The planned mandate of the US National Institutes of Health (NIH) to include both sexes in effectively all preclinical studies could undermine its own objective by wasting resources, slowing down research or even provoking a backlash (see J. A. Clayton and F. S. Collins Nature 509, 282–283; 2014). Instead of a blanket mandate, the NIH should be promoting research into the sex differences that are important to science and in disease.

Duplicating studies to “compare and contrast experimental findings in male and female animals and cells” is rarely practical, affordable, prudent, scientifically warranted or ethically justifiable. Researchers use both sexes because this roughly halves the costs of breeding and maintenance. Sometimes one sex is excluded if results are likely to differ between sexes, and possibly for well-known reasons — for instance, male rats run faster than female rats through a maze. If there is no justification for studying both sexes, then it should not be done.

Clayton and Collins suggest that statistical variability will not be increased by using equal numbers of male and female cells or animals in studies, but this is questionable and undermines the premise for the NIH's argument. If the sexes were not different, there would be no need to use both. Variances are additive, so using both sexes halves sample size while increasing variance, making it less likely that an observed difference not due to sex can be detected at a statistically significant level. Thus, an increased number of samples would be needed to reach firm conclusions.

Understanding gender differences in disease is a goal in itself, but this will not be attained as a by-product of mandating its intrusion into every hypothesis under investigation.