Fewer than half of papers published by NIH-funded researchers analyze or report their data by sex, which could make it harder to know what the results mean for men and women, a new study found.
Over a decade ago, the National Institutes of Health set out to promote sex-inclusivity in study design by introducing the expectation that research it funded consider sex as a biological variable (SABV). The guidelines are broad, asking researchers to consider SABV in their design, analysis, and reporting, without mandating that sex differences be examined in the results.
A new study from Northwestern University, published on Monday in Nature Communications Medicine, looked at a sample of 574 NIH-funded papers published between 2017 and 2024, and found that although a majority — 61% — included both sexes as subjects, only 44% of them reported their results by sex. The lack of sex in results could make research harder to replicate, and make it hard to know whether the results apply equally to men and women, the authors said.
“We see 13% of publications don’t even tell you the sex of their subjects,” said Nicole Woitowich, the executive director of the Northwestern University Clinical and Translational Sciences Institute and the paper’s lead author. Even when the research mentioned including both sexes, she said, information about the specific breakdown can be missing. “They don’t actually say we used 10 female mice and 10 male mice; they’ll say we used mice of both sexes,” she said.
The NIH has promoted the importance of analyzing results by sex since 1993, when it established the Office of Women’s Health Research, and mandated the inclusion of women and racial and ethnic diversity in clinical trials. But research specifically on women’s health, and the inclusion of women in broader studies, has lagged. The creation of the SABV policy was intended to get researchers to be more cognizant of the issue and spur change.
“It was finally recognition that we need to start looking at both sexes,” said Anne Murphy, director of the neuroscience institute at Georgia State University who was on the committee that helped write the SABV policy. She found the results of the new study sad. The SABV policy was meant to foster inclusion, “but that was never the ultimate end goal,” she said. “The end goal was to generate sex-specific knowledge to improve health outcomes for everyone.”
In some cases, the exclusion of one sex is due to the nature of the research — for instance, studies focusing on exclusively female or male conditions — though in others the reason is not evident. However, only a small minority of papers (4%) discuss their rationale for how they approached SABV.
The lack of information about sex is more prominent in non-human subject studies. Among them, 41% included both sexes, and 19% didn’t disclose sex information at all. Only half the non-human subject studies reported the findings by sex, and a mere 28% analyzed the findings by sex.
“A lot of the basic research that we already have is basically based on male models. We realized we need to do more, which is where the SABV comes in and tries to address that and fix that. But it’s not clear that we are fixing it,” said Alina Salganicoff, director for women’s health policy at the health policy think tank KFF, who co-chaired a National Academies committee that assessed NIH women’s health research. That report found that even as the NIH’s budget increased, the portion it was spending on women’s health declined.
The new study documents ways that research has fallen short, but some outside experts said there are further areas that could be analyzed. To see if a study considered gender, the paper grouped together several different kinds of analyses. But previous research has found that scientists often employ the wrong statistical analyses to study sex differences. The paper also does not investigate what portion of papers weigh gender, or how they define sex. Further, the authors don’t note whether there was an increase in the reporting of sex during the timeframe studied or in comparison with the years prior to the introduction of SABV guidelines.
“This is really important work. I just wish they could have given us a little bit more detail, both on the analytic methods … and the operationalization of sex,” said Annika Gompers, an epidemiologist at Emory University and the GenderSci Lab, who researches sex differences, and has done work on SABV. “There’s been a lot of literature around discussing the merits of picking and defining a very specific sex-related variable that is relevant to the context of your research question … that would have been a meaningful innovation.”
Policy changes unrelated to SABV under the Trump administration have threatened progress on women’s health research. An analysis from The Washington Post this month found that the number of NIH grants mentioning “women” has declined the the lowest point in a decade — walking back years of steady increases.
President Trump signed an executive order that discouraged the use of the term “gender” in research. “That’s going to set us back years,” said Tom Babor, a professor emeritus at the University of Connecticut School of Medicine, and one of the researchers who developed the Sex and Gender Equity in Research (SAGER) guidelines — a set of recommendations for reporting sex and gender in studies.
The paper also found when women were the first or senior author, the research was more likely to include both sexes and analyze the results by sex. This likelihood increased even further, in particular when it came to analysis, when both the first and last author of the paper were women: While only 39% of papers with a male lead author analyzed data by sex, 58% of papers with a woman-woman dyad did. But, over the past year, women have also been more likely to have their grants terminated by NIH.
“The NIH can be very influential, and at this point, it does not seem like they are helping the situation with their attempts to restrict the use of the words ‘sex’ and ‘gender,’ especially in scientific communications,” Babor said.
He said that the group that developed the SAGER guidelines is considering issuing guidelines for institutional review boards, asking that sex be discussed in studies on non-human subjects, and both sex and gender in human studies.
