Women in clinical trials statistics: the research funding gap that still shapes care

She sat across from me holding a printout of her new blood-pressure medication. Same dose her husband takes. Same pill color. Within two weeks she had near-fainting spells and a heart rate that dipped into the low 40s. Her husband felt fine. We lowered her dose, her symptoms resolved, and I filed another quiet note in a chart I have written too many times: the label was written on a body that was not hers.

That is not an anecdote about one sensitive patient. It is the clinical face of a research system that, for decades, treated the male body as the default. Women were excluded from early-phase drug trials under a 1977 FDA policy. Congress forced a course correction in 1993. Enrollment has improved on paper. Funding and sex-stratified analysis have not kept pace. Women still experience adverse drug reactions at nearly twice the rate of men, and conditions that predominantly affect women are systematically underfunded relative to the burden they impose.

This article gathers the numbers journalists and clinicians reach for when any women's health story needs context: how exclusion became policy, where female enrollment still lags by therapeutic area, how rarely trial results are broken out by sex, how NIH dollars map onto disease burden, and what that gap costs patients at the pharmacy counter.

How we got here: 1977 to 1993

The modern gap did not start with indifference. It started with fear. After thalidomide caused thousands of limb deformities in babies whose mothers took the drug in pregnancy, regulators tightened the reins. In 1977, the FDA issued guidance recommending that women of childbearing potential be excluded from phase I and early phase II drug trials. Per the NIH Office of Research on Women's Health, the policy was broad. It covered women using contraception, single women, and women whose husbands had undergone vasectomy. Premenopausal women capable of becoming pregnant were treated as a risk class rather than as autonomous research participants.

Sponsors often read the restriction more broadly than the letter required. Even when the formal ban applied only to early phases, women were frequently scarce in later efficacy trials as well. The result was a multi-decade evidence base built largely on male pharmacokinetics, male symptom patterns, and male dose-response curves. Clinicians still work inside that inheritance every time a label lists one dose for adults without sex-specific guidance.

Policy began to shift in the mid-1980s. NIH encouraged inclusion of women in studies in 1986. A 1990 GAO investigation found that policy poorly communicated and inconsistently applied. NIH created the Office of Research on Women's Health that year. In 1991, Bernadine Healy became the first woman to direct NIH and launched the Women's Health Initiative, which eventually enrolled more than 150,000 postmenopausal women.

The decisive legal change came in 1993. The NIH Revitalization Act of 1993 (Public Law 103-43) wrote inclusion into federal statute. Per NIH's current inclusion guidelines, women and minority groups must be included as subjects in clinical research; trials must be designed so analyses can detect differences by sex and by race or ethnicity; and cost is not an acceptable reason for exclusion. Also in 1993, the FDA issued a new guideline on studying gender differences in drug evaluation and reversed the 1977 early-phase ban.

On paper, the ban ended more than thirty years ago. The downstream scientific debt has not.

Female enrollment by therapeutic area

Headline numbers can look reassuring. Steinberg and colleagues, writing in JAMA Network Open in 2021, analyzed 20,020 U.S. interventional trials from 2000 to 2020 with about 5.11 million participants. Sex was reported in 99.2% of those trials, and overall female participation often approaches half. That is real progress from an era when women were simply left out. The useful question is not whether women appear in any trial. It is whether they appear in proportion to who gets the disease.

Steinberg's group compared female enrollment with female disability-adjusted life years (DALYs) for each disease category. Oncology trials enrolled about 43% women while women accounted for roughly 46% of the oncology DALY burden. Neurology: 53% of participants versus 56% of DALYs. Immunology: 46% versus 49%. Nephrology and genitourinary disease: 42% versus 45%. Cardiology showed one of the strongest negative associations with female enrollment in adjusted models (adjusted relative difference approximately −18.7%). Multiplied across decades of trials, those gaps leave the evidence base for leading killers of women still tilted male.

Industry- and government-funded drug and device trials look worse when you zoom into the diseases that dominate adult medicine. Sosinsky and colleagues reviewed 1,433 U.S. phase 1-3 trials registered from 2016 to 2019 with cardiovascular, psychiatric, or cancer endpoints, totaling 302,664 participants. Average female enrollment was 41.2%. Cardiovascular trials enrolled 41.9% women while women represent about 49% of people with cardiovascular disease. Psychiatry trials enrolled 42.0% women while women account for roughly 60% of patients with those conditions. Cancer trials enrolled 41.0% women against a female patient share of about 51%. In every area they analyzed, female participation fell short of the prevalence benchmark.

That matters clinically for conditions such as heart disease in women, where symptom patterns, diagnostic test performance, and drug response already differ by sex. If trial populations under-enroll women, the guidelines built from those trials inherit the skew.

How often sex-stratified results are reported

Including women and analyzing outcomes by sex are different obligations. The Revitalization Act required both. Compliance has been uneven for decades.

Geller and colleagues have tracked this problem across multiple journal samples. In a 2011 analysis of federally funded randomized trials published in 2009, the median enrollment of women in mixed-sex studies was 37%, and 75% of those studies did not report any outcomes by sex. Their 2018 follow-up in Academic Medicine reviewed NIH-funded RCTs published in 14 leading U.S. medical journals in 2015. Among 107 mixed-sex studies, median female enrollment was 46%. Only 28 of 107 (26%) reported at least one outcome by sex or explicitly included sex as a covariate. Sixteen studies (15%) enrolled fewer than 30% women. Of all 142 eligible trials, only 13.4% analyzed or reported outcomes by race or ethnicity. The authors found no statistically significant improvement in sex-based analysis compared with 2004 and 2009.

Nearly three decades after Congress required sex analysis where appropriate, roughly three-quarters of published NIH-funded RCTs in a high-impact sample still did not report outcomes by sex. Enrollment gains can coexist with analytic silence. A trial can list 48% women in the baseline table and then present only pooled results. Journal samples capture elite venues rather than every ClinicalTrials.gov entry, so the broader literature is unlikely to be better. That silence is especially costly for conditions with known sex differences, including autoimmune disease and Alzheimer's disease.

NIH funding relative to disease burden

Enrollment is one pipeline. Dollars are another. Arthur Mirin's 2021 analysis in the Journal of Women's Health compared NIH funding levels with U.S. disease burden measured in disability-adjusted life years. He modeled the funding that would be commensurate with burden, then asked whether sex-dominant diseases fall above or below that line.

The pattern was not subtle. In nearly three-quarters of cases where a disease primarily afflicts one sex, the funding pattern favored males. Sometimes a female-dominant disease was underfunded relative to burden. Sometimes a male-dominant disease was overfunded. The size of the mismatch was nearly twice as large for male-favored diseases as for female-favored ones (chi-square p = 0.015). Mirin's conclusion: NIH applies a disproportionate share of its resources to diseases that affect primarily men, at the expense of those that affect primarily women.

A 2023 Nature feature revisited Mirin's updated NIH data. Female-dominant conditions such as migraine, headaches, endometriosis, and anxiety disorders attracted less funding than their population burden would predict. Male-dominant conditions such as HIV/AIDS and substance misuse attracted more. Underfunding was more severe among underfunded female-dominant diseases: for ME/CFS, the reported ratio of actual funding to burden-commensurate funding was 0.04; for HIV/AIDS, 15.6. That does not mean HIV research is overfunded. It means female-dominant morbidity has been systematically deprioritized. NIH notes that burden is only one input into funding decisions; the residual pattern still tracks sex.

The female-specific condition funding gap

There is a second layer beneath sex-dominant chronic disease: conditions that almost exclusively affect women. Endometriosis, fibroids, PCOS, menopause, heavy menstrual bleeding, preeclampsia, and postpartum complications drive disability, lost work, infertility, and maternal harm. They do not always kill quickly, which makes them easy to undervalue in a mortality-first funding culture.

A National Academies committee reported in late 2024 that women's health research accounted for 8.8% of NIH grant spending across fiscal years 2013 to 2023, falling to 7.9% in FY 2023 even as overall NIH grant funding rose. NIH's own systems have sometimes placed the share closer to 10-11%, depending on definitions. Either figure is small relative to half the population. A WHAM report produced with RAND found that in 2019 only 4.5% (about $20 million) of NIH coronary artery disease research focused specifically on women, even though heart disease is the leading cause of death among U.S. women.

Female-specific conditions fare worse in visibility and funding continuity. Patients with endometriosis wait years for diagnosis. Women navigating menopause and hormone therapy still encounter guidance built on incomplete older trial frames. PCOS research has expanded, yet many women cycle through multiple clinicians before anyone names the syndrome; our PCOS prevalence statistics piece covers how often it is missed. Cycle-tracking tools such as the period calculator and ovulation calculator help document patterns, but documentation is not a substitute for funded treatment trials.

Downstream consequences: adverse drug reactions

When dosing science is built on male bodies and female participants are under-analyzed, the harm shows up as side effects.

Zucker and Prendergast, in a 2020 Biology of Sex Differences paper, reviewed pharmacokinetic and adverse-reaction evidence for 86 FDA-approved medications spanning antidepressants, cardiovascular drugs, anticonvulsants, analgesics, and other classes. Women experience adverse drug reactions nearly twice as often as men. Of the 86 drugs evaluated, 76 showed higher pharmacokinetic values in women. Among 59 drugs with clinically identifiable ADRs, sex-biased pharmacokinetics predicted the direction of sex-biased ADRs in 88% of cases. Ninety-six percent of drugs with female-biased pharmacokinetic values were associated with higher ADR rates in women. Only 29% of male-biased PK patterns predicted male-biased ADRs. Equal milligram doses, given to bodies that clear drugs differently, overmedicate women.

The Ambien (zolpidem) story is the public case study. The drug lingers longer in women's blood. Next-morning impairment led the FDA, in 2013, to cut the recommended dose for women, years after approval. I see the clinical version weekly: dizziness on antihypertensives, excess sedation on sleep aids, bruising on anticoagulants. Many are dose-exposure problems a sex-aware development program would have flagged. The research gap does not stay in journals. It walks into the exam room.

What still needs to change

The 1993 laws fixed a legal exclusion. They did not automatically fix incentives. Sponsors still optimize for speed on a primary endpoint. Journals can accept pooled results without sex tables. Funders can underweight female-dominant conditions year after year.

Practical levers exist: power trials for sex-stratified safety analyses; make sex-disaggregated outcomes a phase 3 funding condition; treat missing sex analysis as a journal limitation; expand NIH investment in female-specific conditions and in sex-difference science inside shared diseases; publish participation-to-prevalence ratios by therapeutic area the way Steinberg and Sosinsky did.

Patients need not wait for perfect policy. Ask whether a recommended drug was studied in women at your age and dose. Report side effects rather than quietly stopping therapy. For cycle questions adjacent to research gaps, tools such as the late period calculator can organize history before a visit, but they cannot replace treatment evidence that only funded research produces.

The short version for a deadline piece is this. Women were written out of early trials and written back in by statute. They remain underrepresented relative to disease burden and under-analyzed in published results. Until enrollment and funding move with the analysis requirement, equal doses will keep producing unequal harm.

Frequently asked questions about women in clinical trials

Why were women excluded from clinical trials?

In 1977, after the thalidomide tragedy, the FDA recommended excluding women of childbearing potential from phase I and early phase II drug trials to avoid fetal risk. The policy was broad, covering women on contraception and even those whose partners had vasectomies. Sponsors often extended exclusion into later phases. The FDA reversed the early-phase ban in 1993, the same year Congress mandated inclusion in NIH-funded research.

What did the NIH Revitalization Act of 1993 require?

The NIH Revitalization Act (Public Law 103-43) required that women and minorities be included in NIH-supported clinical research, that trials be designed to allow analysis of differences by sex and by race or ethnicity, and that cost not be used as a reason for exclusion. It also supported outreach for recruitment and retention and strengthened the Office of Research on Women's Health.

What percentage of clinical trial participants are women today?

It depends on the sample. Across many U.S. trials, overall female enrollment often nears 50%. In a 2016-2019 analysis of 1,433 phase 1-3 cardiovascular, psychiatry, and cancer trials, Sosinsky et al. found average female enrollment of 41.2%, below disease-prevalence benchmarks in each area. Representation remains weakest relative to burden in fields such as cardiology and oncology.

How often are clinical trial results reported by sex?

Less often than the law intends. Geller and colleagues found that only 26% of eligible NIH-funded mixed-sex randomized trials published in 2015 reported at least one outcome by sex or included sex as a covariate. An earlier 2009 sample showed 75% of studies reporting no sex-specific outcomes. Enrollment gains have not reliably translated into sex-stratified evidence.

Is NIH research funding lower for women's health conditions?

Relative to disease burden, yes for many female-dominant conditions. Mirin (2021) found that in nearly three-quarters of sex-dominant diseases, NIH funding patterns favored males. A National Academies analysis estimated women's health research at 8.8% of NIH grant spending from FY 2013 to 2023, declining to 7.9% in FY 2023. Female-specific and female-dominant conditions remain underweighted.

Why do women have more adverse drug reactions?

Women experience ADRs nearly twice as often as men, per Zucker and Prendergast (2020). Sex differences in body composition and drug clearance mean equal doses often produce higher blood levels and longer exposure in women. Among 86 FDA-approved drugs with available data, 76 showed higher pharmacokinetic values in women, and 96% of female-biased PK patterns tracked with higher ADR rates in women.