https://orcid.org/0009-0009-3918-2845
Figures
Abstract
Background
The perspective of sex/gender bias is often missing in tools used to assess study risk of bias in systematic reviews. The aim was to pilot a checklist using an aetiological occupational health research question regarding the impact of sedentary behaviour at the workplace and cardiometabolic health. The checklist examined whether the consideration of sex/gender was associated with different study characteristics.
Methods
A sex/gender checklist developed based on a synopsis of existing instruments with input from the Cochrane Sex/Gender Methods Group was adapted for the present study. This checklist comprises four categories: 1. “Background and conceptual considerations” (3 items), 2. “Study design” (2 items), 3. “Study procedures of investigation or intervention and statistical analysis” (2 items), and 4. “Presentation and interpretation of findings” (3 items). Two independent reviewers evaluated all included studies. Bivariate and multivariable logistic regression analyses were used to examine the consideration of sex/gender across study designs, years of publication, and risk of bias levels (based on the Scottish Intercollegiate Guidelines Network [SIGN]).
Results
Of the 49 studies evaluated with the checklist, none provided detailed information, 69% (n = 34) provided basic information and 31% (n = 15) no information for the consideration of sex/gender. No intervention study provided information for the first two categories. In the third category, all intervention studies (n = 17) and case-control studies (n = 5) provided basic information on sex/gender, while two of the 23 cohort studies did not. In the fourth category, detailed information was found for all study designs (n = 8). Bivariate analyses revealed no association between the consideration of sex/gender and the year of publication (OR per year = 0.89; 95% CI: 0.65, 1.23). A low risk of bias level was not associated with consideration of sex/gender (OR = 0.60; 95% CI: 0.14; 2.50). Compared to intervention studies the odds of considering sex/gender was increased by a factor of 3.6 (95% CI: 1.0, 12.8) in observational studies.
Conclusion
The adapted checklist was applicable to assess the consideration of sex/gender in all studies. None of the primary studies considered sex/gender perspectives in all of the four categories. Further optimisation of the sex/gender tool seems warranted, based on further research on weighting individual categories or items and application of the checklist for occupational epidemiology in general.
Citation: Prigge M, Hegewald J, Reichel K, Backé E, Romero Starke K, Seidler A, et al. (2025) A tool for assessing sex/gender bias in epidemiological studies of occupational health: Pilot testing on studies of sedentary behaviour at the workplace and cardiometabolic health. PLoS One 20(5): e0324391. https://doi.org/10.1371/journal.pone.0324391
Editor: P M. Nabeel, Indian Institute of Technology Madras, INDIA
Received: November 21, 2024; Accepted: April 25, 2025; Published: May 29, 2025
Copyright: © 2025 Prigge et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The analyses described in the manuscript were not funded. However, the underlying systematic review was a scientific project of the Federal Institute for Occupational Safety and Health (BAuA, research project F 2399) and jointly funded by the German Federal Ministry of Education and Research, grant ID FKZ: 01GL1721. The underlying research project (F 2399) was conducted 2017–2019.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: SAGER, Sex and Gender Equity in Research; PROSPERO, International prospective register of systematic reviews; SIGN, Scottish Intercollegiate Guidelines Network; OR, Odds ratio; CI, Confidence interval; n, number of primary studies; n. a., not applicable; RoB, Risk of bias
Introduction
The main goal of systematic reviews is the synthesis of evidence on a topic. While synthesising the results, it is imperative to assess of how bias in the primary studies may have affected the overall results [1]. There are a number of tools that evaluate the risk of bias and quality in occupational epidemiological studies with different study designs [2,3]. However, for the most part, bias resulting from insufficient consideration of sex/gender is not taken into account by these instruments.
Sex is defined as the biological differences between men and women, while gender includes the view of roles and relationships that societies attribute to men and women [4,5]. Up to now, it is mainly the biological sex that is surveyed in epidemiological studies and gender-specific aspects are usually not considered. Thus, since 2010, a “two-step approach” is increasingly being applied, where first the biological sex is asked, followed by a survey of the gender identity affiliation [6].
It is important to consider sex and gender in health research because biological and psychosocial differences may result in divergent risks [7,8] and prevention approaches may need to be customised for men and women [4]. Failure to consider sex/gender could lead to significant omissions in these areas that might otherwise go unnoticed. This requires, that not only the binary understanding of sex/gender is included [6]. The risk of unfavourable health consequences can thus be minimised if studies take sex/gender into account [4]. In 2016, the Sex and Gender Equity in Research (SAGER) guidelines were published [9]. The recommendations were based on the work of a panel of 13 experts from nine countries, and state that data should be routinely presented disaggregated by sex and gender. Sex- and gender-based analyses should be reported regardless of positive or negative outcomes [9].
While some studies have described how sex/gender bias can be avoided in epidemiological studies [10,11], only a few bias assessment tools consider the evaluation of these sex/gender-specific aspects. Doull et al., for example, developed an appraisal tool that primarily evaluates sex/gender-specific aspects of systematic reviews and meta-analyses [4,12]. Briefing notes were developed as a communication tool to inform reviewers where sex/gender should be considered in a systematic review (e. g., background or methods) [4]. Suitable instruments for assessing sex/gender bias in primary studies are available only to a limited extent [8,11,13–15]. Often these tools are not suitable for the assessment of primary studies, as they primarily aim at helping with the design of studies [11] or to serve as examples of (good) sex/gender-sensitive practice [8]. Risk of bias tools (Cochrane ROB-2, SIGN, etc) assess internal validity. That is, they consider sources of bias that could impact the reported effect estimates for the entire study sample (e. g., randomization, allocation, recall bias) [16]. For example, SIGN included criteria for selection of subjects, assessment, confounding and statistical analysis to assess the internal validity of observational studies [17]. Thus, these tools consider sex/gender at most as a confounder and do not explicitly address potential differences in how study methods, such as recruitment, exposure measurements, and outcome detection may be impacted by sex/gender [16]. Thus, this sex/gender checklist addresses these gaps.
Demetriou et al. [13] and colleagues collaborated with international experts, including members of the Cochrane Sex/Gender Methods Group, to develop a practicable sex/gender bias checklist for assessing intervention studies in systematic reviews addressing physical activity and sedentary behaviour outcomes in children and adolescents [13–15]. However, sedentary behaviour or lack of physical activity are also problems in working adults. Based on a meta-analysis Prince et al. [18] found that approximately 60% of employees spend their working and waking time in a sedentary posture [18]. In Germany, approximately 50% of employees are mainly sitting during work [19]. Prolonged sitting is associated with different unfavourable health outcomes, including risks for the cardiovascular system and metabolism, resulting in an increase in diseases (such as hypertension and type II diabetes mellitus) and associated with premature mortality [20].
It is important to consider sex/gender effects in studies examining the association between sedentary behaviour and cardiovascular disease. The Cardiovascular Clinical Study Group describes differences in cardiovascular diseases (e. g., ischaemic heart disease) between men and women. For example, diabetes is associated with a greater increase in coronary artery disease risk in women than in men [5]. Men, on the other hand, suffer acute coronary heart syndrome or myocardial infarction more often than women [5]. However, mortality 30 days after an acute coronary syndrome is greater for women (9.6% in women and 5.3% in men) [21]. Men and women also differ with regard to sedentary behaviour, with approximately 50% of men and 34% of women spending more than 75% of their time sitting at work [22].
A systematic review by Reichel et al. [23] of occupational sedentary behaviour and its association with cardiovascular and metabolic diseases, cardiometabolic mortality, or changes in cardiometabolic risk markers narratively described whether the results differed for men and women [23]. However, no suitable instrument was available to assess sex/gender bias in the primary studies.
Thus, our first aim was to adapt the sex/gender checklist for sedentary behaviour as an outcome developed for children and adolescents [13,14] by modifying the checklist for aetiological occupational health questions and observational study designs with sedentary behaviour as an exposure.
The second aim was to determine whether the consideration of sex/gender identified by this tool was associated with the studies’ results using the following research questions:
- Is the consideration of sex/gender increasing over time in the published research identified?
- Do studies with a high internal validity (low risk of bias assessed with the SIGN) have a greater consideration of sex/gender?
- Is the consideration of sex/gender associated with the study design?
Methods
Demetriou et al. (2019) developed a sex/gender checklist in collaboration with international experts, scientists in this field and members of the Cochrane Sex/Gender Methods Group [13,14]. This checklist was used as a basis and adapted for the research question of the systematic review of Reichel et al. [23].
Sex/gender checklist
The sex/gender checklist consists of four categories that follow the structure of study publications:
- “Background and conceptual considerations”,
- “Study design”,
- “Study procedures of investigation or intervention and statistical analysis”, and
- “Presentation and interpretation of findings”.
Each category consists of two or three items, respectively. The items of the checklist are described and explained with questions and further statements (Table 1).
In consensus discussions, we (MP, EB, KR, UL) adapted the checklist to observational study designs, such as cohort and case-control studies. For observational studies, the assessment of occupational exposure was added to the category “Study procedures of investigation or intervention and statistical analysis”. Although no cross-sectional studies met the inclusion criteria in the systematic review by Reichel et al. [23], it should also be possible to use the sex/gender checklist for cross-sectional studies [23,24].
Minimal modifications (Table 1: specifications were in bold) were made to the original checklist by Demetriou et al. (2019), regarding sex/gender-specific exposure or exposure assessment (e. g., Item 6b: study procedures (assessment of occupational exposure and/or intervention content and delivery, if appl.), to permit the assessment of sex/gender bias among observational studies of sedentary work. For example, the sex/gender checklist was expanded to encompass sex/gender in the description of physical activity/sedentary behaviour exposure in the study background to meet our basic requirements. For instance, van der Ploeg et al. [25] described in their introduction that men and women differ in their sedentary behaviour, referring to Australian National Health Survey data [25]. Furthermore, in the “Study procedures” category, we added the “assessment of occupational exposure” to the study procedure item and specified that sex/gender-related confounders for occupational health (e. g., part-time/full-time working hours, family work with unpaid household and carework) should be considered in the analysis (Table 1, specifications are in bold).
The reviewers independently rated each article using the sex/gender checklist by first assessing the level of reporting for all items using the following ratings: “detailed”, “basic”, “no information provided”, or “not applicable”. The reviewers then summarised the items for each category and assigned the highest score, which was achieved in one of two or two of three items, respectively, to a category. For example, if the first item in the background category was marked “no information” and the other two items were marked “basic”, then “basic” was selected for this category. To simplify the evaluation process, reviewers discussed the evaluation results together at the category level and followed the same procedure for determining the overall quality with which sex/gender was reported in each publication.
The overall score of a study corresponded to the highest score achieved in at least two of the four categories, with each of the four categories considered equally. For example, if “basic” was awarded in the categories “Background” and “Study design”, “no information” in the category “Study procedures” and “detailed” in the category “Presentation and interpretation of finding”, the primary study received an overall score of “basic”. A scoring of “detailed” in one category was insufficient to receive an overall score of “detailed”. In another example using the study by Møller et al. [26], the categories “Background and conceptual considerations” and “Study design” were marked “no information”, and the categories “Study procedures of investigation or intervention and statistical analysis” and “Presentation and interpretation of findings” were ranked “basic”. Thus, in the overall assessment, this study received the rating “basic” (S1 Table). If the overall assessment resulted in “no information”, there was a low consideration of sex/gender. The levels “basic” and “detailed” correspond to at least a fundamental consideration of sex/gender.
Application of the sex/gender checklist and analyses
The adapted checklist was applied to the systematic review with a specific assessment of sex/gender bias. This was described in a PROSPERO protocol [24]. Among other things, this review aimed to summarise the research examining the association between occupational sitting and cardiometabolic risk and to determine whether the associations differ between men and women. The risk of bias was assessed using the Scottish Intercollegiate Guidelines Network (SIGN) checklist for cohort, case-control studies and controlled trials [17]. For the assessment of sex/gender bias, each publication was considered individually regardless of study affiliation (e. g., there were two publications from the Whitehall study), as the assessment of sex/gender bias may differ between publications even though the same population was considered.
First, two reviewers (KR, MP) piloted the sex/gender checklist on two cohort studies, two intervention studies, and one case-control study. Next, the two reviewers (KR, MP) independently rated each article using the sex/gender checklist. Finally, a third reviewer (EB) decided on all category-level and overall assessment disagreements. To examine the usability of the checklist, the reviewers also documented the average processing time of the checklist for each publication.
Statistical analyses
Descriptive statistics were calculated for year of publication, risk of bias based on the SIGN, and study design stratified by overall sex/gender bias of the primary studies. We used bivariate and multivariable logistic regression models to examine the associations between consideration of sex/gender bias (y, dependent variable) and the publication characteristics (xi, independent variables), publication year, study design, and risk of bias based on the SIGN. The dependent variable is assigned the value 1 for at least a basic consideration of sex/gender bias (“detailed”, “basic”) and the value 0 for low or no consideration of sex/gender bias (“no information provided”). No study was rated as “not applicable” in the overall assessment (and this scenario is very unlikely), so the “not applicable” option was not included in the calculation. The model results are reported as odds ratios (ORs) with 95% confidence intervals (CIs). The multivariable regression model included all the variables. The variable for study design was categorised into intervention studies and observational studies (case-control and cohort studies) due to the sparsity of case-control studies. In a sensitivity analysis, we also considered publication year as a continuous variable. The risk of bias assessed with the SIGN was dichotomised in low and high risk of bias (reference). Statistical analyses were performed using IBM SPSS Statistics for Windows, version 29.0 (Armonk, NY: IBM Corp).
Results
The adapted sex/gender bias checklist was pilot tested using the systematic review by Reichel et al. [23], which included 19 intervention studies, 25 cohort studies, and 5 case-control studies (39 studies described in 49 publications) in the data synthesis (S1 Table). Each reviewer needed an average of 20–30 minutes per study (reading and assessing) to complete the sex/gender checklist.
Sex/gender checklist
In summary, 69% (n = 34) of the studies evaluated provided at least basic information and 31% (n = 15) provided no information regarding the consideration of sex/gender. Fig 1 shows the consideration of sex/gender stratified by study design in the four categories and the summary decisions.
Intervention studies.
None of the 19 intervention studies (Fig 1) described sex/gender aspects for the categories “Background and conceptual considerations” or “Study design”. All intervention studies described basic information for the category “Study procedures of investigation or intervention and statistical analysis”. However, this basic information comprised only a description of the number of men and women in the study population. Five intervention studies reported how sex/gender was considered in the statistical analyses [27–31]. In these statistical analyses, sex-adjusted results of four cluster randomised controlled trials [28–31] were presented and one study reported results stratified by sex [27]. This intervention study assessed sex/gender differences in the main outcome (interstitial glucose) [27].
Cohort and case-control studies.
Fig 1 shows that 60% (n = 15) of the cohort studies presented no information on sex/gender terminology, research questions or theoretical or conceptual linkages with sex/gender (“Background and conceptual considerations”) [26,32–45]. In 72% (n = 18) of the cohort studies, no information was given on measurement instruments or sample recruitment (“Study design”) [25,26,32,33,35,37–41,46–53]. Only one study provided detailed information in this category [54].
In 24% (n = 6) of the cohort studies, detailed sex/gender-related information was given for the category “Presentation and interpretation of findings” [33,36,42,43,48,49]. For example, Ferrario et al. [48] included only men, but discussed the results in the context of sex/gender. In ten (40%) cohort studies, a sex-stratified analysis was performed [33,35,36,42,43,45,49–51,55]. Sex-adjusted statistical association analyses were conducted in 13 (52%) cohort studies [26,32,34,37–39,42,44,45,49,51,52,56].
In three (60%) of the five case-control studies, no sex/gender information was provided in the “Study design” [57–59] or “Presentation and interpretation of findings” [57–59] categories (Fig 1). All case-control studies considered components of sex/gender analysis in the presenting of findings, for example, by reporting the number of men and women or by conducting different analyses for men and women. In two case-control studies, a sex-stratified analysis was performed [60,61]. Sex-adjusted association analyses were conducted in three case-control studies [58,60,61]. In the “Presentation and interpretation of findings” category, we found detailed information on differences between sex/gender in relation to the reported primary outcome (acute myocardial infarction) in one (20%) of the case-control studies [60].
Associations between sex/gender bias and study characteristics
According to the bivariate and multivariable models, there was no statistically significant association between publication year as a linear variable and the consideration of sex/gender (bivariate model: OR per year = 0.89; 95% CI: 0.65, 1.23, multivariable model: OR per year = 0.98; 95% CI: 0.70, 1.38). The effect estimates for publication year categories from the bivariate and multivariable models were also not statistically significant (Table 2). The bivariate model showed that sex/gender tended to be considered less frequently in 2015–2016 (OR = 0.80; 95% CI: 0.17; 3.37) and 2017–2018 (OR = 0.56; 95% CI: 0.12; 2.54) than in the reference years (2012–2014). However, in the multivariable model, the odds that sex/gender was considered reversed for 2015–2016 compared to the reference years (2012–2014) (OR = 1.15; 95% CI: 0.19; 6.84).
Compared to those in the intervention studies (reference), the odds of considering sex/gender increased by a factor of 3.60 (95% CI: 1.01, 12.81) in the observational studies. According to the multivariable model, the odds ratio of considering sex/gender increased by a factor of 4.52 (95% CI: 1.03, 19.91) in the observational studies. A low risk of bias (based on the SIGN) was not associated with at least a basic consideration of sex/gender in either the bivariate or the multivariable models (Table 2).
Discussion
We demonstrated the application of a sex/gender checklist for the determination of consideration of sex/gender in aetiological studies in the field of occupational health within a systematic review.
We found that all studies (n = 49) examining occupational sedentary behaviour and cardiometabolic health outcomes included basic information regarding sex/gender in the category “Study procedures of investigation or intervention and statistical analysis”, and a few studies (n = 8) presented detailed information on sex/gender in the category “Presentation and interpretation of findings”. We found deficits among the intervention studies, as no intervention study (n = 19) presented information in the categories “Background and conceptual considerations” and “Study design”, and only 1 of the 19 (6%) intervention studies reported sex-stratified analyses. Cohort studies were more likely to provide detailed (6 of 23 cohort studies) or basic (15 of 23 cohort studies) sex/gender information for the “Presentation and interpretation of findings” than either case-control (1 detailed and 1 basic of 5 case-control studies) or intervention studies (1 detailed and 9 basic of 19 intervention studies). Observational studies were more likely to consider sex/gender than intervention studies. This might be because observational studies will almost always need to consider sex/gender as a confounding factor, whereas in RCTs it is expected that randomisation will remove most confounding. This may have led observational studies to consider and report sex/gender factors more thoroughly. Our finding that observational studies were more than three times more likely to consider sex/gender than intervention studies comprised a relevant difference that should be examined further in future studies. This suggests that sex/gender is currently not well addressed in occupational intervention studies of sedentary work.
Regarding our research questions, we found no temporal trend and no statistically significant association between risk of bias (based on the SIGN) and the consideration of sex/gender.
We expected an increase in consideration of sex/gender over time with growing awareness, however, this was not the case. Possibly, improvements in reporting began before our first included article was published in 2012. In 2000, Niedhammer et al. published a review of publications in occupational health epidemiology from six journals published in 1997 and reported that 11% of the 348 articles reviewed did not specify the sex of the study population, that women were still studied less often than men, and that sex was not investigated as a factor in many mixed-sex studies [62]. In 2022, Antequera et al. reported that overall sex/gender consideration in Cochrane reviews was inadequate [63]. Only 2.7% (14/516) of systematic reviews reported sex/gender in all of the analysed categories (abstract, methods, description of results, analytic results and discussion) [63]. All of the publications (n = 49) we assessed reported the sex of the study population, and we found that 40% of observational studies conducted and reported the results of sex-stratified analysis. Although it is not always possible to stratify by sex/gender when one group is insufficiently represented, epidemiologists should examine whether the dimensions of both sex and gender are adequately reflected in their research design, measures, and interpretation [64]. This includes accounting for dimensions of biological and/or social sex that are relevant to the research questions and identifying possible direct or mediated causal pathways for sex/gender-specific factors [64]. Current developments in research on biological sex and other gender perspectives (e. g., gender identity) must be taken into account [6,64–66]. Bauer recommends distinguishing between the effects of several sex/gender measures (e. g., sex at birth, gender identity) as well as sex-specific and sex-associated factors in the data analysis [64].
According to the recommendations of the SAGER reporting guidelines, the results should be reported stratified by sex and gender [9]. To date, there is no specific guidance from the Cochrane on how to assess primary studies in terms of sex/gender. In the “Handbook for Systematic Reviews of Intervention”, Cochrane recommends for systematic reviews that, for example, biological differences between women and men that may influence the response to an intervention should be made clear in the discussion of the results in the review. This includes a description of whether the included studies included all or only some of these groups and whether significant subgroup effects were found [67]. The unclear handling of sex/gender could be a possible explanation for why primary studies with a low risk of bias (based on the SIGN) did not consider sex/gender aspects.
Since interventions should ideally be tailored to both sexes [68] or better to all sexes, studies should also analyse results for men and women separately so that sex/gender-sensitive aspects can be evaluated. In this way, specific prevention strategies can be developed, when necessary [13]. However, conducting subgroup analyses in systematic reviews is difficult when too few primary studies report sex/gender-stratified results. In their systematic review of intervention studies of sedentary behaviour in boys and girls, Vondung et al. [15] excluded all studies (n = 72) that did not explain how they dealt with sex and gender in the analysis or did not report adjusted or sex-stratified results from their meta-analysis and assessed the included studies (n = 67) with the original sex/gender checklist [15].
Strengths and limitations of the sex/gender checklist
This sex/gender checklist as a first step in the development of systematic reviews to explicitly consider sex/gender in primary studies in order to differentiate and compare sex/gender bias between different study designs and epidemiological questions. The reviewers (KR, MP, EB) found the checklist was useable and it allowed an overall assessment or an assessment at category level. The application of the checklist has shown in which areas future developments and research may be possible. When revising the sex/gender checklist, it is essential to consider current recommendations and developments in epidemiological research. The multidimensionality of sex/gender [64] should also remain an important component of the abridged sex/gender checklist. A further limitation is that only the final consensus results of the assessments were documented electronically, which did not allow calculation of kappa values for interrater reliability. This should be taken into account when evaluating future versions of the sex/gender checklist.
Although the reviewers found the evaluation time (20–30 minutes per study) acceptable, this time requirement might discourage researchers with limited resources to consider an assessment of sex/gender bias. A further limitation of the current checklist, was that the current checklist does not yet provide any guidance for assessing studies where only one sex/gender was included [46–48,53]. Guidance on how to rate such studies would be helpful, because occupational studies sometimes exclude the underrepresented sex/gender when researching occupations comprised predominantly of women or men (horizontal segregation) [69]. In addition, some studies may be designed to examine one sex or the other to fill the research gap created by previous sex/gender biased research. Andersson et al. [70], for example, conducted an intervention study of cardiovascular disease rehabilitation only among women to provide evidence lacking for women [70]. Dealing with this issue should be included in a revised sex/gender checklist. A conscious approach is essential to reflect the reasons for the inclusion or exclusion of gender-specific factors.
Possible future developments/research
Horstmann et al. [6] included 77 different instruments in their scoping review and identified a temporal trend in the number of instruments between 2000 and 2020, and found four self-assessment instruments where participants classified themselves into one or multiple sex and/or gender categories. Even though the published survey instruments considering gender diversity increased from 2010 onwards [6], this was not reflected in the articles we assessed. Gender diversity and the handling is not yet sufficiently reflected in the sex/gender checklist. Future versions of the sex/gender bias checklist may need to be adapted to better assess gender diversity.
We see potential for optimising the sex/gender checklist better reflect how the consideration of sex/gender influenced the results of primary studies. This would require item-level analyses to determine which categories are most relevant or most strongly related with sex/gender-related results and conclusions. An item-level analysis may also show if it is possible to shorten the checklist without minimising its effectiveness.
The processing time of the current sex/gender bias checklist, while acceptable on its own, could be considered a cumbersome addition to the usual assessment of bias, especially when personnel resources and time are lacking or where many primary studies require assessment. A shortened version might be developed, for example one that could be integrated into other risk of bias tools for internal validity, such as those from Cochrane or the SIGN.
Further improvements could be made to how the items and categories are weighted in the summary rating. Currently, the item and categories are weighted equally. With the chosen liberal approach to study assessment studies that at least considered sex/gender basically were discriminated from those that did not. However, this blurred factor pertaining more to reporting quality with factors pertaining to methodological quality. For example, a study that does well with regard to “Background and conceptual considerations” may not have adequately accounted for sex/gender factors in the methods, which could have a greater influence on the results. For example, Jüni et al. [71] included 25 quality assessment scales for clinical trials in a meta-analysis and found that the use of summary scores made it difficult to identify high-quality clinical trials. They also showed that summary scores of quality assessment scale for clinical trials were not significantly associated with treatment effects [71]. To counteract these issues, the sex/gender checklist could be modified so that the items and categories are divided into major and minor domains that give more weight to factors more likely to impact the results. An examination of items or categories in major and minor domains and an improved algorithm for the overall assessment should be advantageous [72]. Also, further research with other examples from the field of occupational safety and health are needed.
Conclusion
The sex/gender checklist for assessing the consideration of sex/gender in aetiological studies of occupational health showed that the sex/gender perspective is not sufficiently taken into account in primary studies of sedentary work behaviour and cardiometabolic health. Applying the checklist to an example systematic review demonstrated that the assessment is possible for different epidemiological study designs. Further improvements of the checklist, regarding its length and suitability for occupational research (e. g., how to assess studies excluding one sex/gender) are necessary. An optimisation of the algorithm for determining the overall sex/gender bias level is considered beneficial. Here, a clearer instrument without a total score and with a central focus on the sex/gender-specific evaluation would be required. Since the consideration of sex/gender has an impact on the interpretation of the aetiological results, as well as on the implementation of sex/gender sensitive prevention strategies, a suitable tool for assessing sex/gender bias is needed to encourage the deliberate consideration of sex/gender in research.
Supporting information
S1 Table. Quality of sex/gender bias in four categories and gender bias overall (stratified by study design).
https://doi.org/10.1371/journal.pone.0324391.s001
(DOCX)
References
- 1.
Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M. Cochrane handbook for systematic reviews of interventions. Glasgow: Wiley-Blackwell. 2019.
- 2. Eick SM, Goin DE, Chartres N, Lam J, Woodruff TJ. Assessing risk of bias in human environmental epidemiology studies using three tools: different conclusions from different tools. Syst Rev. 2020;9(1):249. pmid:33121530
- 3. Romero Starke K, Kofahl M, Freiberg A, Schubert M, Groß ML, Schmauder S, et al. The risk of cytomegalovirus infection in daycare workers: a systematic review and meta-analysis. Int Arch Occup Environ Health. 2020;93(1):11–28. pmid:31359142
- 4. Doull M, Welch V, Puil L, Runnels V, Coen SE, Shea B, et al. Development and evaluation of “briefing notes” as a novel knowledge translation tool to aid the implementation of sex/gender analysis in systematic reviews: a pilot study. PLoS One. 2014;9(11):e110786. pmid:25372876
- 5. , Group CCS, Regitz-Zagrosek V, Oertelt-Prigione S, Prescott E, Franconi F. Gender in cardiovascular diseases: impact on clinical manifestations, management, and outcomes. Eur Heart J. 2016;37(1):24–34.
- 6. Horstmann S, Schmechel C, Palm K, Oertelt-Prigione S, Bolte G. The Operationalisation of Sex and Gender in Quantitative Health-Related Research: A Scoping Review. Int J Environ Res Public Health. 2022;19(12):7493. pmid:35742742
- 7. Ovseiko PV, Greenhalgh T, Adam P, Grant J, Hinrichs-Krapels S, Graham KE, et al. A global call for action to include gender in research impact assessment. Health Res Policy Syst. 2016;14(1):50. pmid:27432056
- 8. Jahn I, Börnhorst C, Günther F, Brand T. Examples of sex/gender sensitivity in epidemiological research: results of an evaluation of original articles published in JECH 2006-2014. Health Res Policy Syst. 2017;15(1):11. pmid:28202078
- 9. Heidari S, Babor TF, De Castro P, Tort S, Curno M. Sex and Gender Equity in Research: rationale for the SAGER guidelines and recommended use. Res Integr Peer Rev. 2016;1:2. pmid:29451543
- 10. Krieger N. Genders, sexes, and health: what are the connections--and why does it matter?. Int J Epidemiol. 2003;32(4):652–7. pmid:12913047
- 11. Eichler M, Fuchs J, Maschewsky-Schneider U. Guidelines to avoid gender bias in health research. Z Gesundheitswiss. 2000;8(4):293–310.
- 12. Doull M, Runnels VE, Tudiver S, Boscoe M. Appraising the evidence: applying sex- and gender-based analysis (SGBA) to Cochrane systematic reviews on cardiovascular diseases. J Womens Health (Larchmt). 2010;19(5):997–1003. pmid:20384450
- 13. Demetriou Y, Vondung C, Bucksch J, Schlund A, Schulze C, Knapp G. Interventions on children’s and adolescents’ physical activity and sedentary behaviour: protocol for a systematic review from a sex/gender perspective. Syst Rev. 2019;8(1):1–7.
- 14. Schlund A, Reimers AK, Bucksch J, Brindley C, Schulze C, Puil L, et al. Do Intervention Studies to Promote Physical Activity and Reduce Sedentary Behavior in Children and Adolescents Take Sex/Gender Into Account? A Systematic Review. J Phys Act Health. 2021;18(4):461–8. pmid:33668018
- 15. Vondung C, Demetriou Y, Reimers AK, Schlund A, Bucksch J. A Sex/Gender Perspective on Interventions to Reduce Sedentary Behaviour in Girls and Boys: Results of the genEffects Systematic Review. Int J Environ Res Public Health. 2020;17(14):5231. pmid:32698412
- 16.
Boutron I, Page M, Higgins J, Altman D, Lundh A, Hróbjartsson A. Considering bias and conflicts of interest among the included studies. Cochrane Handbook for Systematic Reviews of Interventions. Internet: Cochrane. 2024.
- 17. SIGN. Critical appraisal notes and checklists 2014 [cited 27 May 2021]. Available from: https://www.sign.ac.uk/what-we-do/methodology/checklists/.
- 18. Prince SA, Elliott CG, Scott K, Visintini S, Reed JL. Device-measured physical activity, sedentary behaviour and cardiometabolic health and fitness across occupational groups: a systematic review and meta-analysis. Int J Behav Nutr Phys Act. 2019;16(1):1–15.
- 19. Eurostat. Persons performing physical activity when working by type of activity, most frequent activity status, income quantile and degree of urbanization 2021 [cited 23 Jan 2024. ]. Available from:
- 20. Patterson R, McNamara E, Tainio M, de Sá TH, Smith AD, Sharp SJ, et al. Sedentary behaviour and risk of all-cause, cardiovascular and cancer mortality, and incident type 2 diabetes: a systematic review and dose response meta-analysis. Eur J Epidemiol. 2018;33(9):811–29. pmid:29589226
- 21. Berger JS, Elliott L, Gallup D, Roe M, Granger CB, Armstrong PW, et al. Sex differences in mortality following acute coronary syndromes. JAMA. 2009;302(8):874–82. pmid:19706861
- 22. Johansson E, Mathiassen SE, Lund Rasmusse C, Hallman DM. Sitting, standing and moving during work and leisure among male and female office workers of different age: a compositional data analysis. BMC Public Health. 2020;20(1):826. pmid:32487107
- 23. Reichel K, Prigge M, Latza U, Kurth T, Backé E-M. Association of occupational sitting with cardiovascular outcomes and cardiometabolic risk factors: a systematic review with a sex-sensitive/gender-sensitive perspective. BMJ Open. 2022;12(2):e048017. pmid:35135760
- 24. Reichel K, Backé EM, Liebers F, Prigge M. The impact of occupational sitting on cardiometabolic risk factors and cardiometabolic outcomes – a systematic review from a gender-sensitive perspective. PROSPERO. 2018. https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42018079219
- 25. van der Ploeg HP, Møller SV, Hannerz H, van der Beek AJ, Holtermann A. Temporal changes in occupational sitting time in the Danish workforce and associations with all-cause mortality: results from the Danish work environment cohort study. Int J Behav Nutr Phys Act. 2015;12:71. pmid:26031453
- 26. Møller SV, Hannerz H, Hansen AM, Burr H, Holtermann A. Multi-wave cohort study of sedentary work and risk of ischemic heart disease. Scand J Work Environ Health. 2016;42(1):43–51. pmid:26649851
- 27. Brocklebank LA, Andrews RC, Page A, Falconer CL, Leary S, Cooper A. The Acute Effects of Breaking Up Seated Office Work With Standing or Light-Intensity Walking on Interstitial Glucose Concentration: A Randomized Crossover Trial. J Phys Act Health. 2017;14(8):617–25. pmid:28422556
- 28. Danquah IH, Kloster S, Holtermann A, Aadahl M, Bauman A, Ersbøll AK, et al. Take a Stand!-a multi-component intervention aimed at reducing sitting time among office workers-a cluster randomized trial. Int J Epidemiol. 2017;46(1):128–40. pmid:27094749
- 29. Verweij LM, Proper KI, Weel ANH, Hulshof CTJ, van Mechelen W. The application of an occupational health guideline reduces sedentary behaviour and increases fruit intake at work: results from an RCT. Occup Environ Med. 2012;69(7):500–7. pmid:22383591
- 30. Healy GN, Winkler EAH, Eakin EG, Owen N, Lamontagne AD, Moodie M, et al. A Cluster RCT to Reduce Workers’ Sitting Time: Impact on Cardiometabolic Biomarkers. Med Sci Sports Exerc. 2017;49(10):2032–9. pmid:28538025
- 31. Healy GN, Eakin EG, Owen N, Lamontagne AD, Moodie M, Winkler EAH, et al. A Cluster Randomized Controlled Trial to Reduce Office Workers’ Sitting Time: Effect on Activity Outcomes. Med Sci Sports Exerc. 2016;48(9):1787–97. pmid:27526175
- 32. Chau JY, Grunseit A, Midthjell K, Holmen J, Holmen TL, Bauman AE, et al. Sedentary behaviour and risk of mortality from all-causes and cardiometabolic diseases in adults: evidence from the HUNT3 population cohort. Br J Sports Med. 2015;49(11):737–42. pmid:23666019
- 33. Eriksen D, Rosthøj S, Burr H, Holtermann A. Sedentary work--associations between five-year changes in occupational sitting time and body mass index. Prev Med. 2015;73:1–5. pmid:25575798
- 34. Johnsen AM, Alfredsson L, Knutsson A, Westerholm PJM, Fransson EI. Association between occupational physical activity and myocardial infarction: a prospective cohort study. BMJ Open. 2016;6(10):e012692. pmid:27697879
- 35. Kikuchi H, Inoue S, Odagiri Y, Inoue M, Sawada N, Tsugane S, et al. Occupational sitting time and risk of all-cause mortality among Japanese workers. Scand J Work Environ Health. 2015;41(6):519–28. pmid:26575417
- 36. Kim Y, Wilkens LR, Park S-Y, Goodman MT, Monroe KR, Kolonel LN. Association between various sedentary behaviours and all-cause, cardiovascular disease and cancer mortality: the Multiethnic Cohort Study. Int J Epidemiol. 2013;42(4):1040–56. pmid:24062293
- 37. Moe B, Mork PJ, Holtermann A, Nilsen TIL. Occupational physical activity, metabolic syndrome and risk of death from all causes and cardiovascular disease in the HUNT 2 cohort study. Occup Environ Med. 2013;70(2):86–90. pmid:23022656
- 38. Picavet HSJ, Pas LW, van Oostrom SH, van der Ploeg HP, Verschuren WMM, Proper KI. The Relation between Occupational Sitting and Mental, Cardiometabolic, and Musculoskeletal Health over a Period of 15 Years--The Doetinchem Cohort Study. PLoS One. 2016;11(1):e0146639. pmid:26752695
- 39. Pulsford RM, Stamatakis E, Britton AR, Brunner EJ, Hillsdon MM. Sitting behavior and obesity: evidence from the Whitehall II study. Am J Prev Med. 2013;44(2):132–8. pmid:23332328
- 40. Pulsford RM, Stamatakis E, Britton AR, Brunner EJ, Hillsdon M. Associations of sitting behaviours with all-cause mortality over a 16-year follow-up: the Whitehall II study. Int J Epidemiol. 2015;44(6):1909–16. pmid:26454871
- 41. Saidj M, Jorgensen T, Jacobsen R, Linneberg A, Oppert J, Aadahl M. Work and leisure time sitting and inactivity: effects on cardiorespiratory and metabolic health. Eur J Prev Cardiol. 2016;23(12):1321–9.
- 42. Smith P, Ma H, Glazier R, Gilbert-Ouimet M, Mustard C. The relationship between prolonged occupational standing and sitting and incident cardiovascular disease. Eur J Prev Cardiol. 2017;24(2 Suppl 1):21.
- 43. Stamatakis E, Chau JY, Pedisic Z, Bauman A, Macniven R, Coombs N, et al. Are sitting occupations associated with increased all-cause, cancer, and cardiovascular disease mortality risk? A pooled analysis of seven British population cohorts. PLoS One. 2013;8(9):e73753. pmid:24086292
- 44. Stamatakis E, Pulsford RM, Brunner EJ, Britton AR, Bauman AE, Biddle SJ, et al. Sitting behaviour is not associated with incident diabetes over 13 years: the Whitehall II cohort study. Br J Sports Med. 2017;51(10):818–23. pmid:28465446
- 45. Sakaue A, Adachi H, Enomoto M, Fukami A, Kumagai E, Nakamura S, et al. Association between physical activity, occupational sitting time and mortality in a general population: An 18-year prospective survey in Tanushimaru, Japan. Eur J Prev Cardiol. 2020;27(7):758–66. pmid:30396293
- 46. Allesøe K, Holtermann A, Aadahl M, Thomsen JF, Hundrup YA, Søgaard K. High occupational physical activity and risk of ischaemic heart disease in women: the interplay with physical activity during leisure time. Eur J Prev Cardiol. 2015;22(12):1601–8. pmid:25311002
- 47. Allesøe K, Søgaard K, Aadahl M, Boyle E, Holtermann A. Are hypertensive women at additional risk of ischaemic heart disease from physically demanding work?. Eur J Prev Cardiol. 2016;23(10):1054–61. pmid:26876489
- 48. Ferrario MM, Roncaioli M, Veronesi G, Holtermann A, Clays E, Borchini R, et al. Differing associations for sport versus occupational physical activity and cardiovascular risk. Heart. 2018;104(14):1165–72. pmid:29440185
- 49. Hayashi R, Iso H, Cui R, Tamakoshi A. Occupational physical activity in relation to risk of cardiovascular mortality: The Japan Collaborative Cohort Study for Evaluation for Cancer Risk (JACC Study). Prev Med. 2016;89:286–91.
- 50. Lin T, Courtney TK, Lombardi DA, Verma SK. Association Between Sedentary Work and BMI in a U.S. National Longitudinal Survey. Am J Prev Med. 2015;49(6):e117-23. pmid:26437869
- 51. Martin KR, Kuh D, Harris TB, Guralnik JM, Coggon D, Wills AK. Body mass index, occupational activity, and leisure-time physical activity: an exploration of risk factors and modifiers for knee osteoarthritis in the 1946 British birth cohort. BMC Musculoskelet Disord. 2013;14:219. pmid:23883324
- 52. Pinto Pereira SM, Power C. Sedentary behaviours in mid-adulthood and subsequent body mass index. PLoS One. 2013;8(6):e65791. pmid:23762427
- 53. Hall C, Heck JE, Sandler DP, Ritz B, Chen H, Krause N. Occupational and leisure-time physical activity differentially predict 6-year incidence of stroke and transient ischemic attack in women. Scand J Work Environ Health. 2019;45(3):267–79.
- 54. Thompson WG, St Sauver J, Schroeder D. Occupation, Sitting, and Weight Change in a Cohort of Women Employees. J Occup Environ Med. 2018;60(1):44–7. pmid:28902658
- 55. Johnsen A, Alfredsson L, Knutsson A, Westerholm P, Fransson E. Weak associations between occupational physical activity and myocardial infarction. Occup Environ Med. 2016;73(Suppl 1):A198.
- 56. van der Ploeg HP, Hillsdon M. Is sedentary behaviour just physical inactivity by another name?. Int J Behav Nutr Phys Act. 2017;14(1):142. pmid:29058587
- 57. Kumar A, Prasad M, Kathuria P. Sitting occupations are an independent risk factor for ischemic stroke in north Indian population. Int J Neurosci. 2014;124(10):748–54.
- 58. Ma Y, Wang Y-J, Chen B-R, Shi H-J, Wang H, Khurwolah MR, et al. Study on association of working hours and occupational physical activity with the occurrence of coronary heart disease in a Chinese population. PLoS One. 2017;12(10):e0185598. pmid:29049313
- 59. Selim S, Rahman R, Yasmin R, Karim N, Chowdhury SH, Lona H, et al. Risk factors of acute coronary syndrome among Bangladeshi people. Mymensingh Med J. 2013;22(3):513–21. pmid:23982542
- 60. Cheng X, Li W, Guo J, Wang Y, Gu H, Teo K, et al. Physical activity levels, sport activities, and risk of acute myocardial infarction: results of the INTERHEART study in China. Angiology. 2014;65(2):113–21. pmid:23324448
- 61. Held C, Iqbal R, Lear SA, Rosengren A, Islam S, Mathew J, et al. Physical activity levels, ownership of goods promoting sedentary behaviour and risk of myocardial infarction: results of the INTERHEART study. Eur Heart J. 2012;33(4):452–66. pmid:22238330
- 62. Niedhammer I, Saurel-Cubizolles M-J, Piciotti M, Bonenfant S. How is sex considered in recent epidemiological publications on occupational risks?. Occup Environ Med. 2000;57(8):521–7.
- 63. Antequera A, Cuadrado-Conde MA, Roy-Vallejo E, Montoya-Martínez M, León-García M, Madrid-Pascual O, et al. Lack of sex-related analysis and reporting in Cochrane Reviews: a cross-sectional study. Syst Rev. 2022;11(1):281. pmid:36572932
- 64. Bauer GR. Sex and Gender Multidimensionality in Epidemiologic Research. Am J Epidemiol. 2023;192(1):122–32. pmid:36193856
- 65. Döring N. Zur operationalisierung von geschlecht im fragebogen: probleme und lösungsansätze aus sicht von mess-, umfrage-, gender-und queer-theorie. GENDER. 2013;5(2):17–8.
- 66. Horstmann S, Bolte G. Geschlechtliche Vielfalt erfassen. Pflegez. 2022;75(12):52–4.
- 67.
Schünemann H, Vist G, Higgins J, Santesso N, Deeks J, Glasziou P. Interpreting results and drawing conclusions. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane. 2023.
- 68. Wallmann-Sperlich B, Bucksch J, Schneider S, Froboese I. Socio-demographic, behavioural and cognitive correlates of work-related sitting time in German men and women. BMC Public Health. 2014;14:1259. pmid:25496569
- 69.
Kolip P, Lademann J. Family and health. In: Hurrelmann K, Razum O, editors. Handbook health sciences. 6th ed. Weinheim: Beltz Juventa. 2012. p. 517–40.
- 70. Andersson A, Sundel KL, Undén A-L, Schenck-Gustafsson K, Eriksson I. A five-year rehabilitation programme for younger women after a coronary event reduces the need for hospital care. Scand J Public Health. 2010;38(6):566–73. pmid:20639274
- 71. Jüni P, Witschi A, Bloch R, Egger M. The hazards of scoring the quality of clinical trials for meta-analysis. JAMA. 1999;282(11):1054–60. pmid:10493204
- 72. Ijaz S, Verbeek J, Seidler A, Lindbohm M-L, Ojajärvi A, Orsini N. Night-shift work and breast cancer—a systematic review and meta-analysis. Scand J Work Environ Health. 2013;39(5):431–47.