Women enrolled unevenly in trials, then dosed as if the difference didn't exist

Two systematic reviews published within a fortnight of each other address different specialties and land on the same structural problem. One asks whether women are enrolled in trials in proportion to who has the disease. The other asks what happens at the prescribing end when the answer to the first question is not examined.

Exposure at the same dose

The first review, published in the Journal of Personalized Medicine on March 31, searched PubMed for studies from January 2010 to March 2026 reporting sex-stratified pharmacokinetic, pharmacodynamic or safety outcomes for selective serotonin reuptake inhibitors and second-generation antipsychotics [s1]. Twenty-seven studies met inclusion criteria [s1]. Because the methods varied too much to pool, the synthesis is narrative rather than quantitative [s1].

The pattern reported across those studies is that women more frequently showed higher dose-adjusted serum concentrations — the same milligrams producing more drug in the blood — particularly for risperidone and some SSRIs, with age-related increases more evident in women [s1].

On the pharmacodynamic side, the review reports findings suggesting women may reach comparable dopamine D2 receptor occupancy at lower olanzapine doses [s1]. That is a different claim from the concentration finding, and stronger: it is not only that more drug arrives, but that less may be needed for the same target engagement.

Pharmacovigilance analyses covered in the review showed sex-specific adverse event patterns, including greater reporting of endocrine-related effects and of QT prolongation in women [s1].

Against all of this, the review's starting observation is that dosing recommendations for these two drug classes remain largely sex-neutral [s1].

Enrolment against prevalence

The second review, published in Reviews in Cardiovascular Medicine on March 19, searched PubMed, Scopus and EMBASE for atrial fibrillation trials from 1996 to January 1, 2024, covering lifestyle interventions, drug treatments, catheter ablation and device therapies [s2].

It identified 103 trials enrolling 218,322 participants, of whom 39.5% were women [s2].

Raw percentages are the wrong measure here, because atrial fibrillation is not evenly distributed by sex. The review uses the participation-to-prevalence ratio, which compares enrolment share with disease share. Across trials the ratio ranged from 0.00 to 1.73, averaging 1.03 [s2] — which at first glance looks like parity.

The average conceals the distribution. Forty-three per cent of the trials showed female under-representation, defined as a ratio below 0.8 [s2].

Funding source separated the results: university-funded trials had a mean ratio of 0.951, against 0.800 for industry- or government-funded trials [s2].

Why the two findings belong together

A trial that enrols too few women can still produce a valid average treatment effect. What it cannot produce is a reliable sex-stratified analysis — the subgroup is too small, and the analysis is usually not prespecified. So the trial reports one dose, one efficacy estimate, one safety profile.

The psychotropic review describes what is then found later, in therapeutic drug monitoring data and pharmacovigilance databases rather than in the registration trials: concentration differences, occupancy differences, and adverse event patterns that differ by sex [s1].

This is a sequencing problem more than a conspiracy. The information arrives after the label is written, which is the point at which changing it is hardest.

What these reviews do not establish

Neither is an intervention study, and neither demonstrates that sex-adjusted dosing improves outcomes. The psychotropic review's own conclusion is conditional: that integrating sex-aware considerations into dosing strategies could improve therapeutic precision and reduce adverse outcomes [s1]. It synthesises 27 heterogeneous studies narratively [s1], which is a weaker form of evidence than a meta-analysis, itself weaker than a trial.

The atrial fibrillation review measures enrolment, not harm. It does not show that any patient was treated worse because of the enrolment gap; it shows that the data needed to check would in many trials be insufficient [s2].

And "sex differences" in aggregate say nothing about any individual. Between-group differences in drug metabolism are smaller than the variation within either group, which is driven by genotype, age, kidney and liver function, body composition, and interacting medications.

What to watch

Whether regulators begin requiring prespecified sex-stratified analyses rather than post-hoc subgroups, and whether therapeutic drug monitoring — which is where several of these signals were first visible [s1] — moves earlier in the development process.

This article is informational and is not medical advice. Nothing here is a basis for changing a prescribed dose; that decision belongs with a clinician who knows the individual case.

Sources

• Sex Differences in Psychotropic Drug Exposure and Safety: A Systematic Review Toward Personalized Dosing Strategies — Journal of Personalized Medicine, 2026-03-31
• Representation of Women in Atrial Fibrillation Clinical Trials: A Systematic Review — Reviews in Cardiovascular Medicine, 2026-03-19