Genes linked to ADHD went with more memory problems in later life, in 7,761 older adults

ADHD (attention-deficit/hyperactivity disorder) is usually thought of as a condition of childhood: trouble keeping attention, restlessness, acting on impulse. But it often continues into adult life, and several studies have found that adults with an ADHD diagnosis are more likely to develop memory problems and dementia later. Those studies are hard to interpret: people with ADHD differ in many ways, and few older adults were ever diagnosed, because ADHD was rarely recognised when they were young.

A new study took a different route. Instead of diagnoses, it looked at people's DNA, and found that older adults carrying more of the gene variants linked to ADHD were somewhat more likely to develop mild cognitive impairment (memory or thinking problems that are noticeable but don't yet stop a person from managing daily life) or Alzheimer's dementia over the next ten years.

A score built from many genes

ADHD runs strongly in families. No single gene causes it; instead, thousands of common gene variants each nudge the risk up or down a tiny bit. A polygenic risk score adds up these small effects into one number for each person. It does not say whether someone has ADHD; it estimates how much inherited tendency towards it their DNA carries. Its advantage for research is that DNA is fixed from birth and doesn't depend on whether a doctor ever made a diagnosis.

What they did

The team, led from the University of Pittsburgh, used data from the Alzheimer's Disease Sequencing Project, which combines long-running US studies of aging and memory that have read the full DNA of their participants. The analysis included 7,761 people from eight of these studies who had no cognitive problems at the start (average age 71, about two thirds women). They were examined repeatedly, almost five times each on average, and diagnoses were made by clinical teams using standard criteria rather than taken from health records.

The researchers calculated each person's ADHD gene score and followed them for up to ten years. They took into account age, sex, genetic ancestry and study, and in further analyses education, a score of heart and blood vessel risk factors, and APOE4, the best-known gene variant raising Alzheimer's risk. As a check, they repeated everything with gene scores for autism, bipolar disorder, depression and schizophrenia.

What they found

• More new cases with a higher score. During follow-up, 1,347 people developed cognitive impairment (1,163 mild cognitive impairment, 184 Alzheimer's dementia). Over ten years, an estimated 33.8% of those in the top third of ADHD gene scores were affected, against 23.4% in the bottom third: roughly 10 extra cases per 100 people.
• A modest effect per step. After adjustment, each step up the score (one standard deviation) went with about an 11% higher risk (hazard ratio 1.11; 1 would mean no difference). Adding education, vascular risk or APOE4 to the analysis barely changed this.
• Specific to ADHD. The gene scores for the four other psychiatric conditions were not linked to cognitive impairment.
• Not through amyloid. In 1,146 people whose brains or spinal fluid had been tested for amyloid, the protein that builds up in Alzheimer's disease, the ADHD score was not linked to having amyloid, and the link with cognitive impairment was similar with or without it.
• Lower from the start, not faster decline. People with higher scores did slightly worse on memory, planning and language tests from the beginning, but did not decline faster over time.

What to keep in mind

• Links, not causes. The authors note that they could not account for every illness that affects thinking, so other explanations remain possible.
• A rough genetic measure. The ADHD gene score explains only a small part of who actually has ADHD (about 0.7% to 5.5% of the differences, by the estimates the authors cite), and it ignores rare variants and life circumstances.
• Mostly European ancestry. The score was built from studies of mainly European-ancestry people, so it may work less well in other groups.
• Volunteers. People in such studies tend to be healthier; some studies excluded people with serious psychiatric conditions.
• Possible mix-ups. ADHD symptoms such as forgetfulness and poor attention overlap with those of mild cognitive impairment, so some people may have been diagnosed with impairment when the cause was ADHD-like traits. About a third of those diagnosed later tested as unimpaired again, though the link held when the team took that into account.
• Separate effects were small. In each of the eight studies on its own, the link was not statistically significant, though it pointed the same way in all but one.

Why it matters

The finding adds weight, from a different angle, to the idea that ADHD may be one of the things that make the brain more vulnerable in old age, and that this works at least partly outside the amyloid route of Alzheimer's disease. The authors suggest that people with diagnosed ADHD, including older adults whose ADHD was never recognised, might deserve closer watching of memory and thinking, and attention to risk factors that can be changed. How the link works, whether through lower "reserve" from the start or through a lifetime of related health problems, is not yet known.