A systematic review of mental disorder, suicide, and deliberate self harm in lesbian, gay and bisexual people

Why study mental health gaps in lesbian, gay, and bisexual people at all? Because if there’s an extra burden, it changes everything—from the science we pursue to the services we provide. There’s a compelling idea on the table: minority stress. When your daily life includes stigma, discrimination, and the real possibility of rejection or violence, that pressure can seep in. It can shape how safe you feel, how you cope, and, over time, how your mind holds up. Add in the fact that some people internalize shame about their sexuality, and you have a plausible pathway to higher anxiety, more depression, and riskier coping, including alcohol and drugs. That’s the hypothesis. The question is whether the data back it up, and if so, how big the difference is and where it shows up. To answer that, you want population-based comparisons—not clinic samples, not people who happen to be in treatment, but surveys that include both non-heterosexual and heterosexual participants measured the same way, at the same time. King and colleagues went after exactly that. They searched widely—across medical, psychological, and social science databases—pulling studies from the late 1960s through 2005. They ended up with 25 epidemiological studies, which together reported on very large numbers of heterosexual participants and a substantial sample of people identifying as lesbian, gay, or bisexual—LGB. That spread of data matters. It means we’re not just listening to the loudest voices; we’re trying to take the pulse of whole populations. They were fussy about what counted. Sexual orientation had to be defined explicitly—identity, attraction, behavior, or standardized scales—and outcomes had to map onto recognized psychiatric criteria or validated symptom thresholds. The focus was on depression and anxiety, alcohol and other substance misuse or dependence, suicidal behavior and suicidal ideation, and deliberate self-harm. Deliberate self-harm, by the way, means intentional self-poisoning or injury regardless of motive. To keep things clean, studies from specialized clinics or service settings were out. Quality was scored on sampling method, participation rate, the population’s breadth, and whether there were enough LGB participants to say something meaningful. Only one study ticked every box, and seven met three of four. Not great, but not a deal-breaker when you step back and look at the pattern across studies. For synthesis, they did what you’d want them to do: risk ratios from prevalence data, standardized mean differences for continuous outcomes, and, when at least two studies lined up, random-effects meta-analyses with inverse-variance weighting. That model assumes the true effect can vary across studies, which is realistic here. They also tracked heterogeneity—the share of variation that isn’t just random noise—with a statistic called I squared. Most of the data were cross-sectional, and only a handful of cohorts existed. Prospective information, especially about suicide risk over time, was scarce. Let’s start with the starkest outcome: suicide attempts. Across a lifetime, LGB participants were more likely to have ever attempted suicide. The pooled risk ratio was 2.47, which means about two and a half times the risk compared with heterosexual peers. That’s not a tiny bump; that’s a real difference. Looking just at the past year, the combined estimate hovered around 2.56. In other words, in recent time windows, the elevation is still there, roughly two to threefold. Men often showed slightly higher recent risks than women, but the pattern held for both. If you translate that into population terms, the fraction of recent suicide attempts that could be attributed to being in the non-heterosexual group landed in a modest range across studies, and for lifetime attempts it rose into the low double digits in some samples. However you slice it, the signal is consistent. Suicidal ideation followed the same direction. When studies asked, “Have you ever seriously thought about killing yourself?” LGB respondents, on average, were more likely to say yes. The pooled risk ratio for lifetime ideation was 2.04, so about double. There was heterogeneity—definitions varied, and so did who was surveyed—but even in higher-quality subsets, the excess persisted. In school-aged samples, lesbian and bisexual girls stood out with higher cumulative incidence, while among adults, gay and bisexual men tended to show especially elevated lifetime attempt risk and lesbian and bisexual women had higher recent ideation. Those are different manifestations of the same core problem: more people are struggling with thoughts of suicide, and more are acting on them. Now, deliberate self-harm sits next to those outcomes but isn’t identical, because motive isn’t required for the definition. Here, the results were messier. Two studies that broke out women found higher lifetime self-harm in lesbian and bisexual women, with risk ratios around 1.2 in one and 1.7 in another. A pooled estimate tilted higher but was imprecise. For gay and bisexual men, data were thinner and inconsistent. That doesn’t mean there’s no difference; it means we don’t have the same statistical clarity as we do for suicide attempts and ideation. Moving to common mental disorders—depression and anxiety—and you see elevated risk again, but the size depends on the slice. Broadly, LGB groups had about one and a half to two and a half times the risk of mood and anxiety problems compared with heterosexual groups. When researchers looked at twelve-month anxiety across sexes, the combined relative risk was 1.54. Narrow it to men and it jumped to 1.88. Lifetime anxiety painted a similar picture: elevated across both sexes at around 2.28, and clearly higher in men at about 2.40. Women were more variable across studies, and in one small subset the lifetime estimate for women didn’t differ from heterosexual peers, which likely reflects noise from limited data rather than a firm absence of difference. Depression followed suit. Across compilations, lifetime depression was more common in LGB groups overall, with less disagreement between studies than you might expect. Some analyses suggested that women, in particular, showed a stronger signal in certain contexts, including one of the higher-quality studies. The throughline is simple: if you’re lesbian, gay, or bisexual, your odds of having dealt with clinically important depression or anxiety are higher. Substance dependence is where the numbers really open your eyes, and the pattern is not uniform across men and women. Start with alcohol. Across multiple studies, twelve-month alcohol dependence was more common in LGB samples, with relative risks that generally fell between about 1.8 and 3.0. In one of the stronger studies looking at lifetime alcohol dependence, risk was higher in the combined LGB group by a factor of 2.59. But when you split by sex, the asymmetry appears: women had a striking elevation—about 6.5 times the risk—while men didn’t differ clearly from heterosexual peers in that particular analysis. That matches individual studies where lesbian and bisexual women also reported more heavy drinking and alcohol misuse than heterosexual women. Other drugs told a parallel story. For twelve-month drug dependence, the overall relative risk came in around 3.5 across sexes. Looking at lifetime drug dependence in one large study, you see a fourfold elevation in the combined group. Split that again and the gradient sharpens: roughly 2.7 times in men, and, in women, an estimate on the order of 7 to 8 times. Those female estimates came with wide confidence bands, but the direction was the same across datasets. Put plainly, substance problems are more prevalent in LGB communities than in heterosexual ones, and lesbian and bisexual women, in particular, carry a disproportionate share of that burden. A quick pause here, because numbers like that demand a breath. These are not boutique outcomes. We’re talking about the stuff that fills emergency rooms and therapy schedules, that derails schooling and work, that frays families. Seeing twofold increases in suicidality, near doubling in anxiety, and multi-fold jumps in drug dependence isn’t just statistically significant—it’s the kind of pattern that changes how you think about prevention and care. Of course, the evidence isn’t perfect, and King and colleagues are candid about that. Only one included study met all four of their quality indicators. Many relied on non-random samples or had modest participation rates, which can undercut representativeness. In several population surveys, the proportion of respondents identifying as LGB was lower than you’d expect, hinting at under-disclosure. Sexual orientation is also hard to fix in a single tick box: identity, attraction, and behavior don’t always align, and they can shift over time, especially in adolescence and early adulthood. Add to that the cross-sectional nature of most studies—snapshots instead of movies—and you can’t cleanly say what causes what. There was enough heterogeneity across outcomes that some pooled estimates came with big error bars. Because each domain didn’t have dozens of studies, the team couldn’t do fancy bias diagnostics like funnel plots or meta-regressions with much confidence. But step back to the forest view, and the trees point the same way. Across different countries, sampling frames, and measures, LGB groups were at higher risk for suicide attempts and ideation, for depression and anxiety, and for alcohol and other drug dependence. The exact magnitude varies, and women and men show different contours—especially around substances—but the basic elevation is robust. What might be driving it? The review doesn’t claim to nail down causality, but the minority stress framework fits the contours. Chronic exposure to stigma and discrimination is a stressor. It constrains social support, increases vigilance, and encourages concealment. That package can worsen mood, amplify anxiety, and make substances look like a dependable, if costly way to cope. Some studies even point to environmental contributors—like social venues where alcohol is central—that might raise exposure. None of that excuses the heterogeneity or fills the gaps in prospective data, but it gives a coherent map for why we see what we see. So what do you do with this? Clinically, you treat non-heterosexual status as a risk marker for mental health problems, particularly suicidality and substance dependence, while avoiding the trap of assuming every LGB patient has these issues. It’s a prompt to ask, to screen, to create rooms where disclosure feels safe, and to tailor prevention—especially around suicide—to the realities of stigma and isolation. Public health-wise, the pattern argues for resourcing LGB-affirming services and community-based interventions that address both mental health and substance use without forcing people to choose which problem gets attention first. Then there’s the science we still need. We need more prospective, population-based studies that follow people over time, with repeated, nuanced measures of sexual orientation—identity, attraction, behavior—so we can see how changes in those dimensions interact with stress and support. We need adolescence clearly in frame, because that’s where identity formation and risk often spike. We need enough statistical power to slice analyses by sex and orientation without the estimates flying all over the place. The big picture is clear even as the pixels need sharpening. Across multiple lines of evidence, lesbian, gay, and bisexual people face higher risks of depression and anxiety, greater burdens of alcohol and drug dependence, and a twofold elevation in both thinking about and attempting suicide. Those are not abstract disparities. They are lived realities that good measurement helped us see—and that better policy and care can help change.