Commentary

By Meaghan Ray Peters & Teri Slade

---

Introduction to Sex and Gender-based Analyses

On July 8, 2009, the Canadian Institutes of Health Research (CIHR) implemented a requirement that all research applicants integrate sex and gender into their research when appropriate (1). At the most basic level, Sex- and Gender-Based Analysis (SGBA) requires researchers to define sex and gender separately and measure both on, at minimum, a three-point scale (male, female, and intersex for sex category and men, women, and gender diverse people for gender category; 3–7). Without adequate SGBA, differences in physical function or responses to treatment might be misinterpreted as attributable to biological (sex) rather than social (gender) differences, or findings might be generalized to everyone within a sex or gender category when they do not always apply. To receive funding from CIHR, researchers must describe how they will include SGBA in their research (7). This article aims to identify challenges with the current approach to SGBA and propose an alternative strategy to overcome these challenges.

Critical Challenges to Sex and Gender-based Analyses

This approach often leads to methodological and ethical challenges. Firstly, small numbers in the intersex and gender-diverse categories can impair our ability to determine relationships between variables, limiting the analyses of these categories to descriptive statistics. Alternatively, researchers might omit these data or group them with larger categories that may not adequately represent participants’ characteristics. There is also a concern about reporting data from small categories while maintaining anonymity (8).

Recommended Solution: Looking Beyond Global Categorization

To address these challenges, researchers must go one step further and measure specific sex and gender sub-variables such as body composition or gender presentation. This will produce more meaningful data with less confusion and frustration for researchers (9–11).

Sex and gender comprise a constellation of variables as presented in the Tables below (2,4,12–15). We use the terms ‘sex’ and ‘gender’ as a stand-in for the many sub-variables that contribute to them. There are numerous combinations of where someone might fall within these sub-variables and, thus, a wide range of profiles within each category of gender and sex. The solution I propose is to measure not only the global variables of sex and gender but also their relevant sub-variables. This could help show overlap between groups, the differences within groups, and help researchers classify intersex and gender-diverse participants in meaningful ways according to the sub-variables assessed.

Considerations for Sex Sub-Variables

Suppose your research may be affected by variability in the biological sex attributes of your participants, such as a novel treatment for incontinence that may work differently depending on the patient’s pelvic anatomy. In that case, it is essential to capture these attributes more specifically than solely recording global sex categories (10,13). Consider the following list of sub-variables. Identify which ones may independently influence your research and choose 2-3 that would be appropriate and feasible to incorporate in your study. This is not an exhaustive list, but it provides a place to start.

Table 1: Sample sub-variables related to an individual’s sex.

Genetic Variables	Hormone Variables (10,16)
Chromosomal sex determination (generally XX or XY but with many other combinations of one to three X or Y chromosomes) (10,13,16-18) Presence/absence of gametes (10) Epigenetic factors such as early social/gestational environment (17,18)	Sex hormone presence Sex hormone amount Sex hormone variability, range, or cycle Hormone insensitivity or resistance
Genital Variables	Secondary Sex Characteristic Variables
Presence of internal reproductive organs and genitalia (10) The function of internal reproductive organs and genitalia Morphology of external genitalia (10) The function of external genitalia ** note: internal and external genitalia can differ (10,16)	Bone structure (hip width, bone mineral density, length of long bones, etc.) (10) Body composition and/or shape (thigh-to-waist circumference ratio, lean muscle mass, laryngeal prominence (Adam’s apple, facial structure) (10) Voice pitch or range (19,20) Body, face, and head hair (presence or loss without modification or treatment) (10) Breast and mammillary gland development, presence, or function (10) Performance factors (strength, speed, flexibility, explosive ability, etc.)

Considerations for Gender Sub-Variables

Suppose your research may be affected by the gender-based experiences of your participants or yourself. In that case, it is important to capture these experiences more specifically than solely recording global gender categories (14,15). Consider the following list of sub-variables that may independently influence your research and identify which 2-3 would be most relevant to measure. Again, this is not an exhaustive list but provides a place to start.

Table 2: Sample sub-variables related to an individual’s gender.

Gender Identity (9–11,14–16)	Gender Norms and Relations (4,6,9,10,15,16)
Self-reported identity (typically used as the global gender category) Gender-typed personality traits Congruence/incongruence: The extent to which a person’s internal sense of gender aligns with their sex, sexual characteristics, how others perceive them, and how they are identified institutionally Gender presentation: Augmentation or suppression of secondary sex characteristics used to identify a stranger’s gender (clothing, use of make-up or accessories, hairstyle, facial and body hair styling, etc.) Gender expression: Behaviours or actions such as speech patterns, gestures, and body language	Gender Norms Social and cultural attitudes and expectations about what is appropriate for a person of a particular gender. (1) Gender Roles o Social privilege and discrimination o Social attitudes, expectations, and stereotypes o Workplace/ hierarchical roles and employment status o Primary earner status o Familial, parental, or caregiver responsibilities o Household responsibilities (2) Institutionalized gender o Legal protections and discrimination o Structural privilege and discrimination – Wage gap – Educational access o Institutionally documented gender and/or sex – Local, provincial, and national ID – Health care, legal, financial, educational, and employment institutions – Social media platforms o Gender Relations: The aspects of interpersonal interactions that relate to, involve, or reflect gender. – Social network and support – Identifying language, (name, pronouns, gendered terms, inclusive language) – Gender-based violence and trauma
Researcher Interactions (10)
Gender-based researcher interactions can occur during recruitment, data collection, analysis, or presentation of results. –        The gender identity of researchers –        Gender presentation of researchers –        Perceived gender of researchers (typically as perceived by the participants) –        Gendered norms experienced by researchers –        Power/privilege differential between researcher and participant

Recommendations for Measuring Sex and Gender Sub-variables

The first step in accurately measuring an attribute or construct is defining it. When identifying, defining, measuring, analyzing, and reporting sex and gender sub-variables, use neutral and precise language. Avoid attaching a global sex or gender category label to a sub-variable (8,11,16,21,22). For example, if you are talking about Testosterone, refer to it as Testosterone rather than ‘the male sex hormone’. If you need to group characteristics for ease of discussion, avoid using the same labels as the global sex or gender categories. For example, gender presentation can be discussed using the terms masculine, androgynous, and feminine.

Sex sub-variables can be captured in three ways: health record, direct measurement (for example by blood test or scan), or self-report. The method will depend on which variables you are collecting, your study population, and available resources. It is important to keep the measurement of sex sub-variables as objective as possible. Researchers’ perception of a participant’s sex characteristics and categorizing decisions are influenced by researchers’ sociocultural understanding of sex and gender (16). Any degree of perceptual judgement can cause the sex variable to be a representation of gender relations and researcher interactions. To avoid this potential confound, where possible, capture sex (and gender) sub-variables using a continuous scale. Define your endpoints (for example, the range of scores among your participants) and place everyone along a continuum rather than defining arbitrary cut-offs for categories (4,14,15,23,24). Examples of variables where this would work well are voice pitch, hip width, grip strength, or, for gender, ratings of masculinity or femininity.

Gender sub-variables can be captured through a variety of methods: self-report using a purpose-made survey or a validated survey tool, through observation, or secondary data (census data, institutional records, legal reports, etc.). Your method will depend on your research question, study design, and the sub-variables you are measuring. Many validated survey tools cover multiple gender sub-variables, so you can use one measure for all relevant sub-variables. However, the majority of these were developed prior to 2005, and thus their validity could be questioned in the modern social context (21). One recently validated survey tool for gender sub-variables is the Stanford Gender-Related Variables for Health Research (15).

Key Messages 

When you solely identify participants using the global ‘sex’ or ‘gender’ categories, you are assuming that their relevant attributes and experiences conform to discrete societal stereotypes. If you do not measure appropriate sex and gender sub-variables, you risk missing significant data that are key to sound and actionable SGBA. Identify which sub-variables are most relevant to your research and determine the most appropriate data capture method, scale, and language. You may need to access further resources to ensure that your strategy for each variable is as inclusive of all human variation as possible.

I hope you have found this primer useful. For the full document with more information on the global sex and gender categories, sub-variables, measurement, and practical strategies, contact the author at mpeters2@ualberta.ca.

Acknowledgements 

Featured illustration by Alabama Drew rehabINK.

To refer to this article, it can be cited as:

Peters, MR, Slade, T. Enhancing Sex- and Gender-Based Analysis by Using Appropriate Sub-Variables. rehabINK. 2023:Issue15. Available from: https://rehabinkmag.com

---

References

1. Sex and Gender in Health Research – CIHR [Internet]. [cited 2022 Aug 16]. Available from: https://cihr-irsc.gc.ca/e/50833.html
2. Johnson JL, Greaves L, Repta R. Better science with sex and gender: Facilitating the use of a sex and gender-based analysis in health research. Int J Equity Health [Internet]. 2009 May 6 [cited 2022 Jul 25];8(1):1–11. Available from: https://equityhealthj.biomedcentral.com/articles/10.1186/1475-9276-8-14
3. Tannenbaum C, Ellis RP, Eyssel F, Zou J, Schiebinger L. Sex and gender analysis improves science and engineering. Nature. 2019 Nov 7;575(7781):137–46.
4. Cislaghi B, Heise L. Gender norms and social norms: differences, similarities and why they matter in prevention science. Sociol Heal Illn. 2020 Feb 1;42(2):407–22.
5. Slade T, Gross DP, Niwa L, McKillop AB, Guptill C. Sex and gender demographic questions: improving methodological quality, inclusivity, and ethical administration. Int J Soc Res Methodol [Internet]. 2020;00(00):1–12. Available from: https://doi.org/10.1080/13645579.2020.1819518
6. M S-D, S S, Hypatia of Alexandria Charter Group. Incorporating the sex and gender perspective in research content: a toolkit [Internet]. Barcelona (Spain); 2022 [cited 2022 Aug 23]. Available from: https://aquas.gencat.cat/web/.content/minisite/aquas/publicacions/2022/toolkit_perspective_sex_gender_research_aquas2022.pdf
7. How CIHR is supporting the integration of SGBA – CIHR [Internet]. [cited 2022 Aug 16]. Available from: https://cihr-irsc.gc.ca/e/50837.html
8. Lowik A, Cameron JJ, Dame J, Ford J, Pulice-Farrow L, Salway T, et al. Gender & Sex in Methods & Measurement – Research Equity Toolkit. “Tool #3: Sampling Plans & Data Analyses.” 2022.
9. Johnson JL, Greaves L, Repta R. Better science with sex and gender: Facilitating the use of a sex and gender-based analysis in health research. Int J Equity Health. 2009;8.
10. Schiebinger L, Klinge I, Sánchez de Madariaga I, Paik HY, Schraudner M, Stefanick M. Gendered Innovations in Science, Health & Medicine, Engineering and Environment [Internet]. [cited 2022 Aug 16]. Available from: http://genderedinnovations.stanford.edu/index.html
11. Lowik A, Cameron JJ, Dame J, Ford J, PuliceFarrow L, Salway T, et al. Gender and Sex in Methods & Measurement – Research Equity Toolkit. “Tool #1: Determining & Communicating Eligibility,.” 2022.
12. What is gender? What is sex? – CIHR [Internet]. [cited 2022 Aug 16]. Available from: https://cihr-irsc.gc.ca/e/48642.html#d2
13. Arnold AP, Cassis LA, Eghbali M, Reue K, Sandberg K. Sex hormones and sex chromosomes cause sex differences in the development of cardiovascular diseases. Arterioscler Thromb Vasc Biol [Internet]. 2017 May 1 [cited 2022 Mar 15];37(5):746–56. Available from: http://ahajournals.org
14. Hart CG, Saperstein A, Magliozzi D, Westbrook L. Gender and Health: Beyond Binary Categorical Measurement. J Health Soc Behav. 2019;60(1):101–18.
15. Nielsen MW, Stefanick ML, Peragine D, Neilands TB, Ioannidis JPA, Pilote L, et al. Gender-related variables for health research. Biol Sex Differ. 2021 Dec 1;12(1).
16. Rioux C, Paré A, London-Nadeau K, Juster R-P, Weedon S, Levasseur-Puhach S, et al. Sex and gender terminology: a glossary for gender-inclusive epidemiology. J Epidemiol Community Health. 2022;jech-2022-219171.
17. Ratnu VS, Emami MR, Bredy TW. Genetic and epigenetic factors underlying sex differences in the regulation of gene expression in the brain. J Neurosci Res [Internet]. 2017 Jan 1 [cited 2022 Sep 29];95(1–2):301. Available from: /pmc/articles/PMC5120607/
18. Mauvais-Jarvis F, Bairey Merz N, Barnes PJ, Brinton RD, Carrero JJ, DeMeo DL, et al. Sex and gender: modifiers of health, disease, and medicine. Lancet (London, England) [Internet]. 2020 Aug 22 [cited 2022 Sep 29];396(10250):565. Available from: /pmc/articles/PMC7440877/
19. Davies S, Papp VG, Antoni C. Voice and Communication Change for Gender Nonconforming Individuals: Giving Voice to the Person Inside. Int J Transgenderism. 2015;16(3):117–59.
20. Romano TM. The Singing Voice During the First Two Years of Testosterone Therapy: Working with the Trans or Gender Queer Voice [Internet]. 2018. Available from: https://login.ezproxy.library.ualberta.ca/login?url=https://search.ebscohost.com/login.aspx?direct=true&db=ddu&AN=05ADB19E39F35F40&site=eds-live&scope=site
21. Johnson JL, Greaves L, Repta R. Better Science with Sex And Gender: A Primer for Health Research [Internet]. 2007 [cited 2022 Jul 26]. Available from: http://www.whrn.ca
22. Lowik A, Cameron JJ, Dame J, Ford J, Pulice-Farrow L, Salway T, et al. Gender & Sex in Methods & Measurement – Research Equity Toolkit. “Tool #2: Effective Recruitment Strategies.” 2022.
23. Thorne N, Yip AKT, Bouman WP, Marshall E, Arcelus J. The terminology of identities between, outside and beyond the gender binary – A systematic review. Int J Transgenderism [Internet]. 2019 [cited 2022 Jul 14];20(2–3):138. Available from: /pmc/articles/PMC6830980/
24. Bauer GR, Braimoh J, Scheim AI, Dharma C. Transgender-inclusive measures of sex/gender for population surveys: Mixedmethods evaluation and recommendations. PLoS One [Internet]. 2017 [cited 2022 Aug 16];12(5):1–28. Available from: https://doaj-org.login.ezproxy.library.ualberta.ca/article/3109fa7b016849c09c417889385cce85