Cognitive abilities, androgen levels, and body mass index in 5-year-old children
Introduction
There is currently a certain degree of consensus regarding the existence of a relationship between androgen levels and some cognitive abilities, although this relationship has hardly been studied at all in children. A number of studies, mainly carried out with adults, have analyzed the relationship between testosterone levels and some cognitive abilities for which gender differences have been established. Such abilities include, for example, certain spatial abilities, in which men tend to perform better than women. Some of these studies suggest a linear relationship between testosterone levels and visuospatial abilities (Christensen, 1993, Christensen and Knussman, 1987), while others suggest a non-linear relationship in which men with lower testosterone levels perform better in spatial tasks than men with higher testosterone levels, and women with higher testosterone levels perform better than women with lower testosterone levels (Gouchie and Kimura, 1991, Shute et al., 1983). Finally, other researchers failed to find any association between androgens and the cognitive function (McKeever et al., 1987). On the other hand, diverse studies have found an inverse association between BMI and different cognitive abilities (Halkjaer et al., 2003, Hirshman et al., 2004, Teasdale et al., 1992). Teasdale et al. (1992) found a negative relationship between intelligence and BMI, observing that intelligence scores reached their maximum value below the median for the body mass index and declined monotonically thereafter in young men. Similarly, in a sample of Chinese boys with a mean age of 9, Li (1995) found that boys with severe obesity had significantly lower IQ scores than controls.
A number of studies have also analyzed the relationship between androgens and cognitive abilities for which gender differences have not been identified, such as fluid intelligence for example. This type of intelligence has been related to adaptability and flexibility in response to the need to solve problems in unexpected situations and is commonly assessed by means of tasks involving matrices, series, and classifications, etc. Tan and Tan (1998) found that very low or very high levels of testosterone may have a negative effect on fluid intelligence in women, while in men, very low or medium levels of testosterone may have a positive effect, and very high testosterone levels a negative effect on this type of intelligence.
Another interesting line of research in the relationship between hormones and cognition is that connected to studies involving older citizens (the majority receiving exogenous hormone administration). A number of beneficial effects of the administration of replacement testosterone have been observed with regard to spatial cognition (Janowsky et al., 1994) and the visual working memory (Janowsky et al., 2000). For their part, in a study involving older women and men, Wolf and Kirschbaum (2002) found that women with more endogenous testosterone had a better verbal memory, while in men, a negative correlation was found between testosterone and verbal fluidity. Some studies have also analyzed the effects of the administration of DHEA on cognition, although the results are inconsistent (Wolf et al., 1997, Wolf et al., 1998). Certain authors have suggested that the influence of DHEA on some cognitive abilities may be mediated by estrogens and testosterone (Hirshman et al., 2004).
The relationship between androgens and cognitive abilities in prepubescent children has not been widely researched, given that most studies have concentrated on young people and adults. When analyzing the relationship between perinatal testosterone levels and performance in spatial tasks at the age of 6, Jacklin et al. (1988) found that girls with higher levels of perinatal testosterone performed more poorly in the said tasks, while no relationship was observed in the case of boys. Finegan et al. (1992) found a negative relationship in girls between prenatal testosterone levels and performance in a spatial visualization test at the age of 4, while no relationship was observed for boys. In another study, Grimshaw et al. (1995) found a positive correlation for girls between prenatal testosterone levels and mental rotation rate at the age of 7. The results for boys were less clear, although they also suggested a positive relationship.
Traditionally, the early effects of hormones have been thought to be organizational in nature, since it is believed that they influence neural development and as a consequence of this, other physiological and behavioral processes also. Later effects, both activational and/or continuous, act on already existing neural systems. Androgen levels undergo a number of changes during human development (Collaer and Hines, 1995, Cortés-Blanco et al., 2000, Forest, 1989). For example, according to Forest (1989), at 10–12 weeks of gestation, testosterone levels are higher in male fetuses than in female ones. From week 28 onwards, no gender differences have been found at all in relation to the levels of this hormone. During this phase, the temporary pattern for androstenedione is similar to that for testosterone. During the postnatal period, at the beginning of the second week, testosterone levels increase and reach their highest point some time around 20–60 days after birth, when the level is comparable to that found in adults. Afterwards, testosterone levels decrease until the seventh month, at which point all gender differences in this respect disappear again until puberty. During childhood, up to the age of 7, sex steroids in plasma are as low as they will ever be during an individual's life. From this age onwards, until puberty, there is a gradual increase, mainly in DHEA and DHEA-S. In boys, testosterone levels increase 20–30 fold between the ages of 12 and 18 (this increase depends more on pubertal development than on chronological age). In girls, testosterone levels also increase during puberty, although at the end of this period boys have considerably higher levels than girls. In adulthood, testosterone levels are approximately 15 times greater in men than in women, and although there is a certain amount of controversy in relation to this area, a gradual decrease in levels has been observed from the age of 60–70 onwards. In menopausal women, testosterone levels are slightly lower than during the reproductive age.
One interesting aspect of cognition in children, which is particularly relevant from the perspective of social adaptation, is that related to theory of mind abilities. Between the ages of 3 and 5, children acquire a theory of mind, or in other words, the ability to understand other people's mental states, including beliefs, emotions, and desires, etc. (Astington, 1993, Wellman, 1990). Certain authors have found gender differences in relation to some of these abilities, such as in the case of an understanding of emotion, with girls generally performing better in this area than boys (Brown and Dunn, 1996, Dunn et al., 1991), although it is not clear at what precise moment these differences are established. Geary (1999) suggests the possible existence of endocrine correlates for these differences, although no experimental work has been carried out in this field. Nevertheless, other researchers have failed to find any gender differences whatsoever regarding the performance of diverse theory of mind tests (Roazzi and Santana, 1999, Simpson, 2003).
The study presented in this paper forms part of a wider project which aims to explore, from a biosocial perspective, the influence of diverse factors (family, cognitive, endocrine, etc.) on young children's social adjustment to their social environment (peers). The study aims to explore the relationship between diverse cognitive abilities related to crystallized intelligence, fluid intelligence, IQ composite, and theory of mind and testosterone, DHEA, androstenedione, and BMI measurements in children.
Section snippets
Subjects
The subjects were 129 preschool children (60 boys and 69 girls) from eight classrooms in three public schools in San Sebastian, Urnieta, and Puerto Real (Spain). The mean age of subjects in the sample was 5 years 5 months for boys and 5 years 4 months for girls, with the same range of 5 years 0 months to 5 years 11 months for both sexes. The socioeconomic status of subjects in the sample was medium and medium–high. The children's parents had been fully informed of the study and had given their
Gender differences in cognitive abilities, hormonal measurements, and BMI
No significant gender differences were found for any of the cognitive abilities studied. In the case of hormones, gender differences were only found for DHEA, with girls having higher DHEA levels than boys (F = 6.359; P < 0.05). No gender differences were found for BMI.
Correlations between hormone measurements, BMI, and cognitive performance
As a first approximation, all correlations between defined cognitive variables and hormone and BMI measurements were based on combined data for both boys and girls. In this way, significant negative correlations were found
Gender differences
The study failed to find any gender differences in any of the cognitive abilities assessed. In the case of fluid intelligence (Matrices subtest), our results coincide with those found by Cattell (1987), Tan and Tan (1998) and Tan et al. (2003), who also failed to find any gender differences for this type of intelligence.
Our results also failed to show any significant differences in relation to crystallized intelligence (Vocabulary subtest), a result consistent with that obtained by Kaufman and
Acknowledgments
This study was supported by the Basque Country Government Grants (PI99/63; BFI02.118-AE) and the Dirección General de Investigación Científica y Técnica of Spain (Proyect BSO2002-00134). We would like to express our gratitude to the teaching staff, parents, and children of the Egape, Intxaurrondo Hegoa, and El Trocadero schools, to Mita Banerjee for providing us with certain theory of mind tests, and to Mercedes Lopez and Iñaki Verdier for helping us to collect the data.
References (76)
- et al.
The relationship between testosterone levels and cognitive ability patterns
Psychoneuroendocrinology
(1991) - et al.
Salivary testosterone determination in studies of child health and development
Horm. Behav.
(1999) - et al.
Assessing dehydroepiandrosterone in saliva: a simple radioimmunoassay for use in studies of children, adolescents and adults
Psychoneuroendocrinology
(1999) - et al.
Mental rotation at 7 years: relations with prenatal testosterone levels and spatial play experiences
Brain Cogn.
(1995) - et al.
A signal-detection analysis of sex differences in the perception of emotional faces
Brain Cogn.
(2004) - et al.
Evidence than androgenic and estrogenic metabolites contribute to the effects of dehydroepiandrosterone on cognition in postmenopausal women
Horm. Behav.
(2004) - et al.
A curvilinear relationship between testosterone and spatial cognition in humans: possible influence of hand preference
Psychoneuroendocrinology
(1996) - et al.
Estradiol and testosterone effects on lipids in black and white boys aged 10 to 15 years
Metabolism
(2000) Gender differences in structure, means and variances of hierarchically ordered ability dimensions
Learn. Instr.
(1995)- et al.
Salivary testosterone and cortisol in disruptive children: relationship to aggressive, hyperactive, and internalizing behaviors
J. Am. Acad. Child Adolesc. Psych.
(1994)