by Michael Biggs
Michael Biggs is Associate Professor of Sociology at the University of Oxford and Fellow of St Cross College. He researches social movements and collective protest.
Pediatrics just published an article showing that trans-identified children are substantially more likely to report attempted suicide than the general adolescent population. When the results are examined closely, however, we find that the risk extends to kids who identify as lesbian, gay, or bisexual. Emphasis on the exceptional fragility of trans adolescents overlooks the importance of sexual orientation. Indeed, my analysis suggests that gender-nonconforming girls are the most vulnerable, whether they consider themselves to be transgender, bisexual, or lesbian.
Previous evidence on suicide attempts among trans-identified youth has been methodologically flawed, even ignoring the most egregious examples. First, surveys have recruited respondents haphazardly—rather than sampling from a population. Second, respondents have not been asked for their sex, but only for their gender identity. In the United Kingdom, Stonewall’s School Report was marred on both counts.
Toomey, Syvertsen, and Shramko (2018)’s article in Pediatrics provides the first rigorous study of self-reported suicide attempts. They use data on 121,000 adolescents aged from 11 to 19, who were surveyed at schools across the United States. The findings, as reported by LGBTQ magazine The Advocate, are dramatic:
Bear in mind that asking respondents whether they have ever attempted suicide will elicit an overestimate of the actual rate; we know from other studies that more probing questions are needed to distinguish genuine attempts to end life. Therefore we should interpret ‘attempted suicide’ broadly, to include all self-harming behaviors, including those not intended to result in death.
When the original article is examined closely, the results are more complicated than the headline suggests. The authors statistically analyze all the risk factors for attempted suicide, including sex and gender identity, sexual orientation, age, race, and parental education. Surprisingly, perhaps, the biggest single risk factor is actually sexual orientation.
The authors are publishing a companion article on sexual orientation. Until that becomes available, it is possible to estimate (from their Table 2) how the risk of attempted suicide varied according to different combinations of gender identity and sexual orientation—after adjusting for other characteristics like age and race.
The calculation is straightforward for heterosexual, lesbian, gay, and bisexual teens who were not trans-identified. (For simplicity the intermediate categories of ‘mostly heterosexual’ and ‘mostly lesbian or gay’ are omitted.) For each transgender category, I calculate the risk averaged across the observed distribution (from Table 1) of sexual orientations within the category. A caveat is that the these estimates have considerable margin of error because they derive from small numbers: 202 identified as male-to-female, 175 as female-to-male, and 344 as not exclusively male or female (‘nonbinary’ for short). A further 1,052 adolescents were not sure of their gender.
The graph above shows the estimated odds of a student reporting attempted suicide, compared to heterosexual boys. As the Advocate emphasized, teens who identified as female-to-male transgender had the highest risk of attempted suicide; the odds were four times higher than for heterosexual boys. What went unnoticed is that the risk was just as high for bisexual girls.
The next highest rates (triple the odds compared to heterosexual boys) were for bisexual boys, lesbian girls and for kids who identified as nonbinary. The latter’s sex was not recorded, but the majority are likely to be female; other survey evidence suggests that two-thirds of trans-identified adolescents are female (Eisenberg et al. 2017).
The next highest rates (roughly double the odds for heterosexual boys) were for gay boys, for male-to-female transgender kids, and for kids who were unsure of their gender identity (whose sex was not recorded). Finally, heterosexual girls had a significantly higher risk than heterosexual boys.
Stephanie Davies-Arai and Nic Williams’ critique of Stonewall’s School Report suggested that “[t]he ‘transgender’ category may just serve to cover up the scale of suicide attempts and self-harm rates of girls and young women.” Their conjecture is vindicated by this survey evidence from the United States. Over two thirds of the girls who identified as boys were sexually attracted to females (inferred from the proportion calling themselves heterosexual or bisexual), and so arguably are most similar to lesbian and bisexual girls. In sum, then, gender-nonconforming females were the group most likely to report attempted suicide, regardless of whether they identified as male or nonbinary—or as bisexual or lesbian.