PMS, Postpartum, Perimenopause: Why ADHD Hits Women Harder at Every Hormonal Drop
Key Findings
Smári and colleagues (2025) ran a population-based cohort of 5,392 women (535 with ADHD, 4,857 without). Severe perimenopausal symptoms showed up in 54.2% of the ADHD group versus 30.1% of the non-ADHD group. That's about 1.8 times more likely. The psychological symptom load (mood, irritability, brain fog) was 58.6% vs 36.0%. Onset hit earlier too, in the 35 to 39 age band rather than the more typical 40s.
Smári includes Kooij and Wynchank as co-authors, so it's better read as a population-scale extension of the same group's work than as outside replication. The genuinely independent check comes from Chapman 2025 at King's College London (Hunter and Dommett, n=656, no ties to the Kooij network), which lands in the same direction. Earlier perimenopause research on this question used small surveys. A 5,000-plus population sample plus an unrelated KCL group landing the same way is what moves a finding from "interesting hypothesis" to "the field believes it." [T1]
The defensible number on this one comes from Andersson and colleagues (2023, Journal of Affective Disorders), who pulled the Swedish national registry — 773,047 women who had given birth between 2005 and 2013. Women with an ADHD diagnosis prior to pregnancy had postpartum depression at about 5 times the rate of women without (prevalence ratio 5.09, 95% CI 4.68 to 5.54). About 17% of the ADHD group versus roughly 3% of the non-ADHD group. Postpartum anxiety ran similar numbers (PR 5.41).
You'll see higher self-report figures floating around in magazine surveys. The registry data is the cleaner number, and it's still a brutal one. The mechanism makes sense: estrogen drops roughly a hundredfold in the days after delivery. If your dopamine system was already running thin, the floor falls out fast. [T1]
Dorani and colleagues (2021) found premenstrual depressive symptoms in 45% of women with clinically diagnosed ADHD versus 28% in the general female population. Roughly 1.6 times higher. This is the within-the-month version of the same story we covered back in Issue #7, where estrogen dipping in the luteal phase pulls dopamine down with it and ADHD meds start feeling weaker. [T1]
Hinshaw and colleagues (2022) summarized the evidence in their Annual Research Review: girls meet ADHD criteria at roughly half the rate of boys in childhood, but by adulthood the male-to-female ratio is much closer to 1 to 1. The field reads that pattern as evidence the girls were missed, not that they grew into ADHD later.
The presentation is the reason. Girls with ADHD tend toward inattentive, internalizing patterns: anxiety, perfectionism, masking, white-knuckling through school on a "smart but spacey" reputation. That doesn't make a teacher's life hard, so it doesn't get flagged. By the time perimenopause hits and the compensation strategies stop working, a lot of these women are 42 and finally getting diagnosed. [T1]
Ter Beek and colleagues (2023) looked at 300 women presenting to a cardiology clinic. About 35% had lifetime ADHD symptoms, and they presented with cardiac complaints roughly two years younger than the women without ADHD histories.
This is one small hospital sample. We're flagging it as hypothesis-generating, not as proof that ADHD causes heart disease. But it's worth knowing this signal exists, especially if you're a midlife woman whose cardiac symptoms keep getting waved off. [T1]
Why It Matters
Back in Issue #7 we covered why ADHD meds feel weaker premenstrually. Short version: estrogen feeds dopamine, estrogen dips in the luteal phase, dopamine dips with it, the stimulants have less to amplify, focus tanks. That's the monthly version of the story.
Issue #23 is the same story stretched across an entire life. Puberty, every menstrual cycle, postpartum, perimenopause. Same mechanism, longer timeline, more disease consequences. Plus the upstream piece Issue #7 didn't cover: half the women living through this were never diagnosed in the first place, because as girls they were quiet, perfectionist, and "doing fine" by every metric the school cared about.
The connection is what makes this issue land for two different readers at once. If you're the parent of an ND daughter, you now have the language ("inattentive presentation," "internalizing comorbidity," "compensatory perfectionism") to push for an evaluation that doesn't wait until she's hyperactive in a way teachers can't ignore. If you're a woman in your 30s, 40s, or 50s who has been quietly suspecting you have ADHD, this paper is a map of why hormonal transitions have always felt heavier for you than they seem to for other women.
You weren't broken. You were missed. [T1]
The Fine Print
Kooij and 30 co-authors wrote an invited synthesis for Frontiers in Global Women's Health. That format lets the authors pick which studies to include and how to weight them. It's useful for mapping a field. It's not the same level of evidence as a pre-registered systematic review with formal inclusion criteria.
The good news: two independent groups have looked at the broader question. Osianlis 2025 (Monash University, Australia) ran a systematic review covering 1980 through January 2025 and found a consistent direction of effect, strongest for menstrual-cycle and pubertal transitions. Chapman 2025 (King's College London, Hunter and Dommett) surveyed 656 women aged 45 to 60, including 245 with ADHD diagnoses, and found ADHD symptom severity tracked with menopausal anxiety, depression, memory, and concentration complaints. Neither group has institutional ties to the Kooij network. So the synthesis isn't resting on one lab.
⚖️ Our take: Treat the mechanism as well-established. Treat the headline numbers as accurate to the primary studies they're drawn from. Don't treat this as a meta-analysis. It isn't one.
Kooij and Wynchank are both co-authors on Smári 2025. So that n=5,392 study is best read as a population-scale extension of the Kooij group's work, not as someone else checking their homework. Worth saying out loud.
The genuinely independent pieces in the chain are Osianlis 2025 (Monash University, Australia) and Chapman 2025 (Hunter and Dommett at King's College London). Neither group has ties to the Kooij network. Both land in the same direction. The convergence isn't from one lab.
Worth flagging: the 2025 paper itself declares no commercial relationships. But Kooij's career-level disclosures across prior publications include speaking and research engagements, plus unrestricted research and educational grants, with Janssen, Shire, Eli Lilly, and Eurocept. Those companies make the stimulants prescribed for ADHD. She also chairs the DIVA Foundation (which distributes the diagnostic interview she developed) and founded PsyQ Expertise Center for Adult ADHD (a clinical practice). Both organizations benefit when adult-female-ADHD diagnosis grows. None of that invalidates the work, and the same broader-field stake exists for Hinshaw and Young, but readers should have the picture.
⚖️ Our take: Read the Osianlis review and Chapman's KCL work as the independence check. The story doesn't rest on any single lab.
The review cites a Cohen's d of about 3.71 for perimenopausal symptom severity in ADHD vs non-ADHD women. In plain English, that would be an enormous gap. It comes from a small survey-based study and almost certainly overstates what shows up at population scale.
The Smári 2025 prevalence numbers (54.2% vs 30.1% for severe symptoms in 5,392 women) are the more defensible figures. We're leading with those, not the 3.71.
⚖️ Our take: If you see "3.71" floating around, that's the small-survey number. The big-cohort number is "about 1.8 times more likely."
Ter Beek 2023 looked at 300 women in one cardiology clinic. The 35% lifetime-ADHD-symptom number and the "two years younger" finding both come from that one sample. It's worth knowing about. It is not yet a settled population-level fact.
⚖️ Our take: Hypothesis-generating, not actionable cardiac risk advice. Replication needed.
Kooij and colleagues recommend that prescribers consider higher or supplemental stimulant dosing in the low-estrogen luteal phase. The biological logic tracks. The randomized controlled trial evidence isn't there yet.
⚖️ Our take: Worth raising with a sympathetic prescriber as something to try. Not worth treating as standard of care, because it isn't one yet.
The authors flag this themselves. Nobody has tracked the same women with ADHD through their 30s, 40s, and 50s with repeated cognitive testing. Without that, we can't say whether the cognitive dip during perimenopause recovers, persists, or worsens.
⚖️ Our take: The biggest research gap in the paper. Worth knowing if you're trying to plan around your own career trajectory. The data isn't there yet.
What To Do With This
Watch for the "smart but spacey" pattern. The classroom-disruptive boy is the stereotype. The quiet, anxious, perfectionist girl with a strong GPA and a chaotic backpack is also ADHD, just a presentation that doesn't make the teacher's life harder. If your daughter is masking through school on white-knuckle effort, that's not "she's fine." That's a known presentation.
Push for evaluation if any of these are true: persistent inattention, anxiety that ramps up around tests or transitions, perfectionism that costs sleep, social camouflaging that exhausts her, or a family history of late-diagnosed ADHD in women.
Specifically ask the evaluator about inattentive presentation and internalizing comorbidity. Use those words. They're the technical terms for what's getting missed in girls.
If you stopped feeling like yourself somewhere between 38 and 45, get an ADHD evaluation. Especially if your meds, your routines, or your coping strategies all stopped working at the same time. Perimenopause unmasks ADHD that's been compensated for since you were eight.
Track your symptoms by cycle phase before the appointment. Note when focus tanks, when anxiety spikes, when sleep falls apart. Two or three cycles is usually enough to see the pattern. A clinician with a chart in front of them is faster to diagnose than one staring at a verbal description.
Ask the prescriber specifically about cycle-adjusted dosing if your meds feel uneven. The research doesn't have RCTs yet, but the mechanism is well-established and most adult-ADHD specialists are familiar with the conversation.
If your cardiac symptoms are getting waved off and you have an ADHD history, push. The cardiology signal is preliminary, but it's enough that "I have ADHD and chest symptoms started in my 40s" is a sentence worth saying out loud at the appointment.
Logging cycle-phase symptoms alongside daily focus, sleep, and mood is exactly what Brainloot was built for. If you'd rather see the pattern than guess at it, that's the kind of thing the app makes easier.
Screen midlife women presenting with new "depression" or "anxiety" for ADHD history. The unmasking-at-perimenopause pattern is real and not yet on most primary-care radars. A brief lifetime ADHD screen catches a lot of late-diagnosis cases.
Screen for postpartum depression more aggressively in patients with ADHD histories. The Andersson 2023 Swedish registry put the rate at about 5 times the non-ADHD population. That's enough to justify a lower screening threshold.
Inattentive presentation with internalizing comorbidity is the female-typical pattern. Don't anchor on hyperactive-impulsive presentation when evaluating girls and women.
Estrogen feeds dopamine. ADHD brains run lean on dopamine. Every time estrogen drops (premenstrually, postpartum, in perimenopause) ADHD women pay a heavier toll than women without ADHD. About 1.6 times more premenstrual depression. About 5 times the rate of postpartum depression in a Swedish registry of 773,047 women. Almost 2 times more severe perimenopausal symptoms in a 5,392-woman cohort.
And about half of these women were never diagnosed as girls because they were the quiet, anxious, perfectionist ones who didn't make the teacher's life hard. If you stopped feeling like yourself in your 40s, your brain didn't break. Your estrogen did. There's a name for this, and there are clinicians who know it. That alone is worth the read.