Abstract
Objectives
Government-conducted population health surveys are important sources of data on health inequities for gay, bisexual, transgender, and queer men and nonbinary and Two-Spirit people (2S/GBTQ+). There is limited understanding of how vulnerable these surveys are to misclassification bias resulting from participants’ reluctance to disclose their sexual orientation and gender identity. 2S/GBTQ+ people may be more willing to participate in community-based surveys, where they might feel safer disclosing their minority sexual orientation or gender identity than they would on a government survey. We sought to understand whether the proportion of 2S/GBTQ+ people who would disclose their sexual orientation on a government survey changed between 2012 and 2019 survey cycles, as well as the proportion of trans, nonbinary, and Two-Spirit participants who would reveal their gender identity, and the demographic factors associated with both.
Methods
We analysed data from the 2012 and 2019 cycles of Sex Now, a repeated cross-sectional Canada-wide online survey on the health and well-being of 2S/GBTQ+ people conducted by the Community-Based Research Centre. We computed frequencies and prevalence ratios of the likelihood of disclosing sexual orientation and gender identity on a Statistics Canada survey by a variety of demographic variables.
Results
We found that in 2019, 86.0% (95% CI [85.4, 86.7]) of all participants would reveal their sexual orientation, a significant increase from 2012 (69.5%, 95% CI [68.5, 70.4], Δ = 16.6%, 95% CI [15.4, 17.8]). However, participants who identified as bisexual, straight, or heteroflexible; who were in a relationship with a woman; or who were not “out” were less willing to reveal their sexual orientation. We found that 85% of trans men, nonbinary, and Two-Spirit participants would reveal their gender identity, which was more likely among those living with HIV or aged 19–29 years old.
Conclusion
These findings suggest that government datasets may significantly misclassify and underestimate the population size of 2S/GBTQ+ individuals. Persistent mistrust of government institutions within this community may exacerbate underreporting and non-disclosure, underscoring the need for research into methodologies that can enhance trust and improve the accuracy of population estimates. Researchers using existing government datasets should consider using statistical methods to account for potential misclassification error.
Supplementary Information
The online version contains supplementary material available at 10.17269/s41997-025-01127-0.
Keywords: Sexual and gender minorities, Sexuality, Bias, Surveys and questionnaires, Canada
Résumé
Objectifs
Les enquêtes gouvernementales sur la santé des populations sont d’importantes sources de données sur les iniquités en santé chez les hommes gais, bisexuels, transgenre et queer et les personnes non binaires et bispirituelles (2S/GBTQ+). Or, on en sait peu sur la mesure dans laquelle ces enquêtes sont vulnérables au biais d’erreur de classification résultant de la réticence des personnes participantes à dévoiler leur orientation sexuelle et leur identité de genre. Les personnes 2S/GBTQ+ peuvent être plus disposées à participer aux enquêtes locales, où elles ont peut-être moins peur de dévoiler une orientation sexuelle ou une identité de genre minoritaire que dans le cadre d’une enquête gouvernementale. Nous avons cherché à savoir si la proportion des personnes 2S/GBTQ+ qui dévoileraient leur orientation sexuelle pour une enquête gouvernementale a changé entre les cycles d’enquête de 2012 et de 2019, ainsi que la proportion de personnes participantes trans, non binaires et bispirituelles qui révéleraient leur identité de genre, et les facteurs démographiques associés aux deux populations.
Méthodes
Nous avons analysé les données des cycles de 2012 et de 2019 de Sexe au présent, un sondage en ligne transversal pancanadien sur la santé et le bien-être des personnes 2S/GBTQ+ mené par le Centre de recherche communautaire. Nous avons calculé en fonction d’un éventail de variables démographiques la fréquence et les ratios de prévalence de la probabilité de dévoiler son orientation sexuelle et son identité de genre dans le cadre d’une enquête de Statistique Canada.
Résultats
Nous avons constaté qu’en 2019, 86%, IC de 95% [85,4, 86,7], des personnes participantes auraient révélé leur orientation sexuelle, ce qui représente une hausse importante par rapport à 2012 (69,5%, IC de 95% [68,5, 70,4], Δ=16.6%, IC de 95% [15,4, 17,8]). Cependant, les hommes s’identifiant comme étant bisexuels, hétérosexuels ou hétéroflexibles; qui étaient en couple avec une femme; ou qui n’étaient pas « sortis du placard» étaient moins disposés à révéler leur orientation sexuelle. Selon nos constatations, 85% des hommes trans et des personnes non binaires et bispirituelles auraient révélé leur identité de genre, un résultat qui était plus probable chez les personnes vivant avec le VIH ou âgées de 19 à 29 ans.
Conclusion
Ces constats indiquent que les fichiers du gouvernement pourraient significativement mal classifier et sous-estimer la taille de la population 2S/GBTQ+. La méfiance persistante envers les institutions gouvernementales dans cette communauté pourrait exacerber la sous-déclaration et la non-divulgation, d’où la nécessité de faire de la recherche sur les méthodes susceptibles de renforcer la confiance et d’améliorer l’exactitude des estimations de cette population. Les chercheurs et les chercheuses qui utilisent les fichiers gouvernementaux existants devraient songer à utiliser des méthodes statistiques pour tenir compte des erreurs de classification possibles.
Mots-clés: Minorités sexuelles et de genre, Sexualité, Biais, Enquêtes et questionnaires, Canada
Introduction
Government-conducted population surveys are essential for understanding health inequities among marginalized groups, including gay, bisexual, transgender, and queer men and nonbinary and Two-Spirit people (2S/GBTQ+). However, these surveys are vulnerable to several forms of information bias, including measurement bias associated with self-reported sexual orientation and gender identity (Wolff et al., 2017). Selection bias associated with different sampling strategies also impacts study findings. Probability samples of sexual minorities tend to include more married/partnered individuals and fewer individuals who are employed, lesbian/gay-identified, and report higher substance use, suicidal ideation, and sexual partners (Salway et al., 2019). United States (US) government survey samples of trans people have included more individuals who are older, married, heterosexual, and Black or Hispanic compared to non-probability samples (Bauerband et al., 2023; Lett & Everhart, 2021). While sexual orientation and gender identity questions have historically been stigmatized, nonresponse rates for sexual orientation questions have improved over time (Bates et al., 2019).
While there is little research on how trans, nonbinary, and Two-Spirit people respond to government surveys, transphobia and cisnormativity likely shape individuals’ willingness to disclose their gender (Suen et al., 2022). Many population health surveys use the two-step gender measure, which Statistics Canada implemented for the first time in the 2021 Census (Statistics Canada, 2022). This method compares participants’ gender identity and sex assigned at birth to identify trans respondents (Lombardi & Banik, 2016). The Trans PULSE Canada survey asked trans and nonbinary participants to answer sex and gender questions as if completing a Statistics Canada survey and found that 82% of participants would be correctly classified as trans or nonbinary based on their responses (The Trans PULSE Canada Team, 2022). Other studies suggest that while most trans people are willing to disclose their trans identity in government health research, they have concerns regarding question wording and intent (Bauer et al., 2017; Holzberg et al., 2017).
Understanding the prevalence of non-disclosure and misclassification of sexual and gender minority identities on government surveys is important, as these biases can affect study validity and findings (Salway et al., 2015, 2019). For instance, Salway et al. (2015) argued that misclassification errors contribute to underestimating the burden of HIV mortality among men who have sex with men. As government surveys are key authoritative sources of population health data, understanding how sexual and gender minority individuals respond to these surveys is critical to accurately interpreting and applying their findings. If particular demographic groups of 2S/GBTQ+ people are less likely to disclose their sexual and gender identities in government health surveys, these groups’ health needs will not be accurately represented in the data, and 2S/GBTQ+ health programs and policies may not meet their needs. Using data from the 2012 edition of Sex Now, a Canada-wide bilingual periodic cross-sectional survey of GBTQ+ men, nonbinary, and Two-Spirit people, Ferlatte and colleagues (2017) found that 86% of gay participants and 40% of bisexual participants were willing to disclose their sexual orientation on a Statistics Canada survey. However, public acceptance of sexual minorities in Canada has increased considerably since 2012 (Flores, 2021), which may increase 2S/GBTQ+ people’s willingness to disclose their sexual orientation in a government survey. To investigate this possibility, we analysed data from the 2019 edition of Sex Now and compared it with data from the 2012 edition. We also considered additional predictors and added an analysis of disclosure of gender identity. Our research questions were (1) What proportion of participants indicated they would disclose their sexual orientation and gender identity on a Statistics Canada survey, and does this vary by demographic variables?; and (2) Is this proportion significantly different from Sex Now 2012?
Methods
Data
Sex Now is a Canada-wide bilingual periodic cross-sectional survey conducted by the Community-Based Research Centre (CBRC) to promote the health and well-being of 2S/GBTQ+ people. We analysed data from Sex Now 2012 and 2019. In both years, data were collected anonymously using an online survey promoted on dating/sex-seeking websites and apps, social media, word-of-mouth, and community partner organizations (Salway et al., 2016, 2021). In 2012, men who have sex with men aged 18 and over living in Canada were eligible to participate (Ferlatte et al., 2017). In 2019, participants had to (i) identify as a man (including trans man), nonbinary, or Two-Spirit; (ii) identify as gay, bisexual, queer, or another non-heterosexual identity or report having had sex with a man in the last five years; and (iii) be 15 or older and living in Canada (Salway et al., 2021). Trans women were ineligible. Participants provided consent to participate before completing the survey, which was available in English and French.
Dependent variables
Our primary dependent variable was derived from a question that asked participants how likely they would be to reveal their sexual orientation in a Statistics Canada survey. In 2012, this question was phrased as: “Would you reveal your sexual orientation if asked in a Statistics Canada survey?” In 2019, this question was phrased as “How likely or unlikely would you be to reveal each of the following, if asked in a Statistics Canada survey (e.g., Census, Canadian Community Health Survey)?” and had separate questions for “sexual orientation” and “gender identity (i.e., cisgender, transgender, non-binary)”. The response options were “Very likely,” “Likely,” “Unlikely,” and “Very unlikely” in 2019, and similar in 2012 (“Somewhat unlikely” and “Totally unlikely” were used instead). We recoded both variables into a binary “Likely” and “Unlikely” variable for analyses.
Independent variables
Our independent variables of interest were survey year, province of residence, age (at the time of the survey, at the time of coming out, and at first sex with a man), gender identity, trans status, sexual identity, relationship status, rurality, income, education, HIV status, and ethnicity. As question wording and response options in the 2012 and 2019 surveys were not identical, we recoded some variables to create compatible categories (see Supplemental Table 1). We further collapsed some categories for the gender identity disclosure analysis due to the smaller sample size.
Analyses
We merged the 2012 and 2019 datasets into a single dataset and conducted all analyses using SAS 9.4. We computed frequencies of disclosing sexual orientation if asked in a Statistics Canada survey by each of our independent variables and conducted a Pearson’s chi-squared test (α = 0.05) to assess the change between 2012 and 2019 of the percent of participants who said they would disclose their sexual orientation. Next, only for trans, nonbinary, and Two-Spirit participants, we computed frequencies of disclosing gender identity if asked in a Statistics Canada survey by each of our independent variables. We calculated prevalence ratios and their 95% confidence intervals for each of our independent variables using PROC GENMOD’s log-binomial regression (Spiegelman & Hertzmark, 2005). Specifically, we conducted binary contrasts, where we compared each category to the weighted mean of all other categories (Johfre & Freese, 2021).
Results
Prevalence of revealing sexual orientation in 2012
In 2012, 49.8%, 95% CI [48.7, 50.9] (n = 4173), of participants said they were very likely to reveal their sexual orientation if asked in a Statistics Canada survey; 19.7% [18.8, 20.5] (n = 1649) said they were likely to do so, 11.3% [10.7, 12.0] (n = 950) were somewhat unlikely to disclose, and 19.1% [18.4, 20.1] (n = 1610) were totally unlikely to disclose. Likelihood of disclosure differed significantly by sexual orientation, with 85.9% [85.0, 86.4] of gay participants indicating they would disclose, compared with only 38.8% [36.9, 40.6] of bisexual participants and 34.3% [27.3, 41.3] of heterosexual participants. Participants who were currently in a relationship with a woman were much less likely to indicate they would disclose their sexual orientation than participants who were single, in a relationship with a man, or had another relationship status (see Table 1). Further, only 37.0% [35.3, 38.7] of participants who said they were not out about their sexuality to other guys said they would disclose. We also observed differences in willingness to disclose by the age at which participants came out and the age at which they first had sex with a man, with participants who came out or had sex younger being more likely to disclose. Prevalence of disclosing sexual orientation also varied by participants’ province of residence, gender identity, trans status, age, rurality, income, education, HIV status, and ethnicity. The largest differences were based on HIV status and age, with younger participants and participants living with HIV being more likely to disclose.
Table 1.
Self-reported likelihood of disclosing sexual orientation in Statistics Canada survey by demographic variables
| Sex Now 2012 | Sex Now 2019 | Change from 2012 to 2019 | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Variable | Total n | Likelya % (95% CI) | PR of likely (95% CI) | Variable | Total n | Likelya % (95% CI) | PR of likely (95% CI) | Δ likely (95% CI) | χ2 | p |
| Overall | 8382 | 69.5 (68.5, 70.4) | Overall | 10,415 | 86.0 (85.4, 86.7) | 16.6 (15.4, 17.8) | 761.31 | <.0001 | ||
| Province/Territories | Province/Territories | |||||||||
| Alberta | 1064 | 66.8 (64.0, 69.7) | 0.96 (0.90, 1.02) | Alberta | 1573 | 83.5 (81.6, 85.3) | 0.99 (0.96, 1.02) | 16.7 (13.3, 20.0) | 98.59 | <.0001 |
| British Columbia | 1804 | 75.0 (73.0, 77.0) | 1.09 (1.03, 1.14) | British Columbia | 1833 | 87.5 (86.0, 89.0) | 1.04 (1.01, 1.07) | 12.5 (10.0, 15.0) | 93.56 | <.0001 |
| Manitoba | 342 | 62.3 (57.1, 67.4) | 0.89 (0.81, 0.97) | Manitoba | 319 | 85.6 (81.7, 89.4) | 1.01 (0.96, 1.07) | 23.3 (16.9, 29.7) | 46.03 | <.0001 |
| New Brunswick | 105 | 72.4 (63.8, 80.9) | 1.04 (0.92, 1.19) | New Brunswick | 188 | 81.4 (75.8, 86.9) | 0.96 (0.89, 1.03) | 9.0 (−1.2, 19.2) | 3.20 | 0.0737 |
| Newfoundland & Labrador | 85 | 71.8 (62.2, 81.3) | 1.04 (0.90, 1.19) | Newfoundland & Labrador | 175 | 88 (83.2, 92.8) | 1.05 (0.99, 1.11) | 16.2 (5.5, 27.0) | 10.54 | 0.0012 |
| Nova Scotia | 223 | 76.7 (71.1, 82.2) | 1.11 (1.02, 1.21) | Nova Scotia | 463 | 87.3 (84.2, 90.3) | 1.04 (0.99, 1.08) | 10.6 (4.3, 16.9) | 12.41 | 0.0004 |
| Ontario | 3366 | 66.5 (64.9, 68.0) | 0.95 (0.91, 1.00) | Ontario | 3162 | 83.7 (82.5, 85.0) | 0.99 (0.96, 1.02) | 17.3 (15.2, 19.3) | 257.88 | <.0001 |
| Prince Edward Island | 33 | 72.7 (57.5, 87.9) | 1.05 (0.85, 1.30) | Prince Edward Island | 50 | 82 (71.4, 92.6) | 0.97 (0.85, 1.10) | 9.3 (−9.3, 27.8) | 1.01 | 0.3158 |
| Quebec | 1048 | 74.3 (71.7, 77.0) | 1.08 (1.02, 1.14) | Quebec | 2330 | 90.6 (89.4, 91.8) | 1.08 (1.05, 1.11) | 16.3 (13.4, 19.2) | 154.80 | <.0001 |
| Saskatchewan | 289 | 62.6 (57.1, 68.2) | 0.89 (0.81, 0.99) | Saskatchewan | 274 | 81.4 (76.8, 86.0) | 0.96 (0.90, 1.02) | 18.8 (11.5, 26.0) | 24.42 | <.0001 |
| Territories | 23 | 65.2 (45.8, 84.7) | 0.93 (0.69, 1.26) | Territories | 48 | 79.2 (67.7, 90.7) | 0.93 (0.80, 1.08) | 14.0 (−36.6, 8.7) | 1.60 | 0.2061 |
| Gender | Gender | |||||||||
| Man | 8299 | 69.4 (68.4, 70.4) | 0.96 (0.84, 1.10) | Man | 9992 | 85.7 (85.0, 86.4) | 0.92 (0.89, 0.94) | 16.3 (15.1, 17.5) | 709.58 | <.0001 |
| Other | 83 | 72.3 (62.7, 81.9) | 1.04 (0.91, 1.19) | Other | 419 | 93.5 (91.2, 95.9) | 1.09 (1.07, 1.12) | 21.3 (11.4, 31.2) | 34.94 | <.0001 |
| Trans status | Trans status | |||||||||
| Trans | 98 | 77.7 (69.3, 85.8) | 1.12 (1.00, 1.25) | Trans | 914 | 89.4 (87.4, 91.4) | 1.04 (1.02, 1.07) | 11.8 (3.3, 20.3) | 11.95 | 0.0002 |
| Not trans | 8284 | 69.4 (68.4, 70.4) | 0.89 (0.80, 0.996) | Not trans | 9350 | 85.8 (85.1, 86.5) | 0.96 (0.94, 0.98) | 16.4 (15.2, 17.6) | 689.96 | <.0001 |
| Two-Spirit (among Indigenous participants) | ||||||||||
| Indigenous Two-Spirit | 199 | 94 (90.7, 97.3) | 1.16 (1.09, 1.23) | |||||||
| Indigenous non-Two-Spirit | 320 | 81.3 (77.0, 85.5) | 0.86 (0.81, 0.92) | |||||||
| Sexuality | Sexuality | |||||||||
| Gay (homosexual) | 5406 | 85.9 (85.0, 86.4) | 1.81 (1.67, 1.96) | Gay (only) | 6139 | 93 (92.3, 93.6) | 1.29 (1.23, 1.34) | 7.1 (5.9, 8.2) | 154.32 | <.0001 |
| Straight (heterosexual) | 175 | 34.3 (27.3, 41.3) | 0.53 (0.43, 0.66) | Straight | 144 | 62.5 (54.6, 70.4) | 0.76 (0.67, 0.86) | 28.2 (17.6, 38.8) | 25.24 | <.0001 |
| Bi (bisexual) | 2720 | 38.8 (36.9, 40.6) | 0.63 (0.57, 0.69) | Bisexual | 2402 | 66.3 (64.4, 68.2) | 0.82 (0.78, 0.86) | 27.6 (24.9, 30.2) | 388.26 | <.0001 |
| Other | 81 | 80.3 (71.6, 88.9) | 1.65 (1.45, 1.88) | Other | 1696 | 91.3 (89.9, 92.6) | 1.25 (1.20, 1.31) | 11.0 (2.3, 19.8) | 11.22 | <.0001 |
| Asexual | 178 | 84.3 (78.9, 89.6) | 0.98 (0.92, 1.04) | |||||||
| Pansexual | 786 | 86 (83.6, 88.4) | 1.00 (0.97, 1.03) | |||||||
| Queer | 1530 | 94.2 (93.1, 95.4) | 1.11 (1.10, 1.13) | |||||||
| Heteroflexible | 259 | 62.9 (57.1, 68.8) | 0.72 (0.66, 0.80) | |||||||
| Relationship status | Relationship status | |||||||||
| Single | 4222 | 72.5 (71.2, 73.9) | 1.15 (1.10, 1.19) | Single | 4590 | 86.5 (85.5, 87.5) | 1.13 (1.10, 1.16) | 13.9 (12.3, 15.6) | 265.37 | <.0001 |
| Partnered with a man | 2215 | 91.3 (90.2, 92.5) | 1.56 (1.50, 1.62) | Partnered with a man | 4212 | 93.9 (93.2, 94.6) | 1.26 (1.23, 1.29) | 2.5 (1.2, 3.9) | 14.41 | 0.0001 |
| Partnered with a woman | 1798 | 34.4 (32.2, 36.6) | 0.42 (0.40, 0.45) | Partnered with a woman | 1116 | 53 (50.1, 56.0) | 0.56 (0.56, 0.62) | 18.7 (15.0, 22.3) | 98.9038 | <.0001 |
| Other | 147 | 80.3 (73.8, 86.7) | 1.31 (1.21, 1.43) | Other (in relationship with nonbinary person or 1+ person) | 497 | 89.7 (87.1, 92.4) | 1.19 (1.15, 1.23) | 9.5 (2.5, 16.4) | 9.3444 | 0.0022 |
| Age group (in years) | Age group (in years) | |||||||||
| 18 or younger | 132 | 79.5 (72.7, 86.4) | 1.15 (1.05, 1.26) | 18 or younger | 361 | 82.5 (78.6, 86.5) | 0.98 (0.93, 1.03) | 3.0 (−4.91, 10.9) | 0.5841 | 0.4447 |
| 19–29 | 1734 | 78.1 (76.2, 80.1) | 1.12 (1.08, 1.17) | 19–29 | 3274 | 89.6 (88.6, 90.7) | 1.09 (1.06, 1.11) | 11.5 (9.3, 13.7) | 121.34 | <.0001 |
| 30–59 | 5466 | 67.8 (66.6, 69.0) | 0.93 (0.89, 0.97) | 30–59 | 5890 | 85.3 (84.4, 86.3) | 1.02 (0.99, 1.04) | 17.6 (16.0, 19.1) | 491.43 | <.0001 |
| 60 or older | 1050 | 62.5 (59.5, 65.4) | 0.83 (0.79, 0.88) | 60 or older | 890 | 79 (81.7, 23.7) | 0.92 (0.88, 0.96) | 16.5 (12.6, 20.5) | 62.61 | <.0001 |
| Rurality | Rurality | |||||||||
| Rural | 1245 | 60.8 (58.1, 63.5) | 0.86 (0.82, 0.90) | Rural area (< 1000 people) | 424 | 79 (75.1, 82.9) | 0.90 (0.85, 0.94) | 17.1 (15.9, 18.4) | 46.35 | <.0001 |
| Non-rural | 7137 | 71.0 (69.9, 72.0) | 1.17 (1.11, 1.22) | Non-rural (1000+ people) | 8491 | 88.1 (87.4, 88.8) | 1.11 (1.06, 1.17) | 18.2 (13.5, 22.9) | 719.29 | <.0001 |
| Income (annual, CAD) | Income (annual, CAD) | |||||||||
| Under $10,000 | 739 | 74.3 (71.1, 77.4) | 1.08 (1.03, 1.13) | < $10,000 | 786 | 89.2 (87.0, 91.4) | 1.01 (0.98, 1.03) | 14.9 (11.1, 18.7) | 57.18 | <.0001 |
| $10,000—$19,999 | 779 | 73.2 (70.1, 76.3) | 1.06 (1.02, 1.11) | $10,000–$19,999 | 1084 | 90.7 (89.0, 92.4) | 1.02 (1.00, 1.05) | 17.5 (14.0, 21.1) | 100.21 | <.0001 |
| $20,000—$29,999 | 874 | 75.3 (72.4, 78.1) | 1.10 (1.05, 1.14) | $20,000–$29,999 | 926 | 90 (88.0, 91.9) | 1.01 (0.99, 1.04) | 14.7 (11.2, 18.1) | 68.06 | <.0001 |
| $30,000—$49,999 | 1847 | 68.3 (66.2, 70.4) | 0.98 (0.95, 1.02) | $30,000–$49,999 | 1743 | 90.1 (88.7, 91.5) | 1.02 (1.00, 1.04) | 21.9 (19.3, 24.4) | 257.25 | <.0001 |
| $50,000—$59,999 | 968 | 71.5 (68.6, 74.3) | 1.04 (0.99, 1.08) | $50,000–$59,999 | 756 | 88.2 (85.9, 90.5) | 0.99 (0.97, 1.02) | 16.7 (13.1, 20.4) | 71.27 | <.0001 |
| $60,000—$69,999 | 730 | 68.5 (65.1, 71.9) | 0.99 (0.94, 1.04) | $60,000–$69,999 | 613 | 86.9 (84.3, 89.6) | 0.98 (0.95, 1.01) | 18.5 (14.2, 22.8) | 63.93 | <.0001 |
| $70,000—$79,999 | 608 | 66.9 (63.2, 70.7) | 0.96 (0.91, 1.02) | $70,000–$79,999 | 527 | 88.8 (86.1, 91.5) | 1.00 (0.97, 1.03) | 21.9 (17.3, 26.5) | 76.42 | <.0001 |
| $80,000—$89,999 | 484 | 65.7 (61.5, 69.9) | 0.94 (0.88, 1.01) | $80,000–$89,999 | 431 | 91.2 (88.5, 93.9) | 1.03 (1.00, 1.06) | 25.5 (20.5, 30.5) | 85.44 | <.0001 |
| $90,000—$99,999 | 345 | 67.0 (62.0, 71.9) | 0.96 (0.89, 1.04) | $90,000–$99,999 | 306 | 87.3 (83.5, 91.0) | 0.98 (0.94, 1.02) | 20.3 (14.1, 26.5) | 37.16 | <.0001 |
| $100,000+ | 1007 | 63.2 (60.2, 66.1) | 0.90 (0.86, 0.95) | $100,00 | 976 | 86 (83.8, 88.1) | 0.96 (0.94, 0.99) | 22.8 (19.1, 26.5) | 135.27 | <.0001 |
| Education | Education | |||||||||
| Some high school | 348 | 59.8 (54.6, 64.9) | 0.90 (0.82, 0.99) | Did not finish high school | 361 | 82 (78.0, 86.0) | 0.94 (0.90, 0.99) | 22.2 (15.7, 28.7) | 42.58 | <.0001 |
| High school | 1131 | 61.7 (58.9, 64.5) | 0.95 (0.89, 1.01) | High school, or equivalent | 1733 | 85.5 (83.9, 87.2) | 1.00 (0.97, 1.04) | 23.8 (20.5, 27.1) | 213.26 | <.0001 |
| Any post-secondary | 6903 | 71.2 (70.1, 72.3) | 1.17 (1.11, 1.23) | Any post-secondary | 6774 | 88.5 (87.8, 89.3) | 1.06 (1.03, 1.09) | 20.2 (16.0, 18.6) | 634.37 | <.0001 |
| HIV status (self-reported) | HIV status (self-reported) | |||||||||
| HIV negative or unknown | 7715 | 67.8 (66.7, 68.8) | 0.76 (0.74, 0.79) | Never been diagnosed with HIV | 6673 | 88.3 (87.6, 89.1) | 0.94 (0.92, 0.96) | 20.6 (19.3, 21.9) | 862.16 | <.0001 |
| HIV positive | 667 | 89.1 (86.7, 91.4) | 1.31 (1.27, 1.35) | Living with HIV | 609 | 93.6 (91.7, 95.5) | 1.06 (1.04, 1.08) | 4.5 (1.5, 7.6) | 8.20 | 0.0042 |
| Ethnicity | Ethnicity | |||||||||
| Indigenous | 168 | 73.2 (66.5, 79.9) | 1.05 (0.95, 1.16) | Indigenous | 558 | 86.4 (83.5, 89.2) | 1.00 (0.96, 1.04) | 13.2 (5.9, 20.4) | 16.11 | <.0001 |
| ACB | 76 | 67.1 (56.6, 77.7) | 0.94 (0.80, 1.10) | ACB | 361 | 86.4 (82.9, 90.0) | 1.00 (0.96, 1.04) | 19.3 (8.2, 30.5) | 16.66 | <.0001 |
| Asian | 286 | 69.2 (63.9, 74.6) | 0.98 (0.89, 1.07) | Asian | 739 | 85.1 (82.5, 87.7) | 0.98 (0.95, 1.01) | 15.9 (10.0, 21.8) | 33.38 | <.0001 |
| Other POC | 542 | 74.0 (70.3, 77.7) | 1.06 (0.99, 1.14) | Other POC | 600 | 88.3 (85.8, 90.9) | 1.03 (0.99, 1.06) | 14.4 (9.9, 18.9) | 38.92 | <.0001 |
| White only | 7310 | 69.1 (68.0, 70.1) | 0.98 (0.92, 1.03) | White only | 8017 | 86.1 (85.3, 86.8) | 0.99 (0.98, 1.01) | 17.0 (15.7, 18.3) | 642.92 | <.0001 |
| Age when first had sex with man | Age when first had sex with man | |||||||||
| < 16 | 3091 | 72.7 (71.1, 74.2) | 1.08 (1.05, 1.11) | < 16 | 2454 | 86.9 (85.6, 88.3) | 1.02 (1.00, 1.05) | 14.3 (12.2, 16.3) | 167.43 | <.0001 |
| 16–18 | 1896 | 75.4 (73.4, 77.3) | 1.13 (1.10, 1.17) | 16–18 | 3028 | 90.2 (89.2, 91.3) | 1.07 (1.05, 1.09) | 14.9 (12.7, 17.0) | 196.50 | <.0001 |
| 19–22 | 1462 | 73.1 (70.8, 75.4) | 1.09 (1.05, 1.13) | 19–22 | 2318 | 89.3 (88.0, 90.5) | 1.06 (1.04, 1.08) | 16.1 (13.5, 18.7) | 165.62 | <.0001 |
| 23+ | 1897 | 55.6 (53,4, 57.9) | 0.75 (0.72, 0.79) | 23+ | 1717 | 78.9 (77.0, 80.8) | 0.91 (0.88, 0.93) | 23.3 (20.4, 26.3) | 220.29 | <.0001 |
| Never had sex with a man | 366 | 81.4 (77.4, 85.4) | 0.94, 0.90, 0.99) | |||||||
| Age when came out | Age when came out | |||||||||
| < 16 | 489 | 93.9 (91.7, 96.0) | 1.32 (1.28, 1.35) | < 16 | 1512 | 93.5 (92.2, 94.7) | 1.17 (1.14, 1.12) | −0.4 (−2.9, 2.1) | 0.10 | 0.7467 |
| 16–18 | 1281 | 90 (88.4, 91.7) | 1.25 (1.22, 1.28) | 16–18 | 2327 | 94.6 (93.7, 95.5) | 1.19 (1.17, 1.21) | 4.6 (2.7, 6.5) | 27.05 | <.0001 |
| 19–22 | 1342 | 90.6 (89.1, 92.2) | 1.26 (1.23, 1.29) | 19–22 | 2289 | 93.2 (92.2, 94.2) | 1.16 (1.14, 1.18) | 2.6 (0.7, 4.4) | 7.82 | 0.0052 |
| 23+ | 2179 | 84.9 (83.4, 86.4) | 1.16 (1.14, 1.19) | 23+ | 2480 | 91.9 (90.8, 93.0) | 1.14 (1.12, 1.16) | 7.0 (5.2, 8.8) | 56.27 | <.0001 |
| Not out | 3091 | 37 (35.3, 38.7) | 0.41 (0.39, 0.43) | Not out | 1585 | 50.7 (48.3, 53.2) | 0.54 (0.52, 0.57) | 13.7 (10.7, 16.7) | 81.09 | <.0001 |
| Degree of outness | ||||||||||
| 1 (Not at all open/out) | 998 | 34.6 (31.6, 37.5) | 0.42 (0.39, 0.46) | |||||||
| 2 | 748 | 61.6 (58.1, 65.1) | 0.86 (0.81, 0.92) | |||||||
| 3 | 1118 | 85 (82.9, 87.1) | 1.29 (1.24, 1.34) | |||||||
| 4 | 1653 | 92.1 (90.8, 93.4) | 1.43 (1.34, 1.47) | |||||||
| 5 (Open/out to all or most people) | 5881 | 96.4 (95.9, 96.8) | 1.51 (1.47, 1.55) | |||||||
a“Likely” includes participants who indicated they were “Very likely” or “Likely” to disclose their sexual orientation. HIV refers to human immunodeficiency virus. ACB means African, Caribbean, and/or Black. POC means person of colour. PR means prevalence ratio. CI means confidence interval
Bolded text emphasizes results that are statistically significant at p<0.05
Prevalence of revealing sexual orientation in 2019
In 2019, the majority (68.4%, 95% CI [67.5, 69.3], n = 7121) of participants said they were very likely to disclose their sexual orientation if asked in a Statistics Canada survey; 17.7% [16.9, 18.4] (n = 1841) said they were likely to do so, 7.1% [6.6, 7.6] (n = 744) said they were unlikely to do so, and 6.8% [6.3, 7.3] (n = 709) said they were very unlikely to do so. Overall, willingness to disclose increased significantly by 16.6% ([15.4, 17.8], p < 0.0001) from 2012. Bivariate analysis revealed that while the prevalence of disclosing sexual orientation increased across all demographic variables, some differences persisted, especially in terms of sexual orientation and relationship status. Only 62.5% [54.6, 70.4] of straight, 62.9% [57.1, 68.8] of heteroflexible participants, and 66.3% [64.4%, 68.2%] of bisexual participants said that they would disclose their sexual orientation, compared with 93.0% [92.3, 93.6] of participants who identified solely as gay. Similarly, just over half of participants in a relationship with a woman said they would disclose their sexual orientation—much fewer than participants who were single or in a relationship with a man, with a nonbinary person, or with more than one person. Further, only 34.6% [31.6, 37.5] of participants who were not at all open/out about their sexuality said they would disclose their sexual orientation, compared with 96.4% [95.9, 96.8] of those who were open/out to all or most people. Likelihood of disclosing sexual orientation also varied slightly by province of residence, gender identity, trans status, being Two-Spirit, age, rurality, income, educational attainment, age at first sex with a man, age of coming out, and HIV status. Overall, these findings indicate that bisexual, straight, and heteroflexible participants, participants in a relationship with a woman, and participants who were not out were still less likely to reveal their sexual orientation in 2019. However, a comparison of prevalence ratios from 2012 and 2019 indicates that the gap has decreased.
We conducted a post-hoc logistic regression to assess whether a change in participant demographics impacted our findings, and found that the increased willingness to disclose in 2019 persisted after controlling for demographic variables (see Supplementary Information).
Prevalence of revealing gender identity in 2019
In 2019, the majority (62.9%, 95% CI [60.1, 65.7], n = 723) of participants who identified as trans, nonbinary, and/or Two-Spirit said they were very likely to disclose their gender identity if asked in a Statistics Canada survey. Additionally, 22.1% [19.7, 24.5] (n = 254) said they were likely to reveal. Only 9.8% [8.1, 11.5] of participants (n = 113) said they were unlikely to reveal, and a further 5.2% [3.9, 6.5] (n = 60) said they were very unlikely to reveal. Prevalence ratios indicated that participants living with HIV and participants aged 19–29 were slightly more likely to reveal their gender identity (see Table 2). Within the group of trans, nonbinary, and/or Two-Spirit participants analysed here, there were no statistically significant differences by gender identity, being Two-Spirit, or trans status.
Table 2.
Likelihood of disclosing gender identity in Statistics Canada survey by demographic variables, for trans, nonbinary, and/or two-spirit participants in Sex Now 2019
| Total n | Likelya (%) | 95% CI of % | PR of likely (95% CI) | |
|---|---|---|---|---|
| Province/region | ||||
| Atlantic | 34 | 85.3 | (73.4, 97.2) | 1.01 (0.87, 1.18) |
| British Columbia | 234 | 85.0 | (80.5, 89.6) | 1.01 (0.93, 1.10) |
| Ontario | 275 | 85.1 | (80.9, 89.3) | 1.01 (0.93, 1.10) |
| Prairies | 340 | 84.4 | (80.5, 88.2) | 0.99 (0.93, 1.08) |
| Quebec | 152 | 88.2 | (83.0, 93.3) | 1.05 (0.67, 1.15) |
| Territories | 14 | 78.6 | (57.0, 100.0) | 0.92 (0.70, 1.21) |
| Age group (in years) | ||||
| 15–18 | 89 | 79.8 | (71.4, 88.1) | 0.96 (0.86, 1.07) |
| 19–29 | 580 | 88.1 | (85.5, 90.7) | 1.10 (1.03, 1.16) |
| 30–49 | 364 | 83.5 | (79.7, 87.3) | 1.02 (0.96, 1.09) |
| 50+ | 117 | 77.8 | (70.2, 85.3) | 0.93 (0.84, 1.03) |
| Trans | ||||
| No | 200 | 86.5 | (81.8, 91.2) | 1.02 (0.96, 1.09) |
| Yes | 922 | 84.6 | (82.3, 86.9) | 0.98 (0.92, 1.04) |
| Gender | ||||
| Man | 736 | 83.8 | (81.2, 86.5) | 0.96 (0.91, 1.11) |
| Nonbinary | 340 | 86.8 | (83.2, 90.4) | 1.01 (0.95, 1.08) |
| Other | 73 | 87.7 | (80.1, 95.2) | 1.03 (0.94, 1.13) |
| Two-Spirit (among Indigenous participants) | ||||
| Indigenous non-Two-Spirit | 72 | 79.2 | (69.7, 88.6) | 0.90 (0.79, 1.02) |
| Indigenous Two-Spirit | 200 | 88.0 | (83.5, 92.5) | 1.11 (0.98, 1.23) |
| Relationship status (by partner gender) | ||||
| No | 503 | 85.5 | (82.4, 88.6) | 1.01 (0.94, 1.07) |
| Yes, with a man | 325 | 84.3 | (80.3, 88.3) | 1.00 (0.94, 1.06) |
| Yes, with a woman | 144 | 80.6 | (74.1, 87.0) | 0.94 (0.86, 1.02) |
| Yes, with a nonbinary person or more than one person | 178 | 88.2 | (83.5, 92.9) | 1.06 (0.99, 1.13) |
| Rurality | ||||
| Non-rural | 931 | 86.3 | (84.0, 88.5) | 1.09 (0.95, 1.24) |
| Rural | 58 | 79.3 | (68.9, 89.8) | 0.92 (0.80, 1.05) |
| Income (personal, CAD) | ||||
| < $10,000 | 198 | 87.9 | (83.3, 92.4) | 1.03 (0.97, 1.10) |
| $10,000–$19,999 | 240 | 86.3 | (81.9, 90.6) | 1.01 (0.94, 1.07) |
| $20,000–$50,000 | 148 | 81.8 | (75.5, 88.0) | 0.94 (0.86, 1.02) |
| $50,000+ | 113 | 87.6 | (81.5, 93.7) | 1.03 (0.95, 1.11) |
| Education | ||||
| Did not finish high school | 78 | 85.9 | (78.2, 93.6) | 1.00 (0.91, 1.10) |
| High school or equivalent | 325 | 85.5 | (81.7, 89.4) | 1.00 (0.93, 1.06) |
| Any post-secondary education | 574 | 85.9 | (83.0, 88.7) | 1.00 (0.94, 1.06) |
| HIV status (self-reported) | ||||
| No (never been diagnosed with HIV) | 786 | 86.0 | (83.6, 88.4) | 0.92 (0.85, 0.9981) |
| Yes (living with HIV) | 46 | 93.5 | (86.3, 100) | 1.09 (1.00, 1.18) |
| Ethnicity | ||||
| Indigenous | 283 | 85.2 | (69.0, 96.5) | 1.01 (0.94, 1.09) |
| African, Caribbean, Black | 29 | 82.8 | (77.7, 95.9) | 0.98 (0.82, 1.16) |
| Asian (including East, South, and Southeast Asian) | 53 | 86.8 | (81.0, 89.3) | 1.04 (0.91, 1.17) |
| Other POC (Latin American and Arab/West Asian) | 49 | 81.6 | (70.8, 92.5) | 0.96 (0.83, 1.11) |
| White only | 708 | 85.3 | (82.7, 87.9) | 1.01 (0.95, 1.09) |
aLikely includes participants who indicated they were “Very likely” or “Likely” to disclose their gender identity. HIV means human immunodeficiency virus. POC means person of colour. PR means prevalence ratio. CI means confidence interval
Bolded text emphasizes results that are statistically significant at p<0.05
Discussion
We found that the proportion of 2S/GBTQ+ participants who said they would disclose their sexual orientation in a Statistics Canada survey increased significantly between 2012 and 2019, from 70 to 86%. However, 14% of participants were still unlikely to reveal their sexual orientation in 2019. Specifically, bisexual, straight, and heteroflexible participants, and participants who were in a relationship with a woman, and/or not out were much less likely to reveal their sexual orientation. Our findings suggest that estimates based on government survey data that do not account for non-disclosure underestimate the 2S/GBTQ+ population size (Sorge et al., 2023), which has important implications for prevalence estimates and population health responses.
Interpreting our finding that only 63% of straight participants would reveal their sexual orientation requires considering how participants conceptualized sexual orientation. While definitions vary, sexual orientation is a multi-dimensional construct that encompasses identity, attraction, and behaviours. These aspects of sexuality may change for individuals across their lifespan and are not always aligned in normative ways (van Anders, 2015). As the Sex Now surveys did not define sexual orientation, we cannot be certain how participants interpreted this question. However, straight-identified Sex Now participants were likely indicating a reluctance to disclose same-gender attractions or behaviours, rather than a heterosexual identity. Qualitative research could explore how straight-identified men who have sex with men interpret the term “sexual orientation.” Individuals with branched sexualities1 whose sexual identity, behaviour, and attractions do not align in normative ways—such as straight-identified men who have sex with men—may experience unique health risk factors and stressors (Lefevor et al., 2022; Mendelsohn et al., 2022). Other participants who were less likely to disclose their sexual orientation, such as individuals who are bisexual or not out, may also have unique health profiles. Reliable data that capture multiple facets of sexual orientation are needed to better understand the health profiles of our diverse population.
Our findings suggest that single-item sexual identity measures misclassify a significant proportion of straight-identified men who have sex with men as heterosexual. While Statistics Canada recognizes that sexual orientation is not synonymous with sexual attraction or behaviour, the agency’s standard adopted in 2023 only measures “how a person describes their sexuality,” such as “heterosexual, lesbian, gay, bisexual or pansexual” (Statistics Canada, 2023). However, the 2015-2016 edition of their Canadian Community Health Survey included questions about both sexual behaviours and identity. Analyzing this data, Silva and Fetner (2022) found that 20% of men who had ever engaged in sexual behaviours with other men identified as heterosexual. Similarly, Dharma and Bauer (2017) estimated that 16% of individuals who had ever engaged in same-sex sexual behaviour would be classified as heterosexual by Statistics Canada’s single-item sexual identity question. If we interpret our finding that 37% of straight-identified men who have sex with men would not disclose their sexual orientation to Statistics Canada to mean that they would not disclose their same-sex sexual behaviour, likely even fewer of these men would be categorized as sexual minorities than previously estimated.
Our findings underscore the need for additional research on sexual minority health in Canada using more expansive measures of sexual orientation. Researchers should carefully choose sexual orientation measures that align with their research questions. Asking about both sexual identity and behaviours may help better identify straight-identified men who have sex with men, though many may still not disclose their sexual behaviours. To promote trust and transparency, researchers should inform participants of how and why sexual orientation data will be used in analysis and highlight safety and confidentiality protections (Puckett et al., 2020; Suen et al., 2022). Meaningful and reciprocal partnerships with trusted community leaders and organizations may be necessary to ameliorate distrust based on historic harms. When analysing existing government data, our findings about the prevalence of non-disclosure and misclassification of sexual and gender minority individuals can inform statistical adjustments for these biases, such as those developed by Dharma et al. (2025).
In terms of gender, we found that 15% of trans men, nonbinary people, and Two-Spirit Sex Now survey participants were unlikely to reveal their gender identity in a Statistics Canada survey, suggesting that estimates of this population will likely be an undercount. Notably, the 2021 Census introduced a new measure of sex and gender, which asked individuals their sex at birth (male or female) and their gender (male, female, or other, where other includes a write-in text box) (Statistics Canada, 2022). As we only asked participants their general willingness to reveal their gender identity in a Statistics Canada survey, we cannot be sure how participants would have responded to the specific 2021 Census questions (see The Trans PULSE Canada Team, 2022 for this). Some individuals who were willing to disclose their gender identity may have been unwilling to report their sex at birth (Puckett et al., 2020) and might therefore be misclassified as cisgender. Future research should explore disclosure of gender modality, expression, and other sex/gender-related constructs. Further, anti-trans social attitudes and political attacks have risen significantly since 2019, and future research should explore how this shapes willingness to disclose gender identity in government surveys.
Some research indicates that trans participants prefer to be asked their gender identity before being asked their sex at birth (Lombardi & Banik, 2016). However, the limited available evidence has not found that the order of these questions impacts nonresponse bias (DeChants et al., 2021; Saperstein & Westbrook, 2021). The validity of any measure of gender identity will likely be impacted by participants’ understanding of the question’s purpose and planned usage (Holzberg et al., 2017; Suen et al., 2022). Researchers seeking to represent trans people accurately and respectfully in their surveys should refer to Lowik et al.’s guidelines Gender and Sex in Methods and Measurement Research Equity Toolkit (2022).
Another important concern is how Indigenous Two-Spirit people are represented in government health surveys. Per Pruden and Salway (2020), Two-Spirit is a term that serves as a “community organizing tool for Indigenous Peoples of Turtle Island who embody diverse sexualities, gender identities, roles, and/or expressions” that “challenges Western terms of gender and sexual orientation.” As such, it cannot be treated as simply a gender or sexuality (Pruden & Salway, 2020). Neither the sexual orientation nor gender questions currently used by Statistics Canada ask individuals whether they are Two-Spirit. Statistics Canada (2022) indicates that anyone who reported their gender as Two-Spirit in the “other” response option of the 2021 Census is included in the category “non-binary persons.” The sexual orientation standard adopted in 2023 classifies individuals who write in their sexual orientation as Two-Spirit as “sexual orientation, not elsewhere classified” and includes them in the broader LGB+ category. Based on these standards, Statistics Canada survey data does not allow researchers to conduct analyses specifically on Two-Spirit people. While concerns about the identifiability of small populations are important, collapsing Two-Spirit people into larger LGBTQ categories perpetuates the colonial erasure of Two-Spirit people and obscures their distinct experiences and needs (Lowik et al., 2022). Researchers designing surveys should consider Pruden and Salway’s (2020) recommendations to ask about being Two-Spirit after a question about race and ethnicity, and only ask the Two-Spirit question to Indigenous participants.
Limitations
Our research has several limitations. First, Sex Now has historically been promoted as about “sex between guys.” While non-heterosexual trans men and nonbinary people were eligible to participate regardless of recent sex with men, the survey eligibility and recruitment strategies limit representation from the full diversity of trans and nonbinary communities. Similarly, we caution extrapolation of Sex Now data to the entire 2S/LGBTQ+ community given the explicit exclusion of lesbian, bisexual, queer, and trans women. Second, we measured intent to disclose on a future survey rather than past behaviour. As such, we do not know how individuals actually responded to Statistics Canada surveys, which may be impacted by factors such as question wording (e.g., asking sex at birth and gender identity, as noted above). Third, our non-probability sample was largely recruited from social media or dating platforms and may underrepresent individuals who are not connected to online 2S/GBTQ+ communities, though less so than samples recruited in person. Previous research has found that non-probability samples tend to overrepresent some people, such as gay-identified individuals, and underrepresent others, such as married and partnered individuals (Salway et al., 2019). As gay identification was associated with a higher likelihood of disclosure and being in a relationship with a woman with a lower likelihood, the true proportion of 2S/GBTQ+ people who would disclose their sexual orientation on a Statistics Canada survey may be even lower than we have reported. Fourth, we did not run multivariable models to assess whether demographic variables were independently associated with willingness to disclose when controlling for other variables. Finally, our question on willingness to disclose specified collection on a Statistics Canada survey, which may limit generalizability to other federal government bodies, other levels of government, or other countries.
Conclusion
The proportion of 2S/GBTQ+ survey participants who would reveal their sexual orientation in a Statistics Canada survey increased significantly from 2012 to 2019, from 70 to 86%. However, fewer participants who identified as bisexual, straight, or heteroflexible, as well as participants who were in a relationship with a woman or who were not out about their sexuality would reveal their sexual orientation. Additionally, 85% of trans men, nonbinary, and Two-Spirit Sex Now survey participants said they would reveal their gender identity in a Statistics Canada survey, with few significant differences by demographic variables. These findings indicate that Statistics Canada datasets likely underestimate the 2S/GBTQ+ population by approximately 15%. Population health research must attend to information bias introduced by non-disclosure of stigmatized information. This requires both technical adjustments in analyses and proactive trust- and relationship-building.
Contributions to knowledge
What does this study add to existing knowledge?
The proportion of 2S/GBTQ+ survey participants who would disclose their sexual orientation in a federal government survey increased significantly from 2012 to 2019, from 70 to 86%. Participants who identified as bisexual, straight, or heteroflexible; were in relationship with a woman; or were not out about their sexuality were less willing to disclose their sexual orientation.
85% of trans men, nonbinary, and Two-Spirit participants would reveal their gender identity in a federal government survey.
What are the key implications for public health interventions, practice, or policy?
Government datasets that use self-reported sexual identity and gender identity will undercount 2S/GBTQ+ individuals, impacting population size estimates. Willingness to disclose sexual orientation varies by demographic factors that have important implications for measuring, understanding, and intervening in health inequities.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1 (DOCX 19.5 KB)
Author contributions
Jacobsen: methodology, formal analysis, writing—original draft. Babin: methodology, formal analysis, writing—review and editing. Gormezano: supervision, validation, writing—review and editing. Blackwell: methodology, writing—review and editing. Klassen: methodology, project administration, writing—review and editing. Higgins: methodology, project administration, writing—review and editing. Card: conceptualization, methodology, writing—review and editing. Ferlatte: conceptualization, methodology, project administration, writing—review and editing. Lachowsky: conceptualization, methodology, writing—review and editing, supervision, funding acquisition.
Funding
Sex Now 2012 was supported by a grant from Vancouver Foundation. Sex Now 2019 received funding support from Canadian Blood Services MSM Research Fund, funded by the federal government (Health Canada) and the provincial and territorial ministries of health. The views herein do not necessarily reflect the views of Canadian Blood Services or the federal, provincial, or territorial governments of Canada. N.J.L. was supported by a Scholar Award from the Michael Smith Foundation for Health Research (no. 16863). OF is supported by a Scholar Award from the Fonds de Recherche du Québec. This work was supported by the Canadian Institutes for Health Research (PJT-180237).
Data Availability
Partial Sex Now data is publicly available on the OurStats dashboard (https://ourstats.ca/our-dashboard). Data access requests can be made through the protocol outlined at: https://www.cbrc.net/cbrc_data_access_process
Code availability
SAS code for analyses is available upon request.
Declarations
Ethics approval
Ethics approval for Sex Now 2012 was granted by the independent ethics board of the Community-Based Research Centre (CBRC). Ethics approval for Sex Now 2019 was granted by research ethics boards at the University of Victoria and the University of British Columbia (BC17-487).
Consent to participate
Informed consent was obtained from all individual participants prior to data collection in Sex Now 2012 and 2019.
Consent for publication
Not applicable.
Competing interest
The authors declare no competing interests.
Footnotes
We use the terms “branched” and “coincident” rather than “discordant” and “concordant” to avoid using language that implies a value judgement about normative sexualities, as per van Anders (2015).
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- Bates, N., García Trejo, Y. A., & Vines, M. (2019). Are sexual minorities hard-to-survey? Insights from the 2020 Census Barriers, Attitudes, and Motivators Study (CBAMS) survey. Journal of Official Statistics,35(4), 709–729. 10.2478/jos-2019-0030 [Google Scholar]
- Bauer, G. R., Braimoh, J., Scheim, A. I., & Dharma, C. (2017). Transgender-inclusive measures of sex/gender for population surveys: Mixed-methods evaluation and recommendations. PLoS ONE,12(5), Article e0178043. 10.1371/journal.pone.0178043 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bauerband, L. A., Altman, C. E., & Teti, M. (2023). Comparison of transgender and gender diverse demographics in United States Transgender Survey and the Behavioral Risk Factor Surveillance System. Annals of LGBTQ Public and Population HealTh,4(3), 251–259. 10.1891/LGBTQ-2021-0038 [Google Scholar]
- DeChants, J., Green, A. E., Price, M. N., & Davis, C. (2021). Measuring Youth Sexual Orientation and Gender Identity (p. 28). The Trevor Project. https://www.thetrevorproject.org/wp-content/uploads/2021/07/Measuring-Youth-Sexual-Orientation-and-Gender-Identity.pdf. Accessed 12 Apr 2022.
- Dharma, C., & Bauer, G. R. (2017). Understanding sexual orientation and health in Canada: Who are we capturing and who are we missing using the Statistics Canada sexual orientation question? Canadian Journal of Public Health,108(1), e21–e26. 10.17269/CJPH.108.5848 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dharma, C., Smith, P., Salway, T., Gesink, D., Escobar, M., & Landsman, V. (2025). A two-step approach to simultaneously correct for selection and misclassification bias in nonprobability samples from hard-to-reach populations. American Journal of Epidemiology,kwaf132, Article kwaf132. 10.1093/aje/kwaf132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ferlatte, O., Hottes, T. S., Trussler, T., & Marchand, R. (2017). Disclosure of sexual orientation by gay and bisexual men in government-administered probability surveys. Lgbt Health,4(1), 68–71. 10.1089/lgbt.2016.0037 [DOI] [PubMed] [Google Scholar]
- Flores, A. R. (2021). Social acceptance of LGBTI people in 175 countries and locations: 1981 to 2020 (pp. 1–48). Williams Institute. https://williamsinstitute.law.ucla.edu/wp-content/uploads/Global-Acceptance-Index-LGBTI-Nov-2021.pdf. Accessed 15 Jun 2022.
- Holzberg, J., Ellis, R., Virgile, M., Nelson, D., Edgar, J., Phipps, P., & Kaplan, R. (2017). Assessing the feasibility of asking about gender identity in the current population survey: Results from focus groups with members of the transgender population (Census Working Paper Nos. RSM2017–12; p. 66). United States Census Bureau. https://www.census.gov/content/dam/Census/library/working-papers/2018/adrm/rsm2018-05.pdf. Accessed 9 Sept 2024.
- Johfre, S. S., & Freese, J. (2021). Reconsidering the reference category. Sociological Methodology,51(2), 253–269. 10.1177/0081175020982632 [Google Scholar]
- Lefevor, G. T., Park, S. Y., Acevedo, M. J., & Jones, P. J. (2022). Sexual orientation complexity and psychosocial/health outcomes. Journal of Homosexuality,69(1), 190–204. 10.1080/00918369.2020.1815432 [DOI] [PubMed] [Google Scholar]
- Lett, E., & Everhart, A. (2021). Considerations for transgender population health research based on US national surveys. Annals of Epidemiology, S1047279721003173. 10.1016/j.annepidem.2021.10.009 [DOI] [PubMed]
- Lombardi, E., & Banik, S. (2016). The utility of the two-step gender measure within trans and cis populations. Sexuality Research & Social Policy,13(3), 288–296. 10.1007/s13178-016-0220-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lowik, A. J., Cameron, J. J., Dame, J., Ford, J., Pulice-Farrow, L., Salway, T., van Anders, S., & Shannon, K. (2022). Gender and Sex in Methods & Measurement—Research Equity Toolkit. Centre for Gender and Sexual Health Equity, University of British Columbia. https://cgshe.ca/practice/research-toolkits/gender-and-sex-toolkit/tools/. Accessed 12 Jun 2023.
- Mendelsohn, D. M., Omoto, A. M., Tannenbaum, K., & Lamb, C. S. (2022). When sexual identity and sexual behaviors do not align: The prevalence of discordance and its physical and psychological health correlates. Stigma and Health,7(1), 70–79. 10.1037/sah0000338 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pruden, H., & Salway, T. (2020). “What and who is Two-Spirit?” in Health Research (Meet the Methods, pp. 1–2). CIHR Institute of Gender and Health. https://cihr-irsc.gc.ca/e/52214.html. Accessed 26 May 2025.
- Puckett, J. A., Brown, N. C., Dunn, T., Mustanski, B., & Newcomb, M. E. (2020). Perspectives from transgender and gender diverse people on how to ask about gender. LGBT Health,7(6), 305–311. 10.1089/lgbt.2019.0295 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Salway, T., Ferlatte, O., Gilbert, M., Cochran, S. D., & Mays, V. M. (2015). Misclassification and Undersampling of Sexual Minorities in Population Surveys. American Journal of Public Health,105(1), e5–e6. 10.2105/AJPH.2014.302408 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Salway, T., Gesink, D., Ferlatte, O., Brennan, D. J., Rhodes, A. E., Marchand, R., & Trussler, T. (2016). Concealment of Sexual Minority Identities in Interviewer-Administered Government Surveys and Its Impact on Estimates of Suicide Ideation Among Bisexual and Gay Men. Journal of Bisexuality,16(4), 427–453. 10.1080/15299716.2016.1225622 [Google Scholar]
- Salway, T., Juwono, S., Klassen, B., Ferlatte, O., Ablona, A., Pruden, H., Morgan, J., Kwag, M., Card, K., Knight, R., & Lachowsky, N. J. (2021). Experiences with sexual orientation and gender identity conversion therapy practices among sexual minority men in Canada, 2019–2020. PLoS ONE,16(6), Article e0252539. 10.1371/journal.pone.0252539 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Salway, T., Morgan, J., Ferlatte, O., Hawkins, B., Lachowsky, N. J., & Gilbert, M. (2019). A systematic review of characteristics of nonprobability community venue samples of sexual minority individuals and associated methods for assessing selection bias. LGBT Health,6(5), 205–215. 10.1089/lgbt.2018.0241 [DOI] [PubMed] [Google Scholar]
- Saperstein, A., & Westbrook, L. (2021). Categorical and gradational: Alternative survey measures of sex and gender. European Journal of Politics and Gender,4(1), 11–30. 10.1332/251510820X15995647280686 [Google Scholar]
- Silva, T., & Fetner, T. (2022). Sexual identity-behavior discordance in Canada. Canadian Review of Sociology/revue Canadienne De Sociologie. 10.1111/cars.12372 [DOI] [PubMed] [Google Scholar]
- Sorge, J., Colyer, S., Cox, J., Kroch, A., Lachowsky, N., Popovic, N., & Yang, Q. (2023). Estimation of the population size of gay, bisexual and other men who have sex with men in Canada, 2020. Canada Communicable Disease Report,49(11–12), 465–476. 10.14745/ccdr.v49i1112a02 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Spiegelman, D., & Hertzmark, E. (2005). Easy SAS calculations for risk or prevalence ratios and differences. American Journal of Epidemiology,162(3), 199–200. 10.1093/aje/kwi188 [DOI] [PubMed] [Google Scholar]
- Statistics Canada. (2022). Age, Sex at Birth and Gender Reference Guide, Census of Population, 2021. Government of Canada. https://www12.statcan.gc.ca/census-recensement/2021/ref/98-500/014/98-500-x2021014-eng.cfm. Accessed 9 Sept 2024.
- Statistics Canada. (2023). Sexual orientation of person. https://www23.statcan.gc.ca/imdb/p3Var.pl?Function=DEC&Id=1223774. Accessed 9 Sept 2024.
- Suen, L. W., Lunn, M. R., Sevelius, J. M., Flentje, A., Capriotti, M. R., Lubensky, M. E., Hunt, C., Weber, S., Bahati, M., Rescate, A., Dastur, Z., & Obedin-Maliver, J. (2022). Do ask, tell, and show: Contextual factors affecting sexual orientation and gender identity disclosure for sexual and gender minority people. Lgbt Health,9(2), 73–80. 10.1089/lgbt.2021.0159 [DOI] [PMC free article] [PubMed] [Google Scholar]
- The Trans PULSE Canada Team. (2022). QuickStat #5 – How trans and non-binary participants would be counted in a Statistics Canada survey. https://transpulsecanada.ca/results/quickstat-5-how-trans-and-non-binary-participants-would-be-counted-in-a-statistics-canada-survey/. Accessed 15 Mar 2023.
- van Anders, S. M. (2015). Beyond sexual orientation: Integrating gender/sex and diverse sexualities via sexual configurations theory. Archives of Sexual Behavior,44(5), 1177–1213. 10.1007/s10508-015-0490-8 [DOI] [PubMed] [Google Scholar]
- Wolff, M., Wells, B., Ventura-DiPersia, C., Renson, A., & Grov, C. (2017). Measuring sexual orientation: A review and critique of U.S. data collection efforts and implications for health policy. The Journal of Sex Research,54(4–5), 507–531. 10.1080/00224499.2016.1255872 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material 1 (DOCX 19.5 KB)
Data Availability Statement
Partial Sex Now data is publicly available on the OurStats dashboard (https://ourstats.ca/our-dashboard). Data access requests can be made through the protocol outlined at: https://www.cbrc.net/cbrc_data_access_process
SAS code for analyses is available upon request.