Sheri A. Berenbaum

Systematic behavioral studies show that females with congenital adrenal hyperplasia due to 21-hydroxylase deficiency (CAH) are masculinized and defeminized in several ways; compared to their sisters, they play more with boys' toys, are more likely to use aggression when provoked, and show less interest in infants. We studied the extent to which these behavioral changes could be attributed to high levels of androgens in the prenatal vs. postnatal periods in 23 girls with CAH, aged 3-12 yr. Sex-atypical behavior was significantly associated with degree of inferred prenatal, but not postnatal, androgen excess; marked boy-typical play was associated with severe salt-wasting CAH, early age at diagnosis, and moderate genital masculinization at birth, but not with bone age advance, concurrent or cumulative high levels of 17-hydroxyprogesterone, or accelerated growth velocity in early childhood. Aggression and interest in infants were not consistently associated with indicators of prenatal or postnatal androgen excess, probably because those behaviors were measured less reliably than was toy play. The results are consistent with the idea that behavioral masculinization in girls with CAH results from high levels of androgens during fetal development and not in postnatal life.

There are several ongoing studies of psychological outcome in children identified through neonatal screening with phenylketonuria, congenital hypothyroidism or congenital adrenal hyperplasia. These studies document the reduction in mental retardation, learning difficulties and behaviour problems associated with neonatal screening. They also describe other behavioural changes resulting from these disorders. Some behavioural changes are transient or preventable with early diagnosis and treatment, whereas some appear to be irreversible, reflecting permanent brain effects of abnormalities associated with the disease. Despite the variety of disorders studied, there are commonalities in approach and issues, including a developmental neuropsychological perspective resulting in behavioural assessments specific to the disorder and its likely manifestations, a recognition that behaviour may change across time in association with brain development and an understanding of the neural mechanisms underlying the behavioural changes. Assessments go beyond IQ, and include specific intellectual abilities, social behaviour, temperament, behaviour problems and identity. Behavioural changes are related to specific characteristics of the disease and its treatment, including the timing of exposure to abnormal hormones and/or neurotransmitters, the severity of the disease, and the age at initiation and adequacy of treatment. These studies provide information about the ways in which hormones and neurotransmitters affect the development and ongoing function of the brain, and an understanding of the ways in which neonatal screening results in improved psychological outcome. They also provide models for psychological follow-up of other disorders detected through neonatal screening.

There are several ongoing studies of psychological outcome in children identified through neonatal screening with phenylketonuria, congenital hypothyroidism or congenital adrenal hyperplasia. These studies document the reduction in mental retardation, learning difficulties and behaviour problems associated with neonatal screening. They also describe other behavioural changes resulting from these disorders. Some behavioural changes are transient or preventable with early diagnosis and treatment, whereas some appear to be irreversible, reflecting permanent brain effects of abnormalities associated with the disease. Despite the variety of disorders studied, there are commonalities in approach and issues, including a developmental neuropsychological perspective resulting in behavioural assessments specific to the disorder and its likely manifestations, a recognition that behaviour may change across time in association with brain development and an understanding of the neural mechanisms underlying the behavioural changes. Assessments go beyond IQ, and include specific intellectual abilities, social behaviour, temperament, behaviour problems and identity. Behavioural changes are related to specific characteristics of the disease and its treatment, including the timing of exposure to abnormal hormones and/or neurotransmitters, the severity of the disease, and the age at initiation and adequacy of treatment. These studies provide information about the ways in which hormones and neurotransmitters affect the development and ongoing function of the brain, and an understanding of the ways in which neonatal screening results in improved psychological outcome. They also provide models for psychological follow-up of other disorders detected through neonatal screening.

Early androgens have been shown to facilitate male‐typical behavior in people, but little attention has been paid to androgen effects on female‐typical behavior. We studied the effects of early androgen on human interest in infants, attempting to extend studies in rodents and primates that indicate that exposure to high levels of androgen in the prenatal and early postnatal periods reduces the expression of maternal behavior in juvenile and adult animals. Parents completed a questionnaire about the behavior of children with congenital adrenal hyperplasia (CAH) who had been exposed to high levels of androgens early in life, and their unexposed siblings. As hypothesized, girls with CAH were reported to have less interest in infants than their sisters. These results suggest that early androgens may act to suppress some aspects of female‐typical behavior in people, as in other species, and that sex differences in maternal behavior result, in part, from early hormones.

As the articles in this special issue demonstrate, gonadal hormones have powerful effects on the development of the brain and behavior in human beings, as in other species. Both androgens and estrogens affect behavior throughout development, from early prenatal life through adulthood, as demonstrated in studies with a variety of methods in several species. High levels of testosterone and other androgens that are present early in development are shown to facilitate the development of male‐typical characteristics, and to inhibit the development of female‐typical characteristics. Some of these effects are suggested to be mediated by the conversion of androgen to estradiol in the brain. Ovarian hormones are also shown to play an important role in the development and maintenance of female‐typical characteristics, and their effects on the brain appear to extend beyond the prenatal and early postnatal periods. Nevertheless, there are not simple relations between the amount of hormone present and behavior. Behavioral effects of hormones are not uniform across behaviors or across individuals. Variations have been shown to relate to the timing of exposure, the organism's sensitivity to the hormones, the specific hormone involved, and modification by the physical and social environment, although not all factors have been studied in all species, and many have not been studied directly in people. The articles in this special issue also describe attempts to identify the mechanisms—neural and basic behavioral—that mediate hormonal effects on complex human behaviors.

The authors examined the neural and cognitive bases for sex differences in verbal memory in 57 patients who underwent left anterior temporal lobectomy (ATL) for the treatment of intractable seizures. On the California Verbal Learning Test (D. C. Delis, J. H. Kramer, E. Kaplan, & B. A. Ober, 1987), women recalled more words than men both before and after surgery, regardless of the extent of hippocampal damage. Extent of hippocampal sclerosis was related to memory loss in both men and women. Women's superiority in verbal memory appears to result in part from their use of an efficient encoding strategy. Women were more likely than men to use semantic clustering both before and after ATL, and sex differences in word recall were attenuated after scores were adjusted for semantic clustering. There was no effect of ATL on semantic clustering. Taken together, these results suggest that sex differences in verbal memory are not due to differences in the integrity of the left hippocampus.