Introduction

Many behaviors differ between girls and boys, and women and men [1]; however, it is often difficult or even impossible to determine the causes of these differences. For example, if a girl prefers to play with dolls instead of rough-and-tumble wrestling, is this because she was taught by her family and friends to prefer dolls? Perhaps her proclivity for dolls results from actions of her genes and/or hormonal environment during critical periods of her development. Among the vast majority of children and adults who exhibit sexually dimorphic behaviors, influences such as learning/socialization, sex chromosomes, and hormone exposure cannot be teased apart for independent consideration. For this reason, the study of sexually dimorphic behaviors in people affected by disorders of sex development (DSD) is useful. Two categories of DSD that are particularly instructive about the origins of sexually dimorphic behaviors in females are complete androgen insensitivity syndrome (CAIS) and congenital adrenal hyperplasia (CAH) due to 21-hydroxylase deficiency. Therefore, these specific types of DSD are the focus here.

Investigating genetic and hormonal effects on sex differences in cognition

DSD collectively refers to a group of conditions in which discordance between genetic, hormonal, and anatomic sex exists in a person at birth [2]. For example, a person with 46,XY DSD may possess a female phenotype throughout life despite the presence of a Y chromosome in every cell of her body. A specific type of 46,XY DSD, CAIS, results when a person who possesses a male-typical chromosome complement (46,XY) and testes develops a female phenotype due to her end-organ insensitivity to androgens [3]. In contrast, a person with 46,XX DSD may possess an ambiguous or male phenotype despite the possession of a female chromosome complement. One type of 46,XX DSD, CAH due to 21-hydroxylase deficiency, results when a person with ovaries develops a masculinized phenotype caused by excessive androgen production an inability to produce the enzyme 21-hydroxylase [4].

Thus, females with CAIS and CAH allow for the independent assessment of male-typic genetic or hormonal influences on sexually dimorphic behaviors in affected girls and women. In other words, if women with CAIS behave similarly to unaffected women on measures of sexually dimorphic behaviors, then the possession of a Y chromosome alone is insufficient to support the expression of male-typic behaviors in humans. In contrast, if women with CAH behave similarly to men, then early androgen exposure is sufficient to support male-typic behavior. Importantly, by comparing women with CAIS to those with CAH, it is possible to determine the impact of possessing a Y chromosome versus early androgen exposure independent of socialization and learning on the expression of sexually dimorphic behaviors in humans (Table 14.1).

Sexually dimorphic cognitive behaviors

Cognitive behaviors that show strong sex differences in humans include gender identity (GI), gender role (GR), and performance on some visuo-spatial tasks [1]. We will first consider GI and GR, as these reflect some of the largest sex differences observed in human behaviors. GI is the internal experience a person has of being female or male, while GR refers to overt behaviors expressed by people, such as toy preferences, that society categorizes as feminine or masculine. The vast majority of women studied report a female GI [5]; however, GR is more variable among girls and women and standards for feminine and masculine behaviors are continuously evolving over time and across cultures [6]. Most girls and women with a female GI and GR have been socialized as female, possess a 46,XX chromosome complement, and experience minimal early androgen exposure and thus, no phenotypic masculinization. Thus, studies of GI and GR in women with CAIS and CAH are needed to understand if, and how, sex chromosomes and early androgen exposure influence these behaviors apart from learning.

Girls and women with CAIS overwhelmingly report female-typical GI and GR [7–9]. Therefore, female patterns of gender can develop in people despite their possession of a Y chromosome. The majority of girls and women with CAH also report a female GI, but the incidence of male GI development in this group is higher than in unaffected women or women with CAIS. Furthermore, girls and women with CAH are frequently labeled as tomboys because they often prefer male-typic toys and play activities, as well as hobbies and professional interests [10]. Thus, there is a greater impact of early androgen exposure, compared with possessing a Y chromosome in the absence of androgenic effects, on GI. Studies of girls and women with CAH show that the impact of androgens on GR is even greater than the relationship between androgens and GI [11]. However, it is not completely clear whether androgen exposure during critical periods associated with CAH directly masculinizes GR via steroid action at the central nervous system, or whether androgens indirectly impact behavior of affected girls and women by influencing how society responds to their masculinized appearance (i.e., hirsutism and masculinized external genitalia) [12].

Another type of cognition that differs significantly between females and males, and also is thought to be influenced by early androgen exposure, is visuo-spatial processing [13,14]. Examples of visuo-spatial processing that elicit large sex differences in favor of males include judgment of line angle orientation and mental rotation of objects. In a study of visuo-spatial cognition in females affected by CAH, the severity of the condition (i.e., a surrogate marker for excessive prenatal androgen exposure) was positively associated with performance on these male-biased, visuo-spatial tasks. Additionally, excessive postnatal androgen exposure also predicted better visuo-spatial performance of women with CAH [15]. Thus, the influence of androgens on cognition in humans is not restricted to critical periods prior to birth, but can also extend throughout childhood. In contrast to GI and GR, visuo-spatial abilities have largely been understudied in girls and women with CAIS. Experiments that employ animal models of XY females, although also few in number, indicate that XY female rodents perform overwhelmingly like their XX counterparts [16; but see 17]. Therefore, similar to GI and GR, androgen exposure during critical periods of early development exerts a greater influence on male-typic, visuo-spatial performance than possession of a sex chromosome complement.

Although clear evidence exists from nonhuman animal studies and clinical investigations of people with DSD for a role of early androgen exposure to influence performance on visuo-spatial tasks, there is also evidence for environmental and learning influences as well. For example, access to education for girls and women ameliorates the male-biased sex difference in a type of visuo-spatial processing known as line angle judgment [18; but see 19 for contrasting results]. Additionally, the male advantage for mental rotation is eliminated following a series of eight 45-minute training sessions for children in first grade [20], indicating that modest training is sufficient to override visuo-spatial advantages associated with early androgen exposure. To what extent education and training might modulate early androgen influences on GI and GR is less amenable to study; however, the differences observed between masculine and feminine GR across history and cultures imply that they too are impacted by learning as well as hormonal influences.

Sex differences in mental health status

Similar to cognition, aspects of mental health differ in prevalence, incidence, and expression between males and females [21–23]. For example, women are more likely to report and seek treatment for depression [24], whereas men have a greater likelihood of being treated for addiction [25]. Because the female preponderance for depressive symptoms is revealed during adolescence, a role for sex steroids is thought to exist [26]. In contrast, the greater tobacco, alcohol, and illicit drug use evident in men does not occur until later adulthood [27], suggesting that perhaps societal influences over time differentially influence addictive behaviors in men and women.

In our long-term outcomes study of 14 women with CAIS, only 1 participant reported problems with depression and addiction [7]. Additional studies conducted by others also report no increase in depressive symptoms or suicidal ideation associated with having CAIS [28–30]; however other investigations of mental health in this group reveal increased suicidal tendencies [8,31,32]. Thus, while it is inconclusive at this time whether women who possess a 46,XY chromosomal complement as a result of CAIS do, or do not, experience increased depressive symptoms, it seems that a male karyotype does not by itself protect women from depression. The variability in reported depressive symptoms across studies is likely due to methodological problems, such as small sample sizes and participation bias [33]. Studies of addictive behavior in women with 46,XY DSD are needed, but to date have not been performed, to understand what role genes encoded on the Y chromosome may play in addiction.

Mental health studies of women with CAH can reveal whether or not prenatal androgen exposure offers protection against depressive symptoms or predisposes a person to addictive behaviors (the male-typical pattern for each). Similar to CAIS, no studies have been conducted to date on tobacco, alcohol, or illicit drug use in women with CAH. Regarding depressive symptoms, women with CAH are more likely to report a history of suicide attempts and psychopharmacologic treatment for depression than unaffected women [29, 30]. Greater virilization in women with CAH at the time of their diagnosis is positively associated with depressive symptoms, suicide attempts, and psychopharmacologic treatment. Of interest, a similar association between depressive symptoms and polycystic ovary syndrome has also been demonstrated [34], suggesting that signs of virilization such as hirsutism (commonly evident in women with CAH or polycystic ovary syndrome) predispose women to experience depression. Furthermore, surgical treatment and the functional outcomes associated with such procedures are also associated with self-reported depressive symptoms in women with CAH [30]. Thus, it remains to be seen how androgen exposure, either directly or indirectly, exerts an effect on depression in women with CAH. Additionally, surgical treatment and subsequent functional outcomes for women with CAH who are born with ambiguous or even male-typic external genitalia can also impact mental health [30]. For example, a recent study of women with CAH reported distress associated with the timing and type of surgical treatment received for CAH, as well as sexual function following those surgeries [30]. Thus, this is an important reminder that differences noted in cognition or mental health among women with CAH, CAIS, or their unaffected counterparts may be due to medical and/or surgical therapies and not whether a Y chromosome or early androgen exposure is present.