Substance + claimed effects

Menstrual cycle tracking is the systematic recording of cycle-related data over time: cycle-start date and length, bleed days and flow heaviness, cycle-coupled symptoms (cramps, breast tenderness, headache/migraine, mood, sleep, skin, libido), and — for users who want it — fertility-window biomarkers (basal body temperature, cervical fluid texture, urinary LH). Tracking happens in paper charts, free apps (Clue, Flo, Apple Health), or paid stacks combining a wearable temperature sensor (Oura, Tempdrop) with an algorithm (Natural Cycles, Daysy). Claimed effects span four lanes: (1) early detection of hormonal disorders — polycystic ovary syndrome, thyroid dysfunction, hyperprolactinemia, primary ovarian insufficiency, functional hypothalamic amenorrhea; (2) detection of structural and bleeding-pathology disorders — heavy menstrual bleeding from fibroids, adenomyosis, polyps, and endometriosis; (3) contraception via fertility awareness; (4) fertility planning when trying to conceive. ACOG has positioned the cycle as a vital sign with diagnostic weight comparable to pulse or blood pressure on the basis that ovulatory function reflects the integrated state of the hypothalamic-pituitary-ovarian axis (ACOG 2015). Entry scope covers all four lanes plus the cardiometabolic risk signal that long-run cycle pattern carries (Solomon et al. 2001; Wang et al. 2020).

Evidence by addressing question

Mechanism

Each ovulatory cycle is the output of a tightly coupled feedback loop: the hypothalamus releases GnRH in pulses; the pituitary translates those pulses into FSH and LH; the ovary responds with follicular development, an estradiol surge, ovulation, and a corpus-luteum-driven luteal phase dominated by progesterone. Any insult to this loop — undernutrition, excess training load, chronic stress, hyperandrogenism, thyroid dysfunction, prolactinoma, premature follicle depletion — registers as a cycle-level signal: skipped cycles, lengthened cycles, shortened luteal phases, anovulatory bleeding. The cycle is a multi-organ readout; tracking surfaces that readout in a form a clinician can act on. A normal adult ovulatory cycle is 24–38 days bleed-onset to bleed-onset, bleed duration ≤8 days, with a luteal phase of roughly 11–14 days (Munro et al. 2018; Bull et al. 2019). The 28-day figure is a population mode, not a norm — real-world app data on 600,000+ cycles puts the mean cycle at 29.3 days with substantial intra-individual variability (Bull et al. 2019).

For TTC and FAM use, the mechanism is the fertile window: viable sperm survive in fertile cervical fluid for up to ~5 days, the oocyte for ~12–24 hours after ovulation, yielding a ~6-day fertile window that ends on ovulation day (Wilcox et al. 1995). Identifying that window requires combining cycle-day prediction with at least one biomarker — basal body temperature (BBT, retrospective ovulation confirmation), cervical fluid (prospective), urinary LH (prospective). Double-check methods combine two biomarkers.

Evidence

For early detection of cardiometabolic disease: irregular and long cycles across reproductive life are strongly associated with downstream type 2 diabetes and premature mortality. In the Nurses' Health Study II, women with very irregular cycles or cycle length ≥40 days carried a ~2.1-fold hazard for type 2 diabetes versus regular cyclers of 26–31 days, independent of BMI (Solomon et al. 2001). The same cohort followed to mortality endpoints showed that women with consistently irregular or long cycles across reproductive years had a hazard ratio of ~1.30–1.40 for premature all-cause mortality, with cardiovascular causes driving most of the excess (Wang et al. 2020). Mechanistically the excess risk runs through insulin resistance and hyperandrogenism — both detectable years before frank disease — which is why cycle pattern is a cheap leading indicator.

For PCOS: the Rotterdam consensus requires two of three criteria — oligo-/anovulation, clinical or biochemical hyperandrogenism, polycystic ovarian morphology on ultrasound — and oligo-/anovulation is exactly what tracking surfaces (Rotterdam 2004). Real-world prevalence is ~10% of reproductive-age women, the majority undiagnosed until infertility workup.

For structural pathology and bleeding disorders: NICE defines heavy menstrual bleeding (HMB) as bleeding that interferes with quality of life — a subjective threshold tracking makes quantitative (soaking through hourly protection, flooding, clots ≥2.5 cm, bleed duration ≥8 days) (NICE NG88 2018). The FIGO PALM-COEIN system stratifies abnormal uterine bleeding into structural causes (Polyp, Adenomyosis, Leiomyoma, Malignancy) and non-structural (Coagulopathy, Ovulatory dysfunction, Endometrial, Iatrogenic, Not yet classified) buckets — the diagnostic ladder downstream of tracking (Munro et al. 2018). Endometriosis presents primarily as cyclic pelvic pain plus dysmenorrhea; the median diagnostic delay across ten countries is ~6–7 years from symptom onset, much of it attributable to symptom normalisation rather than imaging failure (Nnoaham et al. 2011).

For PMDD: DSM-5 criteria require at least two consecutive prospectively recorded cycles showing severe affective symptoms confined to the luteal phase and remitting within days of bleed onset (Epperson et al. 2012). Retrospective recall over-weights recency and current mood; prospective daily tracking is the diagnostic instrument, full stop. Population prevalence is ~20% for PMS and 2–5% for PMDD (Yonkers et al. 2008).

For contraception: the symptothermal method (BBT + cervical fluid + calendar rules) has a perfect-use Pearl Index of ~0.4 per 100 woman-years and typical-use of ~1.8 per 100 woman-years in a 900-woman German prospective cohort (Frank-Herrmann et al. 2007). Algorithm-mediated app FAM (Natural Cycles) reports perfect-use 1.0 and typical-use 6.9 per 100 woman-years in a ~22,000-user observational cohort (Berglund Scherwitzl et al. 2017). These compare to typical-use failure rates of ~9% for the combined oral contraceptive, ~13% for condoms, ~21% for withdrawal, and ≤1% for IUD/implant (Trussell 2011). FAM done well beats condoms; it is materially worse than LARC.

For TTC: per-cycle conception probability peaks on the two days before ovulation (~30%) and is essentially zero outside the 6-day fertile window (Wilcox et al. 1995). Confining intercourse to the fertile window does not raise cumulative pregnancy rate over intercourse every 1–2 days but shortens time-to-pregnancy and is the standard recommendation when intercourse frequency is limited (Stanford et al. 2002).

Protocol

Minimum-viable tracking captures: cycle-start date (first day of red bleeding, not spotting), bleed length, flow heaviness per bleed day, and the presence/absence of cyclic symptoms (cramps, breast tenderness, headache, mood, sleep changes). This alone supports cycle-as-vital-sign surveillance plus PMDD and HMB workup. ACOG recommends a free app or calendar; recall is unreliable for symptom-cycle correlation (ACOG 2015).

For fertility awareness (TTC or FAM contraception), add at least one ovulation biomarker:

• Basal body temperature (BBT) — first-thing-in-the-morning oral temperature, taken before getting out of bed, with a basal thermometer reading to 0.05 °C. The post-ovulatory progesterone rise lifts BBT by ~0.3 °C; a sustained rise across 3+ days confirms ovulation retrospectively. Wearable continuous temperature (Tempdrop, Oura) improves signal in users with disrupted sleep schedules.
• Cervical fluid observation — daily check; oestrogen-driven fertile fluid (clear, stretchy, slippery, egg-white texture) signals the impending fertile window prospectively.
• Urinary LH testing — daily strip across mid-cycle; the LH surge precedes ovulation by ~24–36 hours, identifying peak fertility prospectively.

Double-check methods (BBT + cervical fluid; LH + BBT) are the canonical FAM contraceptive protocol; single-biomarker methods (calendar-only, Standard Days) carry ~5× the failure rate of double-check methods (Frank-Herrmann et al. 2007; Trussell 2011).

Tracking horizon for clinical use: ACOG recommends ≥3 cycles of baseline data before drawing regularity inferences (ACOG 2015); PMDD diagnosis requires ≥2 consecutive cycles of daily ratings (Epperson et al. 2012); HMB workup is triggered by a single cycle meeting criteria, not after a delay (NICE NG88 2018).

Contraindications

Tracking itself is benign — pen-and-paper or app, no physical intervention. The caveats are about the uses tracking enables:

• FAM as primary contraception is inappropriate when pregnancy must be avoided categorically (teratogenic medications, severe maternal illness). Typical-use failure rates of 1.8–6.9 per 100 woman-years are unacceptable in those settings (Frank-Herrmann et al. 2007; Berglund Scherwitzl et al. 2017).
• FAM signal is degraded during postpartum/lactational amenorrhea, perimenopause, and the first 1–3 cycles after hormonal contraception discontinuation — cycles are unpredictable and biomarkers noisy. Bridging contraception is standard.
• Cycle tracking under hormonal contraception measures the regimen, not the underlying cycle; the HPO axis is suppressed and withdrawal bleeds are pharmacological. Diagnostic conclusions about ovulatory function require ≥3 cycles off hormonal contraception.
• Data privacy is non-trivial in jurisdictions with criminalised abortion access; cycle data sold to third parties or subject to subpoena carries legal exposure post-Dobbs in some US states. Local-only or end-to-end-encrypted apps are preferred for users with that threat model.

Misconceptions

• "28 days is normal; anything else is irregular." 24–38 days is the normal adult range; the population mean is ~29 days; intra-individual variation of ±7 days across a year is common in healthy ovulatory cyclers (Munro et al. 2018; Bull et al. 2019).
• "Ovulation is day 14." Day-14 ovulation requires a 28-day cycle with a 14-day follicular phase, which is one mode among many. The luteal phase is the stable half (~11–14 days); follicular-phase length is what varies. Ovulation can occur anywhere from cycle day 8 to day 25+ across women with regular-by-definition cycles (Wilcox et al. 1995).
• "Bleeding while on the pill is a period." It is a withdrawal bleed driven by the hormone-free interval; absent or skipped withdrawal bleeds during continuous regimens are not pathological.
• "Apps predict ovulation accurately from calendar data alone." Calendar-only prediction has ~25% error against confirmed ovulation; biomarker-confirmed methods (BBT + cervical fluid, or LH + BBT) are required for FAM-grade prediction (Frank-Herrmann et al. 2007).
• "Heavy periods are just what some people get." HMB has identifiable causes (PALM-COEIN) and effective treatments; iron-deficiency anaemia from chronic HMB is a common and reversible source of fatigue (Munro et al. 2018; NICE NG88 2018).
• "Tracking PMS is just navel-gazing." PMDD diagnosis is impossible without two prospective cycles of daily ratings; retrospective recall is biased by current mood (Epperson et al. 2012).
• "Lunar / cycle-syncing diet and workout plans are evidence-based." They are not. Cyclic-phase performance prescriptions sold by wellness influencers are not anchored to controlled trials; the underlying intra-individual signal-to-noise is too low to drive day-level decisions.

Failure modes

• App-predicted "ovulation day" treated as confirmation. Most consumer apps predict ovulation by averaging past cycle length minus 14 days. That is calendar-method-grade prediction, not biomarker-confirmed ovulation.
• Tracking too short. One cycle of data is descriptive, not diagnostic. PCOS oligo-ovulation and PMDD diagnosis require multi-cycle data.
• BBT measurement noise. Shift work, parenting small children, alcohol use, illness, and inconsistent wake times all degrade BBT signal. Wearable continuous measurement smooths some of this, but the underlying signal still requires reasonably stable sleep.
• Cycle suppression interpreted as health. Hypothalamic amenorrhea from undereating or overtraining is silent on the tracking dashboard — there is no bleed to log — but is a leading indicator of bone loss, cardiovascular dysfunction, and infertility (Gordon et al. 2017; Mountjoy et al. 2014).
• Symptom-cycle correlation confused with causation. Many symptoms are cyclic without being driven by the cycle (e.g. coffee intake correlated with workweek). Disentangling requires separate tracking of the candidate cause.
• Data-action gap. Tracking that surfaces irregularity but never makes it to a clinician captures the risk signal without acting on it. ACOG's framing is explicit: the cycle is a vital sign meant to enter the clinical record (ACOG 2015).

Practicalities

Cost: paper chart or basic app (Clue, Flo, Apple Health, paper calendar) is free. Paid apps with FAM-grade algorithms (Natural Cycles ~$90/year plus thermometer; Daysy device ~$330 one-time) sit at $30–$330 inclusive of hardware. Wearable temperature monitors (Oura ~$300 hardware plus ~$70/year subscription; Tempdrop ~$240) overlap with sleep-tracking spend and aren't strictly necessary — a $5 oral basal thermometer plus discipline gives the same retrospective ovulation confirmation. Effort: 30–60 seconds/day for symptom + flow logging; +1–2 minutes/day for BBT and cervical-fluid observation when used.

Data export and clinician hand-off: most apps export a CSV or PDF summary; many gynaecologists accept screenshots. Algorithm-derived "predictions" are not data — the underlying timestamps and counts are. When bringing data to a clinical appointment, bring the raw cycle log, not the app's interpretation.

Audience

• Adolescents and early-20s users establishing the baseline pattern; ACOG explicitly targets this group with the vital-sign framing (ACOG 2015).
• Athletes and high-training-load users at risk of functional hypothalamic amenorrhea and RED-S; absent or lengthening cycles are the canary (Gordon et al. 2017; Mountjoy et al. 2014).
• Users coming off hormonal contraception who plan to conceive — 3–6 cycles of off-method tracking surfaces whether the underlying cycle is ovulatory.
• Users with suspected PMDD or menstrual migraine — diagnosis requires prospective tracking by definition (Epperson et al. 2012).
• Perimenopausal users (typically 40s–early 50s) — cycle-length variability of ≥7 days between consecutive cycles is one of the STRAW+10 markers of the early menopause transition.
• Users with family history of PCOS, endometriosis, fibroids, or early menopause — pretest probability is high enough that early signal warrants action (Rotterdam 2004; Nnoaham et al. 2011).

Alternatives

For ovulation confirmation specifically (not general cycle tracking):

• Serum progesterone at cycle day 21 (or mid-luteal, indexed off the next bleed in non-28-day cyclers) — gold standard for ovulation confirmation in clinical workup.
• Transvaginal ultrasound follicle tracking — gold standard for ovulation timing in IVF/IUI cycles; impractical for self-monitoring.
• Saliva ferning — low-cost ovulation prediction via salivary oestrogen crystal patterns; specificity is poor in head-to-head comparison vs urinary LH.
• Standard Days Method / CycleBeads — calendar-only contraception for users with consistently 26–32 day cycles; typical-use failure ~12% per year (Trussell 2011).

For PMS/PMDD assessment: the DRSP (Daily Record of Severity of Problems) is the validated paper instrument; apps that implement DRSP-compatible logging (e.g. Me v PMDD) carry diagnostic weight.

Stakes

Without tracking, multi-year diagnostic delays dominate the cost: endometriosis median 6–7 years from symptom onset across ten-country sampling (Nnoaham et al. 2011); PCOS typically >2 years and a majority of cases undiagnosed until infertility workup (Rotterdam 2004); PMDD characteristically misattributed to "mood swings" without prospective data (Epperson et al. 2012). Downstream costs of those delays are concrete: chronic pelvic pain, iron-deficiency anaemia from sustained HMB (NICE NG88 2018), bone loss from unrecognised hypothalamic amenorrhea (Gordon et al. 2017), lost fertility window from undiagnosed PCOS, missed cardiometabolic risk signal that would have triggered earlier insulin-sensitivity and lipid monitoring (Solomon et al. 2001; Wang et al. 2020).

Payoff

Within 3–6 cycles: baseline established; abnormal-bleeding criteria triggerable; PMS/PMDD/menstrual-migraine pattern legible; flow-volume data quantitative enough to surface HMB (NICE NG88 2018). Within 6–12 months: fertility-window prediction reliable for TTC users; FAM-as-contraception protocol established for users who chose that lane (Frank-Herrmann et al. 2007). Over years: long-run cardiometabolic risk signal surfaces — chronic irregularity, very long cycles, or oligomenorrhea trigger insulin-sensitivity workup years before frank type 2 diabetes (Solomon et al. 2001). Across reproductive life: perimenopause is identified by tracking pattern shift rather than by symptom severity, opening HRT/MHT discussions earlier (Wang et al. 2020).

Out of scope

Hormonal contraceptive choice itself; specific PCOS / endometriosis / fibroid management protocols; HRT/MHT decisioning for perimenopause; specific IVF/IUI fertility protocols; pelvic-floor and dysmenorrhea-specific physiotherapy. These are downstream of the data tracking generates and warrant their own entries.

The credibility range

Optimist case

Cycle tracking is the cheapest, lowest-friction surveillance any patient does for any organ system. Ovulatory function is a high-information readout of hypothalamic, pituitary, thyroid, adrenal, ovarian, and metabolic health — closer to an integrated systems vital sign than to a single biomarker. ACOG, AAP, FIGO, and SOGC have aligned on this framing for nearly a decade (ACOG 2015). Large prospective cohorts demonstrate that long-run cycle pattern predicts cardiometabolic outcomes years ahead of standard screening (Solomon et al. 2001; Wang et al. 2020). For PMDD and HMB, prospective tracking is the diagnostic instrument; there is no substitute (Epperson et al. 2012; NICE NG88 2018). For FAM contraception, the symptothermal method's perfect-use efficacy beats condoms and approaches the pill (Frank-Herrmann et al. 2007). Wearable continuous biomarkers are turning what used to be a high-discipline practice into something passive. Marginal effort is a minute per day; worst case is no information; best case is a diagnosis a decade earlier.

Skeptic case

Tracking-as-screening assumes the user can and will act on signal. The data-action gap is huge: most observed irregularity goes unaddressed because primary-care gynecology under-allocates time to cycle history under typical reimbursement structures. Consumer apps optimise for engagement and ad/subscription revenue, not diagnostic validity — most app "ovulation day" predictions are calendar-method-grade and should not be treated as confirmation. The Pearl Index numbers for FAM-as-contraception look acceptable in motivated cohorts (Frank-Herrmann et al. 2007) but degrade sharply in unselected populations — typical-use 6.9/100 woman-years for the leading FAM app is ~5× the failure rate of an IUD (Berglund Scherwitzl et al. 2017; Trussell 2011). The supplement-stack ROI on wearable temperature sensors is empirically thin; ovulation can be confirmed for free with a $5 thermometer. Data privacy is a real harm vector post-Dobbs. And no RCT directly joins "consumer tracking adoption" to "better health outcome" — the case rests on (a) tracking → earlier diagnosis and (b) earlier diagnosis → better outcome, both well-supported individually but never joined in a single trial.

Author's call

The baseline case for tracking is solid and uncontested at the guideline level (ACOG 2015); diagnostic value for PMDD, HMB, PCOS evaluation, and perimenopause identification is direct, not inferred. Contested terrain is FAM-as-primary-contraception (acceptable in motivated users with bridging methods available; not acceptable as a categorical pregnancy avoidance) and wearable-supplement-stack ROI (marginal over a free thermometer for most users). The entry recommends baseline tracking strongly; FAM-grade biomarker addition is conditional on the user's TTC or contraceptive goals. Evidence rating 4 — multiple large cohorts, alignment of major guideline bodies, direct diagnostic-criterion dependence for PMDD/HMB; the gap is a direct outcome RCT. Controversy rating 2 — baseline tracking is uncontested; FAM efficacy and wearable ROI are mildly contested at the margins.

Stakeholder + incentive map

• Commercial. Period-tracker apps (Clue, Flo, Stardust) monetise via subscription and historically via data sale; FAM apps (Natural Cycles, Daysy) monetise via subscription plus hardware; wearables (Oura, Apple, Garmin) bundle cycle tracking into broader fitness subscriptions. Engagement is the optimised KPI, not diagnostic accuracy.
• Professional. ACOG, AAP, SOGC, FIGO have aligned on cycle-as-vital-sign; reproductive endocrinologists depend on patient cycle data for workup; menstrual-health epidemiologists (Wilcox, Solomon, Wang, Bull) generated the underlying outcomes literature.
• Counter-incentive. Primary-care gynecology under-allocates time to cycle history under typical reimbursement; LARC providers and hormonal-contraception manufacturers have no incentive to promote FAM; the data-broker industry monetises cycle data sold by free-tier apps.
• Cultural. "Tracking your cycle" carries both a wellness-influencer aesthetic (cycle syncing, lunar tracking, phase-based workouts) — many claims of which are not evidence-based — and a feminist health-literacy frame. Both push adoption; the former erodes signal quality.

Population variability

• Age. Cycle variability is highest in the 2 years post-menarche and across perimenopause; central ovulatory years (mid-20s to late-30s) are the most predictable.
• Hormonal contraception users. The underlying cycle is suppressed; tracking measures the regimen. Diagnostic conclusions about ovulatory function require off-method data (typically ≥3 cycles).
• Athletes. High-training-load users are at elevated risk of functional hypothalamic amenorrhea; cycle silence is the leading indicator of RED-S (Mountjoy et al. 2014; Gordon et al. 2017).
• PCOS. ~10% of reproductive-age women globally; presents with oligo-/anovulation that tracking surfaces (Rotterdam 2004).
• Endometriosis. ~10% of reproductive-age women; tracking surfaces cyclic pelvic pain that is otherwise normalised (Nnoaham et al. 2011).
• Post-pill window. 1–3 months of cycle irregularity after discontinuing hormonal contraception is typical and resolves; persistent irregularity past 3 months warrants workup.
• Postpartum / lactational. Cycle return is variable post-delivery, often anovulatory for several cycles; FAM signal is unreliable across this window.

Knowledge gaps

• No RCT directly linking consumer cycle-tracking adoption to clinical-outcome improvement (earlier diagnosis, fewer years to PCOS/endometriosis identification, reduced HMB burden). The case is built from epidemiology plus diagnostic-criterion dependence — both rigorous, neither head-to-head.
• Algorithmic FAM efficacy varies substantially across apps; comparative head-to-head data is limited. The Natural Cycles data is the most published; other apps' algorithmic claims are largely unaudited.
• Wearable temperature ROI vs $5 oral basal thermometer is empirically thin; manufacturer-funded studies dominate the literature.
• Cycle-related cognitive and athletic performance changes are heavily studied at the population level but per-individual signal-to-noise is unclear; intra-individual effect sizes are small relative to common confounders.
• Data-privacy harm rates post-Dobbs in the US are documented narratively (subpoenas of period-app data, location-correlated targeting) but not yet quantified at the population level.
• Perimenopause prediction from cycle data alone — the STRAW+10 framework relies on cycle-length variability plus bleeding-pattern shift, but day-level prediction of menopause is not currently feasible from cycle data alone.

Scope

Brief named four consequences (hormonal, structural, metabolic disorder detection; contraceptive planning; fertility awareness). All four covered end to end. The dossier surfaced a fifth load-bearing consequence the brief did not name — the cycle-as-cardiometabolic-canary signal (Solomon 2001, Wang 2020) — which is covered as part of the "metabolic disorder detection" lane and drives the longevity score.

Audience scoping

Female, ages 18-39 and 40-59. 60+ excluded because post-menopausal users do not have a cycle to track; the perimenopause case lives at the 40-59 edge and is covered explicitly in audience and payoff.

Hard rating calls

• Mood = 3. Driven by PMDD diagnostic-instrument dependence (Epperson 2012). The population-average effect is smaller than a 3 would imply if read as "every user gets a transformative mood lift," but the rubric is holistic — the substance produces this effect for the affected subset, full stop, and the unlock is a real psychiatric intervention. 3 is the honest call.
• Longevity = 2 (not 3). Cardiometabolic risk signal is real and replicated, but the causal chain runs tracking → diagnosis → intervention → outcome. Holistic effect lands at "small additive on mortality risk" rather than "meaningful disease prevention."
• Sleep = 1, Focus = 1. Tracking does not produce these effects directly. The small score reflects that anticipating cycle-coupled sleep disruption and concentration dips is a real, named benefit even though the floor does not move.
• Beauty (both) = 0. PCOS-driven acne and hirsutism are real and tracking enables their detection, but appearance change is two steps downstream (tracking → diagnosis → treatment → look). Honest zeros.
• Action = test (not do). Considered "do" (daily habit fits). Picked "test" because the semantic intent is home data-gathering — the action is measuring yourself, not behavior change.

Contested terrain handled honestly

• Fertility-awareness as primary contraception. Both the rigorous symptothermal number (Frank-Herrmann 2007: typical-use 1.8/100 woman-years) and the app-based number (Berglund Scherwitzl 2017: typical-use 6.9/100 woman-years) reported. Not recommended as default — flagged as inappropriate when pregnancy must be avoided categorically.
• Wearable continuous temperature ROI. Called out as not strictly necessary over a $10 oral thermometer for most users; worth it for shift workers and parents of small infants.
• Cycle syncing / phase-based wellness prescriptions. Pushed back on explicitly in misconceptions — not anchored to controlled trials.
• Data privacy post-Dobbs. Surfaced as a contraindication-callout topic without jurisdiction-narrowing the rest of the entry. Real harm vector; readers need to know.

Separate-entry candidates (backlog)

• Polycystic ovary syndrome — workup, lifestyle and pharmacological management
• Endometriosis evaluation and management
• Heavy menstrual bleeding pathway (PALM-COEIN workup, treatment options)
• Iron-deficiency anaemia screening and replacement
• Hormonal contraception choice (pill vs IUD vs implant vs ring vs patch)
• Perimenopause and menopause hormone therapy decisioning
• Functional hypothalamic amenorrhea and RED-S in athletes
• Menstrual migraine and cycle-linked headache

These are flagged in the article's closing addressing section as reader-facing forward pointers; they should be wired up as related entries when the corresponding entries land.

Citation discipline

17 citations added — all with real DOIs or stable guideline URLs. No preprints. Treloar 1967 (the classic cycle-variability dataset) considered and skipped — Bull 2019 covers the same ground with much larger n and a real DOI. ACOG Committee Opinion 651 used via URL rather than DOI (more durable for the guideline link).