%20(1).webp)
A 58-year-old woman is referred to you from a memory service. She went there eighteen months ago with word-finding trouble and losing the thread in meetings. Her cognitive screen was normal for her age, her MRI was reported as age-appropriate, and she has been discharged with a letter suggesting ADHD.
Her mother died in 2011. Her school holds no records from before the 1990s. She has no siblings. This is the shape of the late-diagnosis referral after 50, and it tests the standard adult pathway at two points at once: evidencing childhood onset, and ruling out early cognitive decline.
This guide covers both, and what the report needs to say. For the general method of building onset evidence without a parent, see our guide on diagnosing ADHD with no childhood informant. This article focuses on what changes once the patient is over 50.
In older adults, the question is rarely only "is this ADHD?" It is "is this ADHD, and is it only ADHD?"
Yes. The best prevalence data for this age group comes from the Longitudinal Aging Study Amsterdam. Michielsen and colleagues (2012) screened 1,494 adults aged 60 and over, interviewed 231 of them, and estimated the prevalence of syndromatic ADHD at 2.8% and symptomatic ADHD at 4.2%. The Updated European Consensus Statement on adult ADHD (Kooij and colleagues, 2019) cites this study as evidence that ADHD persists beyond 60.
Relatively few older adults in UK services carry the diagnosis. Part of the reason is history: this cohort went to school long before adult ADHD services existed, and the diagnostic criteria still ask them to evidence what happened in that school.
NICE NG87 recommendation 1.3.1 asks for a full clinical and psychosocial assessment, a full developmental and psychiatric history, and observer reports and assessment of the person's mental state. It asks for observer reports; it does not name a parent.
That matters in this age group. A spouse of thirty years can describe the adult pattern in detail, though not the developmental one. The assessment is not automatically unsafe when no childhood informant exists, but the report has to be clear about what each source can and cannot show.
The usual hierarchy applies, from contemporaneous documents down to self-report. What changes after 50 is what survives and how to read it:
The reason for that last point is Mannuzza and colleagues (2002). They followed children with ADHD and controls into adulthood and asked them, at a mean age of 25, to recall their childhood symptoms. When the findings were adjusted for the prevalence of ADHD in the general population, the positive predictive value of the retrospective diagnosis was 0.27. That was fifteen years of recall. At 58 it is more than forty. Our article on the reliability of retrospective recall covers the wider evidence.
This is the differential that makes the over-50 assessment different in kind, not just in difficulty. Callahan and colleagues (Journal of Attention Disorders, 2022) compared 40 adults with ADHD, 29 with mild cognitive impairment (MCI) and 37 controls, all aged 50 to 85, at a memory clinic. Both patient groups showed memory impairment, which is why these referrals are confusing.
The pattern behind the impairment differed. The ADHD group showed an encoding deficit, linked to frontal cortical thinning. The MCI group showed a storage deficit, linked to smaller hippocampi. In clinic, five features usually help separate the two:
Our article on ADHD and memory loss explains the working memory side of this in plainer terms, and is useful to share with patients.
Where the picture is unclear, repeat testing at 12 to 18 months can clarify the trajectory in a way a single session cannot. A stable profile points towards ADHD; a declining one does not.
One finding to hold rather than resolve. In an Israeli cohort of more than 109,000 adults followed from 2003 to 2020, Levine and colleagues (2023) found dementia in 13.2% of those with adult ADHD, against 7.0% of those without. The association remained after adjustment (adjusted hazard ratio 2.77), but the study is observational and does not show cause. Its clinical lesson is narrow: finding ADHD does not end your duty to keep watching for decline.
Inattention that appears or worsens in later life has a longer list of possible causes than in younger adults. Before attributing it to ADHD, check for:
Most of these can coexist with ADHD rather than replace it. The question is whether the lifelong pattern was there before any of them began. Our guide to documenting Criterion E covers how to record that reasoning.
Be open about this in the report as well as in your own reasoning. The DIVA-5, the rating scales and the criteria themselves were developed and tested mainly in younger adults, and have had little validation in people over 60. The symptom descriptions also assume a working life a retired patient may no longer have. "Often has difficulty organising tasks" means something different at 68 than at 28. Our guide to using the DIVA-5 in complex presentations covers how to adapt the interview.
Treatment evidence is also thin. Trials of ADHD medication have rarely included people in this age group, so you cannot promise a patient of 62 the response rates quoted for younger adults. If medication is considered, baseline cardiovascular assessment matters more with age; our guide to the ADHD annual review and physical health monitoring sets out the checks.
A report that says "childhood onset confirmed" on the strength of an adult informant's account is unlikely to survive a second opinion. Match the wording to the evidence:
Our guide to writing the ADHD diagnostic report covers the rest of the structure.
Yes. ADHD persists into older age, and a Dutch study estimated a prevalence of 2.8% in adults aged 60 and over. A late diagnosis still needs evidence that symptoms began in childhood.
No. NICE NG87 asks for observer reports as part of a full assessment. It does not name a parent, so a sibling, partner or documents can contribute, depending on what they can genuinely show.
Mainly by time course. ADHD is lifelong and stable, while cognitive decline has a recent onset and gets worse. Memory testing, the type of memory failure and repeat assessment after 12 to 18 months all help.
The ADHD Specialist Certification™ trains clinicians in adult ADHD assessment, including onset evidence when informants are limited. The ADHD Prescribing and Management course covers treatment decisions in more complex patients, including older adults.
An unresolved assessment with a clear list of what is still needed is a better clinical document than a confident one built on weak evidence. With older adults, that list usually includes a second look at cognition. Taking the time to get both questions right protects the patient whichever way the answer falls.
