What does the research say about late diagnosis of ADHD in women?

The diagnosis gap is visible in health records

Large health-record studies support later recognition in females, not a universal delay for every woman. (doi.org)

Skoglund and colleagues’ Stockholm register study (Wiley, 2024) examined 85,330 children and adults during 2011 to 2021. It measured age at the first recorded ADHD diagnosis or ADHD medication record in that period. (doi.org)

The mean age was 23.5 for females and 19.6 for males, about four years later. This is not a waiting-list estimate or a measure of when symptoms started. (doi.org)

By that point, anxiety or stress-related diagnoses were recorded in 50.4% of females versus 25.9% of males, approximately twice as often. This does not establish that those earlier diagnoses were mistaken. (doi.org)

When bringing this research to an appointment, keep the request open: could ADHD be part of the explanation, and what would make it less likely? The question is about your history and present difficulties, not whether you fit a statistical average.

This page provides education, not a diagnosis. These studies do not establish outcomes for care delivered through Finding Focus.

Presentation differences are patterns, not rules

Symptom visibility may contribute to missed recognition, without defining a single female ADHD profile. (link.springer.com)

Mowlem’s population-based twin study (Springer Nature, 2019) linked parent ratings of 19,804 Swedish nine-year-olds with diagnostic records. (link.springer.com)

The questionnaires did not establish impairment or a full diagnosis, so these are not percentages of confirmed missed cases. Among diagnosed children, presentation differences between girls and boys were not statistically clear. (link.springer.com)

For discussion, separate what other people saw from the effort needed to keep up. If you use the term masking, explain the specific actions you mean: reminders, repeated checking or extra preparation. Include what went wrong despite that effort. These are discussion prompts, not a new diagnostic checklist.

Ask whether the assessment will consider both outward behaviour and difficulties you experienced privately. Bring classroom comments without asking them to tell the entire story.

For a parent of a teen, invite the young person’s account of that effort. Avoid turning the appointment into a debate about whether she looks like someone with ADHD.

Menstrual-cycle studies are small and varied

Small studies suggest cycle-related symptom changes, not a predictable pattern for everyone. (journals.sagepub.com)

Osianlis and colleagues’ systematic review (Sage, 2025) included 11 studies of girls and women with diagnosed ADHD or symptom-scale measures, searched through January 24, 2025. (journals.sagepub.com)

Four menstrual-cycle studies included one to 32 participants each. Interviews, case reports and repeated measurements compared cycle phases within individuals or had no control group. Some reported worsening near menstruation or after ovulation; there was no pooled estimate of the size of change. (journals.sagepub.com)

Puberty findings were mixed and pregnancy evidence sparse. These associations do not show that hormones caused ADHD or delayed its recognition. (journals.sagepub.com)

If you notice a pattern, consider recording dates, the difficulty you noticed, sleep and menstrual timing where relevant. Write down what stayed the same as well as what changed. Bring the notes to a clinician rather than assuming that a difficult week establishes a hormonal cause.

Include ordinary days, not just difficult ones. Ask what other explanations need consideration and what additional information would help distinguish them. You do not need to decide the cause before discussing the pattern.

Midlife findings do not prove newly developed ADHD

Perimenopausal symptom scores were higher in women reporting ADHD than in women without reported ADHD in a recent population study. (pmc.ncbi.nlm.nih.gov)

Jakobsdóttir Smári and colleagues’ SAGA cohort analysis (Cambridge University Press, 2025) included 5,392 Icelandic women aged 35 to 55. It compared 535 reporting an ADHD diagnosis with 4,857 not reporting one. The key symptom comparison was cross-sectional, at the 2024 follow-up. (pmc.ncbi.nlm.nih.gov)

The age-adjusted prevalence of severe symptoms on the Menopause Rating Scale was 54.2% versus 30.1%, about 1.8 times as high in the ADHD group. (pmc.ncbi.nlm.nih.gov)

The scale includes complaints that occur outside menopause, and ADHD diagnoses were self-reported. This was not a hormone experiment or an early-versus-late diagnosis comparison. It cannot establish that hormonal changes caused previously unrecognised ADHD to become apparent. (pmc.ncbi.nlm.nih.gov)

For a midlife assessment discussion, prepare separate timelines for difficulties you remember earlier and changes you noticed recently. Ask how both will be considered. A useful question is whether a new change, a longstanding pattern, or both need attention, rather than choosing between ADHD and menopause yourself.

What the evidence does not show

These findings justify taking late recognition seriously, but they do not provide a complete explanation for an individual woman’s experience. (link.springer.com)

• Exactly why recognition differs. Parent ratings linked to records do not establish referral patterns or separate the contributions of symptom presentation, access and diagnostic expectations. (link.springer.com)
• What earlier diagnosis alone would prevent. The observational studies do not isolate the effect of recognising an otherwise similar woman earlier rather than later. (doi.org)
• A hormonal explanation for everyone. Associations across reproductive stages do not identify the cause of a particular person’s difficulties or provide a reliable individual prediction. (journals.sagepub.com)
• Equal applicability across identities and communities. Register-based categories provide limited gender-identity information; these European samples cannot represent every Canadian community. (doi.org)

As a reading rule, separate the observation from the proposed explanation. When a headline moves from women being diagnosed later to a claim about exactly why that happened, check whether the study measured that explanation. Keep your own questions open rather than treating a research headline as a personal conclusion.

How this applies in Canada

Use the international findings to inform questions, not to assign yourself a diagnosis or a Canadian waiting-time estimate. (doi.org)

CADDRA’s Canadian ADHD Practice Guidelines, edition 4.1 (Canadian ADHD Resource Alliance, 2020), flag under-identification in girls and recommend evaluating symptoms, impairment, childhood history and other explanations. This is practice guidance, not proof of assessment outcomes. (caddra.ca)

For this question, an appointment note could include the following prompts:

• Separate memories from records. Note what you recall, what documents show and where details remain uncertain.
• Describe effort alongside results. Include reminders, preparation and support, plus what still went wrong.
• Put changes in order. Note changes in demands, sleep and relevant reproductive stages without deciding their cause.
• Ask for the reasoning. What supports ADHD, what argues against it, and what happens next if ADHD is not diagnosed?

Keep the note open-ended. It can say that ADHD is one possibility you want considered, not the answer you expect. Ask the clinician to explain the conclusion, including any uncertainty and what further information could help. An assessment can conclude that ADHD is not present.

Before the conversation ends, identify the concern you still need help with, whether or not the explanation is ADHD. Ask what the proposed next step is intended to clarify or address.

Questions about how common ADHD is in Canadian adults and CBT evidence for adults are covered separately.