If you’re trying to compare the positions in Jordan-Young, Fine, and Rippon with mainstream science on the crucial issue of masculinization of the male brain, I recommend two evenhanded reviews of the literature on early androgen exposure written by acknowledged experts in the field. One is “Early Androgen Exposure and Human Gender Development” by Melissa Hines, Mihaela Constantinescu, and Debra Spencer.61 Hines is director of the Gender Development Research Centre at Cambridge University. The other is “Beyond Pink and Blue: The Complexity of Early Androgen Effects on Gender Development” by Sheri A. Berenbaum, professor of psychology and pediatrics at Penn State and member of the Penn State Neuroscience Institute.62
Evidence from sexual disorders. Two intersex conditions in which a person’s biological sex is ambiguous have served as natural experiments about the effects of testosterone.
The first of these conditions is the one you have already encountered, classic CAH (as distinct from late-onset CAH; see Appendix 2 for details), in which female fetuses are exposed to high levels of testosterone. Except for that exposure, they are genetically normal females with two X chromosomes. CAH causes partial masculinization of the genitalia. Because the effect is visible, CAH is normally diagnosed and treated at birth through surgical feminization of the genitalia and correction of the hormonal abnormality. The overall effects on females with CAH were summarized in a review of the literature as follows:
Females with CAH differ from unaffected females (their siblings or age-matched comparisons) in a number of domains, including activity interests, personality, cognitive abilities, handedness, and sexuality. Thus, compared to controls, CAH females are more interested in male-typical activities and less interested in female-typical activities in childhood, adolescence, and adulthood, as measured by observation, self-report, and parent-report. The differences are large and, when multiple measures are used, there is very little overlap between females with CAH and control females.63
A subsequent study specifically focused on how 125 women with CAH compared to their unaffected siblings on the People-Things dimension. The researchers found that females with CAH had more interest in Things versus People than did unaffected females, and variations among females with CAH reflected variations in their degree of androgen exposure. On Prediger’s Things-People dimension, the effect size comparing women with and without CAH was –0.75 (indicating that women with CAH were farther toward the Things end of the continuum).64
In terms of their self-defined sexual identity, almost all adult women with CAH self-identify as women. Only about 1–2 percent choose to live as males, but that is far higher than the one per thousands in women without CAH. About 5 percent experience sex dysphoria, also far above the levels for women without CAH.65
The other sexual disorder is complete androgen insensitivity syndrome (CAIS). Fetuses with CAIS are genetically male, carrying the XY chromosome pair. Their testes produce prenatal testosterone in normal amounts at the proper times—but their androgen receptors are not functional. The testosterone circulates, but it has no effect. Persons with CAIS are born with externally normal female genitalia, reared as girls, are usually indistinguishable from girls behaviorally, and are usually designated as females in the technical literature despite their Y chromosome.66 It is also noteworthy that, despite their Y chromosome, CAIS carriers apparently do not have elevated spatial skills.67
A 2017 study by Swedish neuroscientists identified specific ways in which fetuses with a Y chromosome but affected by CAIS develop brains that are a mix of characteristically “male” and “female” patterns. Omitting the most abstruse results, women with CAIS displayed a characteristically female pattern of thicker parietal and occipital cortices and thinner left temporal cortex than male controls.68 On the other hand, the CAIS women displayed a “male” pattern in cortical thickness with regard to a thinner cortex in the precentral and postcentral gyrus. CAIS women were characteristically “female” with regard to the hippocampus volumes and “male” with regard to the caudate volumes. Add in the more abstruse findings, and the authors felt able to conclude that “the results indeed show considerable support” for the hypotheses they took into the study: The CAIS condition—no effect of testosterone on neural tissue—explains the similarities in brain structure between CAIS women and female controls, while the presence of the Y chromosome and its unique genes explains the similarities between CAIS women and male controls.69
Exploring the Interplay of Biology and Socialization
Accepting the organizational role of hormones does not require that we reject a role for socialization. Intuition tells us that both are probably involved, and scholars have made progress in exploring the balance.
In 2000, Richard Udry used an elegant experimental design to initiate an investigation of the nature-nurture balance. Udry took advantage of the natural variation in androgen levels among women even in the absence of a genetic disorder such as CAH. He assembled a sample of 163 adult women (ages 27–30 at the time of analysis) who had measures of their testosterone and SHBG70 values taken in utero (they were drawn from the Child Health and Development Study conducted in the 1960s). Udry’s researchers also obtained measures of the women’s adult levels of testosterone and SHBG along with a variety of questionnaire information that enabled Udry to assess the participants along four masculinity-femininity continua. The continua involved the importance of home, interests, job status, and personality. The results showed the predicted relationships of both prenatal and adult testosterone and SHBG to adult gendered behavior—but also showed independent relationships of childhood gender socialization to adult gendered behavior. The more interesting finding involved an interaction term: The effects of childhood socialization were confined to women who had low levels of prenatal androgen exposure. In Udry’s words, “if a daughter has natural tendencies to be feminine, encouragement will enhance femininity; but if she has below average femininity in childhood, encouraging her to be more feminine will have no effect.”71 A second interesting finding involved scores on a measure of the importance of spending time with one’s family. In adolescence, the women in the study had been tightly bunched in the middle of the range regardless of their level of prenatal androgen exposure. When interviewed at ages 27–30, those with above-average prenatal androgen exposure had moved to the “not important” position, while those with below-average prenatal androgen exposure had moved to the “very important” position.
In 2015, Shannon Davis and Barbara Risman drew from the same Child Health and Development Study that Udry used, but with a larger sample. In their article with the main title “Feminists Wrestle with Testosterone,” they used an analogous set of instruments, applied them to path analysis, and got complementary results. Their analysis showed effects of both prenatal androgen exposure and childhood socialization, with socialization playing the stronger role in terms of path coefficients.[72] Their finding that the effect of hormones was stronger for masculinity than femininity is consistent with Udry’s finding.
73
Male-Female Differences in Brain Lateralization and Connectivity
So far, I have discussed evidence that the prenatal and neonatal surges of testosterone are causally linked to a variety of phenotypic sex differences, but I haven’t directly addressed Geschwind’s hypothesis that the male brain is more lateralized than the female brain because of the organizing effects of testosterone.
Lateralization
Lateralization refers to the relative localization of a function in one hemisphere or the other. To apply it to the question we are discussing, Geschwind hypothesized that males primarily use the left hemisphere for verbal tasks and the right hemisphere for spatial tasks, whereas women use both hemispheres for both types of tasks. Another way of expressing it is that males exhibit more functional asymmetry.
Even before Geschwind formulated his hypothesis about testosterone and the right hemisphere, scholars had been considering the possibility that sex differences in test scores might
point to sex differences in the use of the left and right hemispheres of the cerebrum. As early as 1980, a review of the evidence in Behavioral and Brain Sciences cautiously concluded that the literature did not “overwhelmingly confirm” greater functional asymmetry in males, but among those studies that did find a sex difference, “the vast majority are compatible with [that] hypothesis.”74 A meta-analysis in 1996 came to the somewhat more confident conclusion that “it appears that sex differences in verbal and spatial abilities can be explained, at least in part, by the fact that men tend to be more lateralized than women.”75
Recall that the default in brain development is female. In the case of language processing, this means that the default is to use both hemispheres. The salient issue in analyzing the effects of testosterone on language processing is that something about the development of the male right brain is crowding out the use of the right hemisphere for language processing (which, by default, would also ordinarily be used). The focus of the right hemisphere on spatial processing, driven by the impact of testosterone on the right hemisphere, is a plausible explanation.
Clinical evidence from brain injuries reinforces the probability of sex differences in lateralization. When women suffer brain damage to the left hemisphere, they are less likely than men to develop language difficulties. Women’s language test scores after brain damage suffer the same effect whether the damage occurred in the left or right hemisphere, whereas men are more affected by damage to the left hemisphere.76 Researchers are also able to investigate this difference by anesthetizing just one hemisphere of the brain. Women lose language fluency no matter which hemisphere is anesthetized; men do so only if it is the left hemisphere.77
Most of the evidence from brain injuries had been developed before the advent of fMRI. Since then, neuroscientists have made major advances in understanding what’s going on inside the brain that produces the circumstantial evidence for greater male lateralization.
Evidence of Sex Differences in Structural Connectivity
Neurons in the brain produce thoughts and behaviors through their interconnections. But there has to be an architecture to those connections, just as there must be an architecture to any kind of network. That architecture is labeled structural connectivity. The map of those connections is called the connectome. The word was created to reflect its kinship to the map of the genome. There are many parallels. Maps of both the genome and connectome don’t answer questions of functionality in themselves, but they do provide a framework within which function can be studied. Both consist of structural elements at different levels of scale—from genetic regulatory networks to genes to base pairs in the genome; from brain regions to neuronal populations to neurons in the connectome.78
Don’t expect to see a fully mapped human connectome anytime soon. Whereas the genome has about 3 billion sites, the human brain has about 86 billion neurons. Whereas the sites of the genome lie sequentially along a strip of DNA, neurons can connect with many other neurons. The complexity of a connectome is such that, as I write, neuroscientists are still struggling to complete the connectome of the fruit fly larva, a brain that contains just 15,000 neurons.79
Neuroscientists are nonetheless able to construct connectomes using regions as the unit of analysis instead of individual neurons. In the early 2000s, it was established that the connectome follows a “small-world” topology.80 The phrase has a precise mathematical description, but the easiest way to understand it is through the popularized concept of “six degrees of separation,” referring to the assertion that you can establish a path between yourself and any other person in the world through no more than five intermediary links.81 The reason you can do it is that your personal cluster of acquaintances includes at least a few people who have a direct connection with some other distant cluster. That’s how the brain is organized: It is characterized by a high degree of local clustering of neurons, forming nodes and hubs, supplemented with random connections that permit direct pathways between distant nodes.
From the mid-1990s onward, new evidence from fMRI accumulated for greater lateralization among males.82 A group of neuroscientists at Beijing Normal University hypothesized that connectivity varies by brain size and by sex. They used DTI tractography to analyze MRI images of 72 healthy, right-handed young adults. They found no relationship between path length and either brain size or sex, which the authors interpreted as suggesting that “the global efficiency of structural networks of the brain is not affected by sex or brain size.”83 But their analysis of the efficiency of local clusters told a strikingly different story: The smaller the brain, the higher the efficiency of clusters—but only for women. The correlation between the authors’ clustering coefficient and brain size for women was a sizable –.53. For men, the correlation was a trivially small –.09. It’s a finding that bears on the similarity of g in males and females despite the larger male brain volumes that I discuss in Appendix 3.
In 2014, the state of knowledge saw a major advance through a study titled “Sex Differences in the Structural Connectome of the Human Brain.” First author was Madhura Ingalhalikar. She and her colleagues at the University of Pennsylvania and at Children’s Hospital of Philadelphia used the same Philadelphia Neurodevelopmental Cohort discussed in chapter 3. Their sample size was large: 428 males and 521 females. The 10 coauthors of the Ingalhalikar study parcellated the cerebrum into 95 regions, then used interregional probabilistic fiber tractography to compute the connection probability between regions, expressed in a 95-network matrix. Analyses were conducted by sex and by each of three age groups corresponding to the developmental stages of childhood (8–13 years), adolescence (13–17 years), and young adulthood (17–22 years).
Here are the main findings:
1. Male brains are structurally optimized for communicating within hemispheres. Female brains are structurally optimized for communicating between hemispheres. “Our analysis overwhelmingly supported this hypothesis at every level.”84
2. No significant age-by-sex interactions were found in the connection-based analysis—which is to say, there was no evidence that environmental forces operated from age eight onward to augment the contrasting male and female differences in connectivity.
3. A sex difference exists in the degree to which the connectome can be divided into distinct, separate modules (modularity), and a sex difference exists in the degree to which a given region is connected to its neighbors (transitivity). Both modularity and transitivity were globally higher in males (p < .0001 in both cases). Both results are consistent with a male brain that on average is wired for localized functionality and a female brain that on average is wired for cross-module functionality.
4. A sex difference exists in the degree to which the connections of regional nodes of the connectome are uniformly distributed across all the lobes of the cerebral cortex. The statistic for measuring this quality, the participation coefficient, was significantly higher for women in numerous regions in the frontal, parietal, and temporal lobes, whereas it was never higher for men in the regional nodes of the cerebral cortex. In contrast, the same coefficient was higher for men in the cerebellum.
A study of connectivity by a team of Swiss scientists published later in 2014 confirmed the pattern of connectivity found by the Ingalhalikar study, but argued that it was not the result of a sex difference per se but a function of brain size: “This pattern of connectivity can also be found within genders when comparing small-brained with large-brained women and small-brained with large-brained men.”85 For our purposes, the source of the distinctive connectivity patterns doesn’t make any difference. A recent meta-analysis of sex differences in brain volumes reported effect sizes for intracranial volume, total brain volume, and the cerebrum of –3.03, –2.10, and –3.35 respectively (i.e., favoring males).86 Effect sizes that large mean that the overlap of the male and female distributions is small. It implies a familiar situation—just as many women are taller than the average man, many women have brain connectivity patterns sim
ilar to those of a man with average brain volume. Nonetheless, a large sex difference in connectivity pattern remains because the sex difference in brain size is so large—a biological difference that is wholly unaffected by the environment.
The authors of the Ingalhalikar study speculated about the relationship of these structural differences to neurocognitive functioning. For males (or for women with unusually large brains), “Greater within-hemispheric supratentorial connectivity combined with greater cross-hemispheric cerebellar connectivity would confer an efficient system for coordinated action” and was consistent with results from fMRI studies showing “greater focal intrahemispheric activation in males on a spatial task, in which they excelled.”87 For females (or males with unusually small brains), greater interhemispheric connectivity “would facilitate integration of the analytical and sequential reasoning modes of the left hemisphere with the spatial, intuitive processing of information of the right hemisphere,” and was consistent with results from fMRI studies “which have reported greater interhemispheric activation in females on a language task, in which they excelled.”88
Evidence for Sex Differences in Functional Connectivity
A year after the Ingalhalikar study of structural connectivity, a companion study looked at sex differences in functional connectivity. The first author was Theodore Satterthwaite, who had also been a coauthor of the Ingalhalikar study.89 Like the Ingalhalikar study, it used the Philadelphia Neurodevelopmental Cohort, restricted to male-female pairs matched on age and in-scanner motion (the results from fMRI studies can easily be contaminated by head motion during the scan). This procedure resulted in a sample consisting of 312 males and 362 females.
Human Diversity Page 14
