It is not uncommon for researchers in gender medicine to exaggerate their findings and let allied media spin the rest. But in this case, the authors of the NHB study themselves promoted a narrative of cause and effect. “In this study,” wrote lead author Wilson Lee on LinkedIn, “we presented causal evidence that enacting state-level anti-transgender legislation increased suicide attempts among transgender and nonbinary young people in the US.” Co-author Ronita Nath told CNN, “We’ve long known that the associations between anti-transgender policies and negative health outcomes for LGBTQ+ young people exist, but this is the first time any study has shown this causal relationship.”

Yet, in what has become routine in this research area, the NHB study’s findings and conclusions later crumbled under scientific reexamination. As a methodological criticism published (to its credit) in the NHB last month—over a year after the original study—shows, the observed elevation in suicide attempts came from a small sample (roughly 100 youth) in a single state (Idaho), at a time when that state’s “anti-transgender” laws were not even in effect. Further, the researchers did not properly control for confounding factors. (The Society for Evidence-Based Gender Medicine has published its own methodological analysis, which is worth reading.)

Predictably, neither the authors nor the media outlets that trumpeted their “findings” have so far bothered to correct the record. In fact, the authors have said that the methodological criticisms “do not alter the interpretation of our findings.” NHB gave the authors of the original study a chance to respond to the criticism. As explained below, not only did their response fall short, but it also included a potentially damning admission. To judge by how these things usually go, the study will almost certainly continue to be cited as settled fact, bolstered through citation laundering in scientific journal publications.

Wilson Lee et al. will likely enter the pantheon of zombie studies that provide “evidence” for common talking points: that the practice of pediatric transition was launched based on good evidence, as asserted by de Vries et al.; that regret is less than 1 percent, per Bustos et al.; that “gender-affirming care” reduces suicidality by 73 percent, per Tordoff et al.; that hormones improve “psychosocial functioning,” per Chen et al.; and so on. That the findings from these studies have been scientifically examined and discredited (see here and here for de Vries; here and here for Bustos; here, here, and here for Tordoff; and here, here, here, and here for Chen) does not diminish from their canonical status among advocates of pediatric transition.

But even by the standards of this literature, the NHB study and its shaky foundations stand out. In fact, a coauthor confessed in an interview soon after the study’s release that the work had an explicit political goal. It’s worth examining the story in full.

The two main problems with the NHB study concern its data source and methodology. To begin with the first, the authors—who were (or are) all affiliated with the Trevor Project, a national advocacy group for suicide prevention among “LGBTQ+ young people”—utilized a Trevor Project survey of LGBTQ youth ages 13–24 between the years 2018 and 2022. The survey is a non-probability (non-representative) sample, and the total number of minor respondents was 35,196. Respondents were recruited through social media.

Surveys of this nature are notoriously unreliable as a basis for scientific studies on cause and effect. Consider, for example, the 2015 U.S. Transgender Survey (USTS-15). The survey has served as fertile ground for gender-affirming researchers to produce studies that claim to find positive mental-health benefits from puberty blockers and cross-sex hormones, all based on respondents’ answers.

As Michael Biggs found, however, almost three-quarters of USTS-15 respondents who said they took puberty blockers said that they started them after age 18, which wouldn’t make sense. Even the survey’s authors acknowledged that “the question may have been misinterpreted by some respondents who confused puberty blockers with the hormone therapy given to adults and older adolescents.”

This was not the only example of confusion or recall bias. In a 2023 study purporting to debunk the “rapid onset gender dysphoria” hypothesis, UCSF psychiatrist Jack Turban and colleagues uncritically accepted the self-report of about 700 USTS-15 respondents who, as adults, recalled realizing they were “transgender or gender diverse” by age two.

It makes sense that bias is baked into the USTS-15. Respondents were recruited through social-advocacy networks and were likely to be more politically engaged and tuned in to transgender-movement politics. Perhaps many recalled realizing they were transgender at age two because they are invested in a “born that way” narrative of transgender identity.

Respondents to the Trevor Project survey were likely to have been similarly motivated to produce answers that the Trevor Project wanted to see. And because the survey was taken against the backdrop of “anti-transgender” laws passing around the country—and a constant drumbeat of alarm about those laws—it’s at least possible that some respondents were motivated to exaggerate or misrepresent their experiences of suicide attempts.

Finally, it’s noteworthy that the NHB authors used a survey of “LGBTQ young people” rather than, say, a survey of gender-clinic-referred youth, to study the link between “anti-transgender laws” and suicide attempts. If we want to know how age-minimum restrictions on medical transition affect suicide attempt rates, it might make sense to focus on data from kids who seek medical transition. By including data from kids who identify as transgender but do not seek medical transition (a larger group than those seeking medical interventions), we risk “contaminating” the sample, as those kids may have different psychological response mechanisms.

Assume, though, that the survey is reliable. Even more serious problems await in the NHB study’s methodology. The authors’ confident use of causal language to describe their findings stems from their reliance on a quasi-experimental method called “Difference-in-Differences” (DiD). In DiD, researchers create two groups (intervention and control), document the changes in each group over time, and then compare the changes to estimate the effect of the intervention.

One problem with the NHB study’s DiD model is that the “intervention” (independent variable) is arguably not a single variable. Of the 48 state laws included in the study, 30 prohibit males from competing against females in female sports and only seven impose age minimums for access to pharmaceutical interventions. By coding all laws as “anti-transgender” and treating them as functionally the same, the authors assume that they have the same psychological impacts on adolescents who identify as transgender.

But why should we allow that assumption? Citizens readily perceive that different laws or regulations affecting a particular social group may have different motivations or justifications. For example, a law that stops racial gerrymandering might not have the same psychological impact on African-Americans as a law that eliminates racial affirmative action in higher education, even if progressive activist groups describe both laws as “racist.” The NHB study authors’ assumption that all “LGBTQ young people”—a category, note, that includes gay, lesbian, bisexual, and gender-nonconforming (“queer”) youth, not just youth who identify as the opposite sex and seek medical interventions—would respond in a similar manner to both kinds of laws covered seems highly dubious.

Key to the reliability of a DiD study design is what is known as the “parallel trends” requirement: absent the intervention, the difference between the two compared groups would remain constant over time. If this requirement is not fulfilled, it is not possible to attribute any observed effect to the intervention. As the methodological critique points out, however, there were good reasons to doubt the existence of parallel trends in the NHB study model. One reason: “Idaho was one of only two states that began to ration care due to COVID-19” during the time period reflected in the survey’s data, which “could plausibly influence mental health.”

The authors of the original study claim that they controlled for Covid impacts. Indeed, the original study states: “For controlling for COVID-19 covariate, population-adjusted COVID-19 death counts by state was used as a proxy for the overall impact of COVID-19 in a given state.”

But this control likely does not address the critique’s concerns. Adolescent mortality rates from Covid were extremely low, and far below those of older adults. Meantime, Covid appears to have had negative mental-health impacts on many adolescents. It seems hardly plausible, then, to use Covid deaths as a proxy for youth mental-health impacts. The fact that Idaho had state-specific Covid impacts and responses very likely invalidates the parallel trends assumption and renders the DiD design untrustworthy.

But there’s an even bigger problem with the study. Despite leading readers and the media to believe that “anti-transgender” laws increased suicide attempts anywhere they were passed, the NHB authors’ model found nothing of the sort. The authors examined survey responses from three post-passage periods (t1, t2, and t3) following passage of “anti-transgender” laws. An increase of attempts relative to pre-legislation years was observed in five states during t1, but the increase was very small and barely reached statistical significance (p=0.049).

The study does find an effect on suicide attempts in periods t2 and t3. But that increase is driven entirely by about 100 respondents in the state of Idaho. Moreover, the only two laws passed in Idaho that qualify as “anti-transgender”—HB 500, concerning school sports, and HB 509, concerning the recording of sex on birth certificates—had little to do with medical interventions. Moreover, both laws were blocked in the courts more than a year before the rise in suicide attempts appeared in the data.

In response to this observation, the original authors conceded that though the laws were not fully in effect (the sports law may have been implemented in some areas of the state despite the judicial block), “the public discourse surrounding the passage of these laws would not have ceased when they were enjoined, providing a further plausible mechanism for sustained effects.” In other words, even if the laws weren’t in effect, the mere discussion of the laws would have had the same mental-health impacts as the laws’ actual provisions.

This maneuver by the authors presents a serious problem for them. As noted in the methodological criticism, if public discourse stressing calamitous mental-health impacts from a law is itself “a mechanism for sustained effects,” then that messaging confounds the observed increase in suicide attempts. “For example, the Idaho sports bill was called ‘incredibly dangerous and discriminatory’ and its passage described as ‘send[ing] a strong message to trans youth that they are less than their peers and not deserving of community and acceptance.’ Bills studied in the [NHB study] have been described as ‘hostile’ and ‘threaten[ing] adolescent lives.’” The methodological critics quote from an important 2023 article by Alison Clayton in this context:

An excessive focus on an exaggerated suicide risk narrative by clinicians and the media may create a damaging nocebo effect (for example, a ‘self-fulfilling prophecy’ effect) whereby suicidality in these vulnerable youths may be further exacerbated.

In other words, in their effort to defend against a criticism, the Trevor Project authors inadvertently concede the potential for “public discourse” (alarmist messaging about teen suicidality) to become self-fulfilling prophecy. That implies that transgender advocates themselves might bear responsibility for the (let us assume real) rise in suicidal behavior.

In what seems like a futile effort to shore up that responsibility, the Trevor authors write that “policy discourse [is] a part of the causal pathway rather than a confounder since . . . policy discussion may be a part of the policy impact itself.” This may be true, but in this context the retort is cynical and self-serving. Just as legislative proposals over sports and medical regulation are a choice, so, too, is responding to such proposals with unsubstantiated threats of an impending suicide epidemic.

The NHB study is just the latest in a wave of research purporting to show that “gender-affirming care is life-saving care.” Advocates of pediatric medical transition have consistently misrepresented the evidence on suicide and suicidality (distinct but frequently conflated concepts), generating widespread panic about the potential effects of shifting away from the “gender-affirming” model. Critics of pediatric medical transition have also been prone to exaggeration and misrepresentation, for instance, when they say that transition has been shown to increase suicide.

All we can say with confidence, based on existing research, is that youth who identify as transgender or are diagnosed with gender dysphoria (the two are not synonymous) appear to be at elevated risk for suicide and suicidality, though the risk of suicide remains, thankfully, very low, and there is no credible evidence that hormonal and surgical interventions are needed to manage that risk. The best available evidence suggests that mental-health comorbidities, common in this population, better account for suicide risk, and that risk of suicide remains elevated relative to controls following transition.

At the same time, suicide prevention experts (and LGBT advocacy groups, until recently) have acknowledged that suicide is a socially contagious behavior. For this reason, they have warned about the risks of improper public discussion of the topic. In the United Kingdom, for example, Professor Louis Appleby, a government suicide prevention expert, issued a report in 2024—just a few months before the NHB study was published—in which he described the suicide narrative commonly deployed by transgender activists as “dangerous,” adding that it “goes against guidance on safe reporting of suicide.”

It goes without saying, then, that trusted scientific authorities, especially prestigious academic journals, should exercise extra caution when publishing studies claiming to find that some law or policy is causing youth to attempt suicide more.

One thing journals could do is tighten up the requirements for reporting conflicts of interest. Even a cursory glance at the NHB study suggests evidence of such conflicts. The authors were all affiliated with or worked for the Trevor Project, a nonprofit whose fundraising prospects are directly tied to funder perceptions about suicide being a serious problem. To their (partial) credit, the study authors list their affiliation with Trevor under “competing interests.” But readers unfamiliar with the organization or its fundraising incentives may not see the full significance of this affiliation.

Evidence of author bias appears throughout the study. The authors, for example, refer to laws that ban males from female sports as “anti-transgender.” But it is a male’s sex, not his transgender identification, that makes his participation in female sports a matter of public concern. That’s why there is virtually no controversy over trans-identifying females competing in male sports. The authors also assert that “gender-affirming healthcare has well-established benefits,” providing as citations only surveys of parental attitudes about state age-restriction laws, a survey of adults on the impact of Covid for access to “gender-affirming” interventions, and the World Professional Association for Transgender Health (WPATH)’s latest “Standards of Care.” (Left unmentioned, of course, is the hard evidence that WPATH suppressed systematic reviews after discovering disappointing findings.) The authors simply ignored the existing systematic reviews that find no credible evidence for mental-health benefits from pediatric gender medicine.

Most strikingly, one of the study’s co-authors openly confessed her team’s agenda in an interview with CNN shortly after the study was published: “We knew if we were able to show this it would be a breakthrough for using definitive scientific evidence to support calls for protective and affirming policies for trans and nonbinary young people and to help, ultimately, save their lives.”

This is an astonishing admission. They went into this project intending to provide “scientific” ammunition to a legal fight. It is hardly the first example of dressing up legal advocacy in scientific garb, but it’s striking to see the strategy acknowledged so frankly.

As anyone who closely follows the subject knows, research in this field is contaminated by activist agendas and misconduct. The findings are fundamentally compromised, and the journals—even the most prestigious among them—often act as willing collaborators in the suppression of scientific debate. Though NHB deserves credit for allowing some debate, the 2024 study on suicide attempts shows, once again, that when it comes to transgender issues, activist researchers can exploit academic journals’ refusal to enforce even the most rudimentary scientific standards.