Abstract
INTRODUCTION: Sex differences in play behavior have been seen in rodents, primates and humans with literature showing some associations with maternal sex steroids. The PreSchool Activities Inventory (PSAI) is a validated tool to assess sexually-dimorphic play behavior. We assessed the associations between prenatal maternal estradiol (E2), free testosterone (fT) and total testosterone (T) and children’s PSAI scores at age 4.
METHODS: Data were collected in The Infant Development and the Environment Study (TIDES), a US-based multi-center pregnancy cohort. The analysis included 399 pregnant women with complete data on sex steroids, PSAI and covariates. Early pregnancy E2 and T were analyzed by liquid chromatography tandem mass spectrometry and fT by equilibrium dialysis using labeled testosterone. PSAI scores (masculine, feminine and composite) were calculated from questionnaires completed by mothers when child was approximately 4 years old. Covariates included: maternal age and education, child age at questionnaire completion, number of older brothers and sisters, infant race and parental attitudes towards sex-atypical play behavior. After demonstrating nonlinear associations between PSAI scores and sex steroids, we used multivariable sex-stratified quadratic regression to model these relationships.
RESULTS: The analysis included 192 boys and 207 girls (mean age: 4.5 years). Sex steroids in serum collected at 6-20 weeks’ gestation (mean: 11 weeks) were similar in mothers of boys and girls. In girls, E2 showed a quadratic relationship with the masculine score (beta (E2) = 0.003, p = 0.002, beta (E22) = -7.3E-7, p <0.001). Masculine scores were lower among girls whose mothers had E2 above the 75th percentile (2160 pg/ml). In boys, we observed a positive linear association of borderline significance (beta (fT) = 9.11, p = 0.06) and a significant negative association for the quadratic term (beta (fT2) = -7.44, p = 0.04). PSAI measures were not associated with E2 in mothers of boys, fT in mothers of girls, or T in either sex.
DISCUSSION: This study did not replicate previously reported associations between T and masculine scores in girls. However, we found that, while most maternal E2 levels were not associated with play behavior in either sex, levels above the 75th percentile were associated with reduced masculine scores in girls at age 4. Our data also suggest a somewhat reduced feminine scores in boys born to mothers with fT above the 75th percentile. These results, based on a small sample using a maternal assessment of play behavior, should be examined in a larger sample using an objective measure of play.
Validity of free testosterone calculation in pregnant women
