For Armin Raznahan, publishing research on sex differences is a fraught proposition. Now chief of the section on developmental neurogenomics at the National Institutes of Health, Raznahan learned early that searching for dissimilarities between men’s and women’s brains can have unintended effects.
“I got my fingers burned when I first started,” Raznahan says. As a PhD student, he published a study that examined structural differences between men’s and women’s brains and how they changed with age. “We observed a particular pattern, and we were very cautious about just describing it, as one should be, not jumping to functional interpretations,” he says. Despite his efforts, The Wall Street Journal soon published an article that cited his study in a defense of single-sex schooling, under the assumption that boys and girls must learn in distinct ways because their brain anatomy is slightly different. “That really threw me,” he says. “The experience has stayed with me.”
Nevertheless, Raznahan has continued to study sex differences, in the hope that they could help us better understand neurodevelopmental disorders. He focuses on people with sex chromosome aneuploidy, or any variation other than XX (typically female) and XY (typically male). People with genetic variations (such as XXY) have an inflated risk of autism spectrum disorder, ADHD, and anxiety, among other ailments. Raznahan’s hope is that uncovering if and how men’s and women’s brains differ—for example, in the sizes of regions or the strengths of the connections among them—could help us figure out why people with aneuploidy are more likely to experience neurodevelopmental and psychiatric concerns. Solving this puzzle could be a step toward unlocking the perplexing mystery of psychiatric illness.
Last Monday, he and his team published a study in Proceedings of the National Academy of Sciences that not only reported reliable sex differences in the volumes of certain parts of men’s and women’s brains but also tied those differences to the direct influence of sex chromosomes. “What we were seeking to do in this study is ask some relatively simple questions that hadn’t quite ever been directly addressed,” Raznahan says. “How reproducible are anatomical sex differences in the human brain? And what sorts of functional systems, if any, do they seem to be co-localized with?”
To attack these questions, Raznahan and his team didn’t collect their own data; rather, they leveraged the Human Connectome Project, which collects brain scans taken from over 1,000 participants at several institutions. Using conventional MRI data, they compared how much gray matter—the tissue containing the cell bodies of most neurons—the sampled men and women had in various brain regions. While most regions looked similar, in some spots, either men or women seemed to have more gray matter.
By comparing their results to another large data set, Raznahan’s team found that these brain regions were associated with areas where genes on the sex chromosomes are disproportionately expressed. For Raznahan, this potential link between chromosomes and brain structure is particularly exciting. “If we can understand the biology of sex better, maybe those pathways are going to help us understand what is happening to put a person at risk of manifesting symptoms of autism spectrum disorder, for example,” he says.
But other scholars question the idea that this sort of research will help us understand mental disorders. Lise Eliot, a professor of neuroscience at Rosalind Franklin University, believes that sex differences in the prevalence of some disorders might be better explained by biases among doctors, or in the diagnostic criteria, rather than by biology. Proponents of sex difference research often note that boys are diagnosed with autism about four times as often as girls, but Eliot questions the relevance of that statistic. “I think a lot of that is just diagnostic bias,” she says. “The definition of the disorder is based on a male stereotype.”
And without an obvious medical benefit, Eliot thinks this type of research will simply reinforce the idea that men and women are fundamentally different, or even justify misogyny—although the authors may not intend such an outcome. This research is “far from having medical value,” she says. Instead, it can “validate the fixed, hardwired, God-given—however you want to put it—differences between the sexes, so that we can get over this idea of real equality.”
Concerns like these are one reason why sex difference research in neuroscience has attracted so much controversy. But worries about consistency have also plagued the discipline. Studies that report sex differences in the sizes of brain regions, or in how strongly some regions are connected to others, often disagree about just where those differences lie. “The longer people have been at it, the muddier it’s gotten,” Eliot says.
This inconsistency might arise from a bias among scientists in favor of reporting studies that demonstrate sex differences rather than similarities. In 2018, a group of researchers from the Meta-Research Innovation Center at Stanford found evidence that scientists are more likely to publish studies that uncover potential sex differences than those that find none. Because studies that only examine a small number of subjects are prone to false positives, the authors warn, a bias in favor of publishing those positives makes it hard to know how much that research can be trusted.
Raznahan and his team were well aware of these shortcomings, so they worked to ensure that any differences they found would reflect real patterns in brain anatomy, not the random quirks of the people in the data set. Observations that hold across a substantial number of subjects are more likely to apply to the population as a whole, which is why they relied on the Human Connectome Project’s large data set. After analyzing this data and correcting for total brain volume (just like men’s bodies are, on average, larger than women’s bodies, so too are their brains), they discovered a number of apparent differences.
Among them was a relative size advantage for men in parts of the occipital lobe (which is associated with vision) and in the amygdala and hippocampus (regions that play important roles in emotion and memory). Women, on the other hand, had more gray matter in parts of the prefrontal cortex (which is associated with decisionmaking and self-control) and the insula (which has been connected with numerous functions, including emotion and the sense of taste). These results might seem to suggest that women have an edge over men in decisionmaking and that men have better memories, but it’s impossible to extrapolate such broad conclusions from Raznahan’s results. “It could be that there's absolutely no behavioral relevance for what we're finding,” he says.
To begin with, it’s not clear what gray matter volume really means for brain function. The brain contains two major types of tissue: gray matter, which holds neuron cell bodies, and white matter, which connects gray matter in “tracts” and allows neurons to send signals to distant areas. They depend on each other to carry out their functions, and it’s not obvious whether having a larger volume of either one is advantageous.
“In no way is more gray matter a good thing, necessarily,” says Margaret McCarthy, a professor of pharmacology at the University of Maryland School of Medicine. “It's just a measure that there's a difference in the way the neurons are, how many synapses they're making, how many there are, possibly, and stuff like that.”
Yet it’s easy for people to jump to the conclusion that size matters, and there is some evidence in favor of that notion. Two decades ago, researchers at University College London made a particularly big splash when they published a study showing that London cab drivers, who must learn how to navigate a large, disorganized city by memory alone, have much larger hippocampuses than most other people. Since the hippocampus is linked to spatial memory, this study seemed to show that the size of a brain region correlates with a particular skill. But showing that learning a skill is associated with the growth of a specific brain region does not imply that the region’s size is correlated with that ability in general. So the study doesn't imply that simply being born with a big hippocampus means a person will have an excellent memory.
And in most cases, ascribing a function like “emotion” or “decisionmaking” to an individual region is a gross oversimplification. In fact, most regions have a number of functions, and most of them demand the cooperation of numerous regions. So to see if the set of regions that showed sex differences in his sample was connected with any particular functions, Raznahan took advantage of a database called Neurosynth, which documents these many-to-many relationships by compiling thousands of human neuroscience studies that report links between brain regions and functions. Comparing his findings to the Neurosynth data, he found only one clear result: Some of the regions he had identified were associated with face processing, which involves recognizing faces or the emotions they are expressing.
Perplexingly, while Raznahan found that the face-processing areas of the brain were larger in men than in women, previous research has shown that women actually tend to perform better than men at face-processing tasks. So it’s difficult to say whether Raznahan’s observations about gray matter volume have anything to do with differences in behavior—no matter how eager some readers might be to arrive at such a conclusion. “It’s super important for us to say that up front,” Raznahan says.
Geert de Vries, director of biology at Georgia State University’s Neuroscience Institute, isn’t terribly surprised that Raznahan and his team didn’t find a straightforward link between anatomical variations and functional differences. “Many of the differences that people find,” he says, “might be done in the first place to allow a male brain to work optimally in a male body, the female brain optimally in the female body.” In other words, a difference in brain structure might act not to generate different behaviors in men and women but to produce more similar behaviors. “Most functions should not be that different between males and females, if most of what we do is simply surviving,” de Vries says. “Males don’t survive that differently from females.”
Raznahan and his colleagues do have a biological theory about what’s behind the gray matter differences they observed. They used data on the brain’s transcriptome, or the genes that are preferentially turned on to make proteins in each brain region, to determine which genes appeared to be more active in the areas where they observed the biggest differences. One pair of chromosomes stood out from all the rest: the sex chromosomes. Unlike the others, they were markedly more active in areas where men had more gray matter than women, as compared with the rest of the brain.
“It certainly doesn’t prove that the sex [chromosomes] are causally relevant for the anatomical sex differences, but it’s more guilt by spatial association,” Raznahan says. Like a detective who always observes the same person at the scene of the crime, Raznahan has reason to be suspicious that the chromosomes’ expression has an influence on differences in gray matter volume when they show up in the same brain region—but no direct proof.
If Raznahan’s hypothesis about the chromosomes' role is correct, it could have a substantial influence on sex difference research. According to de Vries, scientists have long assumed that sex chromosomes don’t have a direct effect on brain structure. Rather, he says, their presumed influence was more circuitous: Genes would give rise to the gonads, which would then secrete hormones that could influence brain development. The going idea, he says, has been “that what genes do is they determine that there's going to be a testis or an ovary. And once you've made that decision, the hormones take over.”
But Raznahan’s observations seem to suggest that sex chromosomes could influence brain anatomy without hormones acting as a mediator. “The fact he’s finding genes that are on sex chromosomes,” McCarthy says, “is really very exciting and very novel, because we so much emphasize the role of gonadal steroids.” (These are hormones like estrogen and testosterone.)
Nevertheless, the study falls short of demonstrating that these chromosomes have an effect on structure. Correlation is simply not the same as causation. Raznahan’s work, says de Vries, “doesn’t necessarily suggest that these sex chromosomal genes cause these differences.”
What else could have played the critical role? Possibly environmental factors, like gender roles or the psychological stressors of belonging to an oppressed group. Just as the London cab drivers experienced remarkable changes in brain anatomy as they learned the complex layout of their city, so too might men’s and women’s brains change in response to the particular demands placed on them by their societies. “No one’s exploring the actual environment side, because it’s a lot harder to do,” Eliot says. But Raznahan, de Vries, and McCarthy agree that an environmental explanation is unlikely.
No matter what the cause might be, all of the researchers agree that it would be hard to generalize from this study to the entire world population. The data sets the researchers used came from the US and the UK, which are both wealthy, predominantly white countries. “The whole of neuroscience and genetics is so heavily skewed in terms of its cultural representation,” Raznahan says. “How these features might vary or remain the same across different societal structures would be a really important thing to look at. It’s challenging because the data just aren't there.”
Where does this leave us? Raznahan’s team found a pattern of sex differences in the brain and associated that pattern with sex chromosome expression. But it remains unclear whether these differences mean anything at all in terms of psychology and behavior. Yet people may still take the differences to indicate that there’s a fundamental, biological distinction between men and women. It’s this link between the brain and real-world consequences—behavior, cognition, emotion—that makes this research so controversial, de Vries argues. “The moment it’s about the brain, something differs, some switch is pulled. And I think it's because it comes so close to what we think defines us,” he says.
Eliot goes even farther: She contends that the research community as a whole, whether consciously or not, is looking in the brain for evidence that men and women have essentially different natures. “Why are there so many studies of human brain sex difference?” she asks. “I challenge you to find some studies on human kidney differences, human lung differences. There are probably as many differences, if not more.”
But for Raznahan, de Vries, and McCarthy, the possibility of making progress toward understanding mental disorders makes studying brain sex differences an ethical imperative, even if others may use their research to defend essentialist or sexist beliefs. “We can’t not make discoveries,” McCarthy says, “because they might be misused.”