ADHD linked to later dementia and mild cognitive impairment in the records and genes of 187,341 US adults

Attention-deficit/hyperactivity disorder (ADHD) is usually discussed as a condition of school and working life. A growing set of studies asks what happens to people who have it when they grow old, and in particular whether they are more likely to develop dementia.

A study published on October 7 in JAMA Psychiatry tests that question in a large American sample, using both medical records and DNA from the same people. Among more than 187,000 adults aged 50 and over, those with an ADHD diagnosis on record were about twice as likely to go on to a diagnosis of dementia or mild cognitive impairment, 6.7% against 3.4%. A genetic score for ADHD, which does not depend on anyone having been diagnosed, pointed the same way, though far more weakly. The authors say plainly that this does not prove ADHD causes dementia.

Why ADHD and dementia are being studied together

ADHD is marked by difficulty sustaining attention, restlessness and acting on impulse. It is classed as a neurodevelopmental disorder, meaning one that arises as the brain develops.

Dementia sits at the other end of life. One research group describes it as a syndrome characterized by dysfunction in daily life due to cognitive impairment. In plain terms, that is a loss of memory and thinking severe enough to get in the way of everyday tasks. Mild cognitive impairment (MCI) is a lesser degree of the same kind of problem. A Swedish team defines it by the presence of impairment in one or more cognitive domains without affecting functional independence.

There are reasons to suspect a connection. Adults with ADHD have more of several things already tied to dementia. An Israeli team lists depression, high blood pressure in midlife, smoking, type 2 diabetes, and low levels of education and physical activity as both consequences of ADHD and risk factors for dementia.

The symptoms overlap as well. The same team notes that adult ADHD appears to mimic some cognitive symptoms of dementia, memory lapses among them, which can make the two hard to tell apart in an older person.

The evidence so far has been limited and hard to combine. A systematic review from 2023 said ADHD has been understudied in relation to dementia risk. Its authors searched five databases, and seven studies (five cohort and two case-control studies) ultimately met inclusion criteria. Those seven differed so much in their methods that their results could not be pooled into one number. The review’s verdict was that the magnitude of the direct effect of ADHD on neurodegeneration is yet to be determined. Neurodegeneration means the gradual loss of nerve cells that underlies diseases such as Alzheimer’s.

How the study tracked ADHD and cognitive impairment

The work was led from the University of Pittsburgh School of Medicine, with Guilherme Bauer-Negrini and Douglas Teixeira Leffa as first authors and Tharick Pascoal as senior author. This cohort study used data from the US All of Us Research Program, a project of the National Institutes of Health whose participants share their medical records and provide DNA.

The analysis took in 187,341 participants aged 50 years and older. Their average age was 56, and 57% were women. Their records covered about 10 years of follow-up on average.

The team looked at two things in each person. The first was whether an ADHD diagnosis appeared in the medical record. The second was a polygenic risk score (PRS) for ADHD. A polygenic risk score adds up the small effects of many common gene variants into a single number. It reflects how strongly a person’s variants are tied to ADHD, regardless of whether they were ever formally diagnosed.

The outcome was what the paper calls cognitive impairment, defined as MCI or dementia, appearing in a person’s record for the first time.

Using both measures in the same people is the point of the design. A diagnosis depends on who seeks help and who gets assessed. The paper accepts that, for the earlier research, concerns regarding diagnostic ascertainment and residual confounding remain. Diagnostic ascertainment means how reliably cases are picked up. Residual confounding means differences between the groups that the statistics have not fully removed. A genetic score avoids the first problem, because it is fixed from conception and can be worked out for everyone.

What the records showed about ADHD and cognitive impairment

A diagnosis was uncommon. Of the 187,341 participants, 3,026, or 1.6%, had ADHD on record.

Among participants with ADHD, 203 (6.7%) developed cognitive impairment, compared with 6259 (3.4%) among those without ADHD. Put another way, about 93 in every 100 people with an ADHD diagnosis did not develop either condition during the study.

The raw gap is roughly twofold. The study’s statistical model, which follows each person over time and adjusts for differences between the groups, put it higher. An adjusted figure can exceed a raw one when the groups differ in age or in how long they were followed. In the university’s summary, researchers found that people with ADHD had more than four times the risk of developing cognitive impairment over time. In the paper’s wording, ADHD was associated with a more than 4-fold higher rate. The estimate was 4.6 times, with a range from 3.99 to 5.29.

That figure did not hold at full strength. Adjustments for comorbidities, educational attainment, and health care utilization attenuated but did not eliminate the association. Comorbidities are other illnesses a person has at the same time. Health care utilization is how much care a person uses. Attenuated means weakened. So part of the link is accounted for by people with ADHD having more other health problems, different schooling and more contact with doctors, and part of it is not.

The researchers also tested an obvious alternative, that some ADHD diagnoses were really the first signs of dementia. One possibility is that early, undetected dementia symptoms are sometimes mistaken for ADHD before a formal cognitive diagnosis is made. To check, they counted only cases in which a long gap separated the two diagnoses. The link held even when the team required at least eight years to pass between an ADHD diagnosis and the onset of cognitive impairment, making that explanation less likely to account for the full effect.

What the genetic score for ADHD added

The second analysis set diagnoses aside and ranked everyone by genetic score. People with the highest genetic risk had a 35% greater chance of developing dementia than those with the lowest risk. The paper’s own figure is for cognitive impairment, meaning dementia or MCI. The comparison is between the top fifth of scores and the bottom fifth, and the estimate ranged from 25% to 46% higher.

Pascoal, an associate professor of psychiatry and neurology at Pittsburgh, gave the reasoning in the university’s statement. “This doesn’t prove that ADHD causes dementia, but seeing the same pattern show up independently in medical records and in people’s genes gives us greater confidence that the association is not simply an artifact of how ADHD gets diagnosed,” Pascoal said. “Two different kinds of evidence are pointing in the same direction.”

The two results are far apart in size, though. A recorded diagnosis went with a rate more than four times as high, and a high genetic score with a rate about a third higher. The figures measure different things and cannot be set side by side, since a score in the top fifth is a long way short of a diagnosis. Even so, the genetic signal is much the smaller. It also says nothing about the route from genes to dementia, which could run through ADHD itself, through the conditions that travel with it, or directly through the aging brain.

An earlier and much smaller study by the same group hints at that last route. It worked out the score in 212 older adults with no ADHD diagnosis and no cognitive problems, then followed them with brain scans, spinal fluid tests and memory tests. Higher ADHD-PRS was associated with greater cognitive decline over 6 years. The authors’ reading was that genetic liability for ADHD increases susceptibility to the harmful effects of amyloid, the protein that builds up in the brain in Alzheimer’s disease.

A Swedish study came at heredity another way. It linked more than 2 million people born between 1980 and 2001 to their parents, grandparents, uncles and aunts, and looked for Alzheimer’s disease (AD) among those relatives. Parents of individuals with ADHD had an increased risk of AD, by 55%. The associations attenuated with decreasing genetic relatedness, suggesting shared familial risk between ADHD and AD.

How this fits with earlier ADHD and dementia studies

The Pittsburgh team presents its result as confirmation. “We’re replicating a pattern that’s been observed in other countries, now in a large U.S. sample, and adding genetic evidence on top of it,” Leffa said.

One of those countries is Israel. Researchers there followed 109,218 members of a nonprofit Israeli health maintenance organization born between 1933 and 1952, from 2003 to early 2020. Dementia occurred among 96 of 730 participants (13.2%) with adult ADHD and in 7.0% of everyone else. After the team adjusted for 18 other factors, the rate was 2.77 times as high.

The other is Sweden. A register study there covered 3,591,689 individuals born between 1932 and 1963, identified from Swedish population-based registers. Adjusted only for sex and year of birth, dementia was 2.92 times as common in people with ADHD, and MCI was 6.21 times as common.

Then the Swedish researchers took other mental health conditions into account, and the picture changed. Additional adjustment for psychiatric disorders (depression, anxiety, substance use disorder, and bipolar disorder) substantially attenuated the associations. The figures fell to 1.62 times for dementia and 2.54 times for MCI. High blood pressure, type 2 diabetes, obesity, sleep disorders, head injuries and education, by contrast, had a limited impact on the association.

That pattern helps in reading the American result, where further adjustments also weakened the association without removing it.

The Israeli study added a detail about treatment. There was no clear increase in the risk of dementia associated with adult ADHD among those who received psychostimulant medication. Psychostimulants are the main class of ADHD medicines. The authors did not treat this as proof that the drugs protect the brain. Only 22.3% of the adults with ADHD in their data had been prescribed them, and the association, they wrote, requires further study before causal inference is plausible.

Living with both ADHD and cognitive impairment

The study also looked at people after cognitive impairment had been diagnosed, comparing those who had ADHD as well with those who did not.

Those with both conditions reported worse mental health, lower quality of life and less social satisfaction than those with dementia alone. They also used about 20% more health care overall in the year following diagnosis.

The paper’s conclusion is that an ADHD diagnosis was associated with worse health outcomes and greater health care utilization in this group. It treats that as an argument for closer follow-up, describing the results as supporting the need for targeted monitoring and integrated care in this population.

The same caution applies as to the main result. People with ADHD may have reported poorer mental health and less satisfaction with their social lives before any cognitive decline began.

What the ADHD study cannot show

Cause. This is an observational study. It shows that ADHD and later cognitive impairment go together in these records. It does not show that one leads to the other, or that treating ADHD would change a person’s chance of dementia.

Who has a diagnosis. Only 1.6% of participants had ADHD on record. Adults who are diagnosed in midlife or later may differ from people with similar traits who never are. They may have more severe symptoms, more mental health conditions or simply more contact with doctors. Record-based studies of this kind, the Pittsburgh group wrote in 2022, rely primarily on electronic health records and can present biased estimates of disease prevalence. The Israeli authors made a similar point about their own work, writing that the rate of adult ADHD may be underascertained in the present study. Underascertained means undercounted.

Which came first. The eight-year check makes it less likely that early dementia was being mislabeled as ADHD. It does not rule it out. Reverse causation challenges the association between adult ADHD and dementia because adult ADHD is accompanied by cognitive impairments that resemble dementia, as the Israeli team put it, and dementia develops silently over many years before it is diagnosed.

The size of the effect. The fourfold figure is the study’s main estimate. Further adjustment for other illnesses, education and health care use weakened it, so the link that remains once those are taken into account is smaller than that figure.

The age of the participants. The average age was 56, and only 3.4% of people without ADHD developed cognitive impairment during the study. That is a sign most participants had not yet reached the ages at which dementia is common.

Who took part. The participants had enrolled in a research program and had both medical records and DNA available. They are not a random sample of older Americans.

Medication. The study did not test whether ADHD medicines raise or lower the risk. The team names it among the open questions: whether ADHD medications, such as stimulants, might lower dementia risk over time. Do not stop or change a prescribed medicine without talking to your doctor.

Whether forgetfulness or poor concentration in an older adult reflects ADHD, early cognitive impairment or something else is assessed by their doctor.

The study finds that ADHD and later cognitive impairment travel together in American medical records and, more faintly, in genes, and it leaves open why.