Menopause

The reanalysis that emerged in the late 2000s and 2010s focused on what came to be called the timing hypothesis. The harms observed in the WHI were concentrated in women who had started HRT well after menopause typically more than ten years after their last menstrual period. Women who had started close to menopause showed a different pattern. The 2007 reanalysis by JoAnn Manson and colleagues, published in JAMA, broke down the WHI data by age and time-since-menopause and found that the cardiovascular risk profile was substantially better for women in their fifties at initiation than for women in their sixties or seventies.

The Danish Osteoporosis Prevention Study, a smaller randomized trial that enrolled women at the actual onset of menopause, supported the same conclusion. Published in 2012, it followed roughly 1,000 women for a decade and found that early initiation of HRT was associated with reduced cardiovascular events and mortality. The trial was too small to be definitive alone, but combined with the WHI subgroup analyses it pointed toward a coherent reading: timing of initiation mattered, and the bulk of the harm in WHI came from women who had started outside the window in which therapy made physiological sense.

Menopause is biologically a single moment twelve months after the last menstrual period but the surrounding transition typically lasts seven to ten years. The perimenopausal years see fluctuating hormone levels, irregular cycles, and the onset of vasomotor and other symptoms. Estradiol declines, progesterone production becomes erratic, and the feedback loops between ovaries and the hypothalamic-pituitary axis become disordered before settling into the postmenopausal pattern. The symptom profile is diverse: hot flashes, night sweats, sleep disruption, mood changes, brain fog, joint pain, urogenital atrophy, accelerated bone loss.

The fraction of women who experience clinically significant symptoms is substantial. Cohort data suggests 60-80% of women report vasomotor symptoms, with about a quarter describing them as severe enough to interfere with daily function. The duration is longer than was previously assumed — median duration of frequent vasomotor symptoms is now estimated at around seven years. Black women in US cohorts report longer and more severe symptoms on average than white women, with the disparity not fully explained by socioeconomic factors.

The structural underrepresentation of women in clinical research is well documented. Until 1993, women were routinely excluded from US drug trials on the rationale that hormonal cycles introduced confounding variables. The 1993 NIH Revitalization Act required inclusion, but compliance has been uneven, and many drugs in current use were approved on trials with limited female representation. The pharmacokinetics of many medications differ between sexes in ways that have only been characterized recently.

Menopause research has been particularly thin. The WHI was, and remains, the largest randomized trial of menopausal hormone therapy ever conducted. The follow-on trials needed to characterize the timing hypothesis precisely have not been funded at comparable scale. Research funding for women's health conditions, as a fraction of NIH funding, has historically been below the corresponding burden of disease. Funding for menopause specifically — a transition affecting half the population — has been smaller still.

The 2002 WHI announcement was a textbook case of how a single trial result, presented without adequate framing of its limitations, can reshape medical practice in directions the underlying data did not actually support. The women who started HRT in their fifties were not the same population as those in the WHI, and their risk-benefit calculus was different. The decade and a half it took for that distinction to be acknowledged in revised guidelines was paid for by the women who went through menopause in the interim with inadequate options.