Brief, Empowerment-Based Programs Show Modest Gains for Trans and Non-Binary Mental Health

Transgender and non-binary people carry a burden of psychological distress and health risks that is disproportionately heavy compared with the general population, yet the evidence base for what actually helps them has remained fragmented across hundreds of small, scattered studies. A new systematic review and meta-analysis published in Nature Mental Health now offers the most comprehensive quantitative synthesis to date of interventions designed to improve mental health and health behaviors in these communities, and its findings carry practical implications for clinicians, public health planners and funders alike. Drawing on 68 studies covering 17,023 participants, and pooling effect sizes from 20 trials involving 5,315 people, the research team led by Vedhalakshmi Rajasankar and colleagues at Brown University School of Public Health set out to answer a deceptively simple question: when programs are built for transgender and non-binary populations, do they work, and what kinds work best?

The answer, at the aggregate level, is a qualified yes. Across nine randomized controlled trials that compared interventions against active control conditions, the pooled effect at the end of treatment was small but statistically reliable: Hedges’ g of 0.21, with a 95 percent confidence interval of 0.06 to 0.35 and a P value of 0.005. In the conventional vocabulary of meta-analysis, a Hedges’ g of that magnitude corresponds to a shift of roughly eight percentage points in the probability that a treated participant fares better than a control participant — modest on its face, but meaningful in a field where interventions contend with entrenched stigma, structural barriers to care and profound social stressors that no brief program can dismantle on its own.

One of the more striking discoveries buried in the heterogeneity statistics is a counterintuitive dose-response relationship: briefer interventions outperformed longer ones, a pattern that reached statistical significance at P = 0.015. The authors point to phenomena long recognized in behavioral science, including the novelty effect, in which engagement and perceived benefit spike when a program is fresh and undemanding, and decline as repetition sets in. There is also a well-documented retention problem in digital and community-based mental health programs; longer courses of treatment accumulate dropout, and participants who disengage contribute nothing to the outcomes being measured. For populations that frequently report distrust of health systems and experiences of discrimination in clinical settings, a program that asks less of participants’ time and delivers skills quickly may simply be easier to complete — and a program that is never finished cannot work, however elegant its design.

The methodological machinery behind the review reflects current best practice in evidence synthesis. The team registered the protocol on PROSPERO, searched bibliographic databases systematically, screened records in duplicate, assessed risk of bias with the revised Cochrane RoB 2 tool for randomized trials, and graded the certainty of evidence using the GRADE framework. Effect sizes were computed as Hedges’ g, a bias-corrected standardized mean difference favored over Cohen’s d in meta-analysis because it adjusts for the small sample sizes that dominate this literature. Publication bias was probed with funnel plots and the trim-and-fill procedure, and the analyses were conducted in R using established meta-analytic packages, with the full dataset and code released publicly on GitHub — an unusually transparent approach for a field in which primary data are often locked away.

Perhaps the most actionable finding emerged from the subgroup analyses. When the researchers separated interventions by their theoretical orientation, one family of approaches stood out: empowerment- and skills-based programs, which aim to build participants’ decision-making capacity, self-efficacy and practical coping repertoire rather than simply transmitting information. Across six pooled estimates, these interventions produced a significant positive effect at post-intervention, with Hedges’ g of 0.19, a 95 percent confidence interval of 0.024 to 0.36 and P = 0.024. The logic of why such programs resonate in this population is grounded in minority stress theory, which holds that the excess psychiatric morbidity observed among gender minorities stems less from identity itself than from the chronic, socially imposed stressors of stigma, rejection and violence that surround it. Skills-based programs work at the level of the individual’s agency — strengthening the capacity to navigate disclosure, negotiate safer sex, access care and resist internalized transnegativity — and thereby target mechanisms that are plausibly modifiable even when the broader social environment remains hostile.

The map of what has been studied is as revealing as the map of what works. Nearly two-thirds of the included studies, 42 of 68, focused on HIV or sexually transmitted infection prevention and care, reflecting decades of public health investment that has treated transgender people primarily as a population at epidemiological risk rather than as whole people with mental health needs spanning depression, anxiety, suicide risk, substance use and quality of life. Nearly half the studies, 30 of 68, centered on transgender women, while transgender men, non-binary individuals and youth were strikingly underrepresented. The authors are explicit that this skew leaves the evidence base weakest precisely where community surveys suggest the need may be greatest: non-binary young people consistently report among the highest rates of psychological distress, yet almost no intervention trials have been designed or powered for them.

The trials that did make it into the quantitative synthesis illustrate the creative range of the field. They include gender-affirming, peer-delivered group programs that improved HIV care engagement among transgender women living with HIV; telehealth and text-messaging interventions that supported pre-exposure prophylaxis adherence; community-driven empowerment curricula for young transgender women; online single-session programs aimed at reducing internalized transnegativity; and mobile app-based sexual health promotion platforms. The common threads are affirmation of gender identity, connection to community, and delivery through channels — peers, digital platforms, community organizations — that bypass the gatekeeping and discrimination many participants report encountering in mainstream healthcare. That such varied vehicles converge on a small but consistent average effect suggests the active ingredients may be general ones: social support, skill-building and identity affirmation, rather than any single content module.

Caveats temper the optimism, and the authors do not shy away from them. The overall certainty of evidence, assessed with GRADE, was limited by small samples, heterogeneous outcome measures and imprecise pooled estimates — a reminder that meta-analyses of many small trials do not always predict the results of subsequent large ones. Follow-up assessments showed that effects at later timepoints were attenuated relative to post-intervention, raising the question of whether brief programs produce durable change or simply a transient boost that fades without reinforcement. Funnel plot asymmetry hinted at possible publication bias, though trim-and-fill adjustments did not erase the signal. None of these limitations undermines the central conclusion, but they collectively argue for larger, pre-registered, adequately powered trials with longer follow-up before any specific program is scaled as a standard of care.

What the review ultimately delivers is both a proof of concept and a research agenda. The proof of concept is that interventions designed for and with transgender and non-binary communities can produce measurable improvements, and that brief, empowerment-oriented, skills-based formats are a promising backbone for future programs. The agenda is equally clear: extend rigorous intervention science to transgender men, non-binary individuals and youth; move beyond the gravitational pull of HIV prevention toward mental health outcomes such as depression, anxiety and suicide risk, as well as sexual wellness and quality of life; and test whether the benefits of short programs can be sustained through boosters, digital reinforcement or stepped-care models. In a period when legislative hostility toward transgender people is intensifying in many jurisdictions, the review is a reminder that evidence-based support is both possible and urgently needed — and that the next generation of studies must finally center the populations that current research has left at the margins.

The historical concentration on HIV prevention reflected in this review is not accidental. Public health funding streams in the United States and globally have long channeled resources toward transgender health almost exclusively through infectious disease frameworks, a pattern documented in prior systematic reviews of HIV prevalence and sexual behavior among transgender populations. That legacy produced a generation of interventionists skilled in community-engaged trial design, but it also meant that constructs such as depression, anxiety and suicide risk were often treated as secondary outcomes, if they were measured at all.

The emphasis on empowerment-based approaches also aligns with a broader shift in intervention science away from deficit-focused models. Earlier generations of programs sought to reduce risk behaviors directly, whereas newer curricula position participants as agents who can build resilience against the internalized stigma that minority stress theory identifies as a key mediator between discrimination and poor mental health. Randomized evaluations of online interventions targeting internalized transnegativity, and group-delivered empowerment programs for young transgender women, exemplify this turn and contributed to the pooled estimates in the present synthesis.

Transparency practices deserve note as well. The public release of the dataset, analysis code and R scripts on GitHub allows independent reanalysis, an increasingly valued safeguard in meta-research where analytic flexibility can shape conclusions. The PROSPERO registration likewise anchors the review against outcome-switching concerns. For researchers planning the next wave of trials, the evidence map organized by GRADE outcome categories offers a practical starting point, revealing which outcome domains remain untested and where adequately powered studies of transgender men, non-binary adults and gender-diverse youth could yield the greatest return on scientific investment.

Subject of Research: Systematic review and meta-analysis of mental health and health behavior intervention efficacy among transgender and non-binary populations

Article Title: Improving mental health and health behaviors among transgender and non-binary populations: a systematic review and meta-analysis

Article References: Rajasankar, V., Chen, C. A., Kang, T. K., Garcia Gutiérrez, J. M., Giovannone, N., Medrano, M., Restar, A. J., Operario, D., Guy, A. A., & Sun, S. (2026). Improving mental health and health behaviors among transgender and non-binary populations: a systematic review and meta-analysis. Nature Mental Health, 4(9), 1439-1451. https://doi.org/10.1038/s44220-026-00713-6

Image Credits: AI Generated

DOI: 10.1038/s44220-026-00713-6

Keywords: transgender health, non-binary, mental health, meta-analysis, systematic review, HIV prevention, empowerment interventions, skills-based interventions, minority stress, Hedges' g, randomized controlled trials, health disparities