Until relatively recently, women of childbearing potential were routinely excluded from early-phase clinical research in several jurisdictions, and animal research was conducted predominantly on male animals.
The rationale was protective — concern about effects on potential pregnancies following historical drug safety disasters — and about controlling for hormonal variation.
The consequence was a body of evidence generalised to a population it hadn't studied.
What changed
Policies requiring inclusion of women in publicly funded research were introduced in several countries from the 1990s onwards, and requirements to consider sex as a biological variable in preclinical research have been added more recently.
Representation in trials has improved substantially. Gaps remain, particularly in early-phase studies, in some therapeutic areas, and in analysis — inclusion doesn't guarantee that results are analysed by sex, and many trials don't report disaggregated outcomes.
Why sex differences matter pharmacologically
Several mechanisms produce genuine differences in how drugs behave.
Body composition differs on average, affecting distribution of drugs depending on whether they're water or fat soluble.
Some drug-metabolising enzymes show differences in activity.
Renal clearance differs.
Hormonal fluctuation across the menstrual cycle can affect drug metabolism for some compounds.
The practical consequence is that a dose established in a predominantly male trial population may be inappropriate. There have been cases where dosing recommendations were revised after post-marketing data showed higher exposure and more adverse effects in women.
Analyses of drug withdrawals have found that adverse events leading to withdrawal disproportionately affected women, which is consistent with dosing derived from a population that didn't represent them.
Diagnosis
A separate mechanism with substantial consequences.
Clinical descriptions of conditions were frequently derived from predominantly male presentations, which means presentations more common in women can be read as atypical.
Cardiovascular disease is the most documented example. Symptoms of myocardial infarction can differ, and studies have found longer delays to treatment and differences in investigation and management for women presenting with cardiac symptoms.
Similar patterns have been examined in other conditions, including some autoimmune diseases and certain neurological presentations.
Diagnostic criteria for autism and ADHD were developed largely from male samples, and there's substantial discussion about whether presentation differs and whether that contributes to later diagnosis in girls and women.
Pain
An area with a considerable literature and consistent findings.
Studies examining pain management have found differences in how pain reports are assessed and treated, with several finding that women's pain is more likely to be attributed to psychological causes and less likely to result in analgesic treatment.
Conditions predominantly affecting women — including several chronic pain conditions — have historically received less research funding relative to disease burden and have longer diagnostic delays.
The research gaps that remain
Several areas remain notably under-researched.
Conditions specific to women. Endometriosis, affecting a substantial proportion of women, has an average diagnostic delay measured in years and comparatively limited research investment.
Menopause, despite affecting half the population, has a thin evidence base in several areas.
Pregnancy and lactation, where exclusion from trials has produced a situation in which pregnant women take medications with limited safety data precisely because studying them was considered too risky. The absence of evidence is itself a harm.
And conditions affecting both sexes but differently, where disaggregated analysis is inconsistent.
What's changing
Funding requirements, journal policies requiring sex-disaggregated reporting, and specific research initiatives have all improved matters.
Regulatory guidance in several jurisdictions now expects analysis by sex where relevant.
Progress is real and uneven, and the historical evidence base cannot be retrospectively corrected — decades of research conducted on male subjects remains the foundation of a great deal of current practice.
Practical implications
For individuals navigating healthcare.
If a symptom is being attributed to stress or anxiety and you don't find that explanation convincing, it's reasonable to ask what else has been considered and what would rule it out.
Asking specifically whether a medication's dosing has been studied in women, particularly for newer drugs, is a legitimate question.
Keeping a record of symptoms, including timing relative to the menstrual cycle where relevant, provides information that may not otherwise be elicited.
And seeking a second opinion where a presentation doesn't fit and isn't being investigated is reasonable rather than difficult behaviour.
None of that should be necessary, and the evidence suggests it sometimes is.
General information only. Any health concerns should be discussed with a qualified healthcare professional.
Cardiac symptoms specifically
Worth stating the practical version, because it is the case where the delay does most damage.
Chest pain remains the most common presentation of a heart attack in everyone. What differs is that additional symptoms — nausea, jaw or back pain, shortness of breath, unusual fatigue, light-headedness — are more frequently present and more frequently the dominant complaint.
Which means a presentation without classic crushing central chest pain can be read as unlikely to be cardiac, and the evidence suggests that happens.
The practical advice from cardiology bodies is consistent: if symptoms suggest a possible cardiac event, seek emergency care and say explicitly that you are concerned about your heart. Being specific about the concern changes the triage pathway.