ADHD Diagnosed Later in Life in Women: What the Register Data Actually Shows
Evidence Review · The Reading Room, Vol. 6 · 11 min read · What the register data establishes about the diagnosis gap, what it cannot establish about you, and the single limitation that changes how to read all of it
A Swedish register of 85,330 diagnoses found women identified roughly four years after men, with symptoms starting at the same age. An evidence review of what the numbers show, and the large limitation nobody mentions.
The question usually arrives sideways. A niece is assessed. A colleague mentions hers. And somewhere in the reading that follows, a description of inattentive presentation in adult women lands with an unpleasant precision on thirty years of coping strategies that were assumed to be personality.
This review sets out what the research establishes about women being diagnosed later, and what it cannot tell you about yourself. It is a reading of published register and clinical data; we are not clinicians, the findings belong to the researchers named below, and only the plain-English rendering is ours.
The short answer: the diagnostic delay is real, it is large, and it has been measured in national data rather than inferred from anecdote. In a Swedish register of 85,330 people with an ADHD diagnosis, women were identified at a median age of 19 and men at 14 — a gap of roughly four to five years.¹ In a separate Spanish clinical sample, the age at which symptoms first appeared did not differ significantly between the sexes.² The delay is in recognition, not in onset. There is a name worth having for the mechanism, because it explains the shape of the whole finding: the disruption filter. Referral is triggered by trouble a child causes other people, not by difficulty she is having on her own, and a filter built that way will find boys first every time.
- 85,330 diagnoses, a four-year gap. Swedish register data: women diagnosed at a median of 19.0 years and a mean of 23.5; men at a median of 14.0 and a mean of 19.6.¹
- 900 patients, same onset, different recognition. In a Spanish clinical sample, mean age at diagnosis was 28.96 for women and 24.13 for men, while the age symptoms appeared did not differ significantly (p = 0.220).²
- By diagnosis, half already carry an anxiety diagnosis. Among women in the Swedish register, 50.4% had a recorded anxiety disorder against 25.9% of men; mood disorders 37.5% versus 19.5%.¹
- Adult prevalence estimates range from 1.6% to 5.0% depending entirely on how you count — registers give the lowest figure, population surveys the highest.³
- The registers can only see people who were eventually diagnosed. No study here can estimate how many women were never assessed at all, which is the group most likely to be reading this.¹
The source for this review
Dr Lotta Borg Skoglund, MD, PhD
A psychiatrist and researcher at Uppsala University and first author of the Swedish register study. Her clinical and research work is specifically on how ADHD presents across a woman’s life — including the parts of that life the diagnostic criteria were not written around — and she has spent years arguing, in print and in Swedish public debate, that the diagnosis gap is a service failure rather than a difference in underlying rates.
What we read: Skoglund, Sundström Poromaa, Leksell, Ekholm Selling, Cars, Giacobini, Young and Kopp Kallner, Journal of Child Psychology and Psychiatry (2024), alongside the Spanish clinical sample and the 2024 prevalence review.¹²³
Where to follow her work: LinkedIn · borgskoglund.se · ADDitude author page
Scope and method of this review
This review covers the documented difference in age at ADHD diagnosis between women and men, and what accompanies it. Sources were selected for scale and for design: a national population register, an independent clinical sample from a different country, and the most recent systematic review of prevalence. Where a figure appears below, it is drawn from the source cited beside it.
What this review does not cover. It does not assess whether ADHD is over- or under-diagnosed as a whole, which is a separate and unusually heated literature. It does not evaluate treatment, medication or their risks. It does not cover ADHD in children except where the register data necessarily includes them. It does not address the widely discussed interaction between hormonal change in midlife and attention symptoms, because the evidence base there is currently too thin to review honestly — a review of a handful of small studies would give that question a confidence it has not earned. And crucially, it cannot address undiagnosed women, for the reason set out in the confidence box below.
The evidence table
| Study | Sample | Finding | Limitation |
|---|---|---|---|
| Skoglund et al., 2024¹ | 85,330 with ADHD (37,591 women, 47,739 men) plus 426,626 controls | Women diagnosed about four years later. Median 19.0 vs 14.0; mean 23.5 vs 19.6 | One county only; sees registered diagnoses and prescriptions, not severity or impairment; private healthcare data missing |
| Skoglund et al., 2024 — comorbidity¹ | Same cohort | Wide gaps at diagnosis. Anxiety 50.4% vs 25.9%; mood disorders 37.5% vs 19.5%; eating disorders 5.6% vs 0.6%; personality disorders 6.3% vs 2.1% | Cross-sectional at diagnosis; cannot show whether the additional condition preceded, followed, or was mistaken for the ADHD |
| Oltra-Arañó et al., 2025² | 900 adult patients (54.9% men, 45.1% women), Spain | Same onset, later diagnosis. Mean age at diagnosis 28.96 (women) vs 24.13 (men); age of symptom onset not significantly different (p = 0.220) | A specialist clinic sample rather than a population; people who reach such a clinic differ systematically from those who do not |
| Prevalence systematic review, 2024³ | Pooled across study designs | Adult prevalence 1.6% to 5.0% depending on method: register 1.6% (95% CI 0.9–3.0), survey 5.0% (2.9–8.6), two-stage clinical 4.8% (4.0–5.8) | Criteria, information source and whether impairment is required all significantly change the estimate; no sex breakdown in the results |
| Agnew-Blais, 2024⁴ | Commentary on the Skoglund register findings | Frames the finding as a service failure rather than a difference in underlying rates | A commentary, not new data; interpretive rather than evidential |
Read the finding column downward. Two countries, two entirely different designs, the same direction. That is the kind of agreement the cortisol literature conspicuously lacks.
If the more likely answer is that the last few years have simply been too heavy and nothing was ever wrong with your attention, the Quiet Audit is ten private minutes on that question. No programme, no advice, nobody sees your answers. Start here.
Key terms
Register study — research using a country’s complete administrative health records rather than a recruited sample. Enormous statistical power and no volunteer bias; the trade-off is that it can only ever see what a clinician wrote down.
Inattentive presentation — the form of ADHD characterised by difficulty sustaining attention, disorganisation and forgetfulness, without prominent visible hyperactivity. It disrupts a classroom far less, which is relevant to who gets referred.
Comorbidity — the presence of an additional diagnosed condition alongside the one being studied. Here it matters because the additional diagnosis frequently arrives first.
Median vs mean age — the median is the midpoint; the mean is the average, which a long tail of late diagnoses pulls upward. Both are reported above because the distance between them is itself informative.
Why the gap exists
Three explanations are usually offered, and the data supports them unevenly.
1. Referral is driven by disruption, not distress. A child who cannot sit still generates a problem for the room. A child who is quietly not following anything generates a problem only for herself, and it shows up as an average report and a slightly disappointing grade. The register finding — a median of 14 for boys against 19 for girls¹ — is what a referral system triggered by external disruption would be expected to produce. This explanation fits the data well, and it is what the disruption filter names.
2. The presenting complaint is something else. By the time women in the Swedish register received an ADHD diagnosis, 50.4% carried an anxiety diagnosis and 37.5% a mood disorder, against 25.9% and 19.5% of men.¹ Something brought them into services, and it was very often not attention. Note carefully what this does and does not establish: the register can show the diagnoses co-occur, not which was primary, nor whether the anxiety was a consequence of undiagnosed difficulty, an independent condition, or the thing the difficulty was mistaken for. All three happen. The data cannot separate them.
3. Compensation is available to the competent. This one is intuitive and the least evidenced. The argument is that a bright, conscientious woman builds enough scaffolding — lists, rules, over-preparation, punishing hours — that the underlying difficulty never surfaces as failure, only as effort. It is a plausible reading of why the Spanish sample found identical symptom onset alongside a five-year diagnostic delay.² It is not directly tested by either study, and it should be held more loosely than the first two.
How strong is this evidence?
Moderate. The finding is consistent across independent samples and the samples are large, but it comes from two national healthcare systems and from clinical records rather than from population screening.
Well established: that among people who receive an ADHD diagnosis, women receive it several years later than men — roughly four to five years across two countries and two designs¹² — and that women arrive at that diagnosis carrying substantially more recorded anxiety and mood diagnoses.¹
Not established: that the underlying rate differs by sex; which condition came first in the comorbidity data; or that any particular mechanism explains the delay.
The limitation that matters most for the reader of this page. Every figure above describes people who were eventually diagnosed. A register cannot count anyone it never saw. If women are systematically under-referred — the leading explanation for the gap — then the women most affected by that under-referral are precisely the ones missing from the denominator, and the true delay is longer than any of these numbers.
And a second, blunter one. The median woman in the Swedish data was diagnosed at 19. The mean in the Spanish sample was 29. Neither study describes a woman recognising something in herself at 53. Applying this literature to that situation is extrapolation, and this review will not dress it up as anything else.
Five things that look like this and are not this
The sections above protect the literature. This one is for the reader, and it is the section the symptom checklists do not write, because it invites you to consider that the answer might be something with a blood test attached to it.
Five ordinary conditions produce exactly the experience that sends a woman to an ADHD checklist at fifty-two. Each is worth excluding before, not after.
- Obstructive sleep apnoea. Chronic fragmentation wrecks attention and working memory. Waking unrefreshed after a full night is the signature, and someone else has usually noticed the breathing first.
- Hypothyroidism. Slowed thinking with cold intolerance, dry skin and weight change. One blood test settles it.
- Iron deficiency, with or without anaemia. Common in the years before the final period, and it produces breathlessness on stairs alongside the fog.
- The menopause transition. Verbal memory and processing speed do measurably dip during it, which is a different mechanism with its own imaging evidence behind it.
- Depression, and chronic sleep debt. Both impair concentration directly, and both are more likely to be dismissed by the person experiencing them than by anyone examining them.
None of these rules ADHD in and none rules it out. They also stack, which is why a demanding year with untreated apnoea in it produces a convincing imitation of a lifelong pattern. This is a list to raise with a clinician in order, not one anyone can clear themselves on.
Who else has measured this
One register is one healthcare system. Three other groups have approached the same question from different directions.
Prof. Susan Young (Psychology Services · ResearchGate · LinkedIn), a clinical psychologist who is also a co-author on the Swedish register paper, led the 2020 expert consensus statement on females with ADHD — a lifespan document written specifically because the identification guidance had been built around a male presentation.⁵ Her position, argued for two decades, is that the sex difference in diagnosis rates is a recognition artefact rather than a biological one.
Josep Antoni Ramos-Quiroga (Vall d’Hebron · Google Scholar · ResearchGate) and colleagues at Vall d’Hebron in Barcelona produced the independent clinical replication: 900 adults, a five-year gap in diagnosis, and no significant difference in when symptoms began.² A different country, a different design, the same direction.
Dr Jessica Agnew-Blais (QMUL faculty page · Google Scholar · Bluesky) at Queen Mary University of London wrote the invited commentary on the register findings, and framed them as a failure of services to identify rather than as evidence about who has the condition.⁴ A commentary is interpretation rather than data, and is flagged as such in the table above.
A national register, a clinical replication, a consensus statement and a commentary. The convergence is on the delay. Nobody in this literature is claiming it tells you about any particular woman.
Not a late bloomer. A late referral.
What this actually changes
Not whether you have it. This review cannot answer that and does not try to; no page can, and the ones that claim to are selling something.
What it changes is the interpretation of your own history. Thirty years of finding everything harder than it looked, while the outputs stayed excellent, has at least two available explanations, and until recently only one of them was ever offered: that you are difficult to satisfy and privately disorganised. The register data puts a second explanation on the table and shows it is common. That is worth something even if it turns out not to be yours.
It also changes what to do next, and the order matters. The evidence describes a recognition failure in a system, which means the useful move is not another checklist. It is a developmental history — the one document a clinician cannot get anywhere else and the one thing no online quiz can fake.
When this belongs with a doctor
An article cannot assess you, and an online questionnaire cannot either. A formal assessment involves developmental history, corroboration where possible, and the exclusion of conditions that produce overlapping symptoms.
That exclusion step is the reason to go through a clinician rather than around one. Thyroid disease, anaemia, sleep apnoea, perimenopausal change, depression, anxiety and chronic sleep deprivation can all impair attention and working memory, and several are straightforward to test for. Attributing a new concentration problem to lifelong undiagnosed ADHD without that step is how a treatable condition gets missed.
Book an appointment rather than reading further if concentration has changed noticeably in the last year rather than being lifelong, if it comes with new physical symptoms, or if low mood or anxiety is present. A changed pattern is a different question from a long-standing one.
What to say at the appointment, and what you will probably hear back
The obstacle is rarely the request. It is the reply that a competent fifty-something with a good job does not look like the referral criteria. Three lines, short enough to read off a phone screen.
Short enough to read off a phone screen
Say: “I have written a developmental history covering school and early work as well as the last two years. I would like to go through whether this is a lifelong pattern or a recent change.”
If you hear “but you have managed very well”: “Managing it took a great deal. That is the part I would like assessed, rather than the outcome.”
Then ask for the exclusions: “Before we go towards assessment, could we check thyroid, ferritin and B12, and consider whether a sleep review is warranted?”
If writing that developmental history is the part you keep not starting, the Quiet Audit is ten private minutes that get the first version of it onto a page. Nobody sees your answers. Start here.
The one move
The one move
Two columns, twenty minutes. Column one: school and early work, before any of the current load existed — reports that said “could do better”, whether reading a chapter required rereading it, how homework actually got done. Column two: the same questions, answered honestly about the last two years.
If column one is blank and column two is full, the question is what changed. If both are full, and the second is the first with the scaffolding removed, that is the pattern worth taking to an assessment — described in specifics rather than adjectives.
One caution about the reading you will do next. There is a large and confident internet on this subject, most of it describing a genuinely different situation — younger women, different demands, different decades. Recognising yourself in a list of symptoms is not evidence; almost everyone recognises themselves in these lists, which is why they are popular. The developmental history is the evidence.
Questions this review is asked
How much later are women diagnosed with ADHD than men?
In Swedish register data covering 85,330 diagnoses, women were diagnosed at a median age of 19.0 and men at 14.0 — about four to five years later.¹ A Spanish clinical sample of 900 patients found mean ages of 28.96 and 24.13 respectively.²
Do symptoms start later in women?
The available evidence says no. In the Spanish sample, the age at which symptoms first appeared did not differ significantly between women and men (p = 0.220), despite the diagnostic gap.² The delay appears to be in recognition and referral rather than in onset.
Why do so many women get an anxiety or depression diagnosis first?
The register data shows the pattern clearly — 50.4% of women had a recorded anxiety disorder at ADHD diagnosis against 25.9% of men.¹ What it cannot show is why: whether the anxiety was a consequence, an independent condition, or the thing the difficulty was read as. All three occur.
Can I be assessed for ADHD in my fifties?
Assessment in midlife is possible and is done. But the studies above describe people diagnosed at 19 and 29, not 53, so they are not evidence about that situation. A clinician will want a developmental history showing the pattern existed long before the current load did.
Could this just be perimenopause, or exhaustion?
It could, and both produce genuine difficulty with attention and working memory. The distinguishing question is whether the pattern is lifelong or recent. Brain fog with a clear starting point is a different question from a difficulty that has always been managed with effort.
Is ADHD becoming more common in adults?
The prevalence review found adult estimates ranging from 1.6% to 5.0% depending purely on how the counting was done, and noted that criteria, information source and impairment requirements all significantly shift the number.³ Rising diagnosis rates and rising prevalence are not the same claim.
Informational, not medical advice. Nothing here can diagnose or rule out a condition, and this review does not attempt to.
Where to go next
In order, and each for a reason.
1. The one question worth asking about brain fog — it separates a lifelong pattern from a recent change faster than any checklist.
2. Then why attention stays behind after a meeting ends, which is a workload mechanism rather than a diagnostic one and explains a great deal of what gets attributed to ADHD.
3. Then what a normal blood panel does and does not measure, because several items on the exclusion list live there.
Whatever the answer turns out to be, the effort was real. It was never invisible to you.
References
1. Skoglund, C., Sundström Poromaa, I., Leksell, D., Ekholm Selling, K., Cars, T., Giacobini, M., Young, S., & Kopp Kallner, H. (2024). Time after time: failure to identify and support females with ADHD — a Swedish population register study. Journal of Child Psychology and Psychiatry. Read the paper
2. Oltra-Arañó, L., Crespí, J. J., Corrales, M., Richarte, V., Fadeuilhe, C., Ramos-Quiroga, J. A., & Amoretti, S. (2025). Sex differences in age at ADHD diagnosis and symptom onset. Presented via the European College of Neuropsychopharmacology. Read the material
3. Prevalence of attention-deficit hyperactivity disorder (ADHD): systematic review and meta-analysis (2024). European Psychiatry. Read the review
4. Agnew-Blais, J. (2024). Hidden in plain sight: delayed ADHD diagnosis among girls and women — a commentary on Skoglund et al. Journal of Child Psychology and Psychiatry. Read the commentary
5. Young, S., Adamo, N., Ásgeirsdóttir, B. B., et al. (2020). Females with ADHD: an expert consensus statement taking a lifespan approach providing guidance for the identification and treatment of attention-deficit/hyperactivity disorder in girls and women. BMC Psychiatry, 20, 404. The lead author’s professional profile
On the researchers. Lotta Borg Skoglund is a physician and researcher; Susan Young is a clinical psychologist; Jessica Agnew-Blais is an epidemiologist. None of them is your clinician, none has reviewed this summary of their work, and nothing in their published research is medical advice for an individual. Where this review describes their positions, it describes what is printed in the papers linked above.
Sourcing, disclosure and disclaimers
This is not medical advice. Blue Leaf Journal publishes general information for a general readership. Nothing here is a diagnosis, a treatment recommendation, or a substitute for care from a clinician who knows your history. No article can diagnose or exclude ADHD.
We are not clinicians. Blue Leaf Journal is an independent publication. We read published research and translate it. The findings belong to the researchers and institutions named and linked above; the plain-English rendering is ours, and so is any error in it.
No affiliation and no endorsement. Blue Leaf Journal is not affiliated with Lotta Borg Skoglund, Susan Young, Jessica Agnew-Blais, Josep Antoni Ramos-Quiroga, Uppsala University, Queen Mary University of London, Vall d’Hebron or any register-holding authority. None of them has reviewed, approved or endorsed this article, and none of them is responsible for it. They are cited because their published work is the evidence for what it says.
No commercial relationship. Nobody named above paid for or was paid for this coverage. There are no affiliate links, no sponsored placements and no gifted products in this article. We do not sell assessments, screeners or referrals.
How this was checked. Every figure above is drawn from the primary sources, each linked in the references, and can be verified there. Where the evidence does not extend to women in midlife, we have said so rather than extrapolate. Sources were checked on 27 August 2026. If you find something we have got wrong, write to miriamalderton@blueleafjournal.com and we will correct it and say that we did.
Miriam Alderton is Research Editor at Blue Leaf Journal. She reads the methods section first and the abstract last, and every figure in this piece is linked to its source above.
Written by Miriam Alderton, Research Editor, for Blue Leaf Journal. Updated: 27 August 2026.
