The symptoms of perimenopause are not imagined, not character failings, and not "part of aging" that must be endured. They are hormonal, measurable, and (for most women) treatable. The 2002 WHI trial scared a generation of clinicians into under-prescribing HRT; the 2013+ reanalyses have nuanced that picture.
What actually happens
Perimenopause is the transitional period (typically 4-10 years) before the final menstrual period, marked by increasingly erratic estrogen and progesterone production. Signals:
- Cycle irregularity: shorter, longer, skipped.
- Vasomotor symptoms: hot flashes, night sweats. Affect ~75% of women during the transition.
- Sleep fragmentation: partially from night sweats, partially independent.
- Mood lability, anxiety, new-onset depression.
- Cognitive fog, word-finding difficulty.
- Vaginal dryness and atrophy.
- Accelerated bone loss once estrogen falls.
- LDL particle and ApoB rise post-menopause.
Average age at final menstrual period in North America: 51. Perimenopause for some starts at 35; "late onset" runs past 55.
Testing
Standard labs (cycle day 3 in pre-menopausal, any day in peri- or post-):
| Phase | Dose | Notes |
|---|---|---|
| FSH | Cycle day 3 | Rises as ovarian reserve falls. >25 = peri; >40 = menopause |
| Estradiol | Cycle day 3 | Widely variable; single draws less useful than pattern |
| LH | Cycle day 3 | Rises with FSH as negative feedback collapses |
| AMH | Any day | Anti-Mullerian Hormone; ovarian reserve proxy, more stable than FSH |
| TSH + free T4 | Annual | Thyroid symptoms overlap; rule out |
| Prolactin | Any day | Rule out pituitary cause of cycle disruption |
| Full lipid panel | Annual, [fasting](/topics/fasting/) | ApoB will rise post-menopause; establish pre-transition baseline |
| DEXA scan | At menopause onset, every 2 yr post | Bone density; earlier if other risk factors |
The HRT story: WHI and after
Women's Health Initiative 2002 was halted mid-trial over breast cancer + cardiovascular concerns with combined HRT (oral conjugated equine estrogens + medroxyprogesterone) ( Rossouw, Anderson, Prentice et al. 2002, n=16608 ). Mass media coverage caused rapid HRT prescribing collapse. What subsequent analysis clarified:
- Average enrollment age was 63, far past typical symptom-driven initiation. The trial didn't cleanly test HRT-for-symptoms.
- Manson 2013 (JAMA, extended WHI follow-up, n=27,347): within 10 years of menopause, HRT had a favorable cardiovascular + mortality profile ( Manson et al. 2013, n=27347 ).
- Risk/benefit reverses past ~10 years post-menopause.
- Formulation matters: transdermal estradiol has lower VTE and stroke risk than oral conjugated equine estrogens.
- Progestogen matters: micronized progesterone (bioidentical) has lower breast-cancer signal than medroxyprogesterone.
Modern protocols typically use:
- Transdermal estradiol (patch 25-100 mcg/day, or 0.06-0.1% gel). VTE risk comparable to not taking HRT.
- Micronized progesterone 100-200 mg at bedtime for women with intact uterus (endometrial protection). Oral progesterone also has mild sleep benefit.
- Testosterone (compounded cream, 1-5 mg/day) for libido specifically where evidence supports it.
On the vitamin D question that often comes up in perimenopausal bone-health discussions: VITAL (Manson 2019, n=25,871) showed D3 supplementation did not reduce cardiovascular or cancer incidence in already-sufficient populations ( Manson et al. (VITAL) 2019, n=25871 ). Correct measured deficiency; do not supplement reflexively.
Non-hormonal options
For women where HRT is contraindicated (breast cancer history, active clotting disorder, certain migraine patterns):
- SSRIs/SNRIs (venlafaxine, paroxetine): 50-60% reduction in hot flash frequency in trials.
- Fezolinetant (FDA approved 2023): NK3R antagonist, ~50% hot flash reduction; new and expensive.
- Gabapentin: 30-50% hot flash reduction; sedating side effect can be useful at bedtime.
- CBT-I: for sleep fragmentation specifically.
- Vaginal estrogen (cream, tablet, ring): local-only treatment for atrophy; minimal systemic absorption; safe even in many breast cancer survivors.
What the pop culture is missing
- Cognitive complaints are real. Estrogen has broad central nervous system effects; brain fog is hormonal, not psychological.
- ApoB rises post-menopause independently of weight. Annual lipid panel essential; statin threshold may shift.
- Bone loss is fastest in the first 5 years post-menopause. Resistance training + adequate protein + DEXA monitoring is non-optional.
- Timing matters for HRT. Starting at symptom onset (perimenopause) vs 10+ years later is a different decision.
Counter-view
Avrum Bluming + Carol Tavris ("Estrogen Matters", 2018) argue HRT is still dramatically under-prescribed and the WHI narrative was a generational health harm. JoAnn Manson agrees on the 10-year window but is more cautious on duration > 5 years. Some endocrinologists argue bioidentical vs synthetic distinctions are overstated; others see them as material. The pragmatic middle is the Manson formulation: within 10 years of menopause, for symptomatic women without specific contraindications, modern HRT is more often indicated than prescribed.
