A new study in the JAMA Network Open claims to find that testosterone therapy “significantly reduce[s] gender dysphoria, depression, and suicidality in transgender and gender-diverse individuals desiring testosterone therapy.” The study, titled “Early Access to Testosterone Therapy in Transgender and Gender-Diverse Adults Seeking Masculinization: A Randomized Clinical Trial,” is being celebrated by advocates of sex trait modification (STM) because it is the first randomized controlled trial (RCT) in this area of research.

Systematic reviews of evidence in Sweden, Finland, and the U.K. have graded the evidence from research on testosterone therapy in minors as “very low.” In large part, that owes to the fact that even the best studies to date have been observational cohort surveys, with poor or no controls. As such, these studies have significant risk of bias.

This new paper is no game-changer. Even assuming that the findings of a study done on adults apply automatically to adolescents, the study did not find that hormone therapy improves mental health. Participants in the treatment group did experience reduced gender dysphoria, depression, and suicidality relative to the control group, which the authors said received “no treatment.” A key detail revealed by the authors is that the study’s very short follow-up time of three months was “insufficient for the development of masculinizing physical characteristics and associated reductions in gender dysphoria.” Assuming this means that no physical changes occurred, improvements in the mental health of the treatment group cannot credibly be attributed to STM and the achievement of “embodiment goals.” They are more likely the result of placebo and Hawthorne effects, regression to the mean, or the antidepressant properties of testosterone itself.

Like a study on social transition in children published in Pediatrics last year, the new JAMA study, if interpreted with caution, might actually provide stronger evidence against the conclusion of its authors than in favor. This follows a well-established tradition of researchers interpreting their own results without genuine curiosity or scientific neutrality.

The new JAMA study was conducted on a group of 64 Australian women seeking masculinization in order to achieve a male-typical appearance (“transgender men”) or a “non-binary” appearance (“gender-diverse”). Advocates of STM argue that RCTs are unethical in this area because hormonal therapy is known to be “medically necessary”—a claim that assumes the very thing in dispute. To overcome this supposed ethical problem, the authors included as their control group women who were on a waiting list to receive testosterone therapy. They randomized the 64 participants into two study arms: one whose members received testosterone immediately, and another whose members had to wait to receive the drug.

The authors compared key mental-health outcomes for 62 of the 64 participants at three months following initiation of hormones in the treatment group. The primary outcome was gender-dysphoria mitigation, and the secondary outcomes were reduced depression and suicidality. They found that gender dysphoria dropped by a mean of 7.2 points (on a scale of 14 to 70) in the treatment group relative to the control group, though the mean scores of the treatment group participants remained within the clinical range for gender dysphoria. Depression scores in the treatment group dropped from a mean of 15.2 (moderately severe) at baseline to a mean of 8.6 (mild) at follow-up, on a 1- to 27-point scale. That represents a mean drop of 5.6 points relative to the control group. Suicidality mean scores for the treatment group were 12 at baseline and 5.6 at follow-up on a 0–50 scale where scores above 21 constitute high risk of suicidal behavior. The authors further reported a “resolution” of suicidality in 11 members of the treatment group (55 percent of that group) compared with only one member of the control group (5 percent).

Do these findings suggest that STM causes mental-health improvement? To measure changes in gender dysphoria, the authors used the Gender Preoccupation and Stability Questionnaire. The GPSQ measures a person’s “preoccupation (time spent thinking, worrying, or being upset) with gender and the degree to which an individual’s gender identity is stable and unwavering.” The traditional tool for measuring gender dysphoria, the Utrecht Gender Dysphoria Scale (UGDS), deals with how a person feels about her body and sex-related physical processes. Because the participants in the Australian study’s treatment group did not experience any significant physical changes, they likely would not have shown improvement on the UGDS. And even if they did, the use of that scale to test dysphoria before and after medicalization is highly dubious.

The GPSQ does not deal with the sources of dysphoric feelings, only with “preoccupation” with “gender” and “stability” of “gender identity.” As such, it does not convey information about the causes of changes in gender-dysphoric feelings over time. In other words, the GPSQ could show a reduction in gender dysphoria due to, say, placebo effects or an individual’s being absorbed in a new job. Moreover, the GPSQ, as explained by one of its developers, “was designed to only assess gender dysphoria in populations over the age of 18 years,” a fact that “prevent[s] it from being used reliably in populations under the age of 18.” This is partly because adolescents are by definition still in a period of physical and psychological development, and partly because some of the life pursuits inhibited by fixation on “gender” (e.g., work) are less relevant for adolescents.

As for depression and suicidality, the study’s findings are even more questionable. Research, including in this specific area of medicine, suggests that testosterone likely has antidepressant properties due to its effects on serotonin reuptake transporter binding. As one group of gender-affirming researchers noted in a 2021 study on adolescent females receiving testosterone treatment, “in addition to psychological improvement due to decreases in distress related to the misalignment of gender identity and gender expression, direct physiological actions of T may attenuate depression.”

Given these antidepressant effects, it is important to discern whether improvement in depression and suicidality is due to the hormone itself, the physical changes the hormone induces, or some placebo/Hawthorne effect. To figure this out, a proper RCT would randomize participants into three study arms: testosterone, standard anti-depressants (SSRI), and placebo. In a study like this one with a very short follow-up time, participation can very well be double-blinded. The researchers can conduct mental-health assessments in all three groups until such time as the testosterone group starts to show physical changes, at which point the study is no longer double blinded. The researchers can then compare the results among the three groups to see the direction and magnitude of mental-health effects.

The JAMA authors say that the three-month follow-up period was “insufficient for the development of masculinizing physical characteristics and associated reductions in gender dysphoria.” If this means that the hormone-cohort participants did not have enough time to see physical changes, then mood improvements are more likely the result of placebo/Hawthorne effects, the antidepressant properties of testosterone, or some combination of the two.

Not only did the authors fail to compare testosterone to an active comparator like SSRI, but they actually excluded a candidate from participation in the control group on the grounds that she was already on antidepressants. In the study’s limitations section, the authors explain that “randomization of participants to no treatment or placebo over longer follow-up is unethical, particularly given preexisting barriers to accessing GAHT [gender-affirming hormone therapy].” They concede, however, that “it is possible that the effect of testosterone or patient knowledge of treatment has been evaluated.”

It is highly significant that the study involved female-to-male transitioners only. Advocates of STM commonly obscure the potentially important differences between giving females testosterone and giving males estrogen. They extrapolate from one group to the other, suggesting an overly narrow focus on the sex-trait modifying effects of these drugs.

A 2022 study by Jack Turban and others, which relies on data from the U.S. Transgender Survey of 2015, claims to find that “people who accessed [cross-sex hormones] during early or late adolescence had a lower odds of past-month suicidal ideation and past-month severe psychological distress in adulthood, when compared to those who desired but did not access [cross-sex hormones], after adjusting for a range of potential confounding variables.” As Michael Biggs has shown in a critique of that study, Turban et al. mistakenly assume that “the effect of estrogen on males is . . . identical to the effect of testosterone on females.” A more rigorous analysis of the 2015 Transgender Survey data shows that “[m]ales who took estrogen are more likely to plan suicide, to attempt suicide, and to require hospitalization for a suicide attempt.” They were also significantly more likely to experience “severe distress” than females who took testosterone.

In short, the JAMA study suffers from problems of bias and generalizability. It did not find that STM causes mental-health improvement, only that testosterone is associated with mental-health improvement. In other words, it did not do what an RCT is supposed to do: randomize, control for confounding factors, and find causal evidence. Even if this problem of bias is ignored, the study’s findings would be limited to adult females seeking masculinization. Extrapolations to minors or even to adult males are unsupported by the study’s purported findings.

It is ironic that advocates of STM are celebrating an RCT despite claiming in recent years that RCTs are unethical. It is even more ironic that the study may actually provide evidence against testosterone as “medically necessary” for improving mental health.

The authors are not careful to describe the findings in this way. Though they report that their findings are relevant to “transgender and gender-diverse adults desiring commencement of testosterone therapy” (i.e., females), they go on to say that “these findings have critical implications for service access and delivery to ensure timely access to gender-affirming hormone therapy.”

Generally speaking, those who do research on STM and find “evidence” of benefits are clinicians who already practice and advocate for STM, and who have pegged their personal reputation and professional careers to the controversial practice. Their research, in other words, is tainted by intellectual and professional conflicts of interest. The JAMA study’s lead author, Brendan Nolan, an endocrinologist and Ph.D. candidate at the University of Melbourne, said that he hopes the study will be “an impetus for more clinicians to provide gender-affirming medical care.”  One suspects, however, that a desire to expand “gender-affirming care” was actually the impetus for the study.

As is to be expected, “gender-affirming care” activists are spinning the study to support hormone therapy for all ages and for both sexes. “New Study finds gender-affirming care halves suicidality among trans people,” reads the headline from Pink News. The Guardian’s headline was: “Fast Access to Hormone Therapy in Transgender Adults ‘Lifesaving,’ Study Finds.” Those who read past the headline would discover that the study dealt only with testosterone, but they would not learn about the confounder problems or the fact that the study produced findings not about suicide but about suicidality (the two are not the same).

What’s the only thing worse than the absence of high-quality research supporting a seriously risky medical intervention? Poor-quality research masquerading as high-quality research to support such an intervention. In a highly polarized environment in which mainstream left-of-center journalists are, with rare exceptions, unwilling to ask hard questions, studies such as these are likely to punch above their deserved weight. The new study will doubtless provide ammunition for those already invested in STM and unwilling to subject the evidence to serious scrutiny.

Photo: erhui1979/iStock


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next