Heritability of Testosterone Levels in 12-year-old Twins

The aim of this study was to estimate the heritability of variation in testosterone levels in 2-year-old children, and to explore the overlap in genetic and environmental influences on circulating testosterone levels and androgen dependent pubertal development. Midday salivary testosterone samples were collected on two consecutive days in a sample of 83 unselected twin pairs. Androgen induced pubertal development was assessed using self report Tanner scales of pubic hair development (boys and girls) and genital development (boys). A significant contribution of genetic effects to the variance in testosterone levels was found. Heritability was approximately 50% in both boys and girls. The remaining proportion of the variance in testosterone levels could be explained by nonshared environmental influences. The relatively high correlation between testosterone levels of opposite sex dizygotic twins suggests that sex differences in genes influencing variation in testosterone levels have not yet developed in preand early puberty. Variance in pubertal development was explained by a large genetic component, moderate shared environmental influences, and a small nonshared environmental effect. Testosterone levels correlated moderately (r=.3) with pubertal development; the covariance between testosterone levels and pubertal development was entirely accounted for by genetic influences. INTRODUCTION Puberty is the hallmark period in which a child undergoes the transition to adulthood. Driven by hormonal changes, major physical changes occur during this period of development. Puberty starts with the activation of the hypothalamic – pituitary-gonadal (HPG) axis by an increase of the pulsatile release of gonadotropin releasing hormone (GnRH) in the hypothalamus (Grumbach & Styne, 2003; Sisk & Foster, 2004). GnRH induces secretion of luteinising hormone and follicle stimulating hormone in the pituitary, that in turn direct the testes and ovaries to produce sperm and eggs, and activate secretion of steroid hormones. Typically a few years prior to these hormonal changes, children experience the adrenarche, a process biologically independent from the activation of the HPGaxis. Adrenarche results from the activation of the hypothalamic-pituitary-adrenal axis, leading to secretion of dehydroepiandrosterone (DHEA), DHEA-sulphate and androstenedione by the adrenal glands (Auchus & Rainey, 2004; Grumbach & Styne, 2003). Extraglandular conversion of androstenedione leads to increasing levels of the androgens testosterone and dihydrotestosterone. The increasing levels of circulating hormones lead to different physical changes in puberty. In girls, breast development is induced by increasing estrogen levels, excreted by the ovaries. The development of pubic hair is mainly influenced by androgens, originating from both the ovaries and the adrenal gland. In boys, both genital and pubic hair development are under androgen control (Grumbach & Styne, 2003). Most of the circulating testosterone levels in boys are secreted by the testes. In addition to direct secretion, a small amount of testosterone is derived from peripheral conversion of androstenedione secreted by the testes and adrenal glands (Grumbach & Styne, 2003). Variation in pubertal timing and individual differences in circulating sex hormone levels are of considerable interest to both the fields of medicine and psychology. Early maturation in girls is found to be associated with internalising symptoms and disorders, and increased rates of alcohol, tobacco and substance use (Graber et al., 2004; Grumbach & Styne, 2003; Hayward & Sanborn, 2002). In boys, late maturation may be a risk factor for deviant behaviour and substance abuse (Graber et al., 2004). Furthermore, individual differences in circulating sex hormone levels are thought to play a role in behavioural problems. An extensive line of research suggests that high testosterone concentrations are linked to (problem) behaviour, particularly aggression (Archer, 99; Archer, 2006; Book et al., 200) and dominance (Archer, 2006; Mazur & Booth, 998). The causality of this relationship is unknown. High testosterone levels could increase the risk of behavioural problems, but behaviour itself can also influence circulating testosterone levels (Archer, 2006; Raine, 2002).