Women's Health

The Research Gap — and Why It Matters

Women's health has been systematically understudied. The NIH only mandated inclusion of women in federally funded clinical trials in 1993. Before that, most foundational drug, nutrition, and exercise research was conducted on male subjects, with findings extrapolated across sex — an assumption that fails routinely when examined.

The consequences are significant: women metabolize medications differently, respond to exercise protocols differently, and have distinct cardiovascular disease presentations that were missed in male-dominated research for decades. This is not a minor gap. It is a reason to be especially attentive to study populations when evaluating evidence in this domain.

The good news: the last decade has seen a substantial increase in high-quality women's health research, particularly on the menstrual cycle, perimenopause, and hormonal contraception. The evidence base is now large enough to draw specific, actionable conclusions.

The Menstrual Cycle as a Biological Variable

The menstrual cycle is a repeating natural experiment that most women do not track systematically enough to understand. Across the roughly 28-day cycle, estrogen, progesterone, LH, and FSH follow predictable trajectories that affect cognition, mood, energy, pain tolerance, and physical performance.

Key findings from cycle-phase research:

Follicular phase (days 1–14, rising estrogen): Some studies report differences in verbal memory, fine motor skill, attention, and hippocampal structure or activity across cycle phases, but findings are not uniform enough to make a simple "best phase" rule. Protopopescu et al. reported hippocampal structural differences and higher verbal declarative memory in a late-follicular/postmenstrual scan compared with a late-luteal/premenstrual scan; this should be treated as mechanistic background, not a direct performance prescription.

Ovulation (day ~14): Pain tolerance is at its monthly peak due to estrogen's analgesic effect. This is the optimal window for procedures, dental work, or high-intensity training with elevated injury risk.

Luteal phase (days 15–28, rising progesterone): Resting metabolic rate increases approximately 100–300 kcal/day. Carbohydrate cravings increase due to progesterone's effect on serotonin production. Sleep architecture shifts: body temperature rises, REM sleep decreases, sleep quality ratings decline in the week before menstruation. Core body temperature remains approximately 0.3–0.5°C elevated throughout luteal phase.

Premenstrual (days 24–28): PMS affects 75% of women to some degree; PMDD (a diagnosable disorder with functional impairment) affects approximately 5–8%. The DSM-5 criteria for PMDD require prospective symptom tracking for two cycles — which is itself an acknowledgment that retrospective report is insufficient.

The key insight for self-experimentation: tracking symptoms, performance, and mood by cycle phase for 2–3 months reveals whether and how much your own performance varies, which is not predictable from population data alone.

Exercise Performance and the Menstrual Cycle

A systematic review and meta-analysis by McNulty et al. (78 studies, over 1,000 women) found large variation between studies and mostly low-quality evidence. Average effects of menstrual-cycle phase on exercise performance were small enough that general training rules should not be formed from the literature alone. The more defensible takeaway is personalized: track your own performance and recovery by cycle phase before changing training.

Strength training adaptations may be modulated by cycle phase. A 2021 RCT by Wikström-Frisén et al. found that women who performed higher volume training in the follicular phase and lower volume in the luteal phase gained significantly more muscle mass than those following a flat weekly schedule (p = 0.04). The sample was small (n = 59) but the mechanistic logic is sound: estrogen is anabolic and has anti-catabolic effects on muscle.

Injury risk: ACL tears are approximately 2–8x more likely during ovulation in female athletes (across multiple prospective studies), likely due to estrogen's effect on ligament laxity. This is a meaningful enough finding to influence training periodization.

Contraception and Biological Measurement

Hormonal contraception — used by approximately 65% of women of reproductive age — alters the endocrine environment substantially. This has direct implications for self-experimentation.

Combined oral contraceptives suppress endogenous estrogen and progesterone and replace them with synthetic analogs. The result: the natural cycle-phase variation in cognition, mood, and performance is largely eliminated. HRV patterns differ from naturally cycling women. SHBG (sex hormone binding globulin) rises, which lowers free testosterone and can reduce libido and mood in a subset of users.

A 2016 Danish cohort study (n = 1,061,997, 13-year follow-up) found combined OCP users had significantly higher rates of antidepressant initiation versus non-users (hazard ratio 1.23 overall, rising to 1.80 for adolescents). The study was observational with confounders, but the size and effect consistency are notable.

For self-experimenters on hormonal contraception: cycle-phase protocols do not apply. Tracking should be week-over-week rather than cycle-phase-relative. This doesn't make measurement less useful — it makes it differently structured.

Perimenopause: The Understudied Transition

Perimenopause — the 4–10 year transition leading to menopause — is one of the most biologically disruptive periods in female physiology, and one of the most poorly managed in clinical practice. The average age of onset is mid-40s; menopause (12 months of amenorrhea) occurs at a median age of 51.

During perimenopause, estrogen fluctuates unpredictably rather than following a predictable cycle. These fluctuations drive the classic symptoms: vasomotor events (hot flashes), sleep disruption, mood instability, cognitive changes (the "brain fog" that women consistently report but was historically dismissed), and changes in body composition.

The cognitive effects are real and measurable. A longitudinal study by Maki et al. followed women through the menopause transition and found objective declines in verbal memory and processing speed during perimenopause, with partial recovery post-menopause — suggesting a transitional, not permanent, effect.

Hormone Replacement Therapy (HRT): The WHI (Women's Health Initiative) study in 2002 caused a major shift away from HRT and estrogen-progestogen therapy. Subsequent analyses and newer research have made the picture more nuanced. The 2022 North American Menopause Society position statement concludes that, for many women younger than 60 or within 10 years of menopause onset and without contraindications, the benefit-risk ratio is favorable for treating vasomotor symptoms and preventing bone loss. For women who start later, the balance is less favorable because absolute risks of coronary heart disease, stroke, venous thromboembolism, and dementia are higher. This is clinical decision territory, not a self-experimentation protocol.

Modern formulations (transdermal estradiol, micronized progesterone) have a better safety profile than the conjugated equine estrogen + medroxyprogesterone acetate used in WHI. This distinction matters for clinical decision-making.

Bone Density, Estrogen, and the Long Game

Bone density peaks around age 30 and declines from there, with a sharp acceleration in the first 5–10 years post-menopause due to estrogen loss. The lifetime fracture risk for women is approximately 50% — higher than for breast cancer, heart attack, and stroke combined.

Resistance training is the most effective non-pharmacological intervention for bone density preservation. A meta-analysis by Zhao et al. (2015, 13 RCTs) found resistance training significantly increased lumbar spine BMD in postmenopausal women (standardized mean difference 0.38, p < 0.001). Impact exercise (jumping protocols) shows larger effects than non-impact resistance training on hip bone density specifically.

DEXA scans (dual-energy X-ray absorptiometry) measure bone mineral density with high precision and are the gold standard for tracking.

What to Measure

• Cycle tracking (Clue, Natural Cycles, or paper log): day of cycle, energy (1–10), mood (1–10), pain, sleep quality — minimum 2–3 cycles for pattern recognition
• Basal body temperature: rises 0.2–0.5°C at ovulation; tracked with a basal thermometer (±0.01°C precision), confirms ovulation timing and luteal phase length
• HRV by cycle phase: document whether HRV follows the expected drop in luteal phase; individual variation is high
• Vasomotor event frequency and severity (for perimenopause): log time, duration, and sleep disruption; establishes baseline for intervention trials
• DEXA scan (every 2–3 years from age 40+): bone density, visceral fat, lean mass — more informative than the scale

What to Experiment With

→ Resistance training periodization by cycle phase (higher volume follicular, lower volume luteal) → 3-month strength progression and subjective recovery rating Tests the Wikström-Frisén protocol on your own data. The experiment requires consistent cycle tracking and workout logging simultaneously.

→ Magnesium glycinate (300–400 mg/day, days 14–28) → PMDD or PMS symptom severity in luteal phase A 1991 RCT by Facchinetti et al. found magnesium supplementation significantly reduced premenstrual mood symptoms (p < 0.01). Small study, but mechanistically plausible and easy to test with cycle-tracked symptom logs.

→ Consistent sleep timing (within 30-minute window) → sleep quality rating and HRV across cycle phases Tests whether circadian consistency modulates the luteal-phase sleep disruption that most cycling women experience; requires 2-cycle baseline before intervention.

→ Time-restricted eating window → fasting glucose, energy, and cycle regularity tracked over 3 months Intermittent fasting research is almost entirely conducted on men; the evidence for women specifically is mixed, with some data suggesting menstrual cycle disruption at aggressive restriction. This experiment tests your personal response rather than assuming male findings apply.

Measuring Your Own Hormonal Patterns

The 28-day cycle is a natural experiment that most women run without tracking the outcome variables. Two cycles of structured daily logging — energy, mood, sleep, pain, performance — provides more personalized information than any population study. Layer in basal body temperature and HRV and you have a multivariate dataset specific to your physiology. This is the foundation of evidence-based self-management in women's health.