature review #1: “Why Do People Migrate? A Review of the Theoretical Literature” ( Theoretical literature review about the development of economic migration theory from the 1950s to today.)
• Example literature review #2: “Literature review as a research methodology: An overview and guidelines” ( Methodological literature review about interdisciplinary knowledge acquisition and production.)
• Example literature review #3: “The Use of Technology in English Language Learning: A Literature Review” ( Thematic literature review about the effects of technology on language acquisition.)
• Example literature review #4: “Learners’ Listening Comprehension Difficulties in English Language Learning: A Literature Review” ( Chronological literature review about how the concept of listening skills has changed over time.)

You can also check out our templates with literature review examples and sample outlines at the links below.

Download Word doc Download Google doc

Before you begin searching for literature, you need a clearly defined topic .

If you are writing the literature review section of a dissertation or research paper, you will search for literature related to your research problem and questions .

Make a list of keywords

Start by creating a list of keywords related to your research question. Include each of the key concepts or variables you’re interested in, and list any synonyms and related terms. You can add to this list as you discover new keywords in the process of your literature search.

• Social media, Facebook, Instagram, Twitter, Snapchat, TikTok
• Body image, self-perception, self-esteem, mental health
• Generation Z, teenagers, adolescents, youth

Search for relevant sources

Use your keywords to begin searching for sources. Some useful databases to search for journals and articles include:

• Your university’s library catalogue
• Google Scholar
• Project Muse (humanities and social sciences)
• Medline (life sciences and biomedicine)
• EconLit (economics)
• Inspec (physics, engineering and computer science)

You can also use boolean operators to help narrow down your search.

Make sure to read the abstract to find out whether an article is relevant to your question. When you find a useful book or article, you can check the bibliography to find other relevant sources.

You likely won’t be able to read absolutely everything that has been written on your topic, so it will be necessary to evaluate which sources are most relevant to your research question.

For each publication, ask yourself:

• What question or problem is the author addressing?
• What are the key concepts and how are they defined?
• What are the key theories, models, and methods?
• Does the research use established frameworks or take an innovative approach?
• What are the results and conclusions of the study?
• How does the publication relate to other literature in the field? Does it confirm, add to, or challenge established knowledge?
• What are the strengths and weaknesses of the research?

Make sure the sources you use are credible , and make sure you read any landmark studies and major theories in your field of research.

You can use our template to summarize and evaluate sources you’re thinking about using. Click on either button below to download.

Take notes and cite your sources

As you read, you should also begin the writing process. Take notes that you can later incorporate into the text of your literature review.

It is important to keep track of your sources with citations to avoid plagiarism . It can be helpful to make an annotated bibliography , where you compile full citation information and write a paragraph of summary and analysis for each source. This helps you remember what you read and saves time later in the process.

Receive feedback on language, structure, and formatting

Professional editors proofread and edit your paper by focusing on:

• Academic style
• Vague sentences
• Style consistency

See an example

To begin organizing your literature review’s argument and structure, be sure you understand the connections and relationships between the sources you’ve read. Based on your reading and notes, you can look for:

• Trends and patterns (in theory, method or results): do certain approaches become more or less popular over time?
• Themes: what questions or concepts recur across the literature?
• Debates, conflicts and contradictions: where do sources disagree?
• Pivotal publications: are there any influential theories or studies that changed the direction of the field?
• Gaps: what is missing from the literature? Are there weaknesses that need to be addressed?

This step will help you work out the structure of your literature review and (if applicable) show how your own research will contribute to existing knowledge.

• Most research has focused on young women.
• There is an increasing interest in the visual aspects of social media.
• But there is still a lack of robust research on highly visual platforms like Instagram and Snapchat—this is a gap that you could address in your own research.

There are various approaches to organizing the body of a literature review. Depending on the length of your literature review, you can combine several of these strategies (for example, your overall structure might be thematic, but each theme is discussed chronologically).

Chronological

The simplest approach is to trace the development of the topic over time. However, if you choose this strategy, be careful to avoid simply listing and summarizing sources in order.

Try to analyze patterns, turning points and key debates that have shaped the direction of the field. Give your interpretation of how and why certain developments occurred.

If you have found some recurring central themes, you can organize your literature review into subsections that address different aspects of the topic.

For example, if you are reviewing literature about inequalities in migrant health outcomes, key themes might include healthcare policy, language barriers, cultural attitudes, legal status, and economic access.

Methodological

If you draw your sources from different disciplines or fields that use a variety of research methods , you might want to compare the results and conclusions that emerge from different approaches. For example:

• Look at what results have emerged in qualitative versus quantitative research
• Discuss how the topic has been approached by empirical versus theoretical scholarship
• Divide the literature into sociological, historical, and cultural sources

Theoretical

A literature review is often the foundation for a theoretical framework . You can use it to discuss various theories, models, and definitions of key concepts.

You might argue for the relevance of a specific theoretical approach, or combine various theoretical concepts to create a framework for your research.

Like any other academic text , your literature review should have an introduction , a main body, and a conclusion . What you include in each depends on the objective of your literature review.

The introduction should clearly establish the focus and purpose of the literature review.

Depending on the length of your literature review, you might want to divide the body into subsections. You can use a subheading for each theme, time period, or methodological approach.

As you write, you can follow these tips:

• Summarize and synthesize: give an overview of the main points of each source and combine them into a coherent whole
• Analyze and interpret: don’t just paraphrase other researchers — add your own interpretations where possible, discussing the significance of findings in relation to the literature as a whole
• Critically evaluate: mention the strengths and weaknesses of your sources
• Write in well-structured paragraphs: use transition words and topic sentences to draw connections, comparisons and contrasts

In the conclusion, you should summarize the key findings you have taken from the literature and emphasize their significance.

When you’ve finished writing and revising your literature review, don’t forget to proofread thoroughly before submitting. Not a language expert? Check out Scribbr’s professional proofreading services !

This article has been adapted into lecture slides that you can use to teach your students about writing a literature review.

Scribbr slides are free to use, customize, and distribute for educational purposes.

Open Google Slides Download PowerPoint

If you want to know more about the research process , methodology , research bias , or statistics , make sure to check out some of our other articles with explanations and examples.

• Sampling methods
• Simple random sampling
• Stratified sampling
• Cluster sampling
• Likert scales
• Reproducibility

 Statistics

• Null hypothesis
• Statistical power
• Probability distribution
• Effect size
• Poisson distribution

Research bias

• Optimism bias
• Cognitive bias
• Implicit bias
• Hawthorne effect
• Anchoring bias
• Explicit bias

A literature review is a survey of scholarly sources (such as books, journal articles, and theses) related to a specific topic or research question .

It is often written as part of a thesis, dissertation , or research paper , in order to situate your work in relation to existing knowledge.

There are several reasons to conduct a literature review at the beginning of a research project:

• To familiarize yourself with the current state of knowledge on your topic
• To ensure that you’re not just repeating what others have already done
• To identify gaps in knowledge and unresolved problems that your research can address
• To develop your theoretical framework and methodology
• To provide an overview of the key findings and debates on the topic

Writing the literature review shows your reader how your work relates to existing research and what new insights it will contribute.

The literature review usually comes near the beginning of your thesis or dissertation . After the introduction , it grounds your research in a scholarly field and leads directly to your theoretical framework or methodology .

A literature review is a survey of credible sources on a topic, often used in dissertations , theses, and research papers . Literature reviews give an overview of knowledge on a subject, helping you identify relevant theories and methods, as well as gaps in existing research. Literature reviews are set up similarly to other  academic texts , with an introduction , a main body, and a conclusion .

An  annotated bibliography is a list of  source references that has a short description (called an annotation ) for each of the sources. It is often assigned as part of the research process for a  paper .  

Cite this Scribbr article

If you want to cite this source, you can copy and paste the citation or click the “Cite this Scribbr article” button to automatically add the citation to our free Citation Generator.

McCombes, S. (2023, September 11). How to Write a Literature Review | Guide, Examples, & Templates. Scribbr. Retrieved November 29, 2023, from https://www.scribbr.com/dissertation/literature-review/

Is this article helpful?

Shona McCombes

Other students also liked, what is a theoretical framework | guide to organizing, what is a research methodology | steps & tips, how to write a research proposal | examples & templates, what is your plagiarism score.

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
• 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: 28 November 2023

Meta-analysis of the role of neutrophil to lymphocyte ratio in neonatal sepsis

• Jingyang Chen 1 ,
• Sanaz Yasrebinia 2 ,
• Arshin Ghaedi 3 , 4 ,
• Monireh Khanzadeh 5 ,
• Stephan Quintin 6 ,
• Abeer Dagra 6 ,
• Rodeania Peart 6 ,
• Brandon Lucke-Wold 6 &
• Shokoufeh Khanzadeh 2  

BMC Infectious Diseases volume  23 , Article number:  837 ( 2023 ) Cite this article

207 Accesses

8 Altmetric

Metrics details

Introduction

The neutrophil to lymphocyte ratio (NLR), an inflammatory biomarker, measures innate-adaptive immune system balance. In this systematic review and meta-analysis, we aim to analyze the current literature to evaluate the diagnostic role of NLR in neonatal sepsis.

PubMed, Web of Science, and Scopus were used to conduct a systematic search for relevant publications published before May 14, 2022.

Thirty studies, including 2328 neonates with sepsis and 1800 neonates in the control group, were included in our meta-analysis. The results indicated that NLR is higher in neonates with sepsis compared to healthy controls (SMD = 1.81, 95% CI = 1.14–2.48, P -value < 0.001) in either prospective (SMD = 2.38, 95% CI = 1.40–3.35, P -value < 0.001) or retrospective studies (SMD = 0.87, 95% CI = 0.63–1.12, P -value < 0.001) with a pooled sensitivity of 79% (95% CI = 62–90%), and a pooled specificity of 91% (95% CI = 73–97%). Also, we found that NLR is higher in neonates with sepsis compared to those who were suspected of sepsis but eventually had negative blood cultures (SMD =1.99, 95% CI = 0.76–3.22, P -value = 0.002) with a pooled sensitivity of 0.79% (95% CI = 0.69–0.86%), and a pooled specificity of 73% (95% CI = 54–85%). In addition, neonates with sepsis had elevated levels of NLR compared to other ICU admitted neonates (SMD = 0.73, 95% CI = 0.63–0.84, P  < 0.001). The pooled sensitivity was 0.65 (95% CI, 0.55–0.80), and the pooled specificity was 0.80 (95% CI, 0.68–0.88).

Our findings support NLR as a promising biomarker that can be readily integrated into clinical settings to aid in diagnosing neonatal sepsis. As evidenced by our results, restoring balance to the innate and adaptive immune system may serve as attractive therapeutic targets. Theoretically, a reduction in NLR values could be used to measure therapeutic efficacy, reflecting the restoration of balance within these systems.

Peer Review reports

Neonatal sepsis is a bloodstream infection that affects newborn infants under 28 days. It is a leading cause of morbidity and mortality in these children [ 1 , 2 ]. There are around 2,200 instances of neonatal sepsis per 100,000 live births, with a death rate of 11 to 19 percent [ 3 ]. The mortality rate varies depending on factors such as birth weight, with rates of 50% in newborns with a birth weight of less than 1,500 g, 23.8% in those with a birth weight of 1,500–2,500 g, and 18.2% overall [ 4 ]. Neonatal sepsis has also several long-term health outcomes such as neurodevelopmental disabilities (like cerebral palsy, visual or hearing impairments, and cognitive problems), respiratory complications (like chronic lung disease and bronchopulmonary dysplasia), nutritional and growth issues, and immunological dysfunction [ 5 , 6 ].

Early-onset sepsis (EOS) and late-onset sepsis (LOS) are two types of neonatal sepsis. Sepsis in neonates that occurs before 72 h of life (some experts use 7 days) is referred to as EOS, while sepsis that occurs after 72 h of life is referred to as LOS [ 7 , 8 ].

The spread of infections from the female genitourinary system to the infant or fetus is the most common cause of EOS [ 1 , 2 , 7 , 9 ]. These infections can contaminate the amniotic fluid or may ascend the vaginal canal, cervix, and uterus. As they pass through the vaginal canal in pregnancy or upon delivery, neonates can become contaminated. Group B Streptococci (GBS) and E. coli are common bacterial infections associated with EOS [ 1 , 2 ].

In addition, birth before 37 weeks and protracted rupture of membranes are all maternal variables that enhance the risk of neonatal sepsis [ 3 ]. Delayed treatment of newborn sepsis is linked to a variety of adverse outcomes, including persistent lung illness and neurodevelopmental concerns [ 1 , 2 ]. Overuse of antibiotics, for prophylactic treatment in sepsis prevention, on the other hand, can raise the risk of severe candidiasis and multidrug-resistant organisms [ 1 , 2 , 7 , 9 ].

In contrast, LOS is commonly caused by pathogens in the environment following delivery, often originating from caregivers or healthcare workers [ 1 , 2 , 7 , 8 ]. Occasionally, it may stem from a delayed manifestation of an infection passed from the mother. Infants who undergo invasive procedures or have intravascular catheters are more susceptible to LOS [ 7 ]. Premature babies are at a greater risk of infection or sepsis and mortality than full-term infants [ 2 , 9 ]. EOS and LOS have some differences in terms of symptoms. EOS represent as respiratory distress, hypoglycemia, apnea, hypoglycemia, and lethargy [ 10 , 11 ]. On the other hand LOS represent as fever, poor feeding, lethargy, and respiratory distress [ 12 , 13 ].

Research indicates that E. coli is related to higher mortality rates compared to GBS [ 8 ]. A lower mortality rate has also been observed with early treatment of clinically suspected neonates [ 10 ].

The immature immune system is the primary cause of increased neonatal sepsis susceptibility [ 10 , 14 , 15 ]. Neutrophils, macrophages, and T lymphocytes cannot carry out a complete inflammatory response in newborns due to their immature function. Furthermore, newborns have a restricted quantity of immunoglobulins at birth [ 10 , 14 , 15 , 16 ]. The premature infant’s limited time in the uterus reduces the transfer of immune globulins to the fetus [ 1 , 2 ].

Several clinical investigations have recently established the efficacy of the neutrophil to lymphocyte ratio (NLR) in predicting newborn sepsis [ 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 ]. This paper provides a systematic review and meta-analysis of these studies.

Search strategy

The preferred reporting items for systematic reviews and meta-analyses (PRISMA) standard was followed for this meta-analysis. From conception through May 14, 2022, a systematic review was undertaken using ISI Web of Science, PubMed, and Scopus. The following key phrases were used in search strategies: ((neutrophil AND lymphocyte AND ratio) OR NLR) AND (Neonat* or infant) AND (sepsis OR septic OR bacteremia). There were no date or language restrictions. Furthermore, potential meta-analyses and reviews were manually reviewed to find any further relevant articles that would be appropriate for this study.

Inclusion and exclusion criteria

According to the PICOS (Participants, Intervention, Comparison, Outcomes, and Study Design) framework, two researchers independently discovered and chose the studies:

Population: Newborns with sepsis

Intervention: NLR

Control: newborns suspected of having sepsis who subsequently had negative blood cultures OR healthy newborns OR newborns admitted to the intensive care unit (ICU)

Outcomes: NLR’s ability as a diagnostic tool for neonatal sepsis

Study Design. case-control or cross-sectional publications

Only the most recent or thorough studies were chosen to represent the reduplicative patient group.

Data extraction

Two researchers separately gathered data from the included papers. We gathered the following information:

1) general characteristics: country, publication year, first author, study subjects, study design (retrospective or prospective); 2) sample size in case and control groups; 3) NLR values; 4) the number of controls and cases and their NLR values; 5) type of control group (neonates admitted to the ICU, healthy neonates, sepsis suspected neonates but with final negative blood culture results; 6) sepsis type: LOS, EOS, both; 7) cut-off value, specificity, and sensitivity of NLR. We contacted the corresponding authors of included articles when further information was needed.

Quality assessment

The quality evaluation of included papers was done with the assistance of the Newcastle-Ottawa Scale (NOS). This scale consists of three parts: comparability, selection, and outcome.

Statistical analysis

With the help of STATA application version 12.0 (StataCorp LP, College Station, TX, USA), the meta-analysis of the correlation between neonatal sepsis and NLR was conducted. NLR differences between cases and controls were measured using Standardized mean difference (SMD) with a 95% confidence interval (CI). We reported SMD to accommodate the differences in NLR measurement techniques across various studies. 95% CI was reported to show the likely range of effect sizes supported by the findings. I 2 metric and Q chi-square test were used to quantify the heterogeneity among studies. When the P value was lower than 0.05, and I 2 was more than 50%, we assumed considerable heterogeneity exists. If we detected significant heterogeneity, we used a random-effect model; otherwise, we used a fixed-effect model. To determine the diagnostic value of NLR for sepsis in newborns, we created a summary receiver operating characteristic (SROC) curve by the “metandi” command. The initial construction of the SROC curve involves plotting the sensitivity (which represents true positivity) and the complement of specificity (1 - specificity) for each study. This curve is widely employed to gauge the accuracy of diagnostic tests. Essentially, the closer the ROC curve gets to the upper left corner of the graph, the higher the test’s accuracy, as this position corresponds to a sensitivity of 1 and a false positive rate of 0 (equivalent to a specificity of 1). Ultimately, we used Egger’s test and funnel plot (visual inspection) to detect possible publication bias. In Egger’s test, we use linear regression to assess the association between the standardized effect estimates and the standard error (SE); so, P -Value < 0.05 can be interpreted as a significant publication bias across studies.

Search and selection of literature

The database search and the manual search of the article citation list turned up 1120 records. Thirty studies were included in the systematic review and meta-analysis [ 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 ] after duplicates and irrelevant records were removed. Included studies had a total of 2328 neonates with sepsis and 1800 neonates in the control group, of which 902 were healthy neonates, 271 were neonates who were initially suspected of having sepsis but ultimately had negative blood cultures, and 627 were other neonates admitted to the ICU. The PRISMA flow diagram, shown in Fig.  1 , describes the inclusion and exclusion processes in detail.

PRISMA 2020 Flow diagram for new systematic reviews which includes searches of databases, registers and other sources

Characteristics of included studies

Table 1 displays the methodological and characteristic qualities of these publications. The overall study quality ranged from 6 to 8 stars. Thirty studies were included in our systematic review and meta-analysis [ 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 ]. Of them, 15 studies had a control group of healthy neonates [ 19 , 24 , 25 , 26 , 27 , 29 , 30 , 32 , 33 , 35 , 36 , 37 , 38 , 39 , 46 ], five studies had a control group of neonates who were suspected of sepsis, but eventually had negative blood cultures [ 20 , 28 , 40 , 41 , 44 ], and ten studies included other ICU admitted neonates as the control group [ 17 , 18 , 21 , 22 , 23 , 31 , 34 , 42 , 43 , 45 ]. Twenty studies were prospective [ 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 27 , 29 , 30 , 32 , 35 , 36 , 37 , 38 , 41 , 42 , 44 ] and 10 studies were retrospective [ 26 , 28 , 31 , 33 , 34 , 39 , 40 , 43 , 45 , 46 ]. Nine studies included neonates with early-onset sepsis [ 17 , 24 , 25 , 27 , 31 , 35 , 37 , 44 , 46 ] and five included neonates with late-onset sepsis [ 20 , 26 , 28 , 29 , 43 ]. Other studies considered both types of sepsis.

The differences in NLR levels between neonates with sepsis and healthy controls

Considering the statistical heterogeneity across studies, the pooled meta-analysis was conducted using a random-effects model (I 2  = 97.0%, P -value < 0.001). The results indicated that neonates with sepsis had elevated levels of NLR compared to healthy controls (SMD = 1.81, 95% CI = 1.14–2.48, P -value < 0.001) (Fig.  2 ).

Meta-analysis of differences in NLR level between neonates with sepsis and healthy controls

Further subgroup analysis stratified by study design indicated that neonates with sepsis had elevated levels of NLR compared to healthy controls in either prospective (SMD = 2.38, 95% CI = 1.40–3.35, P -value < 0.001) or retrospective studies (SMD = 0.87, 95% CI = 0.63–1.12, P -value < 0.001) (Fig.  3 ).

Subgroup analysis of differences in NLR level between neonates with sepsis and healthy controls according to study design

Another subgroup analysis of the type of sepsis indicated that neonates with early-onset sepsis had elevated levels of NLR compared to healthy controls (SMD = 2.50, 95% CI = 1.59–3.40, P -value < 0.001). However, there was no difference in NLR level between neonates with late-onset sepsis and healthy controls (SMD = 0.13, 95% CI = -0.79–1.05, P -value = 0.78) (Fig.  4 ).

Subgroup analysis of differences in NLR level between neonates with sepsis and healthy controls according to the type of sepsis

Diagnostic value of NLR for differentiating between neonates with sepsis and healthy controls

The pooled sensitivity of 12 studies was 0.79 (95% CI, 0.62–0.90), and the pooled specificity was 0.91 (95% CI, 0.73–0.97). The pooled positive likelihood ratio, negative likelihood ratio, DOR of NLR were 8.88 (95%CI = 2.65–29.80), 0.21 (95%CI = 0.10–0.45), and 40.42 (95%CI = 7.48–218.38), respectively (Fig.  5 ).

SROC curve of included studies in the meta-analysis of differences in NLR level between neonates with sepsis and healthy controls

The differences in NLR levels between neonates with sepsis and those who were suspected of sepsis but eventually had negative blood cultures

The random-effects model was applied to the pooled meta-analysis, as statistical heterogeneity existed among studies (I 2  = 97.2%, P -value < 0.001). We found that neonates with sepsis had elevated levels of NLR compared to those who were suspected of sepsis, but eventually had negative blood cultures (SMD = 1.99, 95% CI = 0.76–3.22, P -value = 0.002) (Fig.  6 ).

Meta-analysis of differences in NLR level between neonates with confirmed sepsis and those with suspected sepsis

In the subgroup analysis according to study design, we found that neonates with sepsis had elevated levels of NLR compared to those who were suspected of sepsis, but eventually had negative blood cultures in prospective studies (SMD = 0.72, 95% CI = 0.28–1.16, P -value = 0.001), but not in retrospective studies (SMD = 4.12, 95% CI = -1.11–9.34, P -value = 0.122) (Fig.  7 ).

Subgroup analysis of differences in NLR level between neonates with confirmed sepsis and those with suspected sepsis, according to study design

Diagnostic value of NLR for differentiating between neonates with sepsis and those who were suspected of sepsis, but eventually had negative blood cultures

The pooled sensitivity of five studies was 0.79 (95% CI, 0.69–0.86), and the pooled specificity was 0.73 (95% CI, 0.54–0.85). The pooled positive likelihood ratio, negative likelihood ratio, DOR of NLR were 2.93 (95%CI = 1.58–5.41), 0.28 (95%CI = 0.17–0.46), and 10.20 (95%CI = 3.75–27.70), respectively (Fig.  8 ).

SROC curve of included studies in the meta-analysis of differences in NLR level between neonates with confirmed sepsis and those with suspected sepsis

The differences in NLR levels between neonates with sepsis and other ICU admitted neonates

A fixed-effects model was applied to t pool the data of included studies (I 2  = 57.2%, P -value < 0.01). Neonates with sepsis had elevated levels of NLR compared to other ICU admitted neonates (SMD = 0.73, 95% CI = 0.63–0.84, P  < 0.001) (Fig.  9 ).

Meta-analysis of differences in NLR level between neonates with sepsis and other ICU admitted neonates

Then, we conducted a subgroup analysis according to the study design. The results showed that neonates with sepsis had elevated levels of NLR compared to Other ICU admitted neonates in either prospective (SMD = 0.73, 95% CI = 0.55–0.91, p  < 0.001) or retrospective studies (SMD = 0.74, 95% CI = 0.61–0.86, p  < 0.001) (Fig.  10 ).

Subgroup analysis of differences in NLR level between neonates with sepsis and other ICU admitted neonates, according to study design

Diagnostic value of NLR for differentiating between neonates with sepsis and other ICU admitted neonates

The pooled sensitivity of five studies was 0.65 (95% CI, 0.55–0.80), and the pooled specificity was 0.80 (95% CI, 0.68–0.88). The pooled positive likelihood ratio, negative likelihood ratio, DOR of NLR were 3.51(95%CI = 2.22–5.53), 0.38 (95%CI = 0.26–0.54), and 9.23 (95%CI = 4.90–17.39), respectively (Fig.  11 ).

SROC curve of included studies in the meta-analysis of differences in NLR level between neonates with sepsis and other ICU admitted neonates

Publication bias

As seen in Fig.  12 , there was some indication of publication bias among studies with a control group of healthy neonates (Egger’s test P -value < 0.001) and neonates who were suspected of sepsis (Egger’s test P -value < 0.001). However, studies with a control group including other ICU-admitted neonates had no publication bias (Egger’s test P -value = 0.08).

Funnel plots assessing publication bias among studies; A studies with a control group of healthy controls; B studies with a control group of neonates who were suspected of sepsis; C studies with a control group of other ICU-admitted neonates

Neonatal sepsis can manifest itself clinically in a variety of ways, including feeding intolerance, temperature instability, tachycardia, pneumonia, and respiratory distress [ 10 , 14 , 15 , 16 ]. Because these symptoms are similar to non-infectious diseases, newborn sepsis is difficult to identify clinically. Some neonates with bacteremia might even have no symptoms and present with a regular physical examination [ 2 , 9 ]. This emphasizes the importance of decisive tests with quantitative measures for diagnosing sepsis. Blood culture as a diagnostic marker is currently the gold standard for diagnosing sepsis in neonates, although it has drawbacks such as a long waiting period and the risk of contamination [ 2 ]. CBC with differential and C-reactive protein (CRP) are additional crucial lab tests to get and are routinely collected on a serial basis; however, these indices are weak at diagnosing newborn sepsis and are better suited for ruling it out [ 2 ]. These flaws have emphasized the necessity for a neonatal sepsis marker that can be tested rapidly and easily. As a measure of newborn sepsis, neutropenia has a higher specificity than neutrophilia [ 2 , 47 ]. An elevated immature to total neutrophil (I/T) ratio of more than 0.27 has a very high negative predictive accuracy (99%) but a poor positive predictive value (25%) because it can be elevated in up to 50% of uninfected infants [ 2 ]. These counts can be erroneously raised, especially after a baby is born. However, several clinical investigations have recently established the efficacy of NLR in predicting neonatal sepsis [ 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 ]. Pooling the data of these studies indicated elevated levels of NLR in septic neonates versus healthy controls, found in both retrospective and prospective studies, lend to the accuracy of NLR for diagnosis of sepsis. Sub-group analysis showed this association was primarily seen in the context of EOS and combined studies rather than LOS, whereas studies of NLR in LOS were limited in number. The use of healthy controls as a comparator for indicators of neonatal sepsis may be less clinically relevant to studies comparing acutely ill septic neonates to those which are acutely ill but non-septic. Although not shown by limited retrospective studies, the findings in prospective studies of elevated NLR in septic versus neonates with suspected sepsis may indicate the negative predictive value of elevated NLR in neonatal sepsis. Additionally, findings from both prospective and retrospective cohort studies showed that elevated NLR distinguished neonatal sepsis among other ICU-admitted neonates. These findings are important because hospital admission and subsequent environmental stressors may represent confounders if not controlled for in the study design. Previous studies outlined in this paper showed a potential for false positivity or negativity when using clinical presentation or blood-based biomarkers to differentiate sepsis from non-infectious acute illnesses in neonates [ 2 , 9 ]. In addition, NLR’s accuracy in diagnosing sepsis is an important finding.

Limitations

Significant to consider in the interpretation of currently available studies, including healthy controls or controls with suspected sepsis, is the potential for small study/publication bias indicated here via Eggers linear regression analysis. A high degree of statistical heterogeneity is shown in combined, and subgroup analysis, except for studies of septic versus other ICU admitted neonates. Heterogeneity in these studies may be partially attributed to the differing diagnostic threshold for sepsis as well as differences in sample handling. Similar heterogeneity is seen across other sepsis biomarker studies, including CRP and procalcitonin [ 48 , 49 ]. Sub-group analysis helps to reduce some of the heterogeneity seen in these studies and identify potential study design flaws such as inadequate comparators. The use of comparators with suspected sepsis or that are admitted to the ICU may represent a more clinically relevant approach. However, current studies lack precision, and acutely ill comparators may bring additional confounders, contributing to study heterogeneity. Attenuating this effect may require large multicenter prospective studies.

A challenge to using hematological and acute phase markers for accurate diagnosis of sepsis is the variability of immune response in the context of comorbid conditions, gestational age, post-natal age, timing of sepsis onset, and the nature of infectious agents. Coupling this pathophysiologic variability with diagnostic and procedural differences in clinical practice make real-world use of any single biomarker challenging. While markers such as CRP and I/T ratio have been shown independently to have high sensitivities, these findings depend on timing from the onset of sepsis, either lacking accuracy early on or losing accuracy later in the course of sepsis [ 8 , 50 ]. The findings in this review support the use of NLR to diagnose neonatal sepsis among the limitations of current biomarkers. The use of CRP and NLR as combined markers showed improved accuracy in a recent study on LOS [ 41 ]. More high-precision studies are needed to examine these markers with an emphasis on sensitivity and negative predictive value (NPV).

The results of our study support an association between NLR values and the development of sepsis among neonates. NLR represents a unique inflammatory marker whose elevation in neonatal sepsis implicates immune system imbalance in the pathogenesis of the disease. Further, our findings support NLR as a promising biomarker that can be readily integrated into clinical settings to aid in diagnosing neonatal sepsis. With the development of new biomarkers and therapeutic modalities, we can better prevent and treat neonatal sepsis to decrease long-term morbidity and mortality. The NLR, as an inflammatory biomarker, measures innate-adaptive immune system balance. As evidenced by our results, restoring balance to the innate and adaptive immune system may serve as attractive therapeutic target. Theoretically, a reduction in NLR values could be used to measure therapeutic efficacy, reflecting the restoration of balance within these systems.

Availability of data and materials

The dataset supporting the conclusions of this article is included within the article.

Abbreviations

• Neutrophil to lymphocyte ratio

Intensive care unit

Standardized mean difference

95% Confidence interval

C-reactive protein

Late-onset sepsis

Early-onset sepsis

The Newcastle-Ottawa Quality Assessment Scale

Retrospective

Prospective

Group B Streptoccoci

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Weighted mean difference

Summary receiver operating characteristic

Diagnostic odds ratio

Immature to total neutrophil

Shane AL, Sánchez PJ, Stoll BJ. Neonatal sepsis. Lancet. 2017;390(10104):1770–80.

Article   PubMed   Google Scholar  

Singh M, Alsaleem M, Gray CP. Neonatal Sepsis. [Updated 2022 Sep 29]. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2023. Available from: https://www.ncbi.nlm.nih.gov/books/NBK531478/ .

Fleischmann-Struzek C, Goldfarb DM, Schlattmann P, Schlapbach LJ, Reinhart K, Kissoon N. The global burden of paediatric and neonatal sepsis: a systematic review. Lancet Respir Med. 2018;6(3):223–30.

Turhan EE, Gürsoy T, Ovalı F. Factors which affect mortality in neonatal sepsis. Turk Pediatri Ars. 2015;50(3):170–5.

Article   PubMed   PubMed Central   Google Scholar  

Cai S, Thompson DK, Anderson PJ, Yang JY. Short- and long-term neurodevelopmental outcomes of very preterm infants with neonatal sepsis: a systematic review and meta-analysis. Children (Basel). 2019;6(12):131.

PubMed   Google Scholar  

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. 2022;10(5):e661–72.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Glaser MA, Hughes LM, Jnah A, Newberry D, Harris-Haman PA. Neonatal sepsis: a review of pathophysiology and current management strategies. Adv Neonatal Care. 2021;21(1):49–60.

Celik IH, Hanna M, Canpolat FE, Pammi M. Diagnosis of neonatal sepsis: the past, present and future. Pediatr Res. 2022;91(2):337–50.

Wynn JL, Wong HR. Pathophysiology and treatment of septic shock in neonates. Clin Perinatol. 2010;37(2):439–79.

Simonsen KA, Anderson-Berry AL, Delair SF, Davies HD. Early-onset neonatal sepsis. Clin Microbiol Rev. 2014;27(1):21–47.

Flannery DD, Edwards EM, Puopolo KM, Horbar JD. Early-onset sepsis among very preterm infants. Pediatrics. 2021;148(4):e2021052456.

Bulkowstein S, Ben-Shimol S, Givon-Lavi N, Melamed R, Shany E, Greenberg D. Comparison of early onset sepsis and community-acquired late onset sepsis in infants less than 3 months of age. BMC Pediatr. 2016;16(1):1–8.

Article   Google Scholar  

Ogundare E, Akintayo A, Aladekomo T, Adeyemi L, Ogunlesi T, Oyelami O. Presentation and outcomes of early and late onset neonatal sepsis in a Nigerian hospital. Afr Health Sci. 2019;19(3):2390–9.

Goldstein B, Giroir B, Randolph A. International pediatric sepsis consensus conference: definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med. 2005;6(1):2–8.

Hibbert JE, Currie A, Strunk T. Sepsis-induced immunosuppression in neonates. Front Pediatr. 2018;6:357.

Stoll BJ, Hansen NI, Bell EF, Shankaran S, Laptook AR, Walsh MC, et al. Neonatal outcomes of extremely preterm infants from the NICHD Neonatal Research Network. Pediatrics. 2010;126(3):443–56.

Abdelmoktader A, Hussein S, Ali D, Eisa O. Diagnostic value of eosinopenia and neutrophil to lymphocyte ratio in early onset neonatal sepsis. Fayoum Univ Med J. 2020;7(1):161–71.

Akhmaltdinova L, Kolesnichenko S, Lavrinenko A, Kadyrova I, Avdienko O, Panibratec L. Influence of pathogen type on neonatal sepsis biomarkers. Int J Inflamm. 2021;2021:1009231.

AL Nady N, Ismaiel Y, Assar E, Saleh M. Diagnostic value of neutrophil to lymphocyte ratio on neonatal sepsis in full term and preterm neonates. Benha J Appl Sci. 2021;6(6):169–77.

AlkanOzdemir S, ArunOzer E, Ilhan O, Sutcuoglu S. Can neutrophil to lymphocyte ratio predict late-onset sepsis in preterm infants? J Clin Lab Anal. 2018;32(4):e22338.

Ashour R, Waael Z, Rashwan NI, Fayed HM. The pattern of systemic inflammatory markers response in neonatal sepsis. SVU-Int J Med Sci. 2022;5(1):252–60.

Google Scholar  

Awad MH, AbdElhamid HS, AbdElhamid SF, Gaballah AM. Correlation between serum interleukin-lß and neonatal sepsis in neonatal intensive care unit in Zagazig University Hospitals. Egypt J Hosp Med. 2022;86(1):664–9.

Büyükeren M, Çelik HT, Yiğit Ş, Portakal O, Sancak B, Yurdakök M. The role of the delta neutrophil index in determining the etiology of neonatal sepsis. Turk J Biochem. 2021;46(1):71–7.

Can E, Hamilcikan Ş, Can C. The value of neutrophil to lymphocyte ratio and platelet to lymphocyte ratio for detecting early-onset neonatal sepsis. J Pediatr Hematol Oncol. 2018;40(4):e229–32.

Chen S-J, Zheng X-X, Jin H-X, Chen J-H, He T-F, Chen C-E. Can venous cord blood neutrophil to lymphocyte ratio and platelet to lymphocyte ratio predict early-onset sepsis in preterm infants? Clin Exp Obstet Gynecol. 2021;48(4):828–34.

esin Yalınbaş E, Bilgin H. GEÇ NEONATAL SEPSİS TANISINDA TROMBOSİT PARAMETRELERİ VE NÖTROFİL LENFOSİT ORANLARININ DEĞERLENDİRİLMESİ. Kocatepe Tıp Dergisi. 2020;21(1):104–9.

Ghrahani R, Yuniati T, Judistiani RTD, Setiabudiawan B. Strong positive correlation between neutrophil-to-lymphocyte ratio and C-reactive protein in early onset sepsis. Majalah Kedokteran Bandung. 2019;51(4):246–52.

Goldberg O, Amitai N, Chodick G, Bromiker R, Scheuerman O, Ben-Zvi H, et al. Can we improve early identification of neonatal late-onset sepsis? A validated prediction model. J Perinatol. 2020;40(9):1315–22.

Article   CAS   PubMed   Google Scholar  

Hibbert J, Strunk T, Nathan E, Prosser A, Doherty D, Simmer K, et al. Composition of early life leukocyte populations in preterm infants with and without late-onset sepsis. PLoS One. 2022;17(3):e0264768.

Israr S, Hayat A, Ahmad TM, Majeed N, Naqvi SH, Tehseen S. Comparison of procalcitonin and hemattological ratios in cord blood as early predictive marker of neonatal sepsis. PAFMJ. 2020;70(3):824–9.

Karabulut B, Alatas SO. Diagnostic value of neutrophil to lymphocyte ratio and mean platelet volume on early onset neonatal sepsis on term neonate. J Pediatr Intensive Care. 2021;10(02):143–7.

Khattab AA, El-Mekkawy MS, Helwa MA, Omar ES. Utility of serum resistin in the diagnosis of neonatal sepsis and prediction of disease severity in term and late preterm infants. J Perinat Med. 2018;46(8):919–25.

Kurt A, Tosun MS, Altuntaş N. Diagnostic accuracy of complete blood cell count and neutrophil-to-lymphocyte, lymphocyte-to-monocyte, and platelet-to-lymphocyte ratios for neonatal infection. Asian Biomed. 2022;16(1):43–52.

Li T, Dong G, Zhang M, Xu Z, Hu Y, Xie B, et al. Association of neutrophil–lymphocyte ratio and the presence of neonatal sepsis. J Immunol Res. 2020;2020:7650713.

Mahmoud NM, Abdelhakeem M, Mohamed HH, Baheeg G. Validity of Platelet to Lymphocyte Ratio and Neutrophil to Lymphocyte Ratio in Diagnosing Early-onset Neonatal Sepsis in Full-term Newborns. J Compr Pediatr. 2022;13(1).

Mao Y-Y, Su C, Fang C-C, Fan X-P, Wang L-P, Zhu S-S, et al. Clinical significance of the serum miR-455-5p expression in patients with neonatal sepsis. Bioengineered. 2021;12(1):4174–82.

Nasser MM, Afia AA, EL-Khatib GZ, Ibrahim MI. The value of neutrophil to lymphocyte ratio and platelet to lymphocyte ratio and procalcitonin level for detecting early-onset neonatal sepsis. Al-Azhar J Pediatr. 2020;23(2):852–74.

Omran A, Maaroof A, Saleh MH, Abdelwahab A. Salivary C-reactive protein, mean platelet volume and neutrophil lymphocyte ratio as diagnostic markers for neonatal sepsis. J Pediatr. 2018;94:82–7.

Panda SK, Nayak MK, Rath S, Das P. The utility of the neutrophil-lymphocyte ratio as an early diagnostic marker in neonatal sepsis. Cureus. 2021;13(1):e12891.

PubMed   PubMed Central   Google Scholar  

Ruslie RH, Tjipta DG, Samosir CT, Hasibuan BS. Bacterial pattern and role of laboratory parameters as marker for neonatal sepsis. InIOP Conference Series: Earth and Environmental Science. 2018;125(1):012057. IOP Publishing.

Sumitro KR, Utomo MT, Widodo ADW. Neutrophil-to-lymphocyte ratio as an alternative marker of neonatal sepsis in developing countries. Oman Med J. 2021;36(1):e214.

Tang Z, Qin D, Tao M, Lv K, Chen S, Zhu X, et al. Examining the utility of the CD64 index compared with other conventional indices for early diagnosis of neonatal infection. Sci Rep. 2018;8(1):1–7.

Varal IG, Dogan P. Can neutrophil-lymphocyte ratio be a predictor of late-onset sepsis in preterm infants? Ann Med Res. 2020;27(1):0023–8.

Wilar R. Diagnostic value of eosinopenia and neutrophil to lymphocyte ratio on early onset neonatal sepsis. Korean J Pediatr. 2019;62(6):217.

Yorulmaz A, Yücel M, Sadiye S. Yenidoğan sepsis olgularında hematolojik parametrelerin tanısal değeri. Ortadoğu Tıp Dergisi. 2018;10(3):252–62.

Zhang S, Luan X, Zhang W, Jin Z. Platelet-to-lymphocyte and neutrophil-to-lymphocyte ratio as predictive biomarkers for early-onset neonatal sepsis. Birth. 2021;18(36):0.144.

Djordjevic D, Rondovic G, Surbatovic M, Stanojevic I, Udovicic I, Andjelic T, et al. Neutrophil-to-lymphocyte ratio, monocyte-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and mean platelet volume-to-platelet count ratio as biomarkers in critically ill and injured patients: which ratio to choose to predict outcome and nature of bacteremia? Mediators Inflamm. 2018;2018:3758068.

Pontrelli G, De Crescenzo F, Buzzetti R, Jenkner A, Balduzzi S, Calò Carducci F, et al. Accuracy of serum procalcitonin for the diagnosis of sepsis in neonates and children with systemic inflammatory syndrome: a meta-analysis. BMC Infect Dis. 2017;17(1):1–12.

Vouloumanou EK, Plessa E, Karageorgopoulos DE, Mantadakis E, Falagas ME. Serum procalcitonin as a diagnostic marker for neonatal sepsis: a systematic review and meta-analysis. Intensive Care Med. 2011;37(5):747–62.

Newman TB, Puopolo KM, Wi S, Draper D, Escobar GJ. Interpreting complete blood counts soon after birth in newborns at risk for sepsis. Pediatrics. 2010;126(5):903–9.

Download references

Acknowledgements

Not applicable.

No funding was received to undertake this systematic review.

Author information

Authors and affiliations.

The First Clinical Medical College of Fujian Medical University, Fuzhou, Fujian Province, China

Jingyang Chen

Tabriz University of Medical Sciences, Tabriz, Iran

Sanaz Yasrebinia & Shokoufeh Khanzadeh

Student Research Committee, School of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran

Arshin Ghaedi

Trauma Research Center, Shahid Rajaee (Emtiaz) Trauma Hospital, Shiraz University of Medical Sciences, Shiraz, Iran

Geriatric & Gerontology Department, Medical School, Tehran University of Medical and Health Sciences, Tehran, Iran

Monireh Khanzadeh

Department of Neurosurgery, University of Florida, Gainesville, FL, 32610, USA

Stephan Quintin, Abeer Dagra, Rodeania Peart & Brandon Lucke-Wold

You can also search for this author in PubMed   Google Scholar

Contributions

All authors contributed to the design of the study. ShKh and SY undertook the initial searches and screened articles for inclusion with RP. SQ and AD wrote the initial draft of the manuscript. MKh and B.L-W edited the initial draft. A.Gh and J.Ch revised the manuscript according to the reviewers’ comments. The author(s) read and approved the final manuscript.

Corresponding author

Correspondence to Shokoufeh Khanzadeh .

Ethics declarations

Ethics approval and consent to participate, consent for publication, 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.

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.

Chen, J., Yasrebinia, S., Ghaedi, A. et al. Meta-analysis of the role of neutrophil to lymphocyte ratio in neonatal sepsis. BMC Infect Dis 23 , 837 (2023). https://doi.org/10.1186/s12879-023-08800-0

Download citation

Received : 25 March 2023

Accepted : 07 November 2023

Published : 28 November 2023

DOI : https://doi.org/10.1186/s12879-023-08800-0

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
• Systematic review
• Meta-analysis

BMC Infectious Diseases

ISSN: 1471-2334

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

• Open access
• Published: 28 November 2023

Prevalence and predictors of chronic kidney disease among type 2 diabetic patients worldwide, systematic review and meta-analysis

• Eneyew Talie Fenta 1 ,
• Habitu Birhan Eshetu 2 ,
• Natnael Kebede 3 ,
• Eyob Ketema Bogale 4 ,
• Amare Zewdie 5 ,
• Tadele Derbew Kassie 6 ,
• Tadele Fentabil Anagaw 4 ,
• Elyas Melaku Mazengia 6 &
• Sintayehu Shiferaw Gelaw 6  

Diabetology & Metabolic Syndrome volume  15 , Article number:  245 ( 2023 ) Cite this article

143 Accesses

Metrics details

Diabetes is a complicated, chronic condition that requires ongoing medical attention as well as multiple risk-reduction measures beyond glucose control. The prevalence of chronic kidney disease (CKD) is highly variable in different parts of the world due to various environmental, ethnic, socioeconomic, and rural-urban differences. Diabetes is the leading cause of CKD. This study aimed to estimate the global prevalence of CKD and its associated factors among type 2 diabetes(T2DM) patients, provide scientific evidence for a better understanding of the burden of CKD among diabetes mellitus type 2 patients, and design interventional strategies.

Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist guideline was followed for this review and meta-analysis. The electronic databases (Pub Med, Cochrane Library, Google Scholar, and grey literature) were searched to retrieve articles by using keywords. Joanna Briggs Institute Meta-Analysis of Statistics Assessment and Review Instrument was used to assess the quality of studies. The meta-analysis was conducted using STATA 17 software. The Meta logistic regression was computed to present the pooled prevalence and Odds ratio (OR) of the determinate factors with a 95% confidence interval (CI).

In this systematic review and meta-analysis 20 studies were done in 13 different countries. The pooled magnitude of chronic kidney disease among type 2 DM patients was 27% (95% CI 21%, 33%). The prevalence of chronic kidney disease differs across countries, with the maximum in the USA and the lowest in the United Arab Emirates. Patients with CKD have an elevated risk of severe renal and cardiovascular morbidity and mortality. Renin-angiotensin system inhibitors, sodium-glucose cotransporter-2 inhibitors, glucagon-like peptide 1 receptor agonists, and, more recently, non-steroidal mineralocorticoid receptor antagonists are among the medications that have been demonstrated to slow the progression of CKD. In this systematic review and meta-analysis increased age, obesity, having a history of type 2 diabetes mellitus, smoking history, presence of hypertension, and cardiac heart disease were factors significantly associated with the presence of chronic kidney disease among type 2 diabetic patients.

Conclusions

The prevalence of chronic kidney disease among type 2 diabetes mellitus patients was high based on the included 20 articles. The review reported that old age, hypertension, cardiac disease, smoking, obesity, and duration of diabetes mellitus was predictor variable for chronic kidney disease among type 2 diabetic patients. Therefore, in order to lower the morbidity and mortality from chronic kidney disease among type 2 diabetic patients, it is advised to develop both preventive and curative intervention strategies, such as raising awareness, creating a supportive environment, and prescribing appropriate medication at an early stage.

Introduction

Diabetes mellitus (DM) is a metabolic systemic multifactorial disease, where the body becomes unable to utilize or produce a sufficient amount of insulin to match the body’s needs [ 1 , 2 ]. Diabetes is a complicated, chronic condition that need for ongoing medical attention as well as multiple risk-reduction measures beyond glucose control [ 3 , 4 ]. The World Health Organization(WHO) placed the global targets for diabetes mellitus by 2030, 80% of people living with diabetes mellitus are diagnosed, and that 80% of those living with diabetes mellitus will have good control of glycaemia and blood pressure [ 5 ].

Globally, about one in eleven adults have diabetes mellitus, from this 90% have type 2 diabetes mellitus. The main drivers of the T2DM epidemic are the global rise in obesity, sedentary lifestyles, high caloric diets and population aging, which have quadrupled the incidence and prevalence of T2DM [ 6 , 7 , 8 ].The estimated global prevalence of diabetes in 2015 was 415 million; by 2040, it is expected to rise to 642 million, with more increase in low- and middle-income nations [ 9 ]. DM patients have uncontrolled blood sugar levels and a longer duration of diabetes were considered as independent risk factors of CKD [ 10 ]. Diabetes can cause serious health problems, such as heart disease, stroke, and eye and foot problems. Prediabetes also can cause health problems. The good news is that type 2 diabetes can be delayed or even prevented [ 11 ].

The studies showed that diabetic individuals develop CKD and exposed to renal complications with considerable morbidity, poor quality of life, and increased healthcare costs [ 12 , 13 , 14 ]. The study in Europe stated that the prevalence of CKD in patients with diabetes is expected to increase within increase in cost of treatment [ 15 ]. The prevalence of CKD is highly variable in different parts of the world, due to various environmental, ethnic, socioeconomical, and rural-urban variations [ 16 ]. In the 2015 Global Burden of Disease Study, kidney disease was the 12th most common cause of death, accounting for 1.1 million deaths worldwide. Overall CKD mortality has increased by 31.7% over the last 10 years, making it one of the fastest rising major causes of death, alongside diabetes and dementia [ 17 ].

Chronic kidney disease (CKD) is the abnormalities of kidney structure or function that are present for more than 3 months characterized by low estimated glomerular filtration rate (eGFR) based on serum creatinine measurements. It is one of the complications among T2DM patients [ 12 , 18 ]. The primary issue for individuals in stages 1 and 2 was that it was asymptomatic at first; metabolic abnormalities were apparent in stages 3 through 5. Diabetic kidney disease (CKD) and other vascular disorders are the main reasons. Estimating the glomerular filtration rate (eGFR), which evaluates kidney function, and finding albuminuria, a sign of kidney damage, are the first steps in diagnosing chronic kidney disease (CKD) [ 2 , 19 , 20 ].

All individuals with T2DM should be screened for CKD using the spot urine albumin-to-creatinine ratio and estimated glomerular filtration rate at the time of diagnosis and at least once a year after that. Although the exact cause of chronic kidney disease (CKD) is still unknown, research has shown that a number of risk factors, including obesity, ageing, hypertension, diabetes mellitus, hyperlipidemia, use of nephrotoxic medications, family history of kidney disease, smoking, heavy drinking, HIV infection, electrolyte and acid-base disturbances, low-income occupation, use of traditional medication, low hemoglobin, rapid urbanization, physical inactivity, and rapid population growth [ 21 , 22 , 23 , 24 , 25 ].

The study in Japan revealed that elderly patients with type 2 diabetes, renal dysfunction is characterized by low eGFR [ 26 ]. One of the main challenges of diagnosing CKD in diabetic patients is that it is asymptomatic onset with reduced quality of life, high morbidity and mortality rate [ 2 , 27 ].

CKD is estimated to affect 50% patients with T2DM globally, and its presence and severity markedly influences disease prognosis. CKD is more common in certain patient populations, including the elderly, those with youth-onset diabetes mellitus, those who are obese, certain ethnic groups, and disadvantaged populations. The incident cases of CKD as a result of type 2 diabetes mellitus worldwide in 2017 had increased by 74% compared with 1990 [ 28 , 29 , 30 ].The study showed that prevalence of CKD ranged from 11 to 90% in patients with diabetes, high-risk groups in patients with HIV from 1–46%,and 13–51% in patients with hypertension [ 31 ].

In order to improve patients’ quality of life, early screening for CKD is crucial for diagnosis and treatment. Other interventions include counselling, good glycemic and blood pressure control, and medication use. Renin-angiotensin system inhibitors, sodium-glucose cotransporter-2 inhibitors, glucagon-like peptide 1 receptor agonists, and, more recently, non-steroidal mineralocorticoid receptor antagonists are among the medications that have been demonstrated to slow the progression of CKD [ 24 , 32 , 33 , 34 ].

Diabetes is the leading cause of CKD in all developed and most developing countries, with approximately 20% of people with type 2 diabetes showing evidence of diabetic nephropathy within 20 years of diabetes onset [ 35 ]. The patients with diabetes are likely to be prescribed more than one type of medication which exposed them at a higher risk of CKD due to the effect of polypharmacy [ 36 , 37 ]. Studies showed that early detection and treatment of diabetes and chronic diseases can slow or prevent the progression of CKD [ 38 ] .This study aimed to estimate the global prevalence of CKD and its associated factors among type 2 diabetes patients that can provides a scientific evidence for a better understanding of the burden of CKD among type 2 diabetes mellitus patients and design interventional strategies.

Methodology

Search strategy.

This systematic review and meta-analysis was performed by following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [ 39 ]. A search strategy was implemented using electronic databases (PubMed, Cochrane Library, Google Scholar, and grey literature), which were systematically searched online to retrieve related articles. The literature search was conducted by using the key words “chronic kidney disease” OR (chronic renal failure) AND “Type 2 diabetes mellitus” OR “T2DM starting from May 30, 2018 to May 30, 2023 were included. The retrieved studies references were also screened and checked. The review protocol is available on PROSPERO (ID: CRD42023433892).

Eligibility criteria

For the review, CoCoPop mnemonic (Condition, Context and Population) approach was used. Studies that reported chronic kidney disease prevalence and associated factors among type 2 diabetic patients globally with any type of study design at any health facility level, studies with open or free access to full text and written in English language were included. The studies without abstract & full-text, reports, and qualitative studies, conference summaries were excluded. Articles were assessed for inclusion using their title, abstract and then a full review of articles was done before included to the final review.

Data extraction and management

Eligible studies were imported to Endnote v.9 and duplicates were removed. The two independent reviewers (TFA, EKB,) were do the abstract and full text review and data abstraction. Any disagreements were resolved by consensus and involvement of third reviewer (ETF). Full text articles were retrieved for studies that meet the inclusion criteria. Data extraction used a data collection form and was performed by two blind and independent reviewers. The following data were extracted: author, publication year, study design, place of study, sample size, participants.

Quality assessment

The Joanna Briggs Institute (JBI) critical appraisal check list was used to assess the quality of studies which is freely available at https://jbi.global/critical-appraisal-tools . Using the tool as a protocol, the reviewers (NK, SSG) used the blinded review approach to evaluate the quality of the original articles. Those studies, scores 5 or more in JBI criteria were considered to have good quality and included in the review [ 40 ]. Discrepancies in the quality assessment was resolved through a third author (EMM).

Statistical analysis

The necessary data were extracted from the studies using Microsoft Excel V.2016 and the extracted data were exported to STATA version 17 software for analysis. The articles were summarized by tables and forest plot. The standard error and 95% confidence interva1 for the prevalence of chronic kidney disease was calculated for those studies in which estimates of standard error and 95% confidence interval for the prevalence of chronic kidney disease were not found in their full text of their article.

Meta-analysis was performed, the significance of the pooled Odds Ratio (OR) was determined by Z-test. The pooled estimate for the prevalence of CKD in patients with diabetes was calculated using random-effect models with 95% confidence intervals (CIs). The statistical heterogeneity was checked subjectively by using forest plot, and objectively by Cochrane Q-test and I 2 statistics [ 41 ]. The presence of publication bias was checked by using a funnel plot and Egger’s and Begg’s statistical tests [ 42 ].

In this systematic review and meta-analysis, a total of three thousand two hundred seventy-eight (3,278) record articles were searched from all searched database sources. After removing duplicates (3176), reviewing title and abstract (102), and removal of 39 articles due to not reporting for prevalence of chronic disease among diabetes mellitus type 2 patients, then 63 articles were eligible for full text review and finally 20 articles included for systematic review and meta-analysis (Fig.  1 ).

PRISMA flow diagram of study selection

Study characteristics

In this systematic review and meta-analysis 20 studies were done in 13 different countries globally. Two studies done in Ethiopia [ 43 , 44 ],one study each in Ghana [ 45 ], United Arab Emirates [ 46 ], Australia [ 47 ], Germen [ 48 ], united kingdom, Malaysia, Palestine [ 49 , 50 , 51 ], three studies in USA [ 52 , 53 , 54 ] ,two in China [ 55 , 56 ],two in Spain [ 57 , 58 ], three in Thailand [ 59 , 60 , 61 ], and in six countries in Europe and Asia [ 62 ].in this study most of the studies were hospital based except two are community level. The total sample size was 1,711,926 with maximum sample 1, 177 896 sample in six European and Asian countries [ 62 ] and minimum of 119 sample size in Ethiopia [ 44 ] (Table  1 ).

Meta -analysis

Pooled prevalence of chronic kidney disease among type 2 dm patients.

The pooled magnitude chronic kidney disease among type 2 DM patients was 27% (95% CI 21%, 33%). Based on tau square (between studies variance), tau2 = 25252.96 & I2 = 99.8 with p value < 0.00005 which indicates there is statistically significant heterogeneity among studies (Fig.  2 ).

Forest plot of the pooled magnitude of CKD among type 2 diabetes mellites patients

Publication bias

The result of regression-based egger test for small-study effects showed significant result (p value = 0.2540). The funnel plot’s evidence suggests the current study had a source of publication bias. The results seemed like asymmetrical funnel plots. The result of egger’s test was not statistically significant for the presence of publication bias (P = 0.254). and Begg’s test for small-study effects was significant (p = 0.0012) (Fig.  3 ).

 A funnel plot displaying publication bias across the studies

Factors associated with CKD among type 2 diabetic patients

In this systematic review and meta-analysis increased age, obesity, duration of diabetes mellitus, smoking history, presence of hypertension, and having cardiac heart disease were factors significantly associated with the presence chronic kidney disease among type 2 diabetic patients. Eleven studies had shown the association between increased age and he development of CKD among type 2 diabetic patients. The pooled odds ratio showed that diabetics patients with increased age were 24 times more likely to develop CKD than those with younger (AOR = 24, 95 CI, (0.11–0.37) (Fig.  4 ). Four studies revealed that having history of type 2 diabetics has 33 times increased risk of CKD than non-diabetics patients (AOR = 33,95 CI, (0.18–0.48) (Fig.  5 ).

Association between age and chronic kidney disease among type 2 DM patients

Association between history of type 2 DM and chronic kidney disease among type 2 DM patients

Five studies reported that having cardiac disease with DM has 11 times More likely to develop CKD than without. (AOR = 11,95 CI, (0.7 − 0.14) (Fig.  6 ). Two studies showed that being obesity has 15 times more likely CKF than normal. AOR = 15,95 CI, (0.14–0.16)) (Fig.  7 ).

Association between having cardiac and chronic kidney disease among type 2 DM patients

Association between obesity and chronic kidney disease among type 2 DM patients

Four studies describe that smokers were 13 times more likely to develop CKD than non-smokers (AOR = 13,95 CI, (0.11–0.14) (Fig.  8 ). Nine studies discussed that having hypertension was 13 times more likely to get CKD than non-hypertensive DM patients (AOR = 13,95 CI, (0.05–0.21) (Fig.  9 ).

Association between smoking and chronic kidney disease among type 2 DM patients

Association between hypertension and chronic kidney disease among type 2 DM patients

Chronic kidney disease is associated with several consequences, including anemia, bone and mineral disorders, electrolyte imbalance, acid-base abnormalities, sexual dysfunction, hypertension, cardiovascular diseases [ 63 ]. Globally 850 million individuals live with the chronic kidney disease with all-age prevalence of 29.3% [ 64 , 65 ]. Type 2 diabetes, hypertension, and glomerulonephritis are the most frequent underlying diseases that lead to the development of CKD. Patients with CKD have an elevated risk of severe renal and cardiovascular morbidity and mortality [ 66 , 67 , 68 ].

Among the drugs that have been shown to slow the progression of chronic kidney disease are glucagon-like peptide 1 receptor agonists, sodium-glucose cotransporter-2 inhibitors, renin-angiotensin system inhibitors, and, more recently, non-steroidal mineralocorticoid receptor antagonists. Neither an ACE inhibitor nor an angiotensin receptor blocker is recommended for the primary prevention of chronic kidney disease in diabetics with normal blood pressure, urine albumin-to-creatinine ratio (< 30 mg/g creatinine), and normal estimated glomerular filtration rate. For people with type 2 diabetes and diabetic kidney disease, it is recommended to use a sodium-glucose cotransporter 2 inhibitor to slow the progression of cardiovascular events and chronic kidney disease in patients with estimated glomerular filtration rate ≥ 20 mL/min/1.73 m2 and urinary albumin ≥ 200 mg/g creatinine [ 18 , 36 , 69 ].

End-stage renal disease (ESRD) resulting from diabetic kidney disease has a number of systemic effects that lower quality of life, including electrolyte abnormalities, problems with bone metabolism, and renal anemia. Comprehensive care of hyperglycemia, hypertension, dyslipidemia, and healthy lifestyle choices is typically provided to those with diabetes and chronic kidney disease [ 70 , 71 , 72 ]. Approximately 40% of those with type 2 diabetes will develop CKD, and the risk rises with the duration of the disease. When compared to costs for people with diabetes alone, having CKD results in an average increase in healthcare costs of over 50% [ 73 , 74 ].

The goal of this systematic review and meta-analysis was to generate pooled estimates of the prevalence of chronic kidney disease among individuals who have type 2 diabetes mellitus globally. Based on the review the prevalence of chronic kidney disease among type 2 diabetes mellites was 27% (95% CI 21%, 33%).This was comparable with systematic review and meta-analysis study in Africa, which reported 24.7% pooled prevalence of CKD [ 75 ],in the middle east region the prevalence of CKD was 28.96% [ 76 ], Iran prevalence of nephropathy in patients with type 2 diabetes was 30.6% [ 77 ], in Asia the prevalence of CKD was 31% in adults with DM [ 78 ], and another systematic review and meta-analysis which include cohort studies in Europe, Asia and America report the prevalence of CKD among T2DM was approximately 29.1% [ 79 ].

This systematic review and meta-analysis was lower than study reported a pooled prevalence of end stage renal disease was 73% [ 80 ] and in sub Saharan Africa the overall prevalence of diabetic nephropathy was 35.3% [ 81 ] and the study was higher than the study in Africa among the general population which was 15.8% [ 82 ], In Asia the pooled prevalence of CKD among the general population was 14% [ 78 ],another cohort systematic review and meta-analysis reported that ESRD among on T2DM was 1.1% [ 83 ]. The discrepancy could be differences in study period, sample size, demographics characteristics, diabetic duration, and study design of included studies.

This systematic review and meta-analysis include 20 studies in 13 countries with total 1,711,926 study participants. According to the review, USA had the highest incidence of CKD at 54.8% and the United Arab Emirates had the lowest prevalence at 4% [ 46 , 52 ]. The funnel plot’s evidence the results seemed like asymmetrical funnel plots suggests current study had a source of publication bias. but the result of egger’s test was not statistically significant for the presence of publication bias (P = 0.254).

The study revealed that being old age, hypertension, cardiac disease, smoking, obesity, having type 2 diabetes mellitus was predictor variable for presence of chronic kidney disease among type 2 diabetic patients. This was consistent with the study reported that age, smoking, and DM duration has significant association with CKD development among diabetic patient [ 83 ].Another systematic review showed that duration of diabetics and having hypertension and obesity has increased risk of chronic kidney disease among diabetic patients [ 81 , 84 , 85 ]. The study in China also showed that old age, smoking, and hypertension were predictor variables of CKD with increased morbidity and mortality in T2DM patients [ 86 ].

This systematic review and meta-analysis revealed that the prevalence of chronic kidney disease among type 2 diabetes mellites patients was high based on the included 20 articles. The prevalence of chronic kidney disease varies across country with the highest in USA and the lowest in Ethiopia. The review reported that old age, hypertension, having cardiac disease, smoking, obesity, having history of type 2 diabetes mellitus were predictor variables for chronic kidney disease among diabetic patients. Patients with CKD have an elevated risk of severe renal and cardiovascular morbidity and mortality. Therefore, in order to lower the morbidity and mortality from chronic kidney disease among type 2 diabetic patients, it is advised to develop both preventive and curative intervention strategies, such as raising awareness, creating a supportive environment and prescribing appropriate medication at an early stage.

Goyal R, Jialal I. Diabetes Mellitus Type 2: StatPearls Publishing, Treasure Island (FL); 2022; 2022.

Levey AS. Coresh JJTl. Chronic Kidney Disease. 2012;379(9811):165–80.

Google Scholar  

Poretsky L. Principles of Diabetes Mellitus. Springer; 2010.

Galicia-Garcia U, Benito-Vicente A, Jebari S, Larrea-Sebal A, Siddiqi H, Uribe KB et al. Pathophysiology of type 2 Diabetes Mellitus. 2020;21(17):6275.

Cheng AY, Gomes MB, Kalra S, Kengne A-P, Mathieu C, Shaw JEJNRE. Applying the WHO global targets for diabetes mellitus. 2023:1–7.

Zheng Y, Ley SH, Hu FB. Global aetiology and epidemiology of type 2 Diabetes Mellitus and its Complications. Nat Rev Endocrinol. 2018;14(2):88–98.

Article   PubMed   Google Scholar  

Ginter E. Simko VJDaod, a new insight. Type 2 diabetes mellitus, pandemic in 21st century. 2013:42–50.

Zhou B, Lu Y, Hajifathalian K, Bentham J, Di Cesare M, Danaei G et al. Worldwide trends in Diabetes since 1980: a pooled analysis of 751 population-based studies with 4· 4 million participants. 2016;387(10027):1513–30.

Atlas DJIDA. Brussels, Belgium: International Diabetes Federation. Int Diabetes Federation. 2015;33:2. 7th ed.

Shen Y, Cai R, Sun J, Dong X, Huang R, Tian S et al. Diabetes Mellitus as a risk factor for incident chronic Kidney Disease and end-stage renal Disease in women compared with men: a systematic review and meta-analysis. 2017;55:66–76.

Rasooly RS, Akolkar B, Spain LM, Guill MH, Del Vecchio CT, Carroll LEJCJotASoN. The National Institute of Diabetes and Digestive and kidney Diseases Central repositories: a valuable resource for nephrology research. 2015;10(4):710–5.

Alicic RZ, Rooney MT, Tuttle, KRJCjotASoN. Diabetic Kidney Disease: challenges, progress, and possibilities. 2017;12(12):2032–45.

Heo GY, Koh HB, Kim HW, Park JT, Yoo T-H, Kang S-W, et al. Glycemic control and adverse clinical outcomes in patients with chronic Kidney Disease and type 2 Diabetes Mellitus. Results from KNOW-CKD; 2023.

Chung H, Crowe CL, Kong SX, Singh R, Farej R, Elliott J et al. Descriptive study of the economic burden among patients with type 2 Diabetes Mellitus, chronic Kidney Disease, and chronic Kidney Disease and type 2 Diabetes Mellitus in a large US commercially insured population. 2023;29(1):80–9.

Kainz A, Hronsky M, Stel VS, Jager KJ, Geroldinger A, Dunkler D et al. Prediction of prevalence of chronic Kidney Disease in diabetic patients in countries of the European Union up to 2025. 2015;30(suppl_4):iv113–iv8.

Imtiaz S, Salman B, Qureshi R, Drohlia MF, Ahmad AJSJoKD. Transplantation. A review of the epidemiology of chronic Kidney Disease in Pakistan: a global and regional perspective. 2018;29(6):1441–51.

Neuen BL, Chadban SJ, Demaio AR, Johnson DW, Perkovic V. Chronic Kidney Disease and the global NCDs agenda. 2017;2(2):e000380.

Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW, et al. National Kidney Foundation practice guidelines for chronic Kidney Disease: evaluation, classification, and stratification. Ann Intern Med. 2003;139(2):137–47.

Levey AS, Becker C, Inker LAJJ. Glomerular filtration rate and albuminuria for detection and staging of acute and chronic Kidney Disease in adults: a systematic review. 2015;313(8):837–46.

Valmadrid CT, Klein R, Moss SE, Klein BEJA. The risk of Cardiovascular Disease mortality associated with microalbuminuria and gross proteinuria in persons with older-onset Diabetes Mellitus. 2000;160(8):1093–100.

Chen W, Chen W, Wang H, Dong X, Liu Q, Mao H et al. Prevalence and risk factors associated with chronic Kidney Disease in an adult population from southern China. 2009;24(4):1205–12.

Chukwuonye II, Ogah OS, Anyabolu EN, Ohagwu KA, Nwabuko OC, Onwuchekwa U et al. Prevalence of chronic kidney disease in Nigeria: systematic review of population-based studies. 2018:165 – 72.

Ephraim RK, Biekpe S, Sakyi SA, Adoba P, Agbodjakey H, Antoh EOJC et al. Prevalence of chronic Kidney Disease among the high risk population in South-Western Ghana; a cross sectional study. 2015;2:76.

Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW et al. National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classification, and stratification. 2003;139(2):137 – 47.

Moosa M, Van der Walt I, Naicker S, Meyers AJSA. Important causes of chronic Kidney Disease in South Africa. 2015;105(4):320–7.

Takamatsu KJC, Nephrology E. Renal status in elderly patients with type 2 Diabetes. 2020;24:53–62.

Wyld ML, Morton RL, Clayton P, Wong MG, Jardine M, Polkinghorne K et al. The impact of Progressive chronic Kidney Disease on health-related quality-of-life: a 12-year community cohort study. 2019;28:2081–90.

Thomas MC, Cooper ME, Zimmet PJNRN. Changing epidemiology of type 2 Diabetes Mellitus and associated chronic Kidney Disease. 2016;12(2):73–81.

Li H, Lu W, Wang A, Jiang H, Lyu JJJ. Changing epidemiology of chronic Kidney Disease as a result of type 2 Diabetes Mellitus from 1990 to 2017: estimates from global burden of Disease 2017. 2021;12(3):346–56.

Saran R, Robinson B, Abbott KC, Agodoa LY, Albertus P, Ayanian J, et al. Epidemiol kidney disease United States. 2017;69(3):A7–A8. US renal data system 2016 annual data report:.

ElHafeez SA, Bolignano D, D’Arrigo G, Dounousi E, Tripepi G, Zoccali C. Prevalence and burden of chronic Kidney Disease among the general population and high-risk groups in Africa: a systematic review. 2018;8(1):e015069.

Skolnik NS, Style AJJDT. Importance of early screening and diagnosis of chronic Kidney Disease in patients with type 2 Diabetes. 2021;12(6):1613–30.

Toyama T, Neuen BL, Jun M, Ohkuma T, Neal B, Jardine MJ et al. Effect of SGLT2 inhibitors on cardiovascular, renal and safety outcomes in patients with type 2 Diabetes Mellitus and chronic Kidney Disease: a systematic review and meta-analysis. 2019;21(5):1237–50.

Zelniker TA, Wiviott SD, Raz I, Im K, Goodrich EL, Furtado RH et al. Comparison of the effects of glucagon-like peptide receptor agonists and sodium-glucose cotransporter 2 inhibitors for prevention of major adverse cardiovascular and renal outcomes in type 2 Diabetes Mellitus: systematic review and meta-analysis of cardiovascular outcomes trials. 2019;139(17):2022–31.

Bikbov B, Purcell C, Levey A, Smith M, Abdoli A, Abebe M et al. GBD chronic Kidney Disease collaboration: global, regional, and national burden of chronic Kidney Disease, 1990–2017: a systematic analysis for the global burden of Disease Study 2017. 2020;395(709–733):32061315.

Schmidt IM, Hübner S, Nadal J, Titze S, Schmid M, Bärthlein B et al. Patterns of medication use and the burden of polypharmacy in patients with chronic Kidney Disease: the German chronic Kidney Disease study. 2019;12(5):663–72.

Alwafi H, Wei L, Naser AY, Mongkhon P, Tse G, Man KK et al. Trends in oral anticoagulant prescribing in individuals with type 2 Diabetes Mellitus: a population-based study in the UK. 2020;10(5):e034573.

Eknoyan G, Lameire N, Eckardt K, Kasiske B, Wheeler D, Levin A et al. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic Kidney Disease. 2013;3(1):5–14.

Okpara CV, Anselm AU, Felix TO, Omowale A, Gever VCJH. The moderating role of colour in modelling the effectiveness of COVID-19 YouTube animated cartoons on the health behaviour of social media users in Nigeria. 2021;36(6):1599–609.

Masresha SA, Alen GD, Kidie AA, Dessie AA, Dejene TMJSR. First line antiretroviral treatment failure and its association with drug substitution and sex among children in Ethiopia: systematic review and meta-analysis. 2022;12(1):18294.

Higgins J, Green S. Cochrane handbook for systematic reviews of interventions: cochrane book series. Cochrane handbook for systematic reviews of interventions: cochrane book series. 2008. John Wiley & Sons Location: Hoboken, New Jersy, USA; 2008.

Sterne JA, Egger MJJ. Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis. 2001;54(10):1046–55.

Ahmed MA, Ferede YM, Takele WWJP. Incidence and predictors of chronic Kidney Disease in type-II Diabetes Mellitus patients attending at the Amhara region referral hospitals, Ethiopia: a follow-up study. 2022;17(1):e0263138.

Damtie S, Biadgo B, Baynes HW, Ambachew S, Melak T, Asmelash D et al. Chronic Kidney Disease and associated risk factors assessment among Diabetes Mellitus patients at a tertiary hospital, Northwest Ethiopia. 2018;28(6).

Tannor EK, Sarfo FS, Mobula LM, Sarfo-Kantanka O, Adu‐Gyamfi R, Plange‐Rhule JJTH. Prevalence and predictors of chronic Kidney Disease among Ghanaian patients with Hypertension and Diabetes Mellitus: a multicenter cross‐sectional study. 2019;21(10):1542–50.

Al-Shamsi S, Regmi D, Govender RJP. Chronic Kidney Disease in patients at high risk of Cardiovascular Disease in the United Arab Emirates: a population-based study. 2018;13(6):e0199920.

Sukkar L, Kang A, Hockham C, Young T, Jun M, Foote C et al. Incidence and associations of chronic Kidney Disease in community participants with Diabetes: a 5-year prospective analysis of the EXTEND45 study. 2020;43(5):982–90.

Bramlage P, Lanzinger S, van Mark G, Hess E, Fahrner S, Heyer CH et al. Patient and Disease characteristics of type-2 Diabetes patients with or without chronic Kidney Disease: an analysis of the German DPV and DIVE databases. 2019;18(1):1–12.

González-Pérez A, Saéz ME, Vizcaya D, Lind M, Rodríguez LAGJP. Impact of chronic Kidney Disease definition on assessment of its incidence and risk factors in patients with newly diagnosed type 1 and type 2 Diabetes in the UK: a cohort study using primary care data from the United Kingdom. 2020;14(4):381–7.

Sim R, Chong CW, Loganadan NK, Adam NL, Hussein Z, Lee SWHJC. Comparison of a chronic Kidney Disease predictive model for type 2 Diabetes Mellitus in Malaysia using Cox regression versus machine learning approach. 2023;16(3):549–59.

Nazzal Z, Hamdan Z, Masri D, Abu-Kaf O, Hamad, MJBn. Prevalence and risk factors of chronic Kidney Disease among Palestinian type 2 diabetic patients: a cross-sectional study. 2020;21:1–8.

Medicine TSGJNEJo. Long-term Complications in youth-onset type 2 Diabetes. 2021;385(5):416–26.

Zhang J, Chen Y, Zou L, Gong RJAD. Prognostic nutritional index as a risk factor for diabetic Kidney Disease and mortality in patients with type 2 Diabetes Mellitus. 2023;60(2):235–45.

McGuire DK, Busui RP, Deanfield J, Inzucchi SE, Mann JF, Marx N et al. Effects of oral semaglutide on cardiovascular outcomes in individuals with type 2 Diabetes and established atherosclerotic Cardiovascular Disease and/or chronic Kidney Disease: SOUL, a randomized trial, design and baseline characteristics. 2023.

Duan J, Wang C, Liu D, Qiao Y, Pan S, Jiang D et al. Prevalence and risk factors of chronic Kidney Disease and diabetic Kidney Disease in Chinese rural residents: a cross-sectional survey. 2019;9(1):10408.

Yang L, Chu TK, Lian J, Lo CW, Lau PK, Nan H et al. Risk factors of chronic kidney Diseases in Chinese adults with type 2 Diabetes. 2018;8(1):14686.

Candela JM, González JS, Soidán FJG, Trillo JMM, Espino JD, Bosque DB et al. Chronic renal Disease in Spain: prevalence and related factors in persons with Diabetes Mellitus older than 64 years. 2018;38(4):395–407.

Usó-Talamantes R, González-de-Julián S, Díaz-Carnicero J, Saurí-Ferrer I, Trillo-Mata JL, Carrasco-Pérez M et al. Cost of type 2 Diabetes patients with chronic Kidney Disease based on real-world data: an observational population-based study in Spain. 2021;18(18):9853.

Jitraknatee J, Ruengorn C, Nochaiwong SJS. Prevalence and risk factors of chronic Kidney Disease among type 2 Diabetes patients: a cross-sectional study in primary care practice. 2020;10(1):6205.

Pradidthaprecha A, Muktabhant B, Schelp FP, Tatiyanupanwong S, Thinkhamrop W, Chaichaya N. Renal Disease progression and associated factors in type 2 Diabetes patients: A retrospective cohort study.

Zaman SB, Hossain N, Rahman MJD. journal m. Associations between body mass index and chronic kidney disease in type 2 diabetes mellitus patients: findings from the northeast of Thailand. 2018;42(4):330-7.

Birkeland KI, Bodegard J, Eriksson JW, Norhammar A, Haller H, Linssen GC et al. Heart Failure and chronic Kidney Disease manifestation and mortality risk associations in type 2 Diabetes: a large multinational cohort study. 2020;22(9):1607–18.

Bikbov B, Purcell CA, Levey AS, Smith M, Abdoli A, Abebe M et al. Global, regional, and national burden of chronic Kidney Disease, 1990–2017: a systematic analysis for the global burden of Disease Study 2017. 2020;395(10225):709–33.

Jager KJ, Kovesdy C, Langham R, Rosenberg M, Jha V, Zoccali C. A single number for advocacy and communication—worldwide more than 850 million individuals have kidney Diseases. Nephrol Dialysis Transplantation. 2019;34(11):1803–5.

Article   Google Scholar  

Bikbov B, Purcell CA, Levey AS, Smith M, Abdoli A, Abebe M, et al. Global, regional, and national burden of chronic kidney disease, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. The Lancet. 2020;395(10225):709–33.

Xie Y, Bowe B, Mokdad AH, Xian H, Yan Y, Li T et al. Analysis of the Global Burden of Disease study highlights the global, regional, and national trends of chronic kidney disease epidemiology from 1990 to 2016. 2018;94(3):567 – 81.

Carpio EM, Ashworth M, Asgari E, Shaw C, Schartau P, Durbaba S et al. Hypertension and cardiovascular risk factor management in a multi-ethnic cohort of adults with CKD: a cross sectional study in general practice. 2022:1–10.

Fletcher BR, Damery S, Aiyegbusi OL, Anderson N, Calvert M, Cockwell P et al. Symptom burden and health-related quality of life in chronic Kidney Disease: a global systematic review and meta-analysis. 2022;19(4):e1003954.

ElSayed NA, Aleppo G, Aroda VR, Bannuru RR, Brown FM, Bruemmer D, et al. Chronic Kidney Disease and Risk Management: Standards of Care in Diabetes—2023. 2023;11(Supplement1):191–S202.

Kohan DE, Barton M. Endothelin and endothelin antagonists in chronic Kidney Disease. Kidney Int. 2014;86(5):896–904.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Levey AS, Gansevoort RT, Coresh J, Inker LA, Heerspink HL, Grams ME, et al. Change in Albuminuria and GFR as End points for clinical trials in early stages of CKD: A Scientific Workshop Sponsored by the National Kidney Foundation in collaboration with the US Food and Drug Administration and European Medicines Agency. Am J Kidney Dis. 2020;75(1):84–104.

Article   CAS   PubMed   Google Scholar  

Cherney DZI, Bell A, Girard L, McFarlane P, Moist L, Nessim SJ, et al. Management of type 2 Diabetic Kidney Disease in 2022: a narrative review for specialists and primary care. Can J Kidney Health Disease. 2023;10:20543581221150556.

Li R, Bilik D, Brown MB, Zhang P, Ettner SL, Ackermann RT, et al. Medical costs associated with type 2 Diabetes Complications and comorbidities. Am J Manag Care. 2013;19(5):421–30.

PubMed   PubMed Central   Google Scholar  

International Diabetes Federation. Diabetes Atlas10th edn. https://diabetesatlas.org/en/resources .

Abd ElHafeez S, Bolignano D, D’Arrigo G, Dounousi E, Tripepi G, Zoccali CJB. Prevalence and burden of chronic Kidney Disease among the general population and high-risk groups in Africa: a systematic review. 2018;8(1):e015069.

Naser AY, Alwafi H, Alotaibi B, Salawati E, Samannodi M, Alsairafi Z et al. Prevalence of chronic kidney diseases in patients with diabetes mellitus in the Middle East: A systematic review and meta-analysis. 2021;2021.

Rabieenia E, Jalali R, Mohammadi M. Prevalence of nephropathy in patients with type 2 diabetes in Iran: A systematic review and meta-analysis based on geographic information system (GIS). Diabetes & Metabolic Syndrome: Clinical Research & Reviews. 2020;14(5):1543-50.

Shrestha N, Gautam S, Mishra SR, Virani SS, Dhungana RRJP. Burden of chronic Kidney Disease in the general population and high-risk groups in South Asia: a systematic review and meta-analysis. 2021;16(10):e0258494.

Jiang W, Wang J, Shen X, Lu W, Wang Y, Li W, et al. Establishment and validation of a risk Prediction Model for Early Diabetic Kidney Disease based on a Systematic Review and Meta-analysis of 20 cohorts. Diabetes Care. 2020;43(4):925–33.

Buchan TA, Malik A, Chan C, Chambers J, Suk Y, Zhu JW, et al. Predictive models for cardiovascular and kidney outcomes in patients with type 2 Diabetes: systematic review and meta-analyses. Heart. 2021;107(24):1962–73.

Wagnew F, Eshetie S, Kibret GD, Zegeye A, Dessie G, Mulugeta H, et al. Diabetic Nephropathy and Hypertension in Diabetes patients of sub-saharan countries: a systematic review and meta-analysis. BMC Res Notes. 2018;11(1):565.

Article   PubMed   PubMed Central   Google Scholar  

Kaze AD, Ilori T, Jaar BG, Echouffo-Tcheugui JB. Burden of chronic Kidney Disease on the African continent: a systematic review and meta-analysis. BMC Nephrol. 2018;19(1):125.

Ren Q, Chen D, Liu X, Yang R, Yuan L, Ding M, et al. Derivation and validation of a prediction model of end-stage renal Disease in patients with type 2 Diabetes based on a systematic review and Meta-analysis. Front Endocrinol. 2022;13:825950.

Tomino Y, Gohda TJKD. The prevalence and management of diabetic Nephropathy in Asia. 2015;1(1):52–60.

Garofalo C, Borrelli S, Minutolo R, Chiodini P, De Nicola L. Conte GJKi. A systematic review and meta-analysis suggests obesity predicts onset of chronic Kidney Disease in the general population. 2017;91(5):1224–35.

Liao X, Shi K, Zhang Y, Huang X, Wang N, Zhang L et al. Contribution of CKD to mortality in middle-aged and elderly people with Diabetes: the China Health and Retirement Longitudinal Study. 2023;15(1):1–8.

Download references

This study was not funded by any organization.

Author information

Authors and affiliations.

Department of public health, college of medicine and health sciences, Injibara University, Injibara, Ethiopia

Eneyew Talie Fenta

Department of Health Promotion and Health Behavior, Institute of Public Health, College of Medicine and Health Sciences, University of Gondar, PO. Box.196, Gondar, Ethiopia

Habitu Birhan Eshetu

Department of Health Promotion, School of Public Health, College of Medicine and Health Sciences, Wollo University, Dessie, Ethiopia

Natnael Kebede

Health Promotion and Behavioral science department, College of medicine and health science, Bahir Dar, Houston, Ethiopia

Eyob Ketema Bogale & Tadele Fentabil Anagaw

Department of Public Health, College of Medicine and Health Science, Wolkite University, Wolkite, Ethiopia

Amare Zewdie

Department of Public Health, College of Medicine and Health Science, Debre Markos university, Debre Markos, Ethiopia

Tadele Derbew Kassie, Elyas Melaku Mazengia & Sintayehu Shiferaw Gelaw

You can also search for this author in PubMed   Google Scholar

Contributions

ETF, TFA, EKB, ZA, and SSG search and extract the articles, TDK, TFA EMM, and HBE check the quality of the articles, NK, EKB, ETF, ZA, and TFA do the analysis part and write the result, ZA,TDK, TFA EMM, and HBE review the manuscript. ETF, TFA, EKB, and SSG revised the manuscript. Finally,all authors gave approval of the version to be published; agreed on the journal to which the article had been submitted; and agreed to be accountable for all aspects of the work.

Corresponding author

Correspondence to Eneyew Talie Fenta .

Ethics declarations

Ethical approval.

Not applicable.

Competing interests

The authors declare no competing interests.

Data Availability

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

The authors declare that they have no competing interests.

Additional information

Publisher’s note.

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

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.

Fenta, E.T., Eshetu, H.B., Kebede, N. et al. Prevalence and predictors of chronic kidney disease among type 2 diabetic patients worldwide, systematic review and meta-analysis. Diabetol Metab Syndr 15 , 245 (2023). https://doi.org/10.1186/s13098-023-01202-x

Download citation

Received : 13 July 2023

Accepted : 31 October 2023

Published : 28 November 2023

DOI : https://doi.org/10.1186/s13098-023-01202-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

• Chronic kidney disease
• Meta-analysis
• Type 2 diabetes mellitus

Diabetology & Metabolic Syndrome

ISSN: 1758-5996

• Submission enquiries: [email protected]

• Open access
• Published: 27 November 2023

The underlying relationship between exercise and the prevalence of periodontitis: a systematic review and meta-analysis

• Rongkai Cao 1   na1 ,
• Piaopiao Qiu 1   na1 ,
• Yuan Zhou 1 ,
• Bo Dong 1 ,
• Yucheng Han 1 &
• Zhen Fan 1  

BMC Sports Science, Medicine and Rehabilitation volume  15 , Article number:  161 ( 2023 ) Cite this article

44 Accesses

Metrics details

Although exercise has been proposed as an effective intervention for various lifestyle-related diseases and pathological scenarios, few researches assessed the impact of taking exercise on the prevalence of periodontitis.

This study aimed to perform a comprehensive literature review and meta-analysis from both observational and intervention studies to explore the relationship between exercise and periodontitis and to provide references for future intervention programs aimed at preventing periodontitis.

A systematic literature search was conducted in PubMed/MEDLINE, Web of Science, Cochrane Library, and Scopus for peer-reviewed studies published in English From January 1993 to January 2023 according to the PRISMA guidelines. Articles were selected if subjects were human and studies evaluating the association between exercise and periodontitis.

4098 references were retrieved. After screening the results, 30 studies were selected. Of these, 20 studies indicated an inverse association between exercise and periodontitis, while the remaining 10 did not reach conclusive findings. The meta-analysis demonstrated a risk ratio of 0.84 (95%CI: 0.77, 0.91) between the active group and the inactive group (P < 0.01), which indicated an inverse relationship between exercise and periodontitis.

Overall, the increase in exercise presents an inverse association with the presence and severity of periodontitis. Accordingly, taking exercise might be a potential approach that contributes to improvements in periodontitis.

Peer Review reports

Introduction

Exercise has been shown to have numerous beneficial effects on the general health of human beings and increase life expectancy [ 1 ]. Furthermore, studies have indicated the protective effects and improvement of taking moderate exercise on various lifestyle-related diseases including obesity, arthritis, hypertension, depression, and coronary heart disease [ 2 ]. Taking regularly moderate exercise can alleviate systemic inflammatory reactions and reduce the risk of multiple traditional chronic diseases [ 3 ]. Moreover, regular exercise has a positive impact on the management of stress and anxiety [ 4 ]. In addition to these physical and mental benefits, taking exercise may also help reduce healthcare costs and the need for medications through lower hospital stays and physician visits [ 5 ].

Periodontal health is defined as a state free from inflammation and characterized by the absence of gingival bleeding and shallow pockets [ 6 ]. Periodontitis is a chronic multifactorial inflammatory disease characterized by progressive destruction of the tooth-supporting apparatus and combined with dysbiotic plaque biofilms [7]. This disease is prevalent worldwide, affecting approximately 20–50% of the global population, making it a significant public health concern [ 8 ]. Although periodontitis mainly threatens oral health, its association with systemic conditions such as cardiovascular diseases, diabetes, and arthritis has also been established in the literature [ 9 , 10 ]. However, given the high physical, psychological, social, and economic impacts on individuals and communities, prevention and management of periodontitis are crucial globally. A previous study summarized data related to severe periodontitis and indicated that governments should pay attention to the growing burden of severe periodontitis because of the increasing population associated with a better life expectancy around the world and a reduction of tooth loss [ 11 ].

Various lifestyle factors, including diet quality, alcohol consumption, and smoking, have been found to impact the occurrence and severity of periodontitis [ 8 ]. Substantial evidence suggests that interventions targeting lifestyle behaviors are effective, underscoring the importance of behavioral support in managing periodontitis [ 12 ]. Significant factors such as diet and smoking have been thoroughly investigated regarding their role in the risk of periodontal tissue decay [ 13 , 14 ]. Nevertheless, there is limited research on the potential role of exercise in preventing periodontitis. Most studies regarded exercise as a part of lifestyle variables when evaluating the risk of periodontitis. However, few researchers investigated the level of exercise in detail, and the intervention studies are also limited in current literature.

Given that exercise has been shown to modulate systemic inflammatory reactions and alter inflammatory markers such as protein C-reactive levels, there may be a potential relationship between exercise and periodontitis [ 15 , 16 ]. However, it is unclear whether exercise has a direct impact on periodontitis. Previous studies have tried to summarize the potential relationship between exercise and periodontitis [ 17 , 18 ]. For example, a meta-analysis investigated the influence of physical exercise on periodontitis [ 17 ]. Another study also suggested an association between periodontitis and physical activity [ 18 ]. However, only 6 studies were included in this previous study [ 18 ]. Accordingly, the results should be considered carefully due to the limited research available and the certain risk of bias. Furthermore, only observational researches were included in these previous studies. Thus, the goal of this study was to perform a comprehensive meta-analysis from both observational and intervention studies to investigate the potential relationship between exercise and the prevalence of periodontitis and provide references for future intervention programs aimed at preventing periodontitis.

Materials and methods

The present study was registered at PROSPERO under registration number CRD42023396334 and performed based on the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [ 19 ]. Given that no human or animal subjects were involved in this study, medical ethics committee approval was not required.

The guiding question for this systematic review was formulated using the PICO format, with (P) indicating participants, (I) representing intervention, (C) standing for comparison, and (O) representing outcome [ 20 ]. Specifically, the guiding question was, “For individuals of all ages (P), what is the impact of exercise (I) or a sedentary lifestyle (C) on periodontitis (O)?” Participants of all ages were included. The intervention group contained subjects who engaged in regular or recommended exercise. The comparison group consisted of individuals who engaged in inadequate exercise or a sedentary lifestyle. The primary outcomes were the measurements of periodontitis.

Search strategy

To identify relevant publications in accordance with PRISMA guidelines, a systematic literature search was performed in Web of Science, PubMed/MEDLINE, Scopus, and Cochrane Library databases. The search was limited to English-language publications published From January 1993 to January 2023. The search strategy used predefined search terms related to exercise and periodontitis, and the specific search strategies used for each database are provided in Table  1 . Gray literature searches were also conducted on the International Clinical Trials Registry Platform of the World Health Organization and SciELO. To complement the study, the search in the databases was combined with a manual retrieval of the reference lists from the selected studies.

Eligibility criteria

The following criteria were applied to choose publications: (1) peer-reviewed research articles published in English, (2) studies conducted on human subjects with no age restrictions, and (3) observational or experimental studies that examined the relationship between exercise and periodontitis.

The following exclusion criteria were performed: (1) studies conducted on animals, case reports, letters, editorials, conference abstracts, and review articles; (2) studies with insufficient data to extract relevant information including the exercise measurement and evaluation of periodontitis; and (3) publications written in languages other than English.

Study selection

The information obtained by search strategy in each database was collated, and duplicate entries were eliminated. Two reviewers independently assessed the abstract and title of the retrieved articles based on eligibility criteria. Studies which were considered ineligible by both reviewers were promptly excluded, while studies deemed ineligible by one author but eligible by another were retained for full-text evaluation. The reviewers worked in tandem to analyze articles not excluded in their entirety. Publications that satisfied the eligibility criteria were used for further data extraction. Any discrepancies were resolved by means of discussion.

Data from the included studies were retrieved in detail. Report of the following variables was extracted: author(s), year of publication, study type, country, number of participants, gender, age, exercise assessment, evaluation of periodontitis, main findings, and conclusions.

Quality assessment

The risk of bias for observational studies was evaluated in accordance with the “Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies” and the “Quality Assessment of Case-Control Studies” from the National Institutes of Health (Available online at https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools ). These tools comprise 14 criteria (12 for case-control studies) designed to appraise the internal validity of cross-sectional studies, case-control studies, and cohort studies. Responses to each criterion could be “yes,” “no,” “not applicable,” or “not reported.” Responses other than “yes” indicate a potential risk of bias. After evaluation, articles were defined as good, fair, or poor. Studies rated as “good” had a maximum of three domains that were not answered as “no” or “not applicable.” The domains of “Validity of outcomes” and “Adjustment for confounders” were considered the most critical factors in determining the classification of poor study quality.

Furthermore, the risk of bias for experimental researches was assessed based on the Cochrane Collaboration’s tool, using Review Manager 5.4 (The Nordic Cochrane Center, The Cochrane Collaboration, Copenhagen, Denmark) [ 21 ]. The tool contains the following criteria: allocation concealment, random sequence generation, blinding of outcome assessment, blinding of participants, incomplete outcome data, blinding of operator, and selective reporting. Each study’s risk of bias was classified as low, with some concerns, or high based on these criteria.

The same reviewers independently analyzed the included studies using the above tools. In case of any discrepancies, a consensus decision was made through further discussion with a third review author.

Data synthesis

Data on exercise and periodontitis were extracted and analyzed using Stata 17.0 Software (College Station, TX: StataCorp LLC, USA) to evaluate the underlying relationship between exercise and periodontitis. Specifically, the prevalence of periodontitis and the number of individuals in active and inactive groups were recorded to determine the odds ratio with a 95% confidence interval (95%CI). Heterogeneity was calculated using the I 2 test, which measures the rate of variation among articles caused by heterogeneity instead of probabilistic chance. The fixed-effects model and the random-effects model were used to test the significance of treatment effects in case of absence and high heterogeneity, respectively [ 22 ]. The estimated effect was regarded as significant where P  < 0.05.

4,098 papers were initially obtained through an electronic search, comprising 899 articles from PubMed/MEDLINE, 1,324 articles from Scopus, 1,681 articles from Web of Science, and 194 articles from Cochrane Library. Furthermore, a manual search of manufacturers’ reference lists yielded seven additional articles. None of the 243 references from gray literature was considered eligible. 2,995 studies remained after removing duplicates, and 2959 of them were excluded after assessing the titles and abstracts. After thoroughly examining the full texts, 13 studies were excluded since they did not meet the eligibility criteria. At last, 30 articles were selected for inclusion in this systematic review (Fig.  1 ).

Flow chart of the literature search and results

Characteristics of the studies

Tables  2 , 3 and 4 provide detailed information on the 30 studies included in this review, of which 4 were experimental studies and 26 were observational studies. 20 of the included articles were cross-sectional design and most of them were performed in developed countries. Subjects of all ages were included, with a minimum of 25 and a maximum of 16,726. There was considerable variation in the tools used to assess physical activity, including different versions of questionnaires, strength, and maximal oxygen consumption. 18 studies elaborated on the types of sports, and 14 studies reported the duration time of measurement. The frequency and intensity of exercise have also been included in most selected studies. The assessment of periodontitis included clinical attachment loss and periodontal probing depth in most studies, with additional measures including gingival index, plaque index, bleeding on probing, and community periodontal index. The selected studies also investigated sedentary behavior, diet quality, normal body weight, alcohol consumption, and cigarette smoking as potential co-factors. Overall, from the 30 included studies, 20 studies reported that taking exercise was combined with a low prevalence of periodontitis, while the remaining 10 studies did not reach conclusive findings. For example, a representative cross-sectional study indicated that a higher prevalence of periodontitis was found in those inactive individuals compared to partially active individuals and subjects took recommended exercise [ 25 ]. A low level of exercise combined with a poor diet demonstrated a significant association with increased odds of periodontitis in another included study [ 27 ]. Possible explanations can be obtained from the included intervention studies. One research concluded that taking exercise may alter total salivary antioxidants activity, which could eventually affect periodontal health status [ 44 ]. Another study indicated that leisure-time exercise may protect against an excessive inflammatory response in periodontitis [ 46 ]. In addition, none of the selected studies reported a negative effect. Other main potential co-factors reported in the selected studies include maintaining a normal weight, consuming a healthy diet, sedentary behavior, and cigarette smoking.

Risk of bias

Table  5 presents a summary of the quality assessment of observational studies. Three publications were considered to be of good quality with low risk, providing adequate information about the population and objectives of the study, and how the outcomes and exposures were measured. 18 studies were rated as fair, while four publications were rated as poor. In terms of the experimental studies, two studies showed a high risk of random sequence generation since subjects were not randomly assigned to the experimental and control groups. Most studies also exhibited a high risk of bias in blinding of participants because participants were aware of their grouping under different intervention conditions. See Fig.  2 for a graphical representation of the risk of bias assessment.

Risk of bias of experimental studies using Cochrane Collaboration’s tool

Meta-analysis

Among 30 selected articles, 16 were included in the meta-analysis, while 14 publications were excluded due to variability in study outcomes and methodologies. The results presented high heterogeneity (I2 = 88) with statistical significance. Since the selected studies were not functionally equivalent, Random-effects models were used to generalize the results from meta-analysis. The prevalence of periodontitis was 27.85% in the active group and 33.88% in the inactive group, respectively. The Forest plot demonstrated a risk ratio of 0.84 (95%CI: 0.77, 0.91) between the active group and the inactive group (P < 0.01), which indicated a positive association between exercise and periodontitis (Fig.  3 ). The funnel plot did not reveal any publication bias (Fig.  4 ).

Forest plot of the relationship between exercise and prevalence of periodontitis

Funnel plot of the relationship between exercise and prevalence of periodontitis

Regular exercise is considered an essential aspect of many individuals’ lives, and it has been postulated to positively impact periodontitis through various mechanisms, including improved insulin sensitivity, decreased incidence of lifestyle-related diseases and obesity, stress reduction, and decreased inflammation reactions [ 53 ]. Furthermore, lifestyle and obesity are closely linked to the development of periodontitis [ 54 ]. Despite this, few studies have investigated the relationship between exercise and periodontitis. Two previous published meta-analyses have tried to clarify the underlying relationship between exercise and periodontitis [ 17 , 18 ]. However, the results should be considered carefully to some extent due to the limited research available and certain risk of bias. In addition, these previous meta-analyses only included observational studies. Conversely, after a comprehensive literature review, 30 studies were selected for inclusion in our study. Moreover, in this study, we included both experimental studies and observational studies. Following a systematic review, the included studies demonstrated a significant correlation between exercise and periodontitis. To our knowledge, this is one of the first reviews that focus on the influence of taking exercise on periodontitis.

As the fourth leading risk factor for global mortality, physical inactivity has been identified as a modifiable risk factor for various diseases, such as diabetes mellitus, hypertension, cardiovascular disease, and osteoporosis [ 55 ]. Exercise and physical activity have been shown to enhance life quality and increase life expectancy, resulting in an increasing focus on these activities by organizations [ 56 ]. The terms exercise and physical activity have been currently used interchangeably in the literature [ 57 ]. Therefore, we included both of them in the systematic review to improve the reliability. Previous research has indicated that exercise is linked to a favorable inflammatory profile, which may provide some protection against oral diseases [ 58 ]. The World Health Organization (WHO) recommends a moderate exercise pattern, defined as at least 75 min of running or 150 min of brisk walking per week, which has been associated with a lower prevalence of lifestyle-related diseases [ 3 ]. However, there are still a significant number of individuals who fail to meet these recommendations, which needs to be taken seriously.

Periodontitis refers to pathologic loss of alveolar bone and periodontal ligament, involving complex dynamic interactions among destructive immune responses and specific bacterial pathogens [ 59 ]. Our results showed that taking exercise is considered a health-promoting measure, which results in a decrease in the prevalence and severity of periodontitis. The underlying mechanism accounting for this reduction in prevalence may be associated with the influence of taking exercise on cytokine production and immune modulation [ 60 , 61 ]. C-reactive protein (CRP), a by-product of vitamin K metabolism, is significantly correlated with periodontitis. Studies have revealed that physical exercise can modulate several cytokines, including CRP [ 13 ]. Thus, maintaining regular physical activity should be encouraged to improve periodontitis.

This study performed a literature review and meta-analysis, which confirmed the relationship between exercise and periodontitis. Among the 30 studies included, 20 reported a positive relationship between the level of exercise and periodontitis, the remaining 10 articles did not report significant differences. Notably, none of the included studies reported any negative effects regarding the influence of exercise on periodontitis. The meta-analysis revealed a risk ratio of 0.84 (95% CI: 0.77, 0.91) between the exercise group and the inactive group (P < 0.01), although there was considerable heterogeneity, which indicates that engaging in physical exercise is inversely associated with the presence and severity of periodontitis.

The selected studies proposed several influential factors that may impact periodontitis, including maintaining a healthy weight, consuming a nutritious diet, reducing sedentary behavior, and avoiding cigarette smoking. Among the studies, 7 evaluated diet quality and found a strong association between consuming a high-quality diet and improving periodontitis. In addition, 4 studies suggested that maintaining normal body weight is important in preventing periodontitis, as obesity-related inflammation can induce bacterial proliferation and produce inflammatory markers in fat tissues, which may exacerbate periodontitis. 3 studies analyzed the influence of tobacco smoking habits on periodontitis and concluded that cigarette smoking may affect the periodontal status through total salivary antioxidant activity. Furthermore, higher sedentary behavior was found to be associated with higher odds of periodontitis. Accordingly, in addition to emphasizing exercise, interventions should also focus on the factors mentioned above, including improving diet quality, maintaining normal body weight, reducing sedentary behavior, and avoiding cigarette smoking.

The present study has several limitations. Firstly, the measurement of exercise involved multiple tools, including different versions of questionnaires, strength, and maximal oxygen consumption, resulting in challenges in drawing definitive conclusions. Additionally, the included studies had varying sample sizes, with a significant gap between the smallest and largest numbers, and were mostly conducted in developed countries, which may have affected the quality of data analysis. Finally, most selected studies were observational and did not employ measures to reduce bias such as random allocation. Accordingly, in order to address existing gaps in evidence, future researches in larger population samples with a longer follow-up time are needed to understand the real role of exercise on the prevalence of periodontitis. Moreover, more intervention researches are needed to establish a possible cause-effect association between exercise and the management of periodontitis.

Despite current knowledge gaps, the present review and meta-analysis systematically summarized current epidemiological data, providing evidence of a significant correlation between exercise and a lower occurrence of periodontitis. Therefore, taking exercise is proposed as a critical component of periodontitis management. Future researches in larger population samples with a longer follow-up time are needed to understand the real role of exercise on the prevalence of periodontitis. Moreover, more intervention researches are needed to establish a possible cause-effect association between exercise and the management of periodontitis, to help dental health providers take measures in clinical scenarios.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

American College of Sports Medicine, Chodzko-Zajko WJ, Proctor DN, et al. American College of Sports Medicine position stand. Exercise and physical activity for older adults. Med Sci Sports Exerc. 2009;41(7):1510–30. https://doi.org/10.1249/MSS.0b013e3181a0c95c

Article   Google Scholar  

Williams PT, Franklin B. Vigorous exercise and diabetic, hypertensive, and hypercholesterolemia medication use. Med Sci Sports Exerc. 2007;39(11):1933–41. https://doi.org/10.1249/mss.0b013e318145b337

Article   PubMed   PubMed Central   Google Scholar  

Minich DM, Bland JS. Personalized lifestyle medicine: relevance for nutrition and lifestyle recommendations. ScientificWorldJournal. 2013;2013:129841. https://doi.org/10.1155/2013/129841

Mehta KM, Yaffe K, Brenes GA, et al. Anxiety symptoms and decline in physical function over 5 years in the health, aging and body composition study. J Am Geriatr Soc. 2007;55(2):265–70. https://doi.org/10.1111/j.1532-5415.2007.01041.x

Pratt M, Macera CA, Wang G. Higher direct medical costs associated with Physical Inactivity. Phys Sportsmed. 2000;28(10):63–70. https://doi.org/10.3810/psm.2000.10.1237

Article   CAS   PubMed   Google Scholar  

Mariotti A, Hefti AF. Defining periodontal health. BMC Oral Health. 2015;15 Suppl 1(Suppl 1):S6. https://doi.org/10.1186/1472-6831-15-S1-S6

Papapanou PN, Sanz M, Buduneli N, Dietrich T, Feres M, Fine DH, Flemmig TF, Garcia R, Giannobile WV, Graziani F et al. Periodontitis: Consensus report of workgroup 2 of the 2017 World Workshop on the Classification of Periodontal and Peri-Implant Diseases and Conditions. J Clin Periodontol. 2018;45 Suppl 20:S162-S170. https://doi.org/10.1111/jcpe.12946

Nazir MA. Prevalence of periodontal Disease, its association with systemic Diseases and prevention. Int J Health Sci (Qassim). 2017;11(2):72–80.

PubMed   Google Scholar  

Tonetti MS, Van Dyke TE, working group 1 of the joint EFP/AAP workshop. Periodontitis and atherosclerotic Cardiovascular Disease: consensus report of the Joint EFP/AAP Workshop on Periodontitis and systemic Diseases. J Periodontol. 2013;84(4 Suppl):24–S29. https://doi.org/10.1902/jop.2013.1340019

Tar I, Csősz É, Végh E, et al. Salivary citrullinated proteins in rheumatoid arthritis and associated periodontal Disease. Sci Rep. 2021;11(1):13525. https://doi.org/10.1038/s41598-021-93008-y

Article   CAS   PubMed   PubMed Central   Google Scholar  

Kassebaum NJ, Bernabé E, Dahiya M, et al. Global burden of severe periodontitis in 1990–2010: a systematic review and meta-regression. J Dent Res. 2014;93(11):1045–53. https://doi.org/10.1177/0022034514552491

Ramseier CA, Woelber JP, Kitzmann J, et al. Impact of risk factor control interventions for smoking cessation and promotion of healthy lifestyles in patients with periodontitis: a systematic review. J Clin Periodontol. 2020;47(Suppl 22):90–106. https://doi.org/10.1111/jcpe.13240

Article   PubMed   Google Scholar  

Lieske B, Moszka N, Borof K, et al. Association between an anti-inflammatory dietary score and periodontitis-evidence from the Population-based Hamburg City Health Study. Nutrients. 2023;15(14):3235. https://doi.org/10.3390/nu15143235

Amaranath BJ, Gupta S, Kumar S, et al. Assessment of Periodontal Health Status among the Male Adult Population with a dual habit of Smoking and Gutkha Chewing: a cross-sectional study. J Pharm Bioallied Sci. 2023;15(Suppl 2):1020–S1022. https://doi.org/10.4103/jpbs.jpbs_246_23

Stewart LK, Flynn MG, Campbell WW, et al. The influence of exercise training on inflammatory cytokines and C-reactive protein. Med Sci Sports Exerc. 2007;39(10):1714–9. https://doi.org/10.1249/mss.0b013e31811ece1c

Hammonds TL, Gathright EC, Goldstein CM, et al. Effects of exercise on c-reactive protein in healthy patients and in patients with Heart Disease: a meta-analysis. Heart Lung. 2016;45(3):273–82. https://doi.org/10.1016/j.hrtlng.2016.01.009

Rodríguez-Archilla A, Padrón-Curiel DA. Influence of physical exercise on periodontal Disease: a meta-analysis. Int J Dent Sci. 2022;4(1):21–6.

Google Scholar  

Ferreira RO, Corrêa MG, Magno MB, et al. Physical activity reduces the prevalence of Periodontal Disease: systematic review and Meta-analysis. Front Physiol. 2019;10:234. https://doi.org/10.3389/fphys.2019.00234

Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. https://doi.org/10.1136/bmj.n71

Schardt C, Adams MB, Owens T, et al. Utilization of the PICO framework to improve searching PubMed for clinical questions. BMC Med Inform Decis Mak. 2007;7:16. https://doi.org/10.1186/1472-6947-7-16

Cumpston M, Li T, Page MJ, et al. Updated guidance for trusted systematic reviews: a new edition of the Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Database Syst Rev. 2019;10:ED000142. https://doi.org/10.1002/14651858.ED000142

Borenstein M, Hedges LV, Higgins JP, Rothstein HR. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Synth Methods. 2010;1(2):97–111. https://doi.org/10.1002/jrsm.12

Alkan B, Guzeldemir-Akcakanat E, Odabas-Ozgur B, et al. Effects of exercise on periodontal parameters in obese women. Niger J Clin Pract. 2020;23(10):1345–55. https://doi.org/10.4103/njcp.njcp_627_19

Almohamad M, Krall Kaye E, Mofleh D, et al. The association of sedentary behaviour and physical activity with periodontal Disease in NHANES 2011–2012. J Clin Periodontol. 2022;49(8):758–67. https://doi.org/10.1111/jcpe.13669

Al-Zahrani MS, Borawski EA, Bissada NF. Increased physical activity reduces prevalence of periodontitis. J Dent. 2005;33(9):703–10. https://doi.org/10.1016/j.jdent.2005.01.004

Al-Zahrani MS, Borawski EA, Bissada NF. Periodontitis and three health-enhancing behaviors: maintaining normal weight, engaging in recommended level of exercise, and consuming a high-quality diet. J Periodontol. 2005;76(8):1362–6. https://doi.org/10.1902/jop.2005.76.8.1362

Bawadi HA, Khader YS, Haroun TF, et al. The association between periodontal Disease, physical activity and healthy diet among adults in Jordan. J Periodontal Res. 2011;46(1):74–81. https://doi.org/10.1111/j.1600-0765.2010.01314.x

Bazyar H, Adibmanesh A, Javid AZ, et al. The relationship between metabolic factors and anthropometric indices with periodontal status in type 2 Diabetes Mellitus patients with chronic periodontitis. Obes Med. 2019;16:100138.

Cueto A, Mesa F, Bravo M, et al. Periodontitis as risk factor for acute Myocardial Infarction. A case control study of Spanish adults. J Periodontal Res. 2005;40(1):36–42. https://doi.org/10.1111/j.1600-0765.2004.00766.x

Han DH, Lim SY, Sun BC, et al. The association of metabolic syndrome with periodontal Disease is confounded by age and Smoking in a Korean population: the Shiwha-Banwol Environmental Health Study. J Clin Periodontol. 2010;37(7):609–16. https://doi.org/10.1111/j.1600-051X.2010.01580.x

Han K, Hwang E, Park JB. Excessive consumption of Green Tea as a risk factor for Periodontal Disease among Korean adults. Nutrients. 2016;8(7):408. https://doi.org/10.3390/nu8070408

Han K, Park JB. Association between oral health behavior and periodontal Disease among Korean adults: the Korea national health and nutrition examination survey. Med (Baltim). 2017;96(7):e6176. https://doi.org/10.1097/MD.0000000000006176

Han K, Park YM, Park JB. Evaluation of an association between long sleep duration and periodontal Disease among men and women using nationally representative data. Gac Sanit. 2018;32(2):143–50. https://doi.org/10.1016/j.gaceta.2017.01.013

Han SJ, Bae KH, Lee HJ, et al. Association between regular walking and periodontitis according to socioeconomic status: a cross-sectional study. Sci Rep. 2019;9(1):12969. https://doi.org/10.1038/s41598-019-49505-2

Hasan SMM, Rahman M, Nakamura K, et al. Relationship between Diabetes self-care practices and control of periodontal Disease among type2 Diabetes patients in Bangladesh. PLoS ONE. 2021;16(4):e0249011. https://doi.org/10.1371/journal.pone.0249011

Hoppe CB, Oliveira JAP, Grecca FS, et al. Association between chronic oral inflammatory burden and physical fitness in males: a cross-sectional observational study. Int Endod J. 2017;50(8):740–9. https://doi.org/10.1111/iej.12686

Hwang SY, Jang JH, Park JE. Association between Healthy Lifestyle (Diet Quality, physical activity, normal body weight) and Periodontal Diseases in Korean adults. Int J Environ Res Public Health. 2022;19(7):3871. https://doi.org/10.3390/ijerph19073871

Iwasaki M, Yoshihara A, Suwama K, et al. A cross-sectional study of the association between periodontitis and physical activity in the Japanese population. J Periodontal Res. 2023;58(2):350–9. https://doi.org/10.1111/jre.13095

Kongstad J, Enevold C, Christensen LB, et al. Impact of Periodontitis Case Criteria: a cross-sectional study of Lifestyle. J Periodontol. 2017;88(6):602–9. https://doi.org/10.1902/jop.2017.160426

Marruganti C, Traversi J, Gaeta C, et al. Adherence to Mediterranean diet, physical activity level, and severity of periodontitis: results from a university-based cross-sectional study. J Periodontol. 2022;93(8):1218–32. https://doi.org/10.1002/JPER.21-0643

Marruganti C, Baima G, Grandini S, et al. Leisure-time and occupational physical activity demonstrate divergent associations with periodontitis: a population-based study. J Clin Periodontol. 2023;50(5):559–70. https://doi.org/10.1111/jcpe.13766

Mendoza-Núñez VM, Hernández-Monjaraz B, Santiago-Osorio E, et al. Tai Chi exercise increases SOD activity and total antioxidant status in saliva and is linked to an improvement of periodontal Disease in the elderly. Oxid Med Cell Longev. 2014;2014:603853. https://doi.org/10.1155/2014/603853

Merchant AT, Pitiphat W, Rimm EB, et al. Increased physical activity decreases periodontitis risk in men. Eur J Epidemiol. 2003;18(9):891–8. https://doi.org/10.1023/a:1025622815579

Munther S. The effects of cigarette Smoking and exercise on total salivary antioxidant activity. Saudi Dent J. 2019;31(1):31–8. https://doi.org/10.1016/j.sdentj.2018.09.002

Sakki TK, Knuuttila ML, Vimpari SS, et al. Association of lifestyle with periodontal health. Community Dent Oral Epidemiol. 1995;23(3):155–8. https://doi.org/10.1111/j.1600-0528.1995.tb00220.x

Sanders AE, Slade GD, Fitzsimmons TR, et al. Physical activity, inflammatory biomarkers in gingival crevicular fluid and periodontitis. J Clin Periodontol. 2009;36(5):388–95. https://doi.org/10.1111/j.1600-051X.2009.01394.x

Samnieng P, Ueno M, Zaitsu T, et al. The relationship between seven health practices and oral health status in community-dwelling elderly Thai. Gerodontology. 2013;30(4):254–61. https://doi.org/10.1111/j.1741-2358.2012.00672.x

Schmidt J, Vogel M, Poulain T, et al. Association of Oral Health Conditions in adolescents with social factors and obesity. Int J Environ Res Public Health. 2022;19(5):2905. https://doi.org/10.3390/ijerph19052905

Oliveira JA, Hoppe CB, Gomes MS, et al. Periodontal Disease as a risk indicator for poor physical fitness: a cross-sectional observational study. J Periodontol. 2015;86(1):44–52. https://doi.org/10.1902/jop.2014.140270

Omori S, Uchida F, Oh S, et al. Exercise habituation is effective for improvement of periodontal Disease status: a prospective intervention study. Ther Clin Risk Manag. 2018;14:565–74. https://doi.org/10.2147/TCRM.S153397

Wernicke K, Grischke J, Stiesch M, et al. Influence of physical activity on periodontal health in patients with type 2 Diabetes Mellitus. A blinded, randomized, controlled trial. Clin Oral Investig. 2021;25(11):6101–7. https://doi.org/10.1007/s00784-021-03908-6

Yu YH, Lai YL, Cheung WS, et al. Oral health status and self-reported functional dependence in community-dwelling older adults. J Am Geriatr Soc. 2011;59(3):519–23. https://doi.org/10.1111/j.1532-5415.2010.03311.x

Pischon N, Heng N, Bernimoulin JP, et al. Obesity, inflammation, and periodontal Disease. J Dent Res. 2007;86(5):400–9. https://doi.org/10.1177/154405910708600503

Jepsen S, Suvan J, Deschner J. The association of periodontal Diseases with metabolic syndrome and obesity. Periodontol 2000. 2020;83(1):125–53. https://doi.org/10.1111/prd.12326

Banting LK, Dimmock JA, Grove JR. The impact of automatically activated motivation on exercise-related outcomes. J Sport Exerc Psychol. 2011;33(4):569–85. https://doi.org/10.1123/jsep.33.4.569

Peralta LR, Cinelli RL, Cotton W, et al. The barriers to and facilitators of Physical Activity and Sport for Oceania with Non-European, non-asian (ONENA) Ancestry Children and adolescents: a mixed studies systematic review. Int J Environ Res Public Health. 2022;19(18):11554. https://doi.org/10.3390/ijerph191811554

Speck BJ. From exercise to physical activity. Holist Nurs Pract. 2002;16(5):24–31. https://doi.org/10.1097/00004650-200210000-00007

Mathur N, Pedersen BK. Exercise as a mean to control low-grade systemic inflammation. Mediators Inflamm. 2008;2008:109502. https://doi.org/10.1155/2008/109502

Uchida F, Oh S, Shida T, et al. Effects of Exercise on the oral microbiota and saliva of patients with non-alcoholic fatty Liver Disease. Int J Environ Res Public Health. 2021;18(7):3470. https://doi.org/10.3390/ijerph18073470

Codella R, Della Guardia L, Terruzzi I, et al. Physical activity as a proxy to ameliorate inflammation in patients with type 2 Diabetes and periodontal Disease at high cardiovascular risk. Nutr Metab Cardiovasc Dis. 2021;31(8):2199–209. https://doi.org/10.1016/j.numecd.2021.04.022

Slots J. Periodontitis: facts, fallacies and the future. Periodontol 2000. 2017;75(1):7–23. https://doi.org/10.1111/prd.12221

Download references

Acknowledgements

Not applicable.

This research was funded by National Key Research and Development Program of China, grant number 2021YFC2400400 and Scientific Research Project of Shanghai Municipal Health Commission, grant number 202040094.

Author information

Rongkai Cao and Piaopiao Qiu contributed equally to this work.

Authors and Affiliations

Department of Implantology, Stomatological Hospital and Dental School of Tongji University, Shanghai Engineering Research Center of Tooth Restoration and Regeneration, No. 399, Middle Yanchang Rd, Shanghai, 200072, China

Rongkai Cao, Piaopiao Qiu, Yuan Zhou, Bo Dong, Yucheng Han & Zhen Fan

You can also search for this author in PubMed   Google Scholar

Contributions

R.C. and P.Q. established the research plans. Y.Z., B.D. and Y.H. conducted the data analysis. R.C. wrote the main manuscript. Z.F. revised the manuscript. All authors have approved the final version of the manuscript.

Corresponding author

Correspondence to Zhen Fan .

Ethics declarations

Ethics approval and consent to participate, consent for publication, 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.

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.

Cao, R., Qiu, P., Zhou, Y. et al. The underlying relationship between exercise and the prevalence of periodontitis: a systematic review and meta-analysis. BMC Sports Sci Med Rehabil 15 , 161 (2023). https://doi.org/10.1186/s13102-023-00759-4

Download citation

Received : 28 July 2023

Accepted : 25 October 2023

Published : 27 November 2023

DOI : https://doi.org/10.1186/s13102-023-00759-4

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

• Periodontitis
• Relationship
• Systematic review

BMC Sports Science, Medicine and Rehabilitation

ISSN: 2052-1847

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

• Open access
• Published: 28 November 2023

Clinical characteristics and management of autoimmune enteropathy in children: case reports and literature review

• Meng Jin 1 ,
• Youzhe Gong 1 ,
• Wenwen Liu 1 &
• Xuemei Zhong 1  

BMC Pediatrics volume  23 , Article number:  601 ( 2023 ) Cite this article

55 Accesses

Metrics details

Autoimmune enteropathy (AIE) defined by intractable diarrhoea and nonceliac enteropathy with villous atrophy, is a rare digestive disease. Case reports of this disease are sporadic and the clinical characteristics of AIE is seldom discussed.

We evaluate the clinical, laboratory, histopathological features, response to therapy and outcome of AIE in children.

We conducted a retrospective analysis of five children with AIE in our hospital. A comprehensive search of MEDLINE was performed using PubMed, through keywords of “autoimmune enteropathy, pediatric or children”. The clinical manifestations, endoscopic results, pathological results, and medication therapy of these children were collected and the cases were divided into two groups, infants (≤ 1 year old) and children (> 1 year old).

Five cases treated in our department: one case took eight years to make the final diagnosis; one case was positive for anti-intestinal epithelial cell (AE) antibody; three cases showed crypt apoptosis in histopathology; and two cases showed celiac-like changes. All cases were responsive to glucocorticoid therapy in the early stage of treatment, while three cases required immunosuppressant maintenance. After reviewing the literature, we performed a statistical analysis of 50 cases with a male-to-female ratio of 31:19. Among them, 35 patients (70%) were within 1 year of age, and their clinical manifestations were mainly watery stool (43 cases, 86%), weight loss (28 cases, 56%), abdominal distension (3 cases, 6%), serum AE or anti-goblet cell (AG) antibody positivity (32 cases, 64%), other immune-related antibodies (21 cases, 42%), gene mutations (9 cases, 18%), and family history (21 cases, 42%). All the children showed different degrees of intestinal villous atrophy. Thirty-seven (74%) of the children were treated early, and their clinical symptoms were relieved. Comparing the cases between different age groups, it was found that the mortality rate of children with onset in infancy was higher (P < 0.05), and there was no difference in other autoimmune diseases, AE antibody positivity rates, and other antibodies between the two groups. In addition to survival rate between different age group (P = 0. 005), there was no difference in sex, autoantibody positivity rate, single gene mutation, or family history between the two groups (P > 0.05) through analysis of mortality and clinical remission cases.

Endoscopic examination and mucosal pathological examination should be performed to diagnose AIE in children with watery stool and weight loss who fail to be treated with diet therapy. Immunotherapy is the core of medical management of AIE and can improve prognosis. Children with a poor prognosis in infancy should be actively treated to reduce mortality rates associated with AIE.

Peer Review reports

Introduction

Autoimmune enteropathy (AIE) is a rare intestinal disease characterised by intractable diarrhoea and immune-mediated intestinal mucosal damage [ 1 ]. The incidence of AIE is 0.06 out of 100,000 children; most cases are associated with growth retardation while some children have other comorbid autoimmune diseases [ 2 ]. Clinical manifestations of AIE include intractable diarrhoea, severe nutrient malabsorption, and hypoproteinaemia. Small intestinal villus atrophy is a main histopathological feature, among other manifestations such as lymphocytic infiltration in the lamina propria of mucosa, chronic inflammation characterized by increased or active apoptotic bodies in the crypt epithelium, and neutrophilic cryptitis. Anti-intestinal epithelial (AE) or anti-goblet cell (AG) antibodies are present in the serum. In 1978, McCarth et al. [ 3 ] reported a case of severe intestinal mucosal atrophy in a child and then proposed the concept of AIE. In 1997, based on the clinical data of 14 patients, the Mayo Clinic proposed the diagnostic criteria that are still in use today [ 4 ]: (1) chronic diarrhoea (duration > 6 weeks); (2) manifestations of malabsorption syndrome; (3) manifestations of specific small bowel lesions: partial/complete villous atrophy, deep crypt lymphocytosis, increased crypt apoptosis, and increased intraepithelial lymphocytes; (4) exclusion of other causes of villous atrophy, such as Crohn’s disease, sprue, and small intestinal lymphoma; and (5) positivity for AE and/or AG antibodies. Among them, conditions (1)–(4) are necessary for the diagnosis of AlE. Positive AE or AG antibodies can support the diagnosis; however, an absence of positive antibodies cannot exclude the diagnosis of AIE.

Diagnosis of AIE is challenging, requires small bowel biopsy, and often confused with other caoditions causing glycogenic diarrhea, celiac disease, inflammation bowel disease, gene mutations related immune deficiency disease, and so on.

Given the rare nature of AIE, most articles on this disease have been isolated case reports or case series, lacking multi-centre statistical data, especially the clinical characteristics of children at different ages. Therefore, this article will further deepen our understanding of the disease by comparing the clinical characteristics of children of different ages at time of disease onset.

Detailed clinical data of five children diagnosed with AIE in our hospital from January 2020 to August 2022 were collected. Articles from 1982 to December 2021 were searched on Pubmed with the keyword “autoimmune enteropathy, Paediatric or children”, and the case data of all children diagnosed with AIE were obtained. A total of 69 cases involving children have been reported [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 ]. Except for 24 cases with unknown auxiliary examination data and medication therapy data, a total of 50 children were included as part of the research sample of this paper. According to the age of onset, they were divided into an infant group (≤ 1 year) and a child group (> 1 year). SPSS software was used for statistical analysis and Fisher’s exact test was used for comparison between the two groups. This study was reviewed by the Ethics Committee of the hospital (Ethics Review No: SHERLL2020026), and the guardians of five cases treated in our department signed the informed consent form.

A 2 -year-old girl developed gastroenteritis after an insanitary diet at the beginning of the disease and then developed chronic diarrhoea. The clinical manifestations were watery stools and abdominal distension, which could not be relieved by diet therapy, antibiotic application, or other treatments. The stool was yellow green watery with occasional mucus and no bloody substances. During repeated monitoring, no red or white blood cells were found in the routine examination of the stool. Physical examination on admission revealed severe malnutrition. Laboratory examination:serum AE antibody was positive, decreased serum albumin. Endoscopic examination showed that the duodenal villi were low and flat (see Fig.  1 A, B). The histology suggested that the small intestinal villi were atrophied (see Fig.  2 A, B), apoptotic bodies were present.

A 5-month-old girl presented with increasing frequency of defaecation along with slow weight gain. Feeding the patient a deeply hydrolyzed formula powder was not effective. Colonoscopy showed terminal ileal oedema, a pathology suggestive of active inflammation, and the symptoms were slightly relieved by dexamethasone(1 mg/kg/day)(Due to early onset and ineffective to conventional treatment, VEO-IBD was considered by local hospital, so dexamethasone was added for treatment.). By the age of 9 months, after respiratory tract infection, the child had a large amount of watery stool and progressive weight decline. Physical examination at admission revealed severe malnutrition and a dehydrated appearance. Gastroscopy showed duodenal mucosal oedema, and pathological examination showed obvious atrophy of the duodenal mucosa, lymphocyte infiltration in the lamina propria, goblet cells, and crypt apoptosis.

A 9 -month-old-boy, had intermittent watery stool with slow weight gain 2 months after birth. Repeated antidiarrheal and antibiotic treatments were not effective. No ulcer or erosion was found via gastroscopy or colonoscopy in the other hospitals. Physical examination on admission revealed severe malnutrition and mild dehydration. antibody and gene detection of celiac disease were negative. Gastroscopy showed low duodenal mucosa. Pathological examination showed celiac-like changes with few neutrophil infiltrations (see Fig.  2 C). Diet therapy was ineffective in this case.

A 10-year-old-girl had an increased defaecation frequency, 2–5 times per day, by the time she was 2 years old and slow weight gain with abdominal distension. Serum α-1 antitrypsin level was at the lower limit of normal examined by local hospital. Gastroscopy and capsule endoscopy showed chapped fissure-like changes(see Fig.  1 C, D), the celiac antibody gene was negative, and enteral nutrition and single-element diet were ineffective. Physical examination at admission revealed severe malnutrition, abdominal distension, and obvious hypoproteinaemia. Gastroscopy showed celiac-like changes in the descending duodenum, and the small intestinal villi were flat, with celiac-like changes and few neutrophil infiltrations (see Fig.  2 D). Diet therapy was ineffective in this case.

A 4-year-old girl had chronic diarrhoea, watery stool with obvious weight loss after gastrointestinal infection, and ineffective routine fasting and rehydration treatment at the beginning of the disease. Physical examination at admission revealed mild malnutrition and abdominal dilatation, while gastroscopy showed oedema in the duodenal bulb and descending part, along with sparse villi. Pathological examination showed duodenal mucosa atrophy, obvious lymphocyte infiltration in the lamina propria, and goblet cells with crypt apoptosis.

Endoscopy figures of case 1 and 4 . ( A ) Showed the villi in the descending part of duodenum are low and flat. ( B ) Showed the villi of the terminal ileum are sparse, and no ulcer or erosion is found. ( C ) and ( D ) showed the small intestinal villi are flat, with celiac-like changes

Pathological figures of case 1 , 3 and 4 . ( A ) HE 400 times, ( B ) HE 100 times. Shows that the villi of duodenal bulb, descending part and terminal ileum mucosa are flat or disappear, some crypts are damaged, goblet cells are significantly reduced, apoptotic cells are occasionally seen at the bottom of the crypt, and a large number of lymphocytes infiltrate into the lamina propria. Pathology showed celiac-like changes ( C ) HE 400 times, ( D ) HE 100 times The mucosal villus is severely atrophic, the lamina propria can be seen with more lymphoplasma cells and a small amount of neutrophils infiltration, crypt hyperplasia, no apoptosis, and increased intraepithelial lymphocytes

Five children in our hospital received routine rehydration therapy, parenteral nutrition support, and diet restriction after admission; but diarrhoea symptoms were not relieved. Through further examinations to exclude inflammatory bowel disease, celiac disease, stomatitis diarrhoea, small intestinal lymphoma and other diseases that may cause atrophy of small intestinal villi, a whole exon gene examination was performed, and no mutation of pathogenic site was found, which met Mayo diagnostic criteria [ 4 ], and the final diagnosis was AIE. Glucocorticoid therapy was administered at an initial dose of 2 mg/kg/day of methylprednisolone. Except for cases 4 and 5 , the diarrhoea symptoms were relieved, whereas in other children, it was not immediately beneficial. After the dosage increased to 4 mg in cases 1 – 3 , the symptoms of loose stool in cases 2 and 3 were relieved. These symptoms reappeared in the process of glucocorticoid reduction to 4 mg/day but were relieved once tacrolimus was added to the treatment course. In case 1 , the symptoms were relieved by treatment with glucocorticoid combined with cyclosporine, but the symptoms reappeared in the process of adding normal diet, and rescue treatment with infliximab was ineffective. After adjusting the immunosuppressant dose to include tacrolimus, clinical remission was achieved.

Clinical characteristics of five children in our hospital are shown in Tables  1 and 2 , and 3 . Of the five cases, two cases had onset within 1 year of age, the mean onset age was 1.59 years old and the average time from onset to final diagnosis was 2.4 years. All children had malnutrition, anaemia, and hypoproteinaemia and were responsive to glucocorticoid therapy. One case in which the drug was completely stopped, and the other three cases needed immune drugs to be maintained, there is also one patient with hormone reduction under observation.

Statistical analysis was performed on the clinical data of the 50 children (Table  4 ). The ratio of males to females was 31:19, with boys being more affected by AIE. The average age of onset was 1.6 years old, while 35 cases (70%) experienced onset within 1 year of age, among them 27 were onset within 6 months, and the average age of whom was 2.5 months. Clinical manifestations were mainly watery stools (43 cases, 86%), weight loss (28 cases, 56%), abdominal distension (3 cases, 6%), and positive erum AE or anti-goblet cell (AG) antibody (32 cases, 64%). Other immune-related antibodies were detected in 21 cases (42%): anti-thyroid antibody in 6 cases; anti-nuclear antibody in 10 cases; anti-insulin antibody in 2 cases; anti-Ergot antibody in 1 case; and anti-Saccharomyces cerevisiae antibody in 2 cases. Twenty-one children (42%) had combined family histories of AIE. All children had intestinal mucosal pathology showing different degrees of atrophy.

Gene mutations related disease were found in a total of 9 cases (18%): two cases immunodysregulation polyendocrinopathy enteropathy X-linked (IPEX),three cases hypogammaglobulinemia, one case autoimmune polyendrocrinopathy-cadidiasis-ectodermal dystrophy (APECED), one case Kabuki syndrome, one case X-linked immunodeficiency, and one case hyper-IgE syndrome.

42 cases received drug treatment after diagnosis (Table  5 ), 7 cases of simple glucocorticoid, 18 cases of tacrolimus, 9 cases of azathioprine, 2 cases of cyclosporine, 2 cases of cyclophosphamide, and 4 cases of maintenance therapy with parenteral nutrition. Thirty-seven cases achieved clinical remission after drug treatment (74%), five cases died of infection after adding immunosuppressants, and eight cases died without medication.

The children in our study were divided into infant (≤ 1 year old) and child (> 1 year old) groups. The chi-square test was used for the analysis (Table  6 ). The mortality rate of children with onset in infancy was higher (P = 0.005, < 0.05). There was no significant difference in other autoimmune diseases, in the AE antibody-positive rate, and other antibodies between the two groups. In addition to age group, there was no difference in sex, autoantibody positivity rate, single gene mutation, or family history between the two groups through analysis of death and clinical remission case groups (Table  7 ).

Autoimmune enteropathy (AIE) is an autoimmune disease of unknown aetiology, characterised by villous atrophy of the small intestinal mucosal epithelium. Most of these cause chronic diarrhea and may be life-threating diarrhea in infancy.

The main clinical manifestation of AIE is severe and long-term secretory diarrhoea, which is characterised by a large amount of watery stools, with a diarrhoea volume of up to 5,000 mL/day. The effects of fasting and water deprivation are not typically beneficial and the faecal aetiology is negative. Parenteral nutrition is required to maintain water and electrolyte balance. Intestinal malabsorption can result in low body weight and poor growth [ 18 ]. In addition, children with AIE are often prone to local and systemic infections related to the skin barrier, intestinal barrier, immunotherapy, and malnutrition [ 19 ]. In this paper, five children who were clinically relieved by immunotherapy developed systemic infections and died. Endoscopic manifestations are mostly mild, with sparse or oedema of the small intestinal villi, and some of them show fan-shaped and fissure-like changes in the duodenum, which are difficult to distinguish from celiac disease. AIE can be isolated in the gastrointestinal tract or may be part of a systemic disease. In children, these systemic diseases are mostly primary immunodeficiency diseases that are mostly related to gene mutations [ 5 , 20 ]. Common diseases include immune disorders, multiple endocrine diseases, enteropathy, autologous multiple endocrine diseases, candidiasis, ectodermal dystrophy syndrome, and others [ 21 ].

Laboratory tests for AIE lack specificity, and AE and/or AG antibodies are only present in the serum intestinal autoantibodies of some children. Positive serum AE or AG antibodies increase the possibility of AIE diagnosis, but is not necessary for diagnosis [ 22 ]. It has been reported that serum intestinal autoantibodies appear only after the onset of mucosal injury and decrease or become negative with treatment before histological remission [ 23 ]. Although many experts currently recommend their use as diagnostic criteria for AIE, the retrospective analysis in this paper shows that the positive rate of antibodies is only 64%. The presence of antibodies cannot be detected in all patients with typical AIE and not in other diseases either, such as inflammatory enteropathy, celiac disease, and HIV [ 23 , 24 , 25 , 26 ]. In addition, peripheral blood autoantibodies that can be detected in AIE patients include anti-thyroglobulin, anti-thyroid peroxidase, and anti-Saccharomyces cerevisiae antibodies; however, their significance and diagnostic value in the pathogenesis of AIE are still unclear [ 27 ]. Our retrospective statistical analysis also concluded that there was no difference between the presence of autoantibodies and remission rate and prognosis of treatment.

AIE has no specific histological features; however, all patients have gastrointestinal mucosal injuries. Injury is usually confined to the mucosa, while deep ulcers or transmural inflammatory changes are rarely observed. The pathological diagnostic criteria of AIE in the past were blunt villi, lymphocytosis in the deep crypt, increased apoptosis of crypt cells, and increased lymphocytes in the surface intraepithelium; however, at present, the pathological tissues of AIE can be divided into four types: active duodenitis (characterised by active chronic inflammatory process and neutrophil cryptitis, but without abscess), celiac disease-like, graft-versus-host disease-like (obvious apoptosis can be seen in crypts), and mixed mode. At present, all four types have been reported in paediatric patients [ 5 , 20 , 28 ]. A retrospective study of the mucosal pathology of AIE showed that among 23 children, there were 10 celiac-like (43.3%), eight mixed mode, one acute GVHD mode, and four duodenitis modes, while for patients with immunodeficiency, the prevalence of celiac-like mode could be as high as 72% [ 6 ]. Two of the confirmed cases in our hospital also showed a celiac-like pattern; therefore, this pattern is still very common among children, which is important in early differentiation and diagnosis.

These results suggest that the prognosis of patients with AIE depends on the severity of symptoms and signs of the digestive system, the severity and extent of histological lesions of the gastrointestinal tract, and the existence of parenteral involvement [ 27 , 29 ]. However, through retrospective analysis of case data, it was found that only the age of onset was closely related to mortality (Table  2 ), whereas there was no difference in family history, signs, or other diseases between the age groups. This may also be due to the fact that the number of cases we collected was mostly from the literature, and the purpose of case studies in the literature is specific and the description of pathological changes is different, which has a certain impact.

At present, there is no consensus on the treatment of AIE, because children with AIE often suffer from stunting due to nutritional absorption disorder, and their developmental conditions such as weight and height are mostly lower than those of normal children of the same age. The use of immunosuppressants based on nutritional support is recommended [ 21 , 23 ]. At present, there is no evidence-based treatment scheme owing to the lack of a large-scale sample and controlled trial evaluation. Glucocorticoids are the first choice for immunosuppressive therapy; when glucocorticoids are ineffective, tacrolimus, cyclosporine, tumour necrosis factor α monoclonal antibody, and other second-line treatment schemes can be selected for immunosuppressive therapy. Tacrolimus has been reported to have a high clinical remission rate (approximately 77%) for steroid-ineffective or refractory autoimmune bowel disease [ 23 ], which is similar to the remission rate of the cases we collected (78%). All cases in this study were treated with hormone and immunosuppressive therapy, and most of them had clinical remission. A few children died of septicaemia, but the mortality rate was significantly lower than that of the untreated children. Therefore, early drug treatment can save the lives of children with AIE.

Overall, the clinical and histopathological manifestations of AIE in children, especially in infants, are quite varied. Early diagnosis and medication can reduce mortality and improve growth and development. Therefore, it is helpful for clinicians and pathologists to identify the clinical characteristics and different pathological patterns of AIE in its early stages.

This paper has limitations. Since the clinical data of 45 children were reported in the previous literature, the clinical remission time and follow-up after drug addiction were not fixed, so the overall evaluation and long-term clinical remission rate of drug treatment were uncertain. Therefore, the follow-up will strengthen the follow-up work of the confirmed patients in our hospital and help to evaluate the medication and clinical prognosis.

Data Availability

The data have been listed above in the main paragraph.

Sarah E, Umetsu I, Brown C, Langner, Gregory Y, Lauwers. Autoimmune enteropathies. Virchows Arch. 2018;472(1):55–66. https://doi.org/10.1007/s00428-017-2243-7 .

Article   CAS   Google Scholar  

Mitomi H, Tanabe S, Igarashi M, Katsumata T, Arai N, Kikuchi S, Kiyohashi A, Okayasu I. Autoimmune enteropathy with severe atrophic gastritis and Colitis in an adult: proposal of a generalized autoimmune disorder of the alimentary Tract. Scand J Gastroenterol. 1998;33(7):716–20. https://doi.org/10.1080/00365529850171657 .

Article   CAS   PubMed   Google Scholar  

McCarthy DM, Kaltz SI, Gazze L, Waldmann TA, Nelson DL, Strober W. Selective IgA deficiency associated with total villous atrophy of the small intestine and an organ-specific anti-epithelial cell antibody. J Immunol. 1978;120(3):932–8. PMID:632592.

Corazza GR, Biagi F, Volta U, Andreani ML, De Franceschi L, Gasbarrini G. Autoimmune enteropathy and villous atrophy in adults. Lancet. 1997;350(9071):106–9. https://doi.org/10.1016/S0140-6736(97)01042-8 .

Aatur D, Singhi A, Goyal JM, Davison MD, Regueiro RL, Roche S. Pediatric autoimmune enteropathy: an entity frequently associated with immunodeficiency disorders. Mod Pathol. 2014;27(4):543–53. https://doi.org/10.1038/modpathol.2013.150 . Epub 2013 Sep 20.

Villanacci V, Ravelli VLA, Buscarini E, Salviato T, Lionetti P, Martelossi MSS, De Giacomo C. Diego Falchetti, Gloria Pelizzo, Gabrio Bassotti. Clinical manifestations and gastrointestinal pathology in 40 patients with autoimmune enteropathy. Clin Immunol. 2019;207:10–7. https://doi.org/10.1016/j.clim.2019.07.001 . Epub 2019 Jul 4.

Lars M, Marthinsen H, Scott J, EJderhamn. Autoimmune enteropathy in Swedish children, 1985–2002: a call for strict diagnostic criteria. Scand J Gastroenterol. 2008;43(9):1102–7. https://doi.org/10.1080/00365520802029864 .

Article   Google Scholar  

Lachaux A, Loras-Duclaux I, Bouvier R. Autoimmune enteropathy in infants. Pathological study of the Disease in two familial cases. Virchows Arch. 1998;433(5):481–5. https://doi.org/10.1007/s004280050277 .

Seidman EG, Lacaille F, Russo P, Galeano N, Murphy G, Roy CC. Successful treatment of autoimmune enteropathy with cyclosporine. J Pediatr. 1990;117(6):929–32. https://doi.org/10.1016/s0022-3476(05)80140-4 .

Mirakian R, Richardson A, Milla PJ, Walker-Smith JA, Unsworth J, Savage MO, Bottazzo GF. Protracted Diarrhoea of infancy: evidence in support of an autoimmune variant. Br Med J. 1986;293(6555):1132–6. https://doi.org/10.1136/bmj.293.6555.1132 .

Walker-Smith JA, Unsworth DJ, Hutchins P, Phillips AD, Holborow EJ. Autoantibodies against gut epithelium in child with small-intestine enteropathy. Lancet. 1982;1(8271):566–7. https://doi.org/10.1016/s0140-6736(82)92076-1 .

Fisher SE, Boyle JT, Holztapple P. Chronic protracted diarrhea and jejunal atrophy in an infant. Cimetidine-associated stimulation of jejunal mucosal growth. Dig Dis Sci. 1981;26(2):181–6. https://doi.org/10.1007/BF01312239 .

Savage MO, Mirakian R, Harries JT, Botazzo GF. Could protracted Diarrhoea of infancy have an autoimmune pathogenesis? Lancet. 1982;1(8278):966–7. https://doi.org/10.1016/s0140-6736(82)91965-1 .

Ellis D, Fisher SE, Smith WI Jr, Jaffe R. Familial occurrence of renal and intestinal Disease associated with tissue autoantibodies. Am J Dis Child. 1982;136(4):323–6. https://doi.org/10.1001/archpedi.1982.03970400041013 .

Powell BR, Buist NR, Stenzel P. An X-linked syndrome of diarrhea, polyendocrinopathy, and fatal Infection in infancy. J Pediatr. 1982;100(5):731–7. https://doi.org/10.1016/s0022-3476(82)80573-8 .

Martini A, Scotta MS, Notarangelo LD, Maggiore G, Guarnaccia S, De Giacomo C. Membranous glomerulopathy and chronic small-intestinal enteropathy associated with autoantibodies directed against renal tubular basement membrane and the cytoplasm of intestinal epithelial cells. Acta Paediatr Scand. 1983;72(6):931–4. https://doi.org/10.1111/j.1651-2227.1983.tb09846.x .

Charritat JL, Polonovski C. Pediatric autoimmune enteropathies with anticytoplasmic enterocytic auto-antibodies. Ann Pediatr(Paris). 1987;34(3):195–203. PMID:3296919.

CAS   PubMed   Google Scholar  

Catassi C, Fabiani E, Spagnuolo MI, Barera G, Guarino A. Severe and protracted diarrhea: results of the 3-year SIGEP multicenter survey. Working Group of the Italian Society of Pediatric Gastroenterology and Hepatology (SIGEP). J Pediatr Gastroenterol Nutr. 1999;29(1):63–8. https://doi.org/10.1097/00005176-199907000-00016 .

Cuenod B, Brousse N, Goulet O, De Potter S, Mougenot JF, Ricour C, Guy-Grand D, Cerf-Bensussan N. Classification of intractable diarrhea in infancy using clinical and immunohistological criteria. Gastroenterology. 1990;99(4):1037–43. https://doi.org/10.1016/0016-5085(90)90624-a .

Natacha Patey-Mariaud de Serre, Canioni D. Soléne Ganousse, Frédéric Rieux-Laucat, Olivier Goulet, Frank Ruemmele, Nicole Brousse. Digestive histopathological presentation of IPEX syndrome. Mod Pathol. 2009;22(1):95–102. https://doi.org/10.1038/modpathol.2008.161 . Epub 2008 Sep 26.

Charles B, Chen F, Tahboub T, Plesec M, Kay. Kadakkal Radhakrishnan. A Review of Autoimmune Enteropathy and its Associated syndromes. Dig Dis Sci. 2020;65(11):3079–90. https://doi.org/10.1007/s10620-020-06540-8 . Epub 2020 Aug 24.

Massimo Montalto FD, Onofrio L, Santoro A, Gallo A, Gasbarrini G. Gasbarrini. Autoimmune enteropathy in children and adults. Scand J Gastroenterol. 2009; 44(9): 1029-36. https://doi.org/10.1080/00365520902783691 .

Federico Biagi, Paola I, Bianchi L, Trotta, Gino R, Corazza. Anti-goblet cell antibodies for the diagnosis of autoimmune enteropathy? Am J Gastroenterol. 2009;104(12):3112. https://doi.org/10.1038/ajg.2009.511 .

Fiocchi C, Roche JK, Michener WM. High prevalence of antibodies to intestinal epithelial antigens in patients with inflammatory bowel Disease and their relatives. Ann Intern Med. 1989;110(10):786–94. https://doi.org/10.7326/0003-4819-110-10-786 .

Martín-Villa JM, Camblor S, Costa R, Arnaiz-Villena A. Gut epithelial cell autoantibodies in AIDS pathogenesis. Lancet. 1993;342(8867):380. https://doi.org/10.1016/0140-6736(93)91531-p .

Article   PubMed   Google Scholar  

Skogh T, Heuman R, Tagesson C. Anti-brush border antibodies (ABBA) in Crohn’s Disease. J Clin Lab Immunol. 1982;9(3):147–50. PMID:6762441.

Russo PA, Brochu P, Seidman EG, Roy.Autoimmune CC. Enteropathy Pediatr Dev Pathol. 1999;2(1):65–71. https://doi.org/10.1007/s100249900092 .

Ricard Masia S, Peyton GY, Lauwers I, Brown. Gastrointestinal biopsy findings of autoimmune enteropathy: a review of 25 cases. Am J Surg Pathol. 2014;38(10):1319–29. https://doi.org/10.1097/PAS.0000000000000317 .

Article   PubMed   PubMed Central   Google Scholar  

Hill SM, Milla PJ, Bottazzo GF, Mirakian R. Autoimmune enteropathy and Colitis: is there a generalised autoimmune gut disorder? Gut. 1991;32(1):36–42. https://doi.org/10.1136/gut.32.1.36 .

Article   CAS   PubMed   PubMed Central   Google Scholar  

Download references

Acknowledgements

Not applicable.

This study was supported by the Cultivation Plan Foundation of the Capital Institute of Pediatrics (PY-2019-08).

Author information

Authors and affiliations.

Gastroenterology Department, Children’s Hospital Capital Institute of Pediatrics, Beijing, 100020, China

Meng Jin, Youzhe Gong, Wenwen Liu & Xuemei Zhong

You can also search for this author in PubMed   Google Scholar

Contributions

Meng Jin wrote the main manuscript text and prepared Figs. 1 and 2; Tables 1, 2, 3, 4 and 5. Youzhe Gong and Wenwen Liu collected the clinical data and participated in data analysis. All authors reviewed the manuscript.

Corresponding author

Correspondence to Xuemei Zhong .

Ethics declarations

Ethics approval and consent to participate.

This study was approved by the medical ethics committee of the Children’s Hospital Capital Institute of Pediatrics (Ethics Review No.: SHERLL2020026), Beijing, China. Guardians of all pediatric participants signed the informed consent form. All methods were performed in accordance with the relevant guidelines and regulations.

Consent for publication

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.

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.

Jin, M., Gong, Y., Liu, W. et al. Clinical characteristics and management of autoimmune enteropathy in children: case reports and literature review. BMC Pediatr 23 , 601 (2023). https://doi.org/10.1186/s12887-023-04435-x

Download citation

Received : 25 July 2023

Accepted : 19 November 2023

Published : 28 November 2023

DOI : https://doi.org/10.1186/s12887-023-04435-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

• Autoimmune disease
• Villus atrophy
• Immunotherapy

BMC Pediatrics

ISSN: 1471-2431

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

Meta-analysis of Artificial Intelligence-Assisted Pathology for the Detection of Early Cervical Cancer

• Research Article
• Open access
• Published: 27 November 2023
• volume  16 , Article number:  189 ( 2023 )

You have full access to this open access article

• Chunmei Zhang 2 ,
• Huan Zhou 2 ,
• Xiaohui Yin 2 ,
• Geng Rong 2 ,
• Shixian Zhou 2 ,
• Mingming Wang 2 &
• Zhigang Pei 2  

50 Accesses

Explore all metrics

Cite this article

The objective of this study is to evaluate the accuracy of AI in the diagnosis of early cervical cancer using a systematic evaluation/meta-analysis approach and a comprehensive search of published literature. A comprehensive computer search of foreign language databases such as PubMed/MEDLINE, Embase, Cochrane Library, and IEEE; and Chinese databases such as China Knowledge Network, Wan fang Data Knowledge Platform, and Wipu.com (VIP) was conducted to retrieve reports on diagnostic accuracy of AI in early cervical cancer included between 1946 and December 2022. The literature was screened according to inclusion and exclusion criteria, and the quality of the included literature was evaluated using the QUADAS-2 quality evaluation chart. 2 × 2 diagnostic data in text were extracted and complete data were calculated using Review Manager 5.3. Heterogeneity between studies was analyzed using Stata SE 15.0 software with Meta Di Sc 1.4 and causes of heterogeneity were sought. A total of 42 data sets were included in the study of AI for the identification of benign and malignant cervical vitreous nodules, with a combined Sen value of 0.90; a combined Spe value of 0.90; a combined + LR value of 9.0; a combined −LR combined value was 0.11; DOR combined value was 83; and AUC was 0.96. The Fagan plot suggested a 50% pre-test probability and a 90% post-test probability of confirming diagnosis when the AI model diagnosed a glassy nodule positively, and a 10% probability of misdiagnosing the nodule when the result was negative. A total of 34 data sets were included in the study to determine benignity and malignancy of solid cervical nodules by AI, showing a combined Sen value of 0.92; a combined Spe value of 0.93; a combined + LR value of 13.37; a combined −LR combined value of 0.08; DOR combined value of 164; AUC of 0.97. The Fagan plot suggested a 50% pre-test probability and a 93% post-test probability of confirming the diagnosis of a solid cervical nodule when the AI model was positive, and an 8% probability of misdiagnosing the nodule when the result was negative. The results of likelihood ratio dot plots suggest that the use of an AI model for cervical detection in the clinical setting has a good exclusionary diagnostic power. Summing up the accuracy and specificity of the A1 model for diagnosis of early cervical cancer, accuracy for diagnosis of solid cervical nodules (0.90) > diagnosis of cervical nodules (0.92), and specificity for diagnosis of solid cervical nodules (0.90) > diagnosis of cervical nodules (0.93). The AI model is highly accurate in diagnosing early cervical cancer and has high clinical diagnostic value. The accuracy of the AI model in diagnosing solid nodules in the cervical region was higher than diagnosing ground glass nodules in the cervical region. The labeling method, image pre-processing method, and feature learning method affected the accuracy of the AI model in diagnosing early cervical cancer, while the choice of learning image library and validation database did not usually affect the accuracy of the model.

Avoid common mistakes on your manuscript.

1 Introduction

Cervical cancer is one of the deadliest gynecological malignancies in the world, with more than half a million new cases each year. According to Global Cancer Statistics Report 2020, the number one cancer causing death among women in 157 out of 185 countries in the world is breast cancer, and the second most common cancer is cervical cancer, with more than 600,000 new cases (6.5%) and 340,000 deaths (7.7%) reported worldwide in 2020 [ 1 ]. In China, cervical cancer accounts for about 110,000 cases (5.2%) and 59,000 deaths (5.0%), making it the sixth most prevalent cancer among women [ 2 ]. Due to the heavy burden of this disease, the World Health Organization (WHO) adopted in 2020 a global strategy to eliminate cervical cancer by 2030 through quality cervical screening, i.e., to reduce the incidence of cervical cancer to below 4 per 100,000. The long window period also provides time and opportunity for cervical cancer screening, which can be done through timely and effective cervical screening measures to detect cervical lesions and obtain early treatment [ 3 ]. As cervical cancer prevention and control efforts have been aggressively pursued, more and more precancerous cervical lesions have been detected, reducing the risk of cervical cancer and cervical cancer death by approximately 80–90% [ 4 ]. Over the years, global authorities have continued to update cervical cancer screening methods and ages to find a more integrated approach, with several important guidelines being updated. As early as 1988, the American Cancer Society (ACS) recommended that the initial population for cervical screening should be women with a history of sexual intercourse or who had reached adulthood (18 years of age), and in 2002 ACS recommended screening for women who had been sexually active for at least three years and were over 21 years of age. In 2012, cytology should be recommended every three years between the ages of 21 and 29 years, with combined screening recommended for women over 30 years of age [ 5 ]. In 2016, ACOG stated in its cervical cancer screening and prevention guidelines that FDA-approved HPV testing may be used for primary cervical screening in women aged 25 years and older, but that cytologic screening alone and combined screening is also recommended (Level B evidence). In July 2020, ACS, in its latest guidelines for cervical cancer screening, recognized the importance of primary HPV testing in screening and, for the first time, raised the age of cervical cancer screening to 25 years, with primary HPV testing preferred every 5 years between 25 and 65 years [ 6 ]. 1 primary HPV test every 5 years (strongly recommended) [ 7 ]. In July 2021, WHO recommended a change in the exact mode of initial screening to HPV DNA testing and the age of women starting screening, for the general population, from 30 years of age, using HPV DNA as initial screening and for women with human immunodeficiency virus infection from 25 years of age screening is initiated.

The emergence of AI-based cervical nodule detection systems based on deep learning has become an inevitable trend, which is a big step forward toward precision medicine [ 8 ]. The AI cervical nodule-assisted diagnosis system is based on automatically extracted computer data about cervical nodules and can quickly detect nodules and respond to nodule information including nodule location, density, and size, and predict their benign and malignant risk [ 9 ]. Many reports have shown that cervical nodule-assisted detection systems can improve the detection rate and efficiency of radiologists in detecting cervical nodules, and that use of AI software in medical imaging can reduce pressure on physicians, thus assisting them in accurate diagnosis and treatment [ 10 ]. However, the role of AI in clinical practice is still at a trial and research stage, and its performance deserves further evaluation [ 11 ]. Expanding and diversifying datasets, maintenance, external validation, cross-validation, regulatory approvals, metrics evaluation, benchmarking, clinical trials, peer review, ethical considerations, interpretability, monitoring, robustness testing, expert input, implementation, specific clinical contexts, transparency, and professional education are all part of the validation process for AI models in healthcare. AI is unique in its ability to transform image interpretation from a subjective, qualitative task to an objective, quantitative one, with its efficient mechanism for processing complex images [ 12 ]. However, the achievements of AI in the diagnosis of early cervical cancer are manifold, and the development of AI models has been greatly hampered by confusion caused by the variety of AI models and the diversity of methods. The present study will address these issues by conducting a systematic evaluation and meta-analysis of AI models for diagnosis of early cervical cancer [ 13 ]. The main topics include (1) evaluation of Sen and specificity of the AI model for diagnosis of cervical and cervical solid nodules, respectively, and (2) evaluation of Sen and specificity of the labeling method, image pre-processing method, feature learning method, and image library in diagnosis, respectively.

2 Information and Methods

2.1 search strategy.

Foreign language databases, Chinese databases such as China Knowledge Network (CNKI), Wan Fang, and Wei Pu, to find diagnostic reports published between 1946 and December 2022 on diagnostic accuracy of A1 in early cervical cancer accuracy of report [ 14 ]. There were no language restrictions. We developed a search strategy combining keywords and medical subject terms (Me SH)/free words and will use the following expressions.

#1 “AI” [Mesh].

#2 “Machine learning model”.

#3 “Machine learning algorithms” [Title/Abstract].

#4 “Algorithms” [Title/Abstract].

#5 “Machine learning” [Title/Abstract].

#6 “Deep learning” [Title/Abstract].

#7 “Computational Approach”.

#8 “Automated-computer aided”.

#9 “Convolutional neural network”.

#10 “Artificial neuronal network”.

#11 “Support Vector Machine”.

#12 OR/#1-#11.

#13 cervical* [Mesh] [Title/Abstract].

#14 Pulmonary* [Title/Abstract].

#15 chest* [Title/Abstract].

#16 bronchial* [Title/Abstract].

#17 OR/#13-#16.

#18 nodule* [Mesh] [Title/Abstract].

#19 cancer* [Mesh] [Title/Abstract].

#20 tumor*[Title/Abstract].

#21 neoplasm* [Title/Abstract].

#22 lesion*[Title/Abstract].

#23 Carcinoma* [Title/Abstract].

#24 OR/#18-#23.

#25 #17 AND #24.

#26 #12 AND #25.

2.2 Eligibility and Exclusion Criteria

2.2.1 inclusion criteria.

Independent assessment will be performed by 2 reviewers. Inclusion criteria are as follows: (1) At least one AI model is included to be used for diagnosis of early cervical cancer. (2) The data are complete containing sample size, Sen, specificity, or sufficient information to construct a 2*2 outcome table. (3) The gold standard for Sen and specificity reference is tissue biopsy. (4) The article generally conforms to the STARD (Standards for the Reporting of Diagnostic Accuracy Studies) statement.

2.2.2 Exclusion Criteria

Studies for which data were not available or could not be calculated in text, appendices, or after contacting the lead author were excluded.

Excluded were reviews, case controls, case reports, reviews, conference abstracts, animal studies, and other types of articles.

Articles with cases lacking pathological gold standard confirmation of results were excluded.

2.2.3 Study Selection and Data Extraction

Articles will be independently screened by two researchers based on title and abstract, and those that do not meet the criteria will be eliminated [ 15 ]. If there is a difference between the two researchers’ evaluations, disagreement will be resolved through discussion and, if necessary, submitted to a third researcher for independent review [ 16 ]. Following this initial stage, all remaining articles will be reviewed in full by two fellows independently, and these two fellows will determine the final study by inclusion or exclusion from unsatisfactory literature [ 17 ].

2.3 Quality Assessment

The study used the QUADAS-2 tool to assess the quality of included diagnostic accuracy literature, with two reviewers independently assessing the risk of bias for each included study [ 18 ]. The QUADAS-2 tool evaluates the patient selection, index test, reference standard, and flow/timing domains to assess the quality of diagnostic accuracy studies in systematic reviews. Consider particular elements for each domain. It has an impact on the overall caliber and dependability of the data used in systematic reviews and meta-analyses to support diagnostic accuracy. When using the QUADAS-2 tool, reviewers were asked to evaluate the risk of bias sequentially according to the four components described on the website. In the second stage, the QUADAS-2 tool is used to determine the level of risk of bias, and information in the text is used to determine the level of “High risk”, “Unclear risk”, or “Low risk”. If all items in one of the modules are scored as ‘Yes’, then a low risk of bias can be concluded; if one of the answers in one of the modules is ‘No’, then a possible risk of bias can be assessed and reviewer needs to further use guidelines provided in Stage 2 to determine the risk of bias. The reviewer needs to use the guidelines provided in Stage 2 to determine the risk of bias [ 19 ]. If literature does not provide the necessary information to make a judgment, then an ‘Unclear’ classification is applied. Disagreements, if any, will need to be resolved through discussion and consensus. The final risk of bias and quality score is presented in the form of a quality assessment chart. Decision-making is substantially influenced by quality evaluations in diagnostic trials, with high-quality studies having an impact on clinical practice, guideline creation, health policy, and budget allocation. Making educated decisions based on the information provided is ensured by rigorous and transparent processes.

2.4 Assessment of Heterogeneity

Heterogeneity was checked by visually inspecting forest plots of Sen and specificity for each study, as well as SROC curves associated with individual study results. Finding the ideal cutoff threshold for a diagnostic test is the goal of the threshold effect analysis used in investigations of diagnostic accuracy. The SROC curve shows graphically how test performance changes with thresholds, and the spot on the curve with the highest performance corresponds to the ideal threshold. The extracted data were entered into StataSE15.0 software and the ‘mid as’ command was used to perform relevant statistical analyses. Using bivariate box plots, included studies were tested for heterogeneity using Q -tests and P -statistics, with heterogeneity indicated if p  < 0.1 or 12 > 50%, and insignificant heterogeneity if p  > 0.1 and I 2  < 50%. The ability to recognize study heterogeneity is essential for meta-analyses and systematic reviews. Cochrane's Q statistic, I 2 statistic, forest plots, subgroup analysis, meta-regression, sensitivity analysis, and Baujat plot are examples of common techniques. To analyze the source of heterogeneity, the presence or absence of a threshold effect could be determined by visually determining the presence or absence of a “shoulder–arm shape” in included studies through scatter distribution of the SROC curve. Numerous sources of variance, intricate relationships, limited sample sizes, insufficient reporting, and publication bias are some of the causes of the confusion around the significance of labeling techniques in research heterogeneity. More investigation is required. Meta Di Sc 1.4 was used to calculate Spearman’s correlation coefficient and distribution of the ratio of diagnoses to the ratio of combined diagnoses for rechecking the presence of a threshold effect. In research on diagnostic accuracy, techniques like Spearman’s correlation coefficient and the ratio of diagnoses are employed to assess the threshold effect. The best diagnostic threshold for clinical decision-making is determined by measuring sensitivity and specificity, comparing positive diagnoses at various thresholds, and performing these measurements. If combined results suggest no heterogeneity, then either a fixed-effects model or a random-effects model can be used to validate combined data. If heterogeneity is determined to be due to non-threshold effects, effects can be combined using a random-effects model. Due to its capacity to take into account heterogeneity, offer conservative estimates of effect sizes, and improve analytical reliability against extreme or outlier research, a random-effects model is used for data analysis. It is consistent with the reasonable supposition that genuine impact sizes varied between studies as a result of variations in demographics, locations, or methodology.

2.5 Data Analysis

The data were imported into Stata SE 15.0 and merged using the ‘mid as’ command to integrate Sen, Spe, + LR, −LR, DOR, DS, etc., plotting SROC and calculating diagnostic metrics such as AUC, Fagan plots and likelihood ratio dot plots to determine accuracy. Sen, Spe, + LR, and −LR are common diagnostic test accuracy indicators; DOR values usually range from 0 to infinity, with larger values indicating better test identification. Through standardization, efficiency, improved picture quality, feature extraction, consistency, data integration, machine learning, and AI, and decreased reader fatigue, automatic image processing increases diagnostic accuracy, specifically sensitivity (Spe). It guarantees reliable, repeatable analysis, minimizes human error, and combines data from diverse sources for thorough analysis. When DOR < 1 it means that the test is incorrectly designed, probably due to the presence of more negative tests; the DS value is usually proportional to the AUC value and is often used for comparison with other diagnostic modalities; AUC is calculated by plotting SROC curve to reflect accuracy of diagnostic test and is usually artificially divided into five bands: 0.90 to 1 (excellent), 0.80 to The post-test probabilities were evaluated by Fagan plots to simulate improvement of pre-test probabilities by diagnostic approach; likelihood ratio dot plots were used to visualize likelihood ratios of diagnostic models. 10, LRN < 0.1 can be excluded and confirmed, RUQ: LRP > 10, LRN > 0.1 can be confirmed only, LLQ: LRP < 10, LRN < 0.1 can be excluded only, RLQ: LRP < 10, LRN > 0.1 can be excluded and confirmed neither.

2.6 Publication Bias

Dee k’s funnel plot was plotted by Stata SE 15.0 to assess the publication bias of included studies. A visual tool for meta-analysis known as Dee K’s funnel plot forms a symmetric funnel shape with the x -axis standing for effect size and the y -axis signifying study accuracy. Each point on the scatterplot represents different research. The closer the angle between the line in the plot and the X -axis is to 90°, the more publication bias exists; at the same time, p -value analysis was combined with p  < 0.05 to indicate statistical significance and publication bias.

The search form was imported into databases to be searched on demand, and a total of 4719 subject-related papers were obtained, including 4077 papers in English and 642 papers in Chinese, all papers were imported into EndNote X9 software to eliminate duplicates 346 papers, and then titles and abstracts of remaining papers were browsed to exclude 3901 papers that did not meet research direction, and then full papers were obtained and read through Internet. The final total number of included studies was 68 (in addition to 8 groups of data from included literature), as shown in Fig.  1 . 42 groups of data were published between 2015 and 2020, including 42 groups of data for AI identification of benign and malignant cervical nodules, and 34 groups of data for AI identification of benign and malignant solid cervical nodules [ 20 ]. The data were published between 2015 and 2020, with 42 sets of data for AI to identify benign and 34 sets of data for AI to identify solid cervical nodules [ 20 ].

PRISMA inclusion and exclusion flowchart

The QUADAS-2 graphical display template for assessing the quality of diagnostic trials provided on the QUADAS-2 website was used to evaluate the quality of included studies in turn and to automatically generate. By methodically evaluating patient selection, index test, reference standard, flow, and timing, the QUADAS-2 graphical display template improves diagnostic trial quality evaluation. It offers clear direction, improves openness, calls for expert judgment, and supports well-informed decision-making. The responses were “Yes”, “Unclear” or “No” according to descriptions and landmark questions in each item. The risk of bias and clinical suitability of content is then rated as ‘High risk’, ‘Unclear risk’, or ‘Low risk’, and the graphic display template automatically generates green, orange, and blue colors of the quality assessment chart. The results of the quality assessment show that the overall quality of the included studies is high. The basic characteristics of the included studies are shown in Fig.  2 , Tables 1 and 2 .

Quality evaluation chart of QUADAS-2

A total of 42 studies were included in the study of AI for the identification of benign and malignant cervix. Through a systematic search, screening, full-text review, publication bias detection, data extraction, quality assessment, interpretation, meta-analysis, statistical analysis, inclusion criteria, reporting, and peer review, the AI study for differentiating between benign and malignant cervix conditions uses 42 studies. The calculated four-compartment tables were imported into StataSE15.0 software and correlation analyses were performed using the “mi das” command. The bivariate box plot showed that six studies were located outside the graph (Fig.  3 ) and 12 values for combined Spe were 99.12%, p  = 0; 12 values for combined Sen were 99.36%, P  = 0, suggesting heterogeneity of studies. To analyze the results, it was first necessary to analyze the threshold effect. The results showed that there was no "shoulder–arm" distribution of dotted lines in the SROC curve plan, suggesting that there was no threshold effect in this group of included studies; in addition, Spearman’s correlation coefficient calculated by Meta Di Sc 1.4 was − 0.387, p  = 0.092 ( p  > 0.05), and DOR forest plot showed that distribution of ratio of diagnoses to ratio of combined diagnoses was not linear, so above two results also indicated that there was heterogeneity in included studies due to non-threshold effects. A Diagnostic Odds Ratio (DOR) forest plot, which displays study-specific DORs, an overall DOR, and quantitative indicators of heterogeneity, is crucial in a diagnostic accuracy meta-analysis. In addition to providing numerical insights into the level of heterogeneity and identifying outliers, it graphically evaluates possible heterogeneity and highlights aspects other than threshold effects. In summary, results suggest heterogeneity in the results of included articles and considering possible heterogeneity due to non-threshold effects, we will use a random-effects model to combine effect values in data analysis.

Bivariate box plots for the study of AI in the identification of benign and malignant services

The forest plots are shown in Fig.  4 , yielding a combined Sen value of 0.90 and a combined Spe value of 0.90; Fig.  5 shows a combined DOR value of 83 and a combined DS value of 4.42; Fig.  6 shows a combined + LR value of 9.0 and a combined −LR value of 0.11. The combined + LR value was 9.0 and the combined −LR value was 0.11; the SROC curve was plotted (Fig.  7 ) and the area under the curve AUC was 0.96 (95% CI 0.93, 0.97). The Fagan plot (Fig.  8 ) suggests a 50% pre-test probability and a 90% post-test probability of confirming diagnosis when the AI model diagnoses the cervix as positive, and a 10% probability of misdiagnosing nodule when the result is negative. The Fagan plot is a useful tool for evaluating the effects of the model’s test outcomes on pre-test and post-test probability when used in the context of an AI model for cervix detection. This graphical tool aids medical professionals in diagnosing patients, risk assessment, and outcomes communication. The results of likelihood ratio dot plots (Fig.  9 ) indicate that the use of an AI model for cervical detection in clinical practice has weak exclusionary diagnostic power.

Forest plot of Sen and specificity of AI for identification of benign and malignant cervical studies

Diagnostic advantage of AI in the identification of benign and malignant cervical studies compared to DOR

Likelihood ratios for AI in the identification of benign and malignant cervical studies

SROC curves for AI study on identification of benign and malignant cervix

Fagan diagram for AI study on identification of benign and malignant cervix

Likelihood ratio dot plots for AI in the identification of benign and malignant cervical studies

To further analyze heterogeneity, we explored sources of heterogeneity in the Meta-analysis of AI models for diagnosing cervix from the pre-processing method and model structure of AI models, and conducted meta-regression for labeling method, image pre-processing method, feature learning method, and image library, respectively (Fig.  10 ). The results showed that three variables of the label generation method, image pre-processing method, and feature learning method could generate heterogeneity. To determine how these factors affect outcomes, a meta-regression analysis entails defining the research question, gathering data, standardizing it, performing an initial meta-analysis, performing meta-regression for each factor, assessing heterogeneity, determining confounders, conducting sensitivity analyses, and reporting findings. The labeling method was divided into supervised learning groups, semi-supervised learning groups, and unsupervised learning groups. The results are shown in Table 2 . r was lower in the unsupervised learning group and semi-supervised learning group than before, but Spe—98.07% in the supervised learning group—was not changed from the unsupervised group, so it is not clear whether labeling method whether it was a source of heterogeneity in the study. Notably, results of subgroup analysis showed an improvement in Sen: 0.94 (95% CI 0.91, 0.95) and Spe: 0.96 (95% CI 0.93, 0.98) for the semi-supervised group, as well as improvements in −LR, + LR, DOR, and AUC. The image pre-processing methods were divided into manual and automatic image processing. The results are shown in Table 2 . 12 was lower in the automatic image processing group than in the medium manual processing group, so the image pre-processing method may be a source of study heterogeneity. In this work, meta-analysis, heterogeneity assessment, sensitivity analysis, and detailed reporting are used to compare the effects of low-value automatic image processing and medium manual processing on diagnostic accuracy. Due to elements including picture quality, complexity, algorithmic efficacy, human operator experience, data characteristics, and assessment methodology, automatic image processing only received a lower score of 12 than manual image processing. The automatic image processing group showed an improvement in all diagnostic accuracy indicators, especially in Spe: 0.93 (95% CI 0.89, 0.95). The feature learning method was divided into a deep learning group and a shallow learning group. Deep learning models, similar to deep neural networks, can deal with complicated data patterns and need a lot of training data. While shallow learning methods, such as logistic regression, need less data and human feature building, they are more adaptable to fluctuations. When analyzing heterogeneity, researchers should take into account the kind of their data. The deep learning group included 25 studies and the shallow learning group included 17 studies. The subgroups were tested for heterogeneity in turn and effect sizes were combined, and results are shown in Table 2 . In a meta-analysis, subgroups are identified, heterogeneity tests are run, the results are interpreted, a technique is selected, effect sizes are combined, robustness is evaluated, publication bias is evaluated, and clinical practice-relevant conclusions are drawn. In addition, the deep learning group showed an improvement in all diagnostic accuracy indicators such as Sen, Spe, −LR, + LR, DOR, and AUC. The image library was divided into the LIDC/IDRI group, its database group, and the combined group. The image library is divided into three groups: database group, which contains data from other medical image databases; LIDC/IDRI group, which consists of lung CT scan images for lung nodule analysis; and merged group, which combines data from both groups for thorough medical image analysis and research, guaranteeing data accuracy and dependability. The LIDC/IDRI group included 22 studies, the database group included 12 studies, and the combined group included 7 studies. In image library analysis, using LIDC/IDRI and one’s database groups improves data diversity, model generalization, discovery of unusual conditions, cross-dataset validation, diagnostic precision, research validity, dependability, and clinical application.

Meta-regressions of studies on AI to identify benign and malignant cervix

To further analyze sources of heterogeneity in included studies, Sen analysis was conducted using Stata SE 15.0 (Fig.  11 ). After removing each study individually, the impact analysis plot and outlier detection plot revealed the possibility of heterogeneity in six articles, but the goodness of fit and binary normality suggested that fitted images largely overlapped with the null line, indicating that combined results were relatively stable.

Sen analysis of AI for identification of benign and malignant cervical studies

Publication bias was evaluated for included studies by producing Dee ‘s funnel plot (Fig.  12 ) via StataSE15.0 and results suggested p  = 0.10 ( p  > 0.05) suggesting no publication bias.

Publication bias in studies of AI for identification of benign and malignant cervices

4 Discussion

Cervical cancer is the fourth most common malignancy in women worldwide, and its main pathological types include squamous cell carcinoma of the cervix and adenocarcinoma of the cervix, which has caused disease and death in millions of women over past decades [ 23 ]. About 311,000 women died of cervical cancer in 2018, and about 29,500 women die of cervical cancer each year in China. The age of cervical cancer patients is gradually becoming younger, with precancerous lesions in patients under 35 years of age accounting for almost half of the total [ 22 ]. Although the launch of the bivalent HPV vaccine in China has increased the rate of HPV vaccination in Chinese women and fundamentally improved the future incidence of cervical cancer, cervical cancer is still a major threat to Chinese women due to the short time since its launch and low awareness of vaccination among Chinese women. The main cause of death in cervical cancer patients is the depth of infiltration of lesion and occurrence of lymph node metastasis. To reduce mortality and improve the prognosis of women with cervical cancer, early diagnosis, and accurate prognosis prediction are essential to guide clinical treatment, but treatment of cervical cancer has still not improved in past decades. There is an urgent need to explore an effective indicator to suggest prognosis and pathological features that will facilitate. Tumor stage, size, lymph node involvement, histological type, HPV status, invasion depth, p16 expression, immunological response, patient age, and general health are variables that affect the prognosis of cervical cancer. These variables support illness severity evaluation, treatment planning, and outcome prediction. There is an urgent need to explore an effective indicator to suggest prognostic and pathological features that will facilitate the selection of targeted therapies for cervical cancer and provide strong theoretical support for future cervical cancer treatment and prognosis prediction [ 21 ]. Enhance research quality by reviewing prior studies, enhancing methodologies, carrying out pilot studies, standardizing data, utilizing peer review, implementing open scientific practices, offering training, securing financing, working with specialists, and ensuring ethical oversight. After years of exploration in China, combining our clinical experience and foreign big data studies, an Expert Consensus on Issues Related to Cervical Cancer Screening and Abnormal Management in China was not created until 2017 to guide clinical work. By learning from European and American screening guidelines and big data analysis, CSCCP in China recommended cervical cancer screening for women aged 25–30 years [ 24 ]. Recommendations for cervical cancer screening for women between the ages of 25 and 30 are based on elements including rarity and the self-resolving nature of HPV infections. Early detection might result in needless interventions and higher healthcare expenses. HPV immunization for adolescents is advised. A healthcare professional should be consulted before making any healthcare decisions. However, at this stage, for various reasons, younger women in China, especially those under 25 years of age, are also very active in cervical cancer screening, leading to an increase in number of women under 25 years of age who are screened for cervical cancer compared to previous period, possibly due to following reasons. On one hand, they tend to request cervical cancer screening on their initiative due to their high health expectations and higher compliance than other age groups, increasing the number of universal and opportunistic cervical cancer screenings. On the other hand, it has been documented that women who receive the HPV vaccine have a lower risk of developing high-grade cervical lesions compared to those who do not receive the HPV vaccine and that the HPV vaccine reduces the incidence of cervical lesions and cervical cancer. As more young women learn about the long-term benefits of HPV vaccination for women’s health, more women of the right age want to be vaccinated against HPV, and some of them have a history of sexual activity or are even married with children and have not yet been vaccinated against HPV. Most vaccination facilities require a cervical cancer screening report from the recipient, so they will go to the hospital to be screened for cervical cancer on their initiative. Due to the disparity in medical standards between different institutions in China and the specificity of women under 25 years of age, it is difficult for many health professionals to consistently follow guidelines for post-screening management, or to misinterpret test results and over-treat a transient HPV infection that did not need to be over-treated, while patients who should be referred for immediate colposcopy are hesitant to do so, resulting in patients losing out on treatment. The best time to treat patients is lost. In clinical practice, fear of cervical cancer in young women increases the rate of colposcopy biopsies due to unnecessary referrals for colposcopy, which in turn increases the rate of overdiagnosis. Young women have special characteristics of colposcopy images, and there is currently a lack of multicenter, large sample, and prospective research in China on whether it is possible to minimize colposcopy biopsy rates, quickly and efficiently detect hidden high-grade cervical lesions, and reduce the harm of overtreatment or missed diagnosis to women in their reproductive period. The use of evidence-based recommendations, patient involvement, risk assessment tools, routine screening programs, advanced technology, second opinions, early detection awareness, monitoring, surveillance, education, patient advocacy, and quality assurance are just a few of the strategies being used by healthcare providers to lessen overtreatment and missed diagnoses in women of reproductive age.

Meta-analysis of AI model for diagnosis of cervix showed a combined Sen of 90% and a combined specificity of 90%, indicating a 10% miss rate and a 10% misdiagnosis rate, respectively; combined value of + LR was 9.0, indicating that true positive rate of positive nodules diagnosed by AI model was 9 times higher, and combined value of −LR was 0.11, indicating that negative nodules diagnosed by AI model were 0.11 times more likely to be false negative. The DOR was 83 and AUC was 0.96, indicating that the accuracy of diagnosis of the cervix was high; assuming a pre-test probability of 50%, the post-test probability of a diagnosis of the cervix by AI model showed a 90% probability of confirming diagnosis when test result was positive, and a 10% probability of misdiagnosing nodule when the result was negative, indicating that A1 model was effective in the diagnosis of cervix. Consequences might include overdiagnosis, missed diagnoses, and significant clinical, emotional, and financial effects for AI models with a 90% positive diagnosis rate and a 10% mistake rate. Those making decisions should evaluate this probability and take into account techniques like confirmatory testing and risk tolerance analysis. The likelihood ratios of individual studies were summarized in likelihood ratio dot plots, suggesting that the AI model was more evenly distributed across four quadrants of dot plots and that combined results were in the RLQ quadrant, indicating that the AI model was weak in detecting cervical nodules and prone to false-positive and false-negative results. Due to elements including data consistency, standard data collection, strong AI algorithms, enough sample size, controlled variables, rigorous assessment techniques, and random-effects models, the study’s diagnostic accuracy was not greatly impacted by the database used. The bivariate box plots and r values of Spe suggest heterogeneity in the study. To investigate the source of heterogeneity, Meta-regression was conducted for several underlying variables such as labeling method, image pre-processing method, feature learning method, and image library, respectively. The results showed that three variables, namely the label generation method, image pre-processing method, and feature learning method, were likely to generate heterogeneity.

This study also has some limitations: first, the proportion of malignant tumors in the original study sample included in the article was much higher, which may lead to overfitting and more optimistic results, so results still need prospective large sample studies to confirm conclusions; second, for literature where complete diagnostic data were not available in main text, we were still unable to obtain them by downloading original supplementary data and contacting original authors. Third, the study conducted a comprehensive search of relevant databases, but only included literature in English and Chinese, which had some impact on the systematicity of the study; fourth, original studies were mainly retrospective, and the quality of original studies would affect the quality of systematic evaluation.

5 Conclusions

According to the study, feature learning techniques and picture pre-processing might increase heterogeneity. Due to human label input and repeated validation learning, the diagnostic accuracy of the semi-supervised and autonomous image processing groups increased. Deep learning’s complex algorithms and logical framework led to better diagnostic outcomes. The accuracy of diagnosis was unaffected by database selection. Sen analysis, which excluded trials with higher heterogeneity, indicated steady combined results.

Using 42 data sets, the study examined the utility of AI for differentiating between benign and malignant cervical vitreous lesions. When the AI model was positive, the findings indicated a high accuracy rate of 50% pre-test probability and 90% post-test likelihood of confirming a diagnosis. The study also discovered that the AI model’s exclusionary diagnostic power for cervical detection was high. When detecting solid nodules in the cervical area as opposed to ground glass nodules, the AI model’s accuracy was greater. The labeling approach, picture pre-processing method, and feature learning method all had an impact on the AI model’s accuracy.

Code Availability

Not applicable.

Data Availability

No datasets were generated or analyzed during the current study.

Yu, Y., Wang, J., Tan, Y., Wan, H., Zheng, N., He, Z., Mao, L., Ren, W., Lin, Z., He, G., Chen, Y.: 1136p a clinically applicable cervical cancer artificial intelligence screening system for accurate cytopathological diagnosis: a multicenter population-based study and randomized controlled trial. Ann. Oncol. 32 , S925–S926 (2021)

Article   Google Scholar  

Chen, Y.M.: Artificial intelligence assisted computational tomographic detection of lung nodules for prognostic cancer examination: a large-scale clinical trial. Biomedicines 11 , 147 (2023)

Olson, L.G., Throne, R.D., Rusnak, E.I., Gannon, J.P.: Force-based stiffness mapping for early detection of breast cancer. Inverse Probl. Sci. Eng. 29 , 1–35 (2021)

Article   MathSciNet   MATH   Google Scholar  

Nishimura, H., Ping, T.Y., Oguntade, H., Kennedy, C.E., Narasimhan, M.: Hpv self-sampling for cervical cancer screening: a systematic review of values and preferences. Br. Med. J. Glob. Health 6 , e003743 (2021)

Google Scholar  

Zhang, H., Chen, C., Ma, C., Chen, C., Lv, X.: Feature fusion combined with Raman spectroscopy for early diagnosis of cervical cancer. IEEE Photonics J. 13 , 1–11 (2021)

Senthilkumar, G., Ramakrishnan, J., Frnda, J., Ramachandran, M., Mohammed, M.A.: Incorporating artificial fish swarm in ensemble classification framework for recurrence prediction of cervical cancer. IEEE Access 9 , 83876–83886 (2021)

Guo, C., Wang, J., Wang, Y., Qu, X., Hua, K.: Novel artificial intelligence machine learning approaches to precisely predict survival and site-specific recurrence in cervical cancer: a multi-institutional study. Transl. Oncol. 14 , 101032 (2021)

Munetoshi, A., Hashimoto, K.: Artificial intelligence in gynecologic cancers: current status and future challenges – a systematic review. Artif. Intell. Med. 120 , 102164 (2021)

Mai, A.G., El-Ghoneimey, S., Ibrahim, S.A.S., Mohamedhanafy, S.: Systemic lupus erythematosus as a risk factor for cervical cancer and its precursor conditions: assessment using pap smear and histopathology original research. Gynecol. Obstet. Res.–Open J. 8 , 1–7 (2021)

Al, J.: Cervical cancer diagnosis system using ant-miner for managing the knowledge in medical database. Turkish J. Comput. Math. Educ. (TURCOMAT). 12 , 1728–1737 (2021)

Yang, H., Chen, H., Ni, R., Zhang, G., Huang, Y., Ye, X., Fan, X., Kuang, Y., Zhang, J., Huang, C., Liu, H.: Circulating genetically abnormal cells combined with artificial intelligence for accurate and non-invasive early detection on nsclc. J. Clin. Oncol. 39 , 3056–3056 (2021)

Gupta, S., Saulnier, G., Newll, J., Faaborg-Andersen, C., Kelley, R.S., Vargas, A.: Uterine bioimpedance combined with artificial intelligence as a means of cancer detection. J. Med. Eng. Technol. 45 , 606–613 (2021)

Dan, L., Li, Z., Gu, Y., Ge, S., Mao, Y., Gu, Y., Cao, X.: A novel sers biosensor for ultrasensitive detection of hpv-e7 and opn based on a cascade signal amplification strategy of catalytic hairpin assembly and hybridization chain reaction. Mater. Chem. Front. 6 , 1331–1343 (2022)

Ilyas, Q.M., Ahmad, M.: An enhanced ensemble diagnosis of cervical cancer: a pursuit of machine intelligence towards sustainable health. IEEE Access 9 , 12374–12388 (2021)

Naderan, M.: Review methods for breast cancer detection using artificial intelligence and deep learning methods. Syst. Res. Inf. Technol. (2021). https://doi.org/10.20535/SRIT.2308-8893.2021.1.08

Young, M.R., Abrams, N.F., Ghosh-Janjigian, S., Rinaudo, J., Marquez, G., Koay, E., Kenner, B., Srivastava, S.: Prediagnostic images, electronic health records, and artificial intelligence: early detection of pancreatic cancer. Pancreas 50 , 1111–1111 (2021)

Urushibara, A., Saida, T., Mori, K., Ishiguro, T., Masumoto, T.: Diagnosing uterine cervical cancer on a single t2-weighted image: comparison between deep learning versus radiologists. Eur. J. Radiol. 135 , 109471 (2020)

Begum, V.: Detection of cervical cancer using multiple neuro fuzzy inference technique. Int. J. Multidiscip. Res. Rev. 6 , 725–729 (2019)

Crowell, E.F., Bazin, C., Thurotte, V., Elie, H., Renoouf, A.: Adaptation of cytoprocessor for cervical cancer screening of challenging slides. Diagn. Cytopathol. 47 , 890–897 (2019)

Dong, J.R., Jhingran, A., Kisling, K., Cardenas, C., Court, L.: Automated radiation treatment planning for cervical cancer. Semin. Radiat. Oncol. 30 , 340–347 (2020)

Long, Y., Yao, D.S., Wei, Y.S., Chen, J., Ye, X.Q.: Perineural invasion in early-stage cervical cancer: detection and influence on prognosis. Eur. J. Gynaecol. Oncol. 40 , 452–456 (2019)

Tang, D., Wang, L., Ling, T., Lv, Y., Zou, X.: Development and validation of a real-time artificial intelligence-assisted system for detecting early gastric cancer: a multicenter retrospective diagnostic study. EBioMedicine 62 , 103146 (2020)

Bedell, S.L., Goldstein, L.S., Goldstein, A.R., Goldstein, A.T.: Cervical cancer screening: past, present, and future. Sex. Med. Rev. 8 , 28–37 (2019)

Aiswarya, R.S., Kumar, V., Punitha, P.: The effect of grain size and silicon content on non-oriented grain steel anomalous loss through frequency excitation in medical healthcare by using big data analysis. Tamjeed J. Healthc. Eng. Sci. Technol. 1 , 43–53 (2023)

Download references

The authors did not receive any funding.

Author information

Authors and affiliations.

Information Management and Data Center, Chongqing University Jiangjin Hospital, Chongqing Jiangjin District Dingshan Street Office Jiangzhou Avenue No. 725, Chongqing, 402260, China

Department of Pathology, Chongqing University Jiangjin Hospital, Chongqing Jiangjin District Dingshan Street Office Jiangzhou Avenue No. 725, Chongqing, 402260, China

Chunmei Zhang, Huan Zhou, Xiaohui Yin, Geng Rong, Shixian Zhou, Mingming Wang & Zhigang Pei

You can also search for this author in PubMed   Google Scholar

Contributions

DQ, CZ, HZ, XY, is responsible for designing the framework, analyzing the performance, validating the results, and writing the article. GR, SZ, MW, ZP, is responsible for collecting the information required for the framework, provision of software, critical review, and administering the process.

Corresponding author

Correspondence to Zhigang Pei .

Ethics declarations

Conflict of interest.

Authors do not have any conflicts.

Additional information

Publisher's note.

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

Corresponding author: Zhigang Pei

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

Qin, D., Zhang, C., Zhou, H. et al. Meta-analysis of Artificial Intelligence-Assisted Pathology for the Detection of Early Cervical Cancer. Int J Comput Intell Syst 16 , 189 (2023). https://doi.org/10.1007/s44196-023-00367-7

Download citation

Received : 24 August 2023

Accepted : 01 November 2023

Published : 27 November 2023

DOI : https://doi.org/10.1007/s44196-023-00367-7

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

• Artificial intelligence
• Cervical cancer
• Systematic evaluation/meta-analysis
• Find a journal
• Publish with us

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest content
• Current issue
• Zoom Journal Club
• BMJ Journals More You are viewing from: Google Indexer

You are here

• Online First
• Neoadjuvant chemotherapy prior to radical hysterectomy in locally advanced cervical cancer: a systematic review and meta-analysis
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

• Chiara Borghi 1 ,
• Elena Biagioli 2 ,
• Jessica Mauro 1 , 3 ,
• Anna Roberto 2 ,
• http://orcid.org/0000-0002-6504-9978 Martina Borghese 4 and
• http://orcid.org/0000-0002-7093-6862 Alessandro Buda 1
• 1 Division of Gynecology Oncology , Ospedale Michele e Pietro Ferrero , Verduno , Piedmont , Italy
• 2 Research , Istituto di Ricerche Farmacologiche Mario Negri, IRCCS , Milan , Italy
• 3 University of Udine , Udine , Italy
• 4 Division of Obstetrics and Gynecology , Ospedale Santa Croce , Cuneo , Piemonte , Italy
• Correspondence to Dr Alessandro Buda, Division of Gynecology Oncology, Ospedale Michele e Pietro Ferrero, Verduno, Piedmont, Italy; alebuda1972{at}gmail.com

Objective The objective of this systematic review was to evaluate the effect of different types of neoadjuvant chemotherapy regimens, in terms of optimal pathological response and oncological outcomes, in patients with locally advanced cervical cancer.

Methods A systematic search of the literature was performed. MEDLINE through PubMed and Embase databases were searched from inception to June 2023. The study was registered in PROSPERO (ID number CRD42023389806). All women with a pathological diagnosis of locally advanced cervical cancer (International Federation of Gynecology and Obstetrics (FIGO) 2009 classification stages IB2-IVA), any age or histology, who underwent intravenous neoadjuvant chemotherapy before radical surgery, and articles only in English language, were included. We conducted a meta-analysis for optimal pathological response after surgery and survival outcomes. The risk of bias was assessed using the Newcastle-Ottawa scale and the Risk of Bias 2 (RoB) tools. The review methods and results were reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

Results 25 studies with a total number of 1984 patients fulfilled the eligibility criteria of our review and were included for data extraction and efficacy analysis. When compared with a two-drug regimen, the three-drug combination including cisplatin, paclitaxel, and ifosfamide or anthracyclines showed superior efficacy in terms of optimal pathological response with an odds ratio of 0.38 (95% CI 0.24 to 0.61, p<0.0001), with no difference in disease-free survival (hazard ratio (HR) 0.72, 95% CI 0.50 to 1.03, I 2 =0%, p=0.07) and higher overall survival (HR 0.63, 95% CI 0.41 to 0.97, I 2 =0%, p=0.03).

Conclusions The three-drug combination of cisplatin, paclitaxel, and ifosfamide or anthracyclines showed a higher rate of complete or optimal partial response, with the triple regimens having an advantage over the platinum-based schedules in terms of overall survival. Neoadjuvant chemotherapy followed by radical surgery should not be considered a standard of care in locally advanced cervical cancer.

• Cervical Cancer
• Hysterectomy
• Gynecologic Surgical Procedures

Data availability statement

Data are available upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information.

http://dx.doi.org/10.1136/ijgc-2023-004863

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

The standard treatment of locally advanced cervical cancer is concurrent radio-chemotherapy. International guidelines recommend neoadjuvant chemotherapy only in the context of clinical trials.

WHAT THIS STUDY ADDS

This review provides data in terms of pathological responses after neoadjuvant chemotherapy of different chemotherapy regimens. The three-drug combination of cisplatin, paclitaxel, and ifosfamide or anthracyclines seems to be more effective than other platinum-based combinations in patients with locally advanced cervical cancer.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

This review adds new insight to the available literature on the pathological response after neoadjuvant chemotherapy in locally advanced cervical cancer. However, the evidence is limited, and information mainly retrospective. Our review might be a reference for future prospective studies with well-defined inclusion criteria to try to define the subgroup of patients who may benefit from a neoadjuvant approach in terms of survival.

INTRODUCTION

Concurrent chemoradiation is the standard treatment of patients with locally advanced cervical cancer (International Federation of Gynecology and Obstetrics (FIGO) 2009 IB2-IVA), although definitive surgery can also be performed in patients with stage IB2 or IIA disease. 1 Nevertheless, up to 40% of such patients experience relapse and/or distant failure. Several studies have evaluated the role of neoadjuvant chemotherapy followed by radical surgery as an alternative to reduce the tumor volume, and to sterilize any foci of micrometastasis, thus reserving the use of radiotherapy in case of local recurrence. 2–4 A recent randomized trial showed a better 5-year disease-free survival of chemoradiation compared with neoadjuvant chemotherapy to surgery. 3

Furthermore, the preliminary results of the EORTC-55994 trial, a face-to-face comparison of concurrent chemoradiation with neoadjuvant chemotherapy to radical surgery, showed no statistically significant difference in the 5-year overall survival between the groups. 5 The choice of the best cytotoxic regimen for neoadjuvant chemotherapy represents another issue. Two Italian randomized studies 2 4 showed that the combination of paclitaxel, ifosfamide, and cisplatin achieved a higher rate of responses than the platinum-based combination of cisplatin with ifosfamide, or paclitaxel. However, in both studies, the three-drug combination regimen showed a higher hematological toxicity.

The primary endpoint of this review and meta-analysis was to evaluate the efficacy of different regimens of neoadjuvant treatments, in terms of optimal pathological response after radical surgery in patients with FIGO 2009 stage IB2-IVA cervical cancer. Disease-free survival and overall survival were evaluated as secondary endpoints.

Study Protocol

The research protocol was designed a priori, defining methods for searching the literature. The review methods and results were reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and the study was registered in PROSPERO (ID number CRD42023389806).

Search Strategy and Study Selection

The MEDLINE and Embase databases were searched for English abstracts and keywords of relevant studies published from inception until June 2023. The studies were identified with the use of a combination of the following mesh terms: ‘uterine cervical neoplasms’ and ‘neoadjuvant therapy’. The overall search strategy is included in Online supplemental table S1 .

Supplemental material

We also searched the reference lists of identified studies, reviews, and meta-analysis for other relevant articles.

A two-step process was used to select the eligible papers. As a first step, abstracts were divided into two groups and two reviewers for each group (CB, AB and JM, MB) independently checked the titles, abstracts, and keywords of identified studies. In the second step, two independent reviewers (CB, JM) evaluated the full-text. Disagreements were resolved by consensus.

Inclusion criteria were studies that included adult female patients with a pathological diagnosis of locally advanced cervical cancer stages IB2-IVA (FIGO 2009), who underwent intravenous neoadjuvant chemotherapy before radical surgery. Squamous cells, adenocarcinoma, and other histotypes were included. In case of a heterogeneous population, only those that reported results separately for subgroups were included. Studies of neoadjuvant and surgery in patients with cervical cancer with FIGO stage <IB2 were included only if the latter subgroup had a sample size <5% of the entire population.

The eligible chemotherapy regimens included three different schedules: platinum and taxanes combination, triple combinations (platinum, paclitaxel, and other drugs), and other platinum-based schemes. The scheduled drugs were administered as a dose-dense schedule, with a 7- to 10-day interval, or on intervals between cycles ranging from 3 to 9 weeks. We excluded studies with the same data source (doubles), chemotherapy administration different from intravenous, and outcomes not detailed according to different neoadjuvant regimens. Reasons for exclusion were recorded.

Outcome Measures

The primary outcome was optimal pathological response after chemotherapy. Secondary outcomes were disease-free and overall survival. Pathologic response after neoadjuvant chemotherapy was defined as follows: complete response, with complete disappearance of the tumor in the cervix with negative nodes; optimal partial response, with a residual disease inferior to 3 mm of stromal invasion; and suboptimal partial response, with persistent cervical residual disease with >3 mm stromal invasion. Stable disease was indicated when cancer volume was neither decreasing nor increasing in extent or severity, while progressive disease was when cancer was increasing, spreading, or worsening. The studies were considered eligible if at least the primary outcome was assessed. We excluded studies in which partial response was reported without differentiating between optimal partial and suboptimal partial responses.

Disease-free survival and overall survival were evaluated only in studies with a minimum follow-up of 12 months, or if the disease control rate or progression-free survival rate was reported at 12 months at least.

Quality Assessment and Risk of Bias Within Studies

The Newcastle-Ottawa scale tool was used to evaluate the quality of non-randomized studies included ( Online supplemental file ). 6 7 The Risk of Bias 2 (RoB 2) tool was used to assess the risk of bias of the two randomized studies included in the review. 8

Data Extraction

Information on the type of study, patient and tumor characteristics, clinical and pathological responses, neoadjuvant and adjuvant treatments, type of surgery, and results (point estimates and confidence interval for the outcome data) were extracted from each manuscript. Two reviewers independently recorded all the data using a pre-defined data extraction form. Discrepancies were resolved by a third reviewer.

Statistical Analysis

The percentage of patients with optimal pathological response rate and the relative 95% confidence interval (95% CI) were calculated for each study. To obtain the overall estimates, the Freeman-Tukey double arcsine transformation was performed and then the weighted pooled estimate was derived by applying the Der Simonian Laird random effect model 9 (with the weight of each study calculated as the inverse variance). Lastly, the overall estimate was back-transformed. 10 Analysis was performed by means of the STATA software version 15.1.

For randomized studies, the odds ratio (OR) of the optimal pathological response and the relative 95% CI were calculated and meta-analyzed using a random-effects meta-analysis according to the Der Simonian and Laird method. 9

For each randomized study, the log hazard ratio (HR) and its standard error (SE) for survival outcomes were calculated and meta-analyzed using a random-effects meta-analysis according to the DerSimonian and Laird method. 9 These analyses were done by means of the RevMan 5.4.1 software of the Cochrane Collaboration.

According to hospital rules, ethical approval was not required as only data from previously published studies were retrieved and analyzed. We will provide our data to the Editorial Team for the purposes of additional data analysis or for the reproducibility of this study if such is requested.

Study Selection

Initial research identified 1411 records. One article was added by manual searching, and citation tracking. After removing duplicates, 1405 records remained, of which 1256 articles were discarded following abstract screening and 149 articles were selected for the full-text evaluation. A further 124 articles were excluded. Lastly, 25 articles were included for data extraction and efficacy analysis. 2 4 11–33 The PRISMA flowcharts describe the complete review ( Figure 1 ). 34

• Download figure
• Open in new tab
• Download powerpoint

PRISMA 2020 flow diagram including searches of databases and registered only. 34 PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Studies’ Characteristics

Overall, 1984 patients were included. The publication dates of selected articles ranged from 2000 to 2021. Eighteen studies (72%) 11–28 were single center cohort, and seven (28%) were multicentered. 2 4 29–33 35 Fifteen studies (60%) were experimental, 2 4 11–13 15 19 23 24 27 30 33 eight (32%) had an observational retrospective design, 14 20–22 25 26 29 32 and two were observational prospective. 28 31 Five studies were comparative. 2 14 20 23 25 Two compared different neoadjuvant schedules, 2 14 two compared neoadjuvant chemotherapy to surgery and neoadjuvant chemotherapy with concurrent chemo-radiotherapy, 20 23 and one compared neoadjuvant chemotherapy and surgery to surgery alone. 25 Only two studies were randomized 2 4

Patients’ and Tumor Characteristics

The mean age ranged from 43 to 53 years. Eighty-four percent of patients (1671 women) had squamous cell carcinoma, 269 (13.6%) had adenocarcinoma, and 44 (2.2%) had other histotypes. Mean tumor size at diagnosis ranged from 44–68 mm. Tumor grading at diagnosis was not specified in eight studies. 12 16 17 23–26 29 Table 1 shows the characteristics of the included studies.

• View inline

Characteristics of studies and patients

Treatment Characteristics

Studies were divided into three groups, according to the different types of treatment: the two drug combinations including platinum and taxanes, triple combinations (platinum, taxane, and other drugs), and other platinum-based schemes ( Online supplemental table S2 ). Twelve studies included 528 patients treated with platinum and taxane-based chemotherapy. 4 11–16 18 20 21 25 29 Among them, in five studies chemotherapy was administered as a 3-weekly schedule, 4 11 12 16 21 in six as a dose dense schedule, 13 15 18 20 25 29 whereas in one the schedule was not specified. 14 The mean number of cycles ranged from 2 to 9.

Seven studies, including 526 patients, reported a platinum-based treatment. 2 14 17 23 27 30 33 All but one used 3- to 4-weekly administrations. 2 17 23 27 30 33 In one 14 the schedule was not specified. The mean number of cycles ranged from 3 to 6.

Ten manuscripts, including 838 patients, reported a three-drug combination. 2 4 14 19 22 24 26 28 31 32 Nine of them used a 3-weekly administration schedule, 2 4 14 19 24 26 28 31 32 and one used a 4-weekly schedule. 22 The mean number of cycles ranged from 2.5 to 3.

Clinical response after treatment was described according to RECIST (Response Evaluation Criteria in Solid Tumors) criteria, after magnetic resonance or CT scan, in all studies but six. 2 4 14 15 31 32 Two studies reported patients treated either by simple or radical hysterectomy after neoadjuvant chemotherapy, 2 14

Overall, 1693/1892 (89.5%) patients underwent radical surgery after neoadjuvant chemotherapy. In the platinum and taxane group, 468/528 (88.6%) patients were operated. Among them, complete response was achieved in 0% to 43.2% of cases, while optimal partial response ranged from 5% to 20%; suboptimal partial response ranged from 25% to 83.9%, while stable disease or progression were detected in a percentage ranging from 0% to 50%. The optimal pathological response obtained with a platinum and taxane combination is shown in Online supplemental figure S1 . In the platinum-based group, 442/526 (84.2%) patients underwent surgery, among whom 6.3% to 23% had complete response, 12.5% to 20% had optimal partial response, 35% to 79.4% had suboptimal partial response, and 0% to 11% had stable disease or progression ( Online supplemental figure S2 .) Finally, in the triple group, 783/838 (93.4%) patients underwent surgery; 10% to 25% had complete response, 7% to 50% had optimal partial response, 38.9% to 70% had suboptimal partial response, and 0% to 19.4% had stable or progression of disease ( Table 2 ). The pathological response to the triple regimens is shown in Online supplemental figure S3 .

Surgical pathological response and adjuvant therapy

In the platinum plus taxane group, eight studies assessed the lymph nodal status after surgery. Nodal involvement ranged from 11.7% to 52.3% in the included population and patients needed further adjuvant therapy. 11 12 15 16 18 20 21 In the platinum-based group, four studies reported rates of nodal involvement ranging from 17% to 37%. 17 23 30 33 In the triple group, six studies assessed this topic, with a nodal involvement rate ranging from 11% to 33%. 19 22 24 26 28 31

Quality and Risk of Bias Assessment Within the Studies

The literature scores were distributed between 5 and 7. Seven studies scored 7, and seven studies scored 6. The literature included was of high quality, and in seven studies there was a low risk of bias ( Online supplemental figure S4 ). Applying the Cochrane’s Risk of Bias 2 (RoB2) algorithm the two randomized studies were judged to be at low risk of bias for all domains.

Efficacy Results

Optimal pathological response analysis.

The mean optimal pathological response rate after surgery of patients who received the platinum-paclitaxel scheme was 28.2% (95% CI 22.6% to 34.2%; number of studies 12, number of patients 468) with a quite high heterogeneity of 41% ( Online supplemental figure S1 ). 4 11–16 18 20 21 25 29 The heterogeneity seems mainly due to the study of Mousavi et al 25 that reached a 51% of optimal pathological response rate. The overall optimal pathological response with platinum-based regimens was 29.7% (95% CI 22.2% to 37.7%; number of studies 7, number of patients 442) and with a high heterogeneity of 59.3% ( Online supplemental figure S2 ), 2 15 18 24 28 31 35 mainly due to the two comparative studies of Buda et al 2 14 in which the optimal pathological response rate was 23% and 19%. Finally, in women receiving a three-drug combination the overall optimal pathological response was 39% (95% CI 32% to 46%; number of studies 10, number of patients=783) with a huge heterogeneity of 68.3% ( Online supplemental figure S3 ). 2 4 14 19 22 24 26 28 31 32

Survival Outcome Analysis

Median follow-up time of studies reporting survival outcomes ranged from 12 to 95 months. 2 11 12 14 17 18 21 24 26 32 33 Overall survival rate according to the different neoadjuvant regimens are reported in Table 3 . If the rate at these timepoints (24 and 60 months) was not reported, the value was extrapolated from Kaplan-Meier curves using the WebPlotDigitizer 4.6 free tool.

Overall survival according to different neoadjuvant chemotherapy groups

Comparative Analyses of Optimal Pathological Response and Clinical Outcomes

Only two studies 2 4 can be included in the comparative analysis to evaluate the efficacy of the different chemotherapy regimens in terms of OR for optimal pathological response and HR for disease-free survival and overall survival ( Figure 2 ). The triple regimens showed superior efficacy in terms of optimal pathological response, with an OR of 0.38 (95% CI 0.24 to 0.61, p<0.0001; number of studies 2, number of patients 327, heterogeneity 0%), with a not statistically significant efficacy in terms of disease-free survival (HR 0.72, 95% CI 0.50 to 1.03, p=0.07; number of studies 2; patients 327, heterogeneity 0%) and a statistically significant superior efficacy in terms of overall survival (HR 0.63, 95% CI 0.41 to 0.97, p=0.03; number of studies 2, patients 327; heterogeneity 0%).

Optimal pathological response (A), disease-free survival (B), and overall survival (C) plots in randomized studies comparing triple therapy with a platinum-based association.

Summary of Main Results

This review showed that neoadjuvant chemotherapy including a three-drug combination of cisplatin, paclitaxel, and ifosfamide for squamous histology, or anthracyclines in the presence of adenocarcinoma type, is more active in terms of optimal pathological response than the two-drug platinum-based regimens with or without paclitaxel. In addition, the achievement of a complete response or optimal partial response, with a residual disease <3 mm, may benefit in terms of oncological outcome showing a statistically significant advantage of triple over the platinum-based therapies in terms of overall survival.

Results in the Context of Published Literature

The aim of neoadjuvant chemotherapy in locally advanced cervical cancer is to shrink the cervical lesions and reduce the tumor volume, allowing wider uninvolved surgical margins to be obtained in responsive patients. 2 4 Several studies aimed to find out which neoadjuvant chemotherapy regimens and dose-intensity schedules could be considered the most effective in terms of pathological response, toxicity profile, and oncological outcomes. This review’s analysis of the available literature showed that these parameters vary widely, thus not allowing any firm conclusions to be made. 35–37 Indeed, it has been shown that the timing and dose intensity of platinum-based association seems to have an important impact on whether it is beneficial in this setting. 38 A literature review showed that a pathological complete response can be achieved in 9% to 27% of patients after neoadjuvant chemotherapy, with an incidence of nodal metastases around 23%. 39 Two Italian prospective studies showed that the combination of cisplatin, ifosfamide, and paclitaxel was superior in terms of optimal pathological response to both the two drug combinations of cisplatin and ifosfamide, or paclitaxel. The triple regimen was associated with higher hematological toxicity.

In the SNAP 01 study, 2 the 3-year disease-free survival rate was 63% and overall survival was 71%, whereas in the SNAP 02 study, 4 for women treated with cisplatin, ifosfamide, and paclitaxel, the 5-year disease-free survival rate was 71% and overall survival was 78%. 2 4 A recent meta-analysis of 22 studies, including 3632 women with stage IB2 to IVA cervical cancer treated with dose-intense cisplatin-based neoadjuvant chemotherapy, the authors concluded that dose-dense cisplatin, at a dose over 72.5 mg/m 2 /3 weeks, was associated with better overall and progression-free survival. Nevertheless, the meta-analysis did not find any significant difference in terms of survival when comparing the triple with the non-triple cisplatin-based chemotherapy, and in comparative meta-analyses the effect was in line among all three indicators. 40

Strengths and Weaknesses

The strict methodology with defined inclusion criteria and the careful selection of the included studies represent a strength of this review. The first limitation of our study is that most of the manuscripts included in the review were from retrospective series. Furthermore, the presence of a heterogeneity of approaches and patient populations does not allow definitive conclusions to be drawn on the effectiveness of neoadjuvant chemotherapy in this setting. Finally, this review did not analyze the morbidity of the different combinations, due to the lack of safety data in the selected articles.

Implications for Practice and Future Research

Currently, definitive radiation therapy, with concurrent cisplatin-based chemotherapy, is considered the standard of care of locally advanced cervical cancer. Only two prospective randomized studies, comparing concurrent chemo-radiation with neoadjuvant chemotherapy, reported survival outcomes and toxicity. 3 41 These two studies were matched by Marchetti et al, 42 in a meta-analysis that reported similar overall survival rates between treatments; however, concurrent chemo-radiation was superior in terms of disease-free survival with a lower rate of drug toxicity. Despite this, the authors found the studies to be of low quality.

Neoadjuvant chemotherapy prior to surgery cannot be considered a standard of care in locally advanced cervical cancer. The subgroup of women who may benefit from neoadjuvant chemotherapy to radical surgery, in terms of toxicity, quality of life and oncological outcomes, has not yet been identified. Future studies, also including immunotherapy in a neoadjuvant setting, 43 may define the most effective chemotherapy regimen that can have an impact on survival outcomes in this subgroup of women. Neoadjuvant chemotherapy followed by radical surgery should be reserved for only a well-defined subgroup of women, and only in the context of clinical trials. 44

Ethics statements

Patient consent for publication.

Not applicable.

Acknowledgments

We want to thank the Fondazione Ospedale Alba Bra ONLUS for the support in the activities of the Division of Gynecologic Oncology of Michele and Pietro Ferrero Hospital.

• Motzer RJ ,
• Jonasch E ,
• Agarwal N , et al
• Fossati R ,
• Colombo N , et al
• Maheshwari A ,
• Parab P , et al
• Lissoni AA ,
• Colombo N ,
• Pellegrino A , et al
• Vergote I , et al
• Sterne JAC ,
• Savović J ,
• Page MJ , et al
• DerSimonian R ,
• Angioli R ,
• Montera R , et al
• Benedetti Panici P ,
• Bellati F ,
• Pastore M , et al
• Marchetti C , et al
• Floriani I , et al
• Huang H-J ,
• Lin H , et al
• Dueñas-Gonzalez A ,
• López-Graniel C ,
• González-Enciso A , et al
• Lopez-Graniel C ,
• Gonzalez A , et al
• Ferrandina G ,
• Palluzzi E ,
• Gallotta V , et al
• Distefano MG ,
• De Vincenzo R , et al
• Gadducci A ,
• Barsotti C ,
• Laliscia C , et al
• Huang H , et al
• Rao R , et al
• Chanona G , et al
• Lorusso D ,
• Ramondino S ,
• Mancini M , et al
• Mousavi A ,
• Modarres Gilani M ,
• Akhavan S , et al
• Scandurra G ,
• Scibilia G ,
• Banna GL , et al
• Vallejo CT ,
• Domínguez ME , et al
• Vitobello D ,
• Pirovano C , et al
• DI Martino G ,
• Ferrari D , et al
• Di Vagno G ,
• Pignata S , et al
• Pellegrino A ,
• Milani R , et al
• Corrado G ,
• Zanagnolo V , et al
• Zanaboni F ,
• Grijuela B ,
• Giudici S , et al
• McKenzie JE ,
• Bossuyt PM , et al
• Winata IGS ,
• Budiana ING ,
• Jawi IM , et al
• Katsumata N , et al
• Cai J , et al
• Neoadjuvant Chemotherapy for Locally Advanced Cervical Cancer Meta-analysis Collaboration
• Serkies K ,
• Nguyen VT ,
• Winterman S ,
• Playe M , et al
• Marchetti C ,
• Fagotti A ,
• Tombolini V , et al
• Upadhyay A ,
• Raspollini MR ,
• Planchamp F , et al

Supplementary materials

Supplementary data.

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

• Data supplement 1
• Data supplement 2
• Data supplement 3
• Data supplement 4
• Data supplement 5
• Data supplement 6
• Data supplement 7

Twitter @jess393

Contributors Conceptualization: AB, EB. Data curation, all authors. Formal analysis: EB, AR. Investigation: CB, JM, AB, MB. Methodology: EB, AB, AR. Supervision: AB, EB. Writing - original draft: CB, AB. Writing - review and editing: all authors. AB is responsible for the overall content as guarantor.

Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests None declared.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Read the full text or download the PDF: