What Mental Illness Has the Highest Mortality Rate?

Anorexia nervosa and substance use disorders consistently top the list. A large meta-review of existing research found that these two categories carried the highest all-cause mortality ratios of any mental illness, outpacing schizophrenia, bipolar disorder, and major depression.¹ But the answer shifts depending on what you mean by “mortality rate,” and the distinction matters more than most people realize.

Anorexia Nervosa Stands Apart

Anorexia nervosa has the highest standardized mortality ratio of any single psychiatric diagnosis. A standardized mortality ratio compares how many people with a condition die versus how many deaths you would expect in a matched group from the general population. For anorexia nervosa, an updated meta-analysis covering studies from 2010 to 2024 found that figure to be roughly five times the expected rate.² An earlier meta-analysis of 36 studies reported a similar figure, about six times the expected rate, over an average follow-up of about 14 years.³

The causes of death in anorexia are not limited to starvation. Cardiac complications, organ failure, electrolyte imbalances, and suicide all contribute. One longitudinal study found that the risk of premature death was highest in the first decade after diagnosis, with the annual death rate dropping substantially after that initial window.⁴ That pattern suggests that the most medically dangerous phase coincides with the period when the disorder is most severe, and that sustained recovery substantially lowers the risk.

Other eating disorders carry elevated mortality too, but not nearly as high. Bulimia nervosa shows a mortality ratio roughly two to two-and-a-half times the expected rate, and binge eating disorder is lower still.² Anorexia’s extreme mortality stands out even within its own diagnostic family.

Substance Use Disorders and Years of Life Lost

When researchers look not just at how many people die relative to expectations, but at how many years of life are lost, substance use disorders move to the front of the line. A systematic review and meta-analysis found that people with substance use disorders had the shortest average life expectancy of any mental health category, around 57 years. They also showed the greatest number of years of potential life lost, roughly 20 years, compared with about 17 for eating disorders and 15 for schizophrenia-spectrum disorders.⁵

The reason substance use disorders claim so many years has to do with timing. Addiction often begins in adolescence or early adulthood, and the medical consequences pile up over decades: liver disease, overdose, infection, accidents, and cardiovascular damage. A nationwide cohort study found that when all-cause mortality was ranked across mental disorders, the highest risk fell on intellectual disability, followed by schizophrenia-spectrum disorders, and then alcohol use disorder.⁶ So depending on the metric, substance use disorders compete with or surpass anorexia nervosa for the top spot.

Schizophrenia and the Quiet Killers

Schizophrenia shortens life by roughly 15 to 20 years in most countries. The standardized mortality rate hovers around two-and-a-half times the general population.⁷ While suicide is a real and significant risk, it is not the main reason people with schizophrenia die earlier than expected. Natural causes, especially heart disease, account for the bigger share of that gap.⁸

Cardiovascular disease accounts for roughly 40 to 50 percent of deaths in most studies of people with schizophrenia.⁷ Several factors feed into that. People with schizophrenia smoke at very high rates, tend to have poor diets, get less physical activity, and often have limited access to preventive medical care. On top of lifestyle, the medications used to treat the illness can compound the problem. Second-generation antipsychotics in particular are associated with weight gain and metabolic disruptions like elevated blood sugar and abnormal cholesterol, which further increase cardiovascular risk.⁹

There is an uncomfortable paradox here. A large registry study found that consistent antipsychotic use was the single strongest factor reducing all-cause mortality in people with schizophrenia, cutting the risk roughly in half. But having cardiovascular disease more than doubled it, and history of substance abuse raised it further.¹⁰ Treating the psychiatric illness keeps people alive; managing its medical side effects is a separate and equally urgent challenge.

Bipolar Disorder

Bipolar disorder’s mortality story is dominated by two threats: suicide and cardiovascular disease. A large meta-analysis found that suicide risk was roughly 12 times higher than in the general population, the highest cause-specific mortality ratio for this diagnosis. Risk of death from respiratory disease was about three times higher, and cardiovascular death was nearly double.¹¹

A Swedish cohort study added detail to these patterns. Both men and women with bipolar disorder had elevated mortality from heart disease, diabetes, and chronic lung disease. More surprisingly, risk of death from influenza or pneumonia was nearly four times higher in both sexes. Suicide risk was eight- to ten-fold higher depending on sex, even after adjusting for demographic factors.¹² Bipolar disorder does not have the extreme mortality ratio of anorexia, but it affects far more people, so its total burden on life-years lost across a population is substantial. The same systematic review that ranked years of potential life lost placed it in a cluster with mood disorders generally, which trailed only substance use disorders and eating disorders.

Depression and the Heart

Major depression is the most common mental illness worldwide, and its link to mortality runs mainly through the heart. A systematic review and meta-analysis covering over 1.9 million people found that depression increased the risk of all-cause mortality by about 43 percent. Cardiovascular mortality specifically was elevated by a similar margin. The largest effect appeared for heart failure, where the risk of dying was roughly tripled.¹³

The relationship between depression severity and cardiac death appears to follow a dose-response pattern. A study of U.S. adults found that moderate to severe depressive symptoms were associated with roughly double the risk of dying from ischemic heart disease compared with people without depression.¹⁴ Mild depression elevated cardiovascular mortality too, but to a lesser degree. Depression alters stress hormones, inflammation, sleep, and motivation to exercise or take medications, all of which feed into heart disease over time.

Because depression is so common, these relatively modest individual risk increases translate into enormous population-level effects. Most people would guess that depression’s main mortality threat is suicide, and suicide is certainly part of the picture. But the cardiovascular burden likely accounts for more total deaths.

Borderline Personality Disorder and Suicide

Borderline personality disorder (BPD) occupies an unusual spot in the mortality landscape. Its overall standardized mortality ratio is not as extreme as anorexia nervosa’s or schizophrenia’s, but its suicide rate is strikingly high. Up to one in ten people with BPD die by suicide.¹⁵ A 24-year prospective study found that about 6 percent of borderline patients died by suicide over the follow-up period, and an additional 14 percent died from other causes.¹⁶

A notable finding from that study was that the vast majority of BPD patients who died, whether by suicide or other causes, had not achieved clinical recovery before death. Prior psychiatric hospitalizations were the strongest predictor of both suicide and premature death from other causes. A meta-analysis confirmed that suicide was more common in the earlier phases of illness than in the chronic stage.¹⁷ Interestingly, a large nationwide cohort study found that personality disorders as a group actually carried the highest suicide risk of any diagnostic category, ahead of bipolar disorder and schizophrenia.⁶ That finding challenges the common assumption that psychotic disorders are always the highest-risk category for self-inflicted death.

Why Comorbidity Changes the Numbers Dramatically

Ranking individual diagnoses can be misleading because mental illnesses rarely travel alone. When someone with schizophrenia, bipolar disorder, or depression also has a substance use disorder, the mortality risk jumps sharply. A large Danish study quantified this clearly. Among people with schizophrenia, those who also had a substance use disorder had a standardized mortality ratio of about 8.5, compared with 3.6 for those without. In bipolar disorder, the ratio jumped from about 2.9 to 6.5 with substance use comorbidity. In depression, it went from about 1.9 to 6.1.¹⁸

Personality disorders show a similar pattern. A study examining total and cause-specific mortality in personality disorder patients found that comorbid substance use was associated with nearly four times the risk of death compared with a personality disorder alone. Young women with this combination had the highest relative risk.¹⁹ Deaths from poisoning and suicide drove much of the excess. The practical takeaway is that any mental illness combined with active addiction creates a risk profile far worse than either condition alone.

Diagnostic Overshadowing and Why Physical Symptoms Get Missed

One factor that amplifies mortality across all mental illnesses is something clinicians call diagnostic overshadowing. It happens when a health-care provider attributes a patient’s physical symptoms to their known psychiatric condition and misses a separate medical problem. Someone with schizophrenia who reports chest pain might have it written off as anxiety. A person with BPD who complains of abdominal pain might be assumed to be seeking attention. This pattern leads to delayed diagnoses and worsened outcomes.²⁰

A mixed-methods systematic review confirmed that diagnostic overshadowing exacerbates health inequities for people with mental disorders and compromises their access to appropriate care for co-occurring physical illnesses.²¹ At one community hospital, a quality-improvement project revealed that physical health concerns were frequently minimized or misattributed to psychiatric illness in the emergency department, with patients initially cleared for psychiatric admission who later deteriorated medically.²² The problem is systemic, not anecdotal, and it likely contributes to the cardiovascular and metabolic deaths that dominate the mortality statistics for schizophrenia and bipolar disorder.

The Role of Lifestyle and Biological Aging

People with severe mental illness tend to age biologically faster than their peers. One study measured telomere length, a marker associated with cellular aging, in patients with severe mental disorders. Those who had unhealthy lifestyles had substantially shorter telomeres than those with healthier habits, a difference that corresponded to roughly six additional years of biological age.²³ The finding held after adjusting for diagnosis, medication, trauma history, and education, suggesting that the lifestyle factors themselves, not just the illness, are driving the accelerated aging.

Those lifestyle factors are familiar: smoking, poor diet, sedentary habits, heavy alcohol use. But framing them as “choices” misses the reality. Psychotic disorders, severe depression, and active substance use all impair motivation, energy, and the ability to plan ahead. Many psychiatric medications increase appetite and cause weight gain. Poverty, unstable housing, and social isolation, which are disproportionately common among people with serious mental illness, further limit access to healthy food and safe places to exercise. The mortality gap between people with mental illness and the general population is partly a health-care access gap and partly a biology gap, and the two reinforce each other.

The Mortality Gap Is Getting Wider, Not Narrower

Perhaps the most unsettling finding in this field is that the excess mortality associated with mental illness has been increasing over time, not decreasing. A systematic review and meta-analysis found that more recent studies reported higher mortality rates than older ones. In other words, as life expectancy in the general population has risen, people with mental disorders have not shared in those gains at the same rate.²⁴ This trend was first documented for schizophrenia, but the same review suggested it applies across mood disorders, anxiety disorders, and other categories as well.

The reasons likely involve a combination of factors. Medical advances in cardiovascular care, cancer screening, and diabetes management have driven down mortality in the general population. People with serious mental illness, who have less consistent access to primary care and face the diagnostic overshadowing described above, benefit less from those advances. Meanwhile, second-generation antipsychotic medications, which became widespread in the 1990s and 2000s, carry metabolic side effects that may have partially offset gains from better psychiatric treatment. The overall pattern points to a systemic failure to integrate physical and mental health care, rather than any single cause.

Gender and Suicide Risk

The relationship between mental illness and mortality plays out differently for men and women. A large European cross-sectional study found that women were consistently more likely to experience internalizing disorders like depression and anxiety, while men had higher rates of externalizing disorders like substance abuse. Lifetime suicide attempts were more common in women, but the gender difference varied by country, and men died by suicide at higher rates overall.²⁵

Within specific diagnoses, gender matters too. The 24-year BPD follow-up study found that male sex was one of the strongest predictors of premature non-suicide death, tripling the risk.¹⁶ In bipolar disorder, the Swedish cohort found that women had somewhat higher relative risks for certain specific causes of death, including stroke and colon cancer, while men had higher absolute rates of suicide.¹² These patterns mean that broad mortality statistics for a diagnosis can mask important variation. The risk profile for a young woman with BPD and active substance use looks very different from that of an older man with stable bipolar disorder, even though both fall under the umbrella of “mental illness mortality.”

When Dementia Enters the Picture

Dementia is often categorized as a neurological condition rather than a psychiatric one, but in clinical practice the line gets blurry. Many patients with dementia receive care in psychiatric settings, and dementia carries extreme mortality rates. A study of older patients with dementia in psychiatric care found that roughly 16 percent died within one year, 44 percent within three years, and nearly two-thirds within five years. Male sex, higher age, more somatic illnesses, and inpatient psychiatric care all independently predicted earlier death.²⁶

Dementia does not usually appear in the rankings alongside anorexia or schizophrenia because of how diagnostic categories are drawn. It is classified under neurological disease in most systems. But if you include it, its raw mortality rate dwarfs nearly everything else on the list, though much of that reflects the fact that dementia primarily affects the elderly, making direct comparison with conditions that strike younger people somewhat misleading. The years-of-life-lost framework handles this better: an illness that kills at 80 takes fewer life-years than one that kills at 30, even if the raw death rate is higher.