This feature is cross-posted on Cosmologics, an online magazine project of the Program for Science, Religion and Culture at the Harvard Divinity School.
Back in May, the NIH announced their intention to draft new policies to address gender bias in preclinical research. The majority of model organisms and cell lines used in preclinical research are male, a bias that obscures potentially significant differences between the sexes. Sex, the NIH argues, should be treated as a fundamental variable in biomedical research. By revamping inclusion and reporting policies, the NIH hopes that sex-based differences will be identified earlier in the research process.
|Image from TZU-YEN FU via flickr creative commons.|
This policy change is not without precedent. In 1987, the NIH changed its grant guidelines to require equal numbers of women and men in clinical research. Before this moment, clinical trial participants were almost exclusively male. This preference for male bodies was justified by the argument that females’ constantly cycling hormones would add too much noise to experimental data, making it more difficult for researchers to observe the effect of the intervention being studied.
Douglas Fields, a neuroscientist at the NIH, published his critique of the NIH’s new policy in last month’s Scientific American. In his article, Fields laments that the new policy is “about politics, not science.” Twenty years ago, when the inclusion of women in clinical research was first proposed, many scientists made the same complaint. Today, by Fields’ own admission, there is “no debate” about the importance of ensuring diversity among clinical trial participants. As the gradual acceptance of clinical inclusion policies show, the ideals of “good” scientific practice change over time, a process that is inevitably political.
Fields’ critique is two-fold. First, he argues that implementing the new policies will result in unnecessary and wasteful expenditures. Second, he contends that including both male and female research subjects in every study will produce unacceptable levels of experimental variation.
Let’s look at the issue of variation first. According to Fields, the reason why most preclinical researchers conduct single-sex experiments is that including both sexes would increase variation while simultaneously cutting sample size in half. This would make it more difficult for researchers to detect significant differences between the experimental and control groups. Fields rejects the explanation that researchers favor male animals and cells because they exhibit less hormonal variation. He does not offer an alternative explanation, however, for why a majority of single-sex experiments use only male animals.
The way around the problem of variation, it would seem, would be to increase the overall number of test subjects, so that there is a sufficiently large number of animals or cells of each gender. The larger sample size would smooth out the variation so that researchers could still detect significant differences between experimental conditions. Alternatively, researchers could perform separate statistical analyses on male and female populations, which would be useful in identifying the differences between these populations. Let’s say, for example, that female rats react positively to a candidate drug, but male rats do not. That seems like essential information to have before a drug advances into human clinical trials.
Back to Fields’ primary objection: experiments with larger subject pools cost more money. I think that Fields is probably right, even if Janine Clayton and Francis Collins insist that inclusion “need not be difficult or costly.” The NIH policy of preclinical inclusion may add to the already substantial cost of biomedical research. Is it worth it? That depends on what one thinks the goal of preclinical research should be. If we want experiments that are cheap, clean, and produce unambiguous results, then perhaps we should stick to a single-sex model. But if we hope to gain insight into how a patient’s gender (among other biological factors) might play out in a clinical situation, inclusion seems necessary—even if it comes at a cost. Fields actually suggests that using both male and female animals could cut down on laboratory costs, as labs could breed costly transgenic rats instead of purchasing new ones. And this week, the NIH announced 10 million dollars in “administrative supplements” to encourage gender balance in projects already funded by the agency. It seems that if the scientific community wants to make inclusion a priority, additional resources and solutions are available to ease its integration into preclinical research.
Fields accepts that sex difference is an important object of biomedical inquiry. In lieu of mandatory inclusion policies, he suggests earmarking NIH funds for researchers for whom the study of sex difference is a major focus. This forgets that sex differences often crop up in unexpected places—from processes of cellular death to the effects of daily aspirin on heart disease. The NIH policy need not transform every study into a major investigation of sex differences. But if scientists aren’t on the look-out for sex differences in their own research, how will we know where to look? Inclusion policies will provide new leads for researchers who want to delve deeper into the mechanisms behind sex-based variation.
I do, however, have my own reservations about expanding inclusion policies to include preclinical research. My main fear is that such policies will encourage us to see sex-based difference where none exists, or where such a difference would be irrelevant for clinical practice. I am convinced that biological sex can be can be an important variable in the study of a variety of biomedical phenomena. But I also imagine that there are many instances in which variation between the sexes is negligible—and that’s ok. Another danger of lionizing biological difference is that we might neglect the cultural factors that influence sex differences at the clinical level. We need to maintain a flexible understanding of the relationship between biological sex markers (either gonadal or chromosomal) and the lived experiences of all genders.
I also admit that I am less convinced about the necessity of inclusion at the cellular level than among laboratory animals. Maybe this is because sex difference at the cellular level boils down to an X or a Y, which bears little resemblance to the complex manifestations of sex or gender at the organismal level. Maybe it is because I am uncomfortable with personifying cells that are bought, sold, or left to divide in perpetuity, even though I know those cells once came from a human being with a gender identity of his or her own. Historian Hannah Landecker’s work on the HeLa cell line has shown how the racial identity of the donor has been projected back onto the cells themselves, often in highly problematic ways. For example, scientists wrote of HeLa’s tendency towards “miscegenation” through the aggressive contamination of other cell lines. By focusing on the sex of the cells (or even animals) we study, will we be able to avoid slipping into gendered language? How might our projection of gender onto cells change the way we think about or study them?
Only time will tell if the NIH’s new policies will enrich our understanding of biological differences between the sexes. There is a fine line, however, between appreciating sex differences where they do exist and expecting that biological sex will shape every aspect of clinical practice. Thinking critically about the ways in which sex factors into preclinical research is a necessary first step in ensuring that both women and men benefit equally from the fruits of biomedical research.