ADHD linked to 70% higher death rates in a review of 19 studies of nearly 16 million people

Most talk about ADHD is about school reports, missed deadlines and lost keys. A quieter strand of research has been asking a blunter question: whether people with the condition die younger, and of what.

A review published on October 19 in JAMA Network Open pulls that research together. Lucy Corrigan of the University of Liverpool and colleagues at University College London, with Joshua Stott and Amber John as senior authors, reviewed 19 studies covering close to 16 million study participants in North America, Europe and Asia. Pooling the nine that reported overall deaths, they found the rate of death from any cause ran 70% higher among people diagnosed with ADHD than among people without the diagnosis. Suicide and injuries accounted for the largest gaps, but deaths from illness were raised as well, and in a smaller set of studies, people taking ADHD medication had lower death rates than those who were not. Every study in the review was observational, so none of this shows that ADHD, or its treatment, is what changed the numbers.

What the new ADHD mortality review adds to earlier work

ADHD, short for attention-deficit/hyperactivity disorder, is a neurodevelopmental condition characterized by persistent inattention, hyperactivity or impulsivity that gets in the way of daily life. It was long thought of as a childhood problem, and the authors point out that it is now recognized as frequently persisting across the lifespan, into old age for many people. That shift is part of why the team turned to mortality: once ADHD is understood as lifelong, how long people with it live becomes a question for doctors and not only for teachers.

Studies of single countries had already pointed to a gap. In 2015, a team at Aarhus University followed 1.92 million Danes from their first birthday. For every 10,000 years of life the study tracked, it counted 5.85 deaths among individuals with ADHD compared with 2.21 among those without. Accidents were the most common cause of death in that cohort. In 2025, Stott’s own group used UK primary care records to estimate that adults with diagnosed ADHD were living shorter lives than the general population, with an apparent reduction in life expectancy of 6.78 years for men and 8.64 years for women.

The one earlier attempt to pool the evidence, led by Ferrán Catalá-López and published in 2022, combined ADHD with autism. It found higher death rates with ADHD overall, mostly from unnatural causes, but deaths from natural causes were not significantly increased, meaning the small rise it saw could have been chance, and it rated its confidence in the ADHD estimates as low. Corrigan’s team argues that lumping the two conditions together could hide patterns that belong to one of them, and they also had several years of newer studies to work with. To the authors’ knowledge, nobody had pooled death rates for ADHD on its own before, and their analysis also gives separate estimates for suicide and injury deaths and looks at medication, which Catalá-López’s team did not assess.

How 19 studies covering 16 million people were combined

The team searched six research databases up to July 2025 and screened more than 9,000 titles and abstracts. Nineteen studies made the cut, published between 2012 and 2025 and ranging from 839 people to nearly 3 million. Four studies were conducted in the US, three each came from Canada, Sweden and Taiwan, and two each from the UK, Finland and Denmark. Most relied on health records and identified ADHD through diagnostic codes. One used self-reported ADHD and one used a genetic score that estimates a person’s inherited tendency toward ADHD. Some of them, according to the University College London release, followed people for over 20 years.

Twelve of the studies reported results in a form that could be combined, and the authors pooled them with a method that allows each study to have its own true effect, a sensible choice when the data come from different countries and medical systems. They rated the quality of each study on a standard checklist that asks how the groups were chosen, how comparable they were and how deaths were recorded. Seventeen were judged good and two fair. Not every study reported every cause of death, so each pooled figure rests on a different subset: nine studies for deaths from any cause, five for suicide, four for injuries, three for natural causes and three for medication. Where possible they used each study’s fully adjusted estimate, the version that accounts for the most other differences between people, although which differences were counted, income among them, varied from study to study.

That design has a ceiling. Pooling observational studies can show how consistently ADHD goes with higher death rates, and roughly how large the gap is on average. It cannot show why, and it inherits whatever the original studies did or did not measure.

Suicide, injuries and illness: where the extra deaths in ADHD came from

The overall gap was large and uneven. The 70% figure pooled nine studies, and it came with a wide range, from about a third higher to more than double, and the individual studies disagreed sharply about the size of the gap. They did not disagree about its direction. In the authors’ words, “all studies consistently showed higher mortality in the ADHD groups,” which suggests the spread reflects differences in magnitude across countries and study designs rather than doubt about whether the link exists.

Broken down by cause, across five studies the rate of suicide was 90% higher in people with ADHD. Deaths from injuries, whether intentional or accidental, were 55% higher across four. The researchers link the suicide figure to the higher rates of depression, anxiety and substance use that often come with ADHD, and to the societal exclusion, disadvantage, and stigma that people with the condition are more likely to face. For injuries they point to differences in attention and impulse control, and to evidence that ADHD is associated with an increased risk of involvement in serious transport accidents.

The smallest number may be the most telling. Deaths from illness, such as heart disease, are classed as natural causes, and the death rate from natural causes was 27% higher in people with ADHD. Only three studies fed that figure, but their results lined up almost exactly, the only cause of death on which the studies agreed so closely. Where the 2022 analysis found no clear increase, the new one found a modest association with natural-cause mortality, possibly because of the newer studies it included.

Why would ADHD go with more deaths from illness? The authors offer several routes, none of which their analysis was designed to test. They include health risk behaviors more common in people with ADHD, such as smoking and drinking, difficulty keeping up with treatment plans, and challenges with engagement and access to health care services. Higher rates of heart disease and metabolic disorders, and the effects of poverty and lower education, may add to it. Stott’s 2025 UK study came to a similar view. Its authors judged the lost years unlikely to be because of ADHD itself, and blamed smoking, unmet support and other things that can be changed. Stott described the new results in the same terms: “We found that people with ADHD face an increased risk not only of death by suicide and injury, but also of deaths by natural causes, adding to evidence that health inequalities faced by people with ADHD are broad and wide-ranging.”

Is ADHD medication linked to fewer deaths?

Prescriptions for ADHD have been climbing. In England, the number of people getting ADHD drugs has more than doubled during the past 6 years, Corrigan’s team writes, reaching about 278,000 people in 2023-24.

Three of the studies compared people with ADHD who took medication with those who did not, and the treated group had a 25% lower death rate. Corrigan was careful about what that can carry. “Our findings provide early evidence to suggest that medication may play an important role in supporting the health of people with ADHD, but this is based on a small number of studies and further research is needed to clarify this link,” Corrigan said.

The obvious problem is that people who start medication are not a random sample. They have been diagnosed, they have a prescriber, and they may differ from untreated people in income, the severity of their symptoms and other illnesses. The authors flag both residual confounding and confounding by indication: the reasons a person gets a treatment can also shape how they fare. They write that the result should not be interpreted as evidence of a causal protective effect of the medicine.

Other studies have tried harder to get around that problem, and they point the same way. A Swedish study of 148,578 people newly diagnosed with ADHD was set up to mimic a trial, comparing those who started medication within three months of diagnosis with those who did not. Over two years, 39.1 per 10,000 died in the group that started treatment and 48.1 per 10,000 in the group that did not. The difference came mostly from deaths due to injuries, poisonings and suicide; deaths from illness were not clearly different. In Quebec, a study of 217,192 young people found that periods on ADHD medication were associated with reduced all-cause mortality, meaning deaths from any cause. Those periods also went with fewer accidental injuries that ended in an emergency room or a hospital stay, although the Quebec authors say leftover confounding could not be ruled out. A 2026 review by Yasmin Rahali and colleagues at Cardiff University concluded that the evidence is clearer for stimulant than for non-stimulant medication.

What the ADHD mortality figures leave open

The size of the gap is the least certain part of the result. The included studies were cross-cultural, and countries differ in who gets diagnosed with ADHD, what care they receive and how society treats them. The authors say their pooled figures should be read as average associations across diverse populations rather than as a risk that applies everywhere. Rahali’s review found reported risks that varied from 1.07 to 4.44 times the rate in people without ADHD, with higher estimates in studies that followed people to older ages, in women, in people diagnosed later in life and in those with other psychiatric conditions.

Corrigan’s team could not test those differences itself. Too few studies reported results by sex, age or other traits, so reliable subgroup analyses or meta-regression were not possible, meaning there was no way to work out statistically which study features explained the spread. Most of the studies included more men than women, which may hide differences for women, who are more often diagnosed late or missed.

Diagnosis is a limit of its own. Almost all the studies identified ADHD through medical records, and most adults with ADHD in the UK have never been diagnosed. Stott’s 2025 study estimated that about 1 in 9 of all adults with ADHD had a diagnosis in UK primary care. People who reach a diagnosis may be more affected, or more connected to health care, than those who never do, so the figures describe diagnosed ADHD and not the condition as a whole.

The paper also gives no absolute death rates. A 70% higher rate sounds alarming, but most people with ADHD, like most people without it, do not die young, as the earlier Danish study’s count of fewer than six deaths for every 10,000 years of life shows. The medication result rests on three studies, which is too few for the usual checks on whether studies that found no effect went unpublished.

What the ADHD death-rate results mean for care

The authors frame the results as an argument for earlier and better care. “ADHD is not by definition a life-limiting condition as the right support and treatment can help people with ADHD to live long and healthy lives,” Stott said. Amber John, the other senior author, called “equitable access to evidence-based assessment and treatment” vital. The paper’s discussion calls for “long-term, comprehensive care for individuals with ADHD, with particular attention to identifying and addressing modifiable factors” that may add to the risk.

Treatment guidance already exists in some countries. In the UK, the National Institute for Health and Care Excellence (NICE) guideline advises clinicians to offer medication to adults with ADHD whose symptoms still cause significant problems after changes to their surroundings have been tried. For people who have made an informed choice not to take medication, or who cannot tolerate it, it suggests non-drug treatment. The new paper does not change that guidance, and its medication result is too thin to settle whether treatment adds years to anyone’s life. Whether a particular treatment suits a particular person is assessed by their doctor.

Across 19 observational studies the gap for diagnosed ADHD pointed the same way every time, and it ran through illness as well as through injury and suicide, though how much treatment can narrow it is still unknown.