By David Chen, MD

A recent study indicating a positive relationship between anti-bullying laws and reductions in suicidal tendencies among youth who identify as gay, lesbian, bisexual, or unsure about their sexuality (LGBQ) indicates an opportunity to find common ground and work in solidarity with this community to develop interventions that effectively improve health outcomes while avoiding treatment strategies that many find immoral.

Since the advent of William of Ockham’s nominalism and Roger Bacon’s empiricism, experimental science has worked toward the scientific proof of causation. The historic first double-blind control trial in 1943 and the first randomized trial in 1946 marked a foundational breakthrough because these approaches permitted the exclusion of confounding, cause-effect relationship confusion, and design biases, especially in sampling. Since then, double-blind randomized placebo-controlled trials (DBRPCTs)—as opposed to observational studies, which by their design can only identify factors that predict but do not necessarily cause an outcome—have become the gold standard for understanding cause-effect relationships in scientific research.

In the 1850s, John Snow developed a form of case-control study aimed at evaluating the community effect after the public introduction of a certain causal policy. This form of observational study eventually became known as the difference-in-difference case-control association (DID) study. [1] In May 2023, JAMA Pediatrics published a DID study to investigate whether state anti-bullying laws (ABLs) reduced self-reported suicidal behaviors (ideation, planning, and attempts) among LGBQ youths.

The authors used over half a million case-control subjects (76,188 LGBQ cases, 496,069 heterosexual controls) from the 2009–2017 Youth Risk Behavior Survey. The authors found that ABLs were associated with a “25.0% reduction in the odds of planning suicide (odds ratio [OR], 0.75; 95% CI, 0.61–0.92) and a 27.9% reduction in the odds of attempting suicide (OR, 0.72; 95% CI, 0.53–0.99).” Liang and colleagues concluded that ABLs protected self-reported LGBQ students from suicidal behavior, including ideation and planning generally and suicide attempts for a subset of the sample. The results supporting these conclusions seemed to evidence valid parallel trend assumptions (viz., that the difference between the cases and controls would have remained consistent over time if there had been no intervention), an important premise for the DID design. The authors endorsed limitations to their study, including the inability to determine sexual attraction fluidity, because of the cross-sectional, non-longitudinal nature of the data.

The conclusions drawn by Liang and colleagues raise intriguing and exciting implications. The first is the relevance of their findings to suicidality among persons who identify as transgender. As mentioned earlier, the study data precluded including this population in the study (and the authors could not confirm that none of the participants identified as transgender). Additionally, one can make competing claims about the generalizability of findings such as these, given evidence for the co-occurrence of uncertainty regarding one’s gender identity and sexual attraction as well as for the different distribution of sexual attraction based on gender identity. (If nothing else, this points to the need for more research in this area.) With all this in mind, however, let us presume that the parallel trend seen between the LGBQ cases and heterosexual controls also holds for youths who identify as transgender, and therefore that Liang and colleagues’ findings can be applied to them.

If this is the case, the study by Liang and colleagues appears to initiate a new argument for DBRPCTs looking at states and school districts with and without existing ABLs so as to prove ABLs cause a protective effect against suicidal behavior for self-reported LGBTQ youth. If effective, interventions like ABLs could reduce the frequency at which individuals seek out so-called gender-affirming surgeries to prevent suicidality—surgeries which violate their physical integrity or otherwise promote a disintegrated sense of identity (see the pieces by Lawrence Mayer and Paul McHugh, Patrick Hunter, and Paul McHugh). In other words, DBRPCTs proving the conclusions of Liang and colleagues might provide strong evidence that gender-affirming surgeries are not only morally illicit, but no longer needed to achieve a significant portion of their stated goals.

This is important because of the potential for ABLs to be an area of common ground for acting in solidarity with LGBTQ youths by providing them with the sacred space they need to evaluate their feelings and their sense of self. From a Catholic perspective, no one should be bullied or mistreated, regardless of sexual or gender identity. ABLs, unlike other promoted options today, are not intrinsically immoral. From a psychiatric perspective, unlike affirmative treatments, these less-invasive interventions like ABLs can address multiple causations (e.g., safety, development, spirituality, psychology, and other familial social causes of distress) and are more in line with best practices because they have a better likelihood of addressing the potential causal factor. Further studies confirming Liang and colleagues’ findings should be completed in order to confirm this new discovery, develop a larger common ground with the LGBTQ community, and more fully realize our Catholic value that no one should be bullied or mistreated.

[1] DID studies are aimed at demonstrating the effect (here the reduction of suicidal behavior in self-reported LGBQ students) after a certain causal policy (here the anti-bullying laws) has been introduced. To prove causality beyond doubt, confounding, cause-effect relationship confusion, and design biases, especially in sampling, have to be excluded. For example, the protective effect detected by Liang and colleagues could have been caused by a yet unidentified factor. In other words, a dominant cultural movement causing the protection effect could also have led lawmakers to codify ABLs. An observational study investigating the circumstances surrounding ABL introductions could sufficiently dispel the causality reversal concern and clarify whether confounding exists. Though these observational studies would not prove causality, they would argue for conducting DBRPCTs. Additionally, the authors did not control the suicidal severity between the cases and controls, exposing the study to the risk of a mean-regression bias. By not controlling suicidality between cases and controls, such sampling might have biased the potential suicidality of the control group by the removal of a substantial portion of the most suicidal individuals in the population who just so happen to self-report as LGBQ. This bias would inflate the effect size. This concern could be alleviated by the authors’ publishing an erratum supplemental table comparing the suicidality between cases and controls. This could also enumerate data on gender in this study.

David T.W. Chen, MD, MS-Bioethics, DFAPA, is the chair of the Catholic Medical Association Boot Camp Committee and part of the adjunct faculties at George Washington University School of Medicine and Children's National Hospital in Washington, DC.