Listening to Our Stakeholders On Considering Sex as a Biological Variable


Dr. Janine Clayton Janine Austin Clayton, M.D., is NIH’s Associate Director for Research on Women’s Health, and the Director of the NIH Office of Research on Women’s Health.

One year ago, NIH announced a plan to adopt a new policy requiring a deliberate approach to the consideration of sex as a biological variable (SABV) in preclinical research. (Read the article, co-authored by Janine Clayton and NIH Director Francis Collins, here.) Since that moment, we have been working diligently and collaboratively inside and outside NIH to craft meaningful policy that promotes the best science.

One of the most important first steps we took was to issue a Request for Information (RFI): Consideration of Sex as a Biological Variable in Biomedical Research (NOT-OD-14-128) to gather input from the research community and other interested stakeholders with the following questions:

  • Does considering SABV affect the reproducibility, rigor, and/or generalizability of research findings?
  • What are the areas of science or phases of research that might benefit from consideration of SABV?
  • What are the main impediments of considering SABV?
  • How can NIH facilitate considering SABV?

What we learned from this outreach was gratifying (see more details here). The vast majority of respondents agreed that consideration of sex as a biological variable is an issue affecting the reproducibility, rigor, and/or generalizability of research findings. We also learned that despite overwhelming agreement that SABV is good science, scientists and other stakeholders are concerned about practical matters. In an era of biomedical fiscal austerity, researchers are worried about cost as well as constraints on methodological and experimental design. As such, more than half of the people we heard from suggested that NIH could offer tangible resources to help with SABV policy implementation.

The NIH Office of Research on Women’s Health and the NIH Office of Extramural Research have taken into account input from scientists and the public in developing SABV policy, and we understand the need for more information about studying both sexes. Above all, our goal is to fund research that takes into account essential biological variables — of which sex is a very important one. Our unwavering goal is to support the best science that underpins health advances for women and for men.

We are working hard to develop SABV training resources and scientific tools such as courses, workshops, and online resources to help applicants, reviewers, and NIH program staff to be prepared for the forthcoming policy. Stay tuned!


  1. It sounds by the nature of your questions that you have already decided that sex as a biological variable is a good thing. Has that question been answered yet? And if it’s not considered, Would it invalidate any prior research?

  2. I’m compelled to share in regards to gender as a Biological Variable. I’m not an educated gal yet as life has unfolded, the many years of bad judgement and learning curves whether age being the testing points for growth and experiences – many years of pushing the bar in competitive pride as well as examining all areas of 7 deadly sins within the rate of self determined objective, the jaunting halt has revealed much. Over the years, obvious chemical changes took place ‘literally biologically as a reproducing female vessel goes through. Many different facets than the male for sure. Since working with substance abusing male and female genders, there are many characteristics of differences – as the lack of knowledge within recognition of gender affecting the causes and conditions of substance abuse – therefore defining some of the manifested risky behaviors resulting in medication solutions all too often. All in all, to simplify the previous statement is in the understanding of identity – on the latter of the point of male (in the moment more often than female). Actually, this is way bigger than just gender as obvious for biomedical research – there are too many variances that affect the chemical makeup within – as circumstances, experiences, perceptions, discernment capacity, ad infinitum are abundant enough for a Biological Variable for sure. Thanks for reading…

  3. Opposite-sex twin pairs are possibly THE best possible design for studying SABV. We have established an International Network of Twin Registries (see Buchwald et al. Twin Res Hum Genet 2014) that makes possible global studies of SABV across a wide range of traits using existing data – and more importantly – facilitating new collaborative studies. As Chair of the INTR Working Group, I would be very keen to talk with NIH staff about how this potential might be realised.

  4. Dr. Clayton, It is really heartening to see this initiative moving forward and I thank you and your collaborators for keeping the ball moving up the field. There are certainly some thorny practical considerations, and as a department head I think about the increased time it takes to collect, interpret and publish data from human studies when they are sufficiently powered to intentionally assess sex differences. There are implications for a host of non-trivial items, including tenure and promotion for young faculty. But all of the issues are soluble if there is a true commitment to making this happen. Keep it going!!!

  5. While considering sex as a biological variable is essential for good biomedical science, it is a bad idea to apply this concept to the early stages of research. Not only is duplicating studies in both sexes expensive and would invariably lead to compromises elsewhere in the study design (such as #animals/treatment), early stages of research are commonly conducted in rodent models, and rodents are particularly bad species for modeling human differences in sexes. That is, rodents have sex-specific patterns in growth hormone secretion that regulate many organs . Humans lack these sex-specific growth hormone patterns and the associated tissue differences. Thus, SABV studies in rodents likely will lead to misconceptions about nonexistent differences between the sexes in human. It seems better to identify an experimental condition or response in the most appropriate sex of a model animal and then, once established, determine if that response occurs in humans and if it varies with sex, age, BMI, environmental factors, etc. The idea that the presence or absence of differential responses based on sex in animal models can be readily extrapolated to humans is a flawed one that will lead to more bad information than true knowledge.

Before submitting your comment, please review our blog comment policies.

Leave a Reply to Professor John Hopper Cancel reply

Your email address will not be published. Required fields are marked *