Included is not analyzed

It is easy to see a trial population that is 60% or 70% female and conclude that the results apply to women. That conclusion is only partly justified. Being included in a trial is not the same as being analyzed in a way that answers sex-specific questions.

A well-powered overall result tells you what happened across the whole group. It does not tell you whether the effect was the same in women and men. If women and men respond differently, the overall average can hide clinically meaningful variation. The only way to detect that variation is to analyze results by sex and to test for a treatment-by-sex interaction.

This problem shows up repeatedly in peptide-medicine research. The obesity trials for semaglutide and tirzepatide enrolled large numbers of women, in part because obesity trials often recruit more women than men. Yet the public trial reports and regulatory reviews still flag limitations: subgroup analyses may be prespecified but underpowered, life-stage data such as menopausal status may be missing, and contraceptive use is rarely described in detail.

The FDA statistical review of tirzepatide illustrates the point. It noted a statistically significant interaction between sex and treatment in SURMOUNT-1, with weight reduction appearing more favorable in females than in males. At the same time, the review described the interaction as quantitative-both sexes benefited compared with placebo-and called for further investigation. That is exactly the kind of nuanced finding that overall enrollment numbers cannot capture.

When you read a study, do not stop at the baseline demographics table. Look for the subgroup section, the interaction test, and the forest plot. If those are absent or labeled exploratory, the trial has told you who was enrolled but not whether the findings are reliably specific to women.