- Gender is a multidimensional concept which includes gender norms, gender identity and gender relations (Nielsen et al., 2021). Gender norms consist of spoken and unspoken rules produced through social institutions, such as the family and workplace, and cultural products, such as technology and social media. Gender identity refers to how individuals and groups perceive and present themselves in relation to gender norms. Gender relations refer to how individuals interact with other people and institutions in specific sociocultural contexts. In health, all cases may apply, but frequently one dimension is more relevant than the others for the specific question asked.
-
Researchers should first identify which dimensions of gender are most relevant to their work and select appropriate instruments to capture any that are relevant. In recent years, some instruments have been developed which group two or more dimensions (Tate, et al., 2014; Pelletier, et al., 2015; Nielsen et al., 2021). There is marginal overlap between the instruments currently available, so researchers need to make an informed choice. Sometimes a single in-depth instrument might be a better option than a more comprehensive one that offers less differentiated output. Intersectional aspects will most probably play a role as well. Depending on the research question, factors intersecting with sex and gender, such as ethnicity or socioeconomic status, should be considered.
- Gender identity of the participants. A broad array of gender identities exists, and they are evolving over time. Researchers need to find a balance between inclusion and practicality. Three categories (man, woman, non-binary or gender diverse) may be more suitable for analysis than six or seven; however, making this choice may exclude some participants. Researchers should spend some time defining whether and how gender-diverse populations will be identified for recruitment while avoiding stereotyping and discrimination. Some consideration should also be given to the study population and their potential resistance to modern concepts of gender as asked in the study. Which questionnaires are acceptable for which population? Extensive piloting with the target population is advisable when new methods are employed; some degree of resistance can be expected.
- Gender norms and relations In choosing the type of study, some thought should be given to the potentially dynamic nature of gendered norms and behaviors. Gender norms and behaviors change over time and depend on the societal and cultural context. The experiences of a gay black man in an urban context differ from those of a straight Asian woman in a rural setting. Nevertheless, the gender-related reality of both these individuals has changed over the last two decades. Is this potentially relevant to the research, and can researchers account for this?
- Gender identities and norms of the researchers. Finally, the gender identity of investigators and the gender norms they experience could also be a factor. Gender norms and relations play out in interactions between participants and researchers. Collecting data only for participants may limit opportunities for investigation.
-
The impact of gender on the research question can be addressed in quantitative, qualitative or mixed-method studies. Quantitative data will offer highly standardized and comparable responses across different studies. If deeper insight into motivations, contextual factors and situated experiences and realities is needed, qualitative data might be a better option. Qualitative data may supplement quantitative findings to better understand outliers or missing data, or to explore the accuracy of the quantitative tools in certain subpopulations. Especially in exploring gender-diverse populations, whose health needs have largely been ignored, qualitative research or a mixed-method design might be a better fit.
- Explanation of gender terms. Participants might not be familiar with the concept of gender and might question its use. It might be advisable to explain why gender is a research component.
The fundamental requirement for statistical analysis is an adequately large sample and an appropriate number of outcomes to guarantee the power of the analysis. Researchers should reflect on the distribution of outcomes and consider oversampling, if outcomes are not equally distributed across genders. Specifically, a low number of participating gender-diverse subjects will affect the analysis of this subgroup. These individuals are frequently treated as outliers and excluded from the analysis. Ways to meaningfully include information about their experiences should be considered a priori.- Multiple instruments. If multiple instruments will be used to determine different dimensions of gender (identity, norms, relations), these might have to be weighted for analysis. Choice of instruments will also influence whether variables will be nominal or continuous. This will affect the type of analysis performed since many methods traditionally used in biomedicine are designed for dichotomous outcomes. This also applies when aggregate scores are being used. The hierarchical order of this analysis should be defined a priori to avoid the introduction of potential sources of bias.
Sex and gender interact. A possible interaction between sex and gender can be expected in human subjects. Will this be part of the analysis and will this interaction be explicitly analyzed? In addition to the interaction of sex and gender, other factors intersecting with gender may need to be included in the analysis. In many cases, gender will play a role in both explaining the results and acting as a potential modulator for other causal pathways. Nevertheless, most studies will allow only for the description of correlation rather than causality. The structural interactions of other dimensions with gender, i.e. work, education, (dis)ability and ethnicity should be considered, and analysis should reflect how these have been taken into account. The design of these analyses can vary from multiple stratifications to the inclusion of interacting terms to complex factorial designs.Qualitative data. Qualitative data analysis can be exploratory or confirmatory in nature, based on the study design and research objectives. Gender needs to be considered as an important modulator in all research contexts. It is important to consider how gender might not just modify relations between individuals in isolation, but also interact with age, hierarchies, socioeconomic means, etc. Identifying outliers in qualitative analyses, i.e. individuals who do not conform to the majority of the population, often provides a better understanding of the gender norms and relations at play. These aspects can also be identified by addressing underlying domains, such as gendered language and gendered themes. How participants and patients describe their symptoms and experiences can differ significantly depending on prevailing gender norms and can have an effect on the diagnostic and therapeutic options provided.
Mixed methods. In mixed-method sequential designs—where qualitative approaches are followed by quantitative ones—qualitative data analysis may be used to identify which gender variables to include in the quantitative part. In sequential designs where qualitative approaches are used to explore findings from quantitative data, such analysis aims to bring a better understanding of drivers of observed statistical differences. For example, if quantitative data reveal that health outcomes differ between partnered women with children and single women with children, qualitative data may be used to understand what dimensions of “partnership” explain the observed difference.
Other analysis. Analysis of verbal and non-verbal gendered body language, tone of voice and interactions may be appropriate. A social science perspective is particularly helpful for analyzing gender through qualitative methods.
Secondary data analysis. The growing number of internationally available cohort studies offers an opportunity for post-hoc analysis of gender-correlated variables. Although this option is limited compared to the strategic inclusion of gender variables when designing research, it still allows for pilot data that may be valuable for identifying follow-up studies. Some options for the development of proxy measures based on existing variables within cohort studies have been reported (Pelletier, 2015; Nielsen, 2021) and more are underway. The most basic indicators used are education, professional status, and individual or household income. However, more complex measures, such as care responsibilities and their timing, role behavior, stress, anxiety and depression have also been suggested.
Reporting and Disseminating results
Formats for reporting. Some granting agencies and peer-reviewed journals require reporting gender in addition to sex (see Policy Recommendations), yet there is currently no standard for reporting gender. Reporting should offer enough detail to support reproducibility. Researchers should consider following the SAGER guidelines (Heidari et al., 2016) for general publications and the PRISMA and CONSORT Equity guidelines (Welch et al., 2016; Welch et al., 2017) when reporting clinical trials. The latter have been designed for the reporting of sex and potential intersecting factors but can be applied to the reporting of gender. Report null findings. Researchers should report when gender differences (main or interaction effects) are not detected in their analyses in order to reduce publication bias, an important consideration in meta-analyses (IOM, 2012). If gender-diverse individuals are not oversampled, they may form a small subgroup of the study population and preclude comprehensive statistical testing because of a lack of power. In this case, descriptive statistics should still be reported to allow for potential pooled analysis in the future.
| Choice of instrument |
|
| Application of the instrument |
|
| Study type |
|
| Study protocol |
|
| Analysis |
|
| Reporting |
|
Works Cited
Clayton, J. A. (2016). Studying both sexes: a guiding principle for biomedicine. FASEB J., 30, 519–524.
Clayton, J. A., & Tannenbaum, C. (2016). Reporting sex, gender, or both in clinical research? JAMA, 316(18), 1863-1864. Dahlgren, G., & Whitehead, M. (1991). Policies and Strategies to Promote Social Equity in Health. Stockholm: Institute for Futures Studies. Heidari, S., Babor, T. F., De Castro, P., Tort, S., & Curno, M. (2016). Sex and gender equity in research: rationale for the SAGER guidelines and recommended use. Research Integrity and Peer Review, 1(1), 2. Institute of Medicine (IOM). (2010). Women’s Health Research: Progress, Pitfalls, and Promise. Washington, D.C.: National Academies Press. Moerman, C., Deurenberg, R., & Haafkens, J. (2009). Locating sex-specific evidence on clinical questions in MEDLINE: a search filter for use on OvidSP. BioMed Central Medicine Medical Research Methodology, 9(1), 25. Nielsen, M. W., Stefanick, M. L., Peragine, D., Neilands, T. B., Ioannidis, J. P., Pilote, L., ... & Schiebinger, L. (2021). Gender-related variables for health research. Biology of sex differences, 12(1), 1-16. Oertelt-Prigione, S., Parol, R., Krohn, S., Preissner, R., & Regitz-Zagrosek, V. (2010). Analysis of sex and gender-specific research reveals a common increase in publications and marked differences between disciplines. BioMed Central Medicine, 8, 70-80. Oertelt-Prigione, S.& Regitz-Zagrosek, V. (Eds.) (2012). Sex and Gender Aspects in Clinical Medicine. London: Springer Verlag. Pelletier, R., Ditto, B., & Pilote, L. (2015). A composite measure of gender and its association with risk factors in patients with premature acute coronary syndrome. Psychosomatic medicine, 77(5), 517-526. Schenck-Gustafsson, K., DeCola, P., Pfaff, D. & Pisetkey, D. (Eds.) (2012). Handbook of Clinical Gender Medicine. Basel: Karger. Tate, C. C., Youssef, C. P., & Bettergarcia, J. N. (2014). Integrating the study of transgender spectrum and cisgender experiences of self-categorization from a personality perspective. Review of General Psychology, 18(4), 302-312. Tannenbaum, C., Schwarz, J. M., Clayton, J. A., de Vries, G. J., & Sullivan, C. (2016). Evaluating sex as a biological variable in preclinical research: the devil in the details. Biology of sex differences, 7(1), 13. Tannenbaum, C., Ellis, R. P., Eyssel, F., Zou, J., & Schiebinger, L. (2019). Sex and gender analysis improves science and engineering. Nature, 575(7781), 137-146. Welch, V., Petticrew, M., Petkovic, J., Moher, D., Waters, E., White, H., ... & PRISMA-Equity Bellagio group. (2016). Extending the PRISMA statement to equity-focused systematic reviews (PRISMA-E 2012): explanation and elaboration. Journal of Development Effectiveness, 8(2), 287-324. Welch, V. A., Norheim, O. F., Jull, J., Cookson, R., Sommerfelt, H., & Tugwell, P. (2017). CONSORT-Equity 2017 extension and elaboration for better reporting of health equity in randomised trials. British Medical Journal, 359. doi: https://doi.org/10.1136/bmj.j5085
