Not evenly shared. Who suicide touches most — and what the data say about why.

An 8-minute read · Reid Robison, MD · Suicide Prevention Month 2026

Suicide can touch any family. But risk is not spread evenly: it clusters by age, by sex, by place, by illness, and by how a community treats the people in it. Knowing where the weight falls isn't about labeling anyone. It tells us where to look first, and — more hopefully — what protects.

Explore the data.

Age and sex: two different stories.

Men account for most deaths. The male suicide rate has run three to four times the female rate for the past decade, and the single highest rate of any group is among men 75 and older (40.1 per 100,000 in 2024)¹. Older men combine several risks at once: loss of a spouse, retirement, chronic illness and pain, isolation, and a generation that was taught not to talk about feelings.

Women report more suicidal thoughts and attempts, but die by suicide less often — a pattern sometimes called the gender paradox. Women's rates peak in midlife (8.1 per 100,000 at ages 45–64)¹.

Young people carry an outsized share of the thoughts. Suicide is the second leading cause of death at ages 10–14, 15–24 and 25–34². Serious suicidal thoughts are most common among young adults 18–25 (12.6% in 2024), who also have the highest attempt rate². In 2023, 20% of US high school students seriously considered suicide and 9% attempted — including 27% of girls who seriously considered it — though several measures improved between 2021 and 2023⁷.

Place: where you live matters.

In 2024, state rates ranged from 5.7 per 100,000 in Washington, DC to 29.7 in Alaska¹. The highest-rate states are Alaska and the Mountain West — Alaska, Montana, Wyoming, Idaho and New Mexico — and my home state of Utah, at 20.7, sits well above the national 14.7³. Rural Americans die by suicide at about 20 per 100,000, compared with about 13 in urban areas⁴.

Globally, 73% of suicides happen in low- and middle-income countries, where access to mental health care and control of highly lethal pesticides are major levers⁵.

Conditions: real risk, not destiny.

Psychological autopsy studies — interviews with families after a death — have long reported that around 90% of people who die by suicide had a diagnosable mental disorder¹². That figure is debated: the method relies on hindsight, and it can make suicide seem like purely a medical event when social crises matter too¹³. What is clear is that some conditions raise risk sharply (see the Conditions tab above): borderline personality disorder (about 45 times the general population), anorexia nervosa (31×), depression (20×), bipolar disorder (17×), opioid use disorder (14×) and schizophrenia (13×)⁶.

In anorexia nervosa — the field I've spent much of my career in — about one in five deaths is by suicide¹⁴. And the weeks after leaving a psychiatric hospital are among the highest-risk periods known, which is why follow-up in those first days matters so much¹⁵.

The flip side is just as important: the large majority of people living with each of these conditions will never die by suicide, and every one of them is treatable.

LGBTQ+ people: the risk is in the environment.

The numbers are stark. In 2023, 41% of LGBTQ+ high school students seriously considered suicide, compared with 13% of their cisgender, heterosexual classmates; 65% reported persistent sadness or hopelessness, vs 31%⁷. In the Trevor Project's 2025 national survey of more than 16,000 LGBTQ+ young people, 36% had seriously considered suicide in the past year and 10% had attempted — rising to 45% and 15% among 13-to-17-year-olds⁹.

For decades, researchers have asked why. The best-supported answer is minority stress^8,16: being LGBTQ+ is not a risk factor in itself, but living with rejection, discrimination, harassment, the need to hide, and the expectation of all of these creates chronic stress — which drives depression, anxiety, substance use and suicidal thinking. In Joiner's language from the first post in this series, it is a direct assault on belonging.

Several kinds of evidence point the same way:

• Supportive environments, fewer attempts. Among nearly 32,000 Oregon 11th graders, LGB youth attempted suicide at more than five times the rate of heterosexual peers (21.5% vs 4.2%) — and those living in less supportive counties and schools were 20% more likely to have attempted than those in supportive ones¹⁷.
• A natural experiment. As states legalized same-sex marriage, adolescent suicide attempts fell by 7% overall and by 14% among sexual-minority students, compared with states that hadn't yet changed their laws¹⁸.
• Families matter enormously. Young adults who experienced high levels of family rejection were 8.4 times more likely to report a suicide attempt than those with little or no rejection¹⁹.
• One accepting adult. LGBTQ+ young people with at least one accepting adult were 40% less likely to report a past-year suicide attempt²⁰.
• Something as simple as a name. Among transgender youth, being able to use their chosen name at home, school, work and with friends was linked to 71% fewer symptoms of severe depression, a 34% decrease in suicidal thoughts, and a 65% decrease in attempts²¹.

And a harder road to help

The same forces that raise risk also get in the way of care. In 2025, 84% of LGBTQ+ young people wanted mental health care, but 44% of those who wanted it couldn't get it. Top reasons: cost (46%), being afraid to talk about their mental health (39%), fear of not being taken seriously (37%), needing a parent's permission (33%), and fear of police involvement or hospitalization (29%)⁹. Ninety percent said recent anti-LGBTQ+ laws and policy debates had caused them stress or anxiety⁹.

A few honest caveats. Much of this evidence is cross-sectional, and the Trevor Project surveys are large but not random samples. We also know less about deaths than about thoughts and attempts, because sexual orientation and gender identity aren't recorded on death certificates²². But the population-based surveys, the natural experiments and the clinical studies all converge: the gap is driven largely by how LGBTQ+ people are treated, which means it can be closed.

What protects — for everyone.

Across every group in this post, the same protections keep appearing:

• Connection. Feeling close to people at school, at home, at work. Only 55% of US high school students felt close to people at their school in 2023 — and LGBTQ+ students were less likely to⁷.
• Access to care — insurance, broadband for telehealth, clinicians who are nearby and trusted⁴.
• Safer environments during a crisis, especially safe storage of firearms and medications^10,24.
• Acceptance. For LGBTQ+ people, the evidence is unusually clear that acceptance from family, school and community saves lives^17,20.
• A plan. If you or someone you love is in one of the groups above, the safety plan builder in the previous post takes about fifteen minutes.

Risk is not destiny. Almost every number in this post has moved — up and down — within my career. The ones that went down, went down because people did something.

Written following the Recommendations for Reporting on Suicide²⁵ and the Action Alliance Framework for Successful Messaging²⁶. Data are the latest available as of September 2026. Educational only — not medical advice.