Sex bias in science
A decade ago, a landmark study analyzed a slew of scientific publications, and it showed a bias in the sex of the human or animal subjects that were considered by those research papers. So 10 years on, have things got better? Chris Smith heard how Niki Woitowich took it upon herself to reroll that original study and find out...
Niki - About 10 years ago, this study came out that looked at sex bias in biomedical research. And what this study found was that overwhelmingly biomedical researchers tend to use males in their research. And I thought to myself, you know, this is coming up on the 10 year anniversary of when this data was collected originally. I just wanted to know simply had there been a change. And so I reached out to the original authors, Annalise Berry, and Irving Zucker. And I asked if they would be willing to share their original data with me and would be interested in doing this 10 year follow up.
Chris - Now, when you say there was a male bias, do you mean as in animals, just humans, where was the bias?
Niki - The bias was actually across all research that used mammals. So from male mice, rats, primates, dogs, cats, and humans.
Chris - Wasn't legislation passed though in the interim that this shouldn't be the case? Because people did acknowledge. Look, there is an issue here. We may be missing important biomedical facts by not studying the right groups of individuals and actually it became law.
Niki - Yes. In the United States in 1993, there was a law that was passed that requires the inclusion of women and underrepresented groups in clinical research. So that accounts for human research. However, no such law existed for basic science research. So in 2016, the National Institutes of Health in the U S created a policy requiring investigators to consider sex as a biological variable.
Chris - Right, so given that we've now got legislation there and it doesn't just apply to humans, what one would hypothesise is that, if you do compare the situation in 2019 with the situation in 2009, there should have been an improvement?
Niki - The good news is that compared to 2009, many more studies are including both sexes. But one of the things that we found, which was a bit disheartening to hear, was that while inclusion has increased, the number of studies which analyse data by sex has not changed in 10 years.
Chris - Can you just clarify what that means in practical terms then, what you just said?
Niki - So when scientists have a study and they use both males and females, one might imagine that they would analyse their data based on the sex of their subjects. But what we're finding is instead most studies combine their data and look at it as one homogenous group. This really becomes an issue. If we're trying to determine things related to health and disease, we know the influence of sex impacts health and disease from disease severity, disease progression to incidence. So, you know, I think it's really important that if we're doing these basic science studies that are designed to be the early steps in, you know, the development of drugs and treatments and therapies that even in those early stages, we are considering the influence of sex at that early stage.
Chris - Do you think then that what you're seeing is just a nod to the rules to get the legislators off people's backs? Or do you think that actually there's something going on that can be fixed because it's an oversight?
Niki - That's a really good question. I would say, I think it's a combination of both. I do think that there are researchers who are, you know, stuck in their ways, thinking that this is the way I was trained. This was the way it's always been done. This is the way my colleagues do it. They might see these policies as simply another box to check in order to get grant funding. Alternatively one can say that maybe there's a lack of education and awareness about the influences of sex on health and disease. Because for many years it was assumed that there were no differences between males and females outside of the reproductive tract, but we know that to be false yet some of these behaviours and beliefs might persist within the biomedical research community and impact the way we conduct science today.
Chris - We've mentioned that the rules apply in the US. Are there equivalent rules governing other countries?
Niki - So I must admit, I am less familiar with the policies. I know there are similar pushes towards sex inclusion in Canada, but I am less familiar with the European research policies.
Chris - Where I was going with that is if your sample did not consider exclusively US data subject to that legislation, then it could be that the American side of things is performing beautifully, but actually it's other people who are not feeling the pressure legally to comply that are diluting the message.
Niki - Ah, yes, that's a very good point. And I think we can take the countries of origin of where the research studies are coming from. I think that's something I personally would like to look into in greater detail, but I think because of the high degree of international collaboration among scientists, I think this can't be a single country issue. I think the entire biomedical research community needs to come together and hold ourselves to higher standards. And ultimately I think that comes from journals themselves. Publishers, editors, and peer reviewers can really ensure that we have rigorous science being conducted by ensuring reporting and analysis based on sex.
Chris - You've highlighted with your study an important deficiency that is in current practice. So that's the challenge. How do you think it should be approached? Or what is the solution to this so that when you come back in 2029 and you do this again, what will you hope to see has happened?
Niki - So I think there are three strategies I would like to see happen. Number one is that all funders can be a gatekeeper and they can stipulate that you will not receive funding unless you provide justification for the use of single-sex studies. In addition to funders, publishers have a responsibility. All of these studies reviewed in this paper were from peer reviewed literature. And then ultimately I think it's a matter of education. This may not be on the radar of certain scientists, and if we cannot change the hearts and minds of seasoned investigators, then I think we can start with educating our trainees in the biomedical sciences.