A study of 27,000 mice shows that it depends on which trait you measure

First Published on Medium Nov 19, 2020 · 6 min read

Despite awful pop stereotypes about scientists slaving away to uncover monolithic truths about the universe, disagreeing only occasionally when a new theory comes along, reality turns out to be far more complicated. We scientists have to break problems down into smaller parts, dissecting one aspect of the observable world at a time. Occasionally, researchers working on different problems, or coming from different backgrounds, discover just how different they are and just how messy the reality they are observing can be. The encounter often undermines old certainties, and it can leave scars.

The study of sex differences is one of those areas where bruising encounters between research traditions, methods, and world views seem to have become the norm. My own field of evolutionary biology considers how and why sex differences arise. We ask why male spiders are so tiny, whereas females of the same species can be hundreds or thousands of times bigger. Why do lionesses live in prides with their sisters and aunts, hunting together and raising cubs, while lions come and go, wandering large territories and fighting off rivals? Mostly, the answer comes down to males and females having different ways of maximizing their reproductive success.

In the interests of social progress, some thinkers prefer to ignore biology

Apply this kind of thinking to humans, however, and you’d best be ready for a fight. Women and men do differ in many traits, but the complex origins of those differences in biology, socialization, culture, and economics remain notoriously difficult to unpick.

For centuries, some people have interpreted the sex roles they observe in other animals as parables justifying their preferred form of human social arrangements. Unsurprisingly, these parables have tended to favor the interests of wealthy, powerful men over those of other groups. In the interests of social progress, some thinkers prefer to ignore biology, treating it as irrelevant, and those who study it as abetting oppression.

Medicine for men?

Biomedical research took a different path than the science of society. The plain fact that females and males differ in some traits resulted in a pragmatic but flawed approach of studying only one sex and ignoring the complexities inherent to studying both sexes. Often, researchers focussed on males, arguing that reproductive hormone cycles in females add bothersome ‘noise,’ blurring the crisp lines of scientific evidence.

The staggering chauvinism involved in writing off as noise the natural cycles experienced by half a species cannot be escaped. More than that, medical interventions based on studies of male animals, and optimized in clinical trials on men, leave women’s and girls’ health behind.

The last decade or so has brought a great reckoning in medical research, in which studies, especially clinical trials, have had to grapple with sex differences. Research organizations and funding agencies no longer support work or accept results obtained on only one sex. Evolutionary biologists stand by, amazed, as the biomedical community has rediscovered sex and the importance of understanding sex differences.

Averages only take us so far

Despite a welcome renaissance in the study of sex differences, most of the studies involved consider only differences between the male and female averages. Men are, on average, taller than women. We all know that’s not the whole story. A moderately tall woman is likely to be taller than a modestly short man. The expression ‘men are taller than women’ is only true under the assumption that we are talking about the two groups’ average properties.

We don’t really know whether males are more variable than females, or the reverse

One of the reasons so much research was done on male animals, and tested on men, is the assumption that female animals, including human females, would be more variable. Hormonal changes during the estrus cycle were expected to make females more variable than males, making it harder to discern whether drugs and other interventions are ‘working.’

As any evolutionary biologist will point out, however, there are two good reasons to expect males to vary more dramatically than females. First, female mammals, including humans, have two X-chromosomes, whereas males have only one X and a Y. There aren’t many genes on the Y, apart from the instruction that turns a female embryo into a male one. But the X chromosome contains genes that influence many different traits. The two copies of these genes in females tend to average one another out. With only one X and thus one copy of each gene, Males are more likely to have extreme values of any trait that those genes influence.

The second reason is that the traits males use to attract females and compete against rivals are, in many species, more extreme than matching traits in females. A male who grew up in good circumstances into a vigorous adult is likely to outperform females who had a similarly good start, but the opposite is true for males and females who grew up in poor conditions. This seems to happen in animals, from earwigs to deer.

The important bit, especially for medicine, is that we don’t really know whether males are more variable than females or the reverse. Today, however, we are a little closer to knowing the answer.

Of mice and (wo)men

A team, led by my brilliant colleagues Susi Zajitschek, Shinichi Nakagawa, and Dan Noble, set about analyzing a massive dataset compiled by the International Mouse Phenotyping Consortium. (Disclosure: I was lucky enough to be part of their large research team). Using data from 26,918 mice from nine different genetic strains studied in 11 different institutions, we assessed variation among males and among females in 218 different traits. Those traits included almost anything you might consider measuring, including anatomic structures, physiological function, and behavior. By applying sophisticated statistical approaches to analyzing data from many sources, the question of which sex varies more could finally be settled.

The answer? As is so often the case in science: “It depends.” In this case, it depends on which traits. According to the paper, published in the journal eLife, males tended to vary more than females in the size-related measures. Females varied more in traits related to immunity and in eye traits. There weren’t many differences in variation in behavior, metabolism, physiology, blood traits, heart measures, or hearing.

If you want to know more about sex differences — both in averages and in variation — check out the cool online tool that my colleagues built to accompany the paper. We hope that this resource will make it easier for researchers to design their experiments on males and females, building in some prior knowledge of how much the trait they are interested in varies in each sex.

Embrace the complicated

Sex-dependent variation has serious real-world consequences beyond the biomedical sphere. A study led by Shinichi’s former Ph.D. student Rose O’Dea tested the often-repeated assertion that boys vary more than girls in mathematics and science performance. If that were true, then even if the sexes performed no differently on average, you’d see more boys at the extremes, among the worst performers and the geniuses. Instead, in a dataset of 1.6 million students, maths and science performance was no different, on average or in variability. That’s a serious data-driven rejection of the claim that greater variability pushes more males to the top of the STEM classes.

Studying both males and females, and staying open to the fact that they might differ is important. Add to that the insight that differences in variation are also worth considering, keeping track of, and understanding. Simplification and reductionism play important roles in science. Still, occasionally we scientists need to look outward again, to the complications of the real world, and the uncomfortable insights of other disciplines, and embrace them.