PMDD and ADHD: the week before your period

What PMDD actually is

Premenstrual Dysphoric Disorder is a luteal-phase depressive disorder in Section II, Depressive Disorders of the DSM-5-TR — moved there from the appendix in 2013. The criteria are specific: at least five of eleven listed symptoms in the final week before menses, with at least one from the mood cluster (marked affective lability, irritability, depressed mood, or anxiety); symptoms substantially absent in the post-menstrual week; functional impairment that interferes with work, relationships, or usual activities; and the whole pattern confirmed by prospective daily ratings across at least two symptomatic cycles. Until confirmed prospectively, the diagnosis is provisional.

PMDD is distinct from PMS, which is a non-clinical term covering common cyclical discomfort and isn’t in the DSM at all. Roughly 75% of menstruating people experience some premenstrual syndrome (PMS); only about 3 to 8% meet PMDD criteria. PMDD is also not a hormone imbalance — the Schmidt NIMH studies (1998⁴, replicated 2017³) showed women with and without PMDD have identical circulating hormone levels. Suppressing the cycle with leuprolide eliminates symptoms in PMDD women, and adding back the same hormone doses brings the symptoms back only in the PMDD group. PMDD is a brain-sensitivity disorder, not a hormone-level one.

The PMDD × ADHD overlap

The cleanest dataset on this comorbidity is Dorani, Bijlenga, Beekman, van Someren & Kooij (2021)², an Amsterdam survey comparing women with ADHD against non-ADHD controls.

• Lifetime PMDD: 45.5% in ADHD women vs 28.7% in controls.
• Current PMDD: 31.7% vs 8.3%.
• Lifetime postpartum depression (in those who’d had a pregnancy): 16.8% vs 7.3%.
• Climacteric depression: 51.7% vs 23.5%.

Caveats: single site, self-report screening rather than two months of daily ratings. That likely overestimates strict-criteria PMDD prevalence in both groups. The relative effect — roughly three times the current-PMDD rate in ADHD women — is what holds.

Mechanistically the overlap is plausible. Estrogen modulates dopamine release, D2 receptor density, and transporter expression in prefrontal cortex and striatum (Jacobs & D’Esposito, 2011). ADHD already sits on compromised tonic dopaminergic signaling. The late-luteal estrogen drop removes a compensatory amplification that non-ADHD women weren’t leaning on as heavily. Same hormonal shift, larger symptomatic effect — the same backbone as the perimenopause crash, compressed into 7 to 10 days every month.

PME vs PMDD — the differential

Premenstrual Exacerbation (PME) is when an existing disorder gets worse in the luteal phase but is also present the rest of the month. For ADHD: the emotional dysregulation, the rejection sensitive dysphoria (RSD) on feedback, the executive collapse — all amplify late in the cycle, but none of them were absent in the follicular week. They were just more manageable.

PMDD requires the cluster of luteal symptoms to be substantially absent in the follicular week. The follicular phase has to look meaningfully different, not just less intense. Hartlage et al. (2014)⁶ showed that of women presenting with cyclical symptoms, the majority have PME of an underlying disorder rather than pure PMDD. For ADHD women specifically, this is the entire clinical puzzle — and the way you tell is two months of prospective daily ratings using the Daily Record of Severity of Problems (DRSP) or the IAPMD tracker. Retrospective recall doesn’t resolve it; women routinely overestimate the luteal-follicular difference when asked at the appointment.

Both can be true. A woman can have PME of her ADHD and also meet PMDD criteria. The Dorani 31.7% current-PMDD figure is on top of whatever cycle-related ADHD symptom variation she also has. The point of naming PME explicitly is that most clinicians who know about PMDD will assume that’s what an ADHD woman is reporting, and most who don’t will say “that’s just your ADHD.”

The luteal suicidality window

This part doesn’t get omitted. Eisenlohr-Moul et al. (2022)⁷, studying patients with prospectively confirmed PMDD, found a lifetime suicide-attempt prevalence around 30%, with clustering in the late luteal phase. IAPMD community surveys (not peer-reviewed, flagged accordingly) report over 70% of PMDD respondents experiencing passive suicidal ideation tied to luteal phase and roughly 30% reporting active.

For ADHD women the picture is layered: RSD-driven suicidal ideation and PMDD-cyclical suicidal ideation can coexist and present similarly. The cyclical timing is the diagnostic clue. A monthly pattern of dark windows that resolve at menses is information for a prescriber — that pattern is not normal ADHD, not normal depression, and not normal PMS.

Treatment, in order

First-line: selective serotonin reuptake inhibitor (SSRI), luteal-phase or continuous. Sertraline, fluoxetine, paroxetine, and escitalopram all have evidence; fluoxetine and sertraline have the most PMDD-specific data. In PMDD the SSRI effect is faster than in major depression — often within days — and lower doses (sertraline 25–50 mg) are frequently sufficient. Luteal-only dosing (cycle day 14 through menses) works as well as continuous in mild-to-moderate cases and reduces side-effects. The Cochrane review (Marjoribanks et al., 2013)⁸ finds that roughly one in five women treated will see clear benefit specifically from the drug; the 2023 ACOG guideline endorses SSRI as first-line.

Second-line: combined oral contraceptive pill (OCP) with drospirenone (Yaz/Yasmin), Food and Drug Administration (FDA) approved specifically for PMDD on the 24/4 regimen. Suppresses ovulation; drospirenone has antimineralocorticoid and antiandrogenic activity. Caveat: hormonal contraceptives interact with ADHD symptoms in their own right — some ADHD women report a flat-but-lower mood floor on combined OCPs. Trade-off worth making explicit with the prescriber.

Third-line: gonadotropin-releasing hormone (GnRH) analogs (leuprolide and similar) with add-back hormone replacement therapy (HRT). Chemical menopause; used when SSRI and OCP fail. Add-back estrogen and progesterone are required to prevent bone-density loss, and the add-back itself can trigger symptoms in PMDD women, requiring titration. Used as both treatment and diagnostic confirmation in the Schmidt protocol. Last-line: bilateral oophorectomy — curative but irreversibly surgical menopause, reserved for severe, treatment-refractory cases after a successful GnRH trial.

What does not have good evidence: calcium, vitamin B6, magnesium, chasteberry, evening primrose oil — small trials, weak signals, not substitutes. Progesterone supplementation often makes PMDD worse, because the disorder is a sensitivity to luteal-phase hormones, not a deficiency of them. “Cycle syncing” lifestyle frameworks are not harmful and not treatment for PMDD.

Stimulants are not PMDD treatment. They are the ADHD treatment, and if the ADHD is well-treated, the woman is better positioned to identify the residual luteal-phase depressive disorder cleanly. If the ADHD is undertreated, everything in the luteal phase looks worse and the differential becomes impossible.

The two-prescriber problem

The obstetrician-gynaecologist (OB-GYN) is positioned to diagnose PMDD but typically doesn’t ask about ADHD. Most gynae intake uses the Premenstrual Symptoms Screening Tool (PSST) or a single-cycle recall, not two cycles of DRSP, and doesn’t screen for ADHD. PMDD gets diagnosed, ADHD gets missed, SSRI gets prescribed, ADHD symptom load is unaffected.

The psychiatrist or ADHD prescriber is positioned to diagnose ADHD but typically doesn’t ask about cycle timing. ADHD intake asks about persistent symptoms; PMDD presents cyclically. ADHD gets diagnosed, PMDD gets missed, the stimulant gets prescribed, the luteal depressive episodes are unaddressed and sometimes worsened. No US guideline currently requires either prescriber to screen for the other.

What to bring to the appointment: a two-cycle DRSP record (the IAPMD tracker is free), the Dorani 2021 prevalence figures, and one specific ask — “I’d like to discuss whether this is PMDD or PME of my ADHD, and what two months of prospective tracking would clarify.” The practical payload of this article is mostly that single sentence.

What we don’t know yet

The general-population PMDD prevalence range is genuinely wide (1.8 to 8% strict criteria, broader by other thresholds) because diagnostic methods differ across studies. The Dorani ADHD-specific figures are the best we have but they aren’t US population data and aren’t prospective. Why some women are CNS-sensitive to ovarian-steroid withdrawal and others aren’t is the current research frontier — ESR1 polymorphisms and GABA-A receptor variants are candidates, not answers. PMDD-after-tubal- ligation and PMDD-after-IUD-insertion are real community patterns with thin published evidence. ADHD-medication adjustment in the luteal phase has minimal trial support; the practice is grounded in mechanistic reasoning and clinical experience, not the same evidence level as the SSRI treatment.

Reading honestly: the comorbidity is real, the underdiagnosis is real, the two-prescriber gap is real, and the mechanism story is most of the way there but not finished. The thing that doesn’t change with more data is the practical move — two cycles of prospective ratings, one specific question at the appointment, and an honest accounting of whether what’s happening on day 23 belongs to the ADHD, the cycle, or something the prescribers haven’t named yet.