Amniotic testosterone and psychological sex differences: A systematic review of the extreme male brain theory

Abstract

Background

Baron-Cohen (2002) proposed the Extreme Male Brain Theory (EMB) to suggest that foetal testosterone (FT) (1) is a component of the complex neurobiological aetiology of Autism Spectrum Disorder (ASD) and (2) accounts for its high male prevalence. The theory suggests that ASD is more common in males to an extreme manifestation of psychological maleness due to heightened testosterone exposure in the foetus.

Aim

To assess the EMB theory by reviewing cohort studies that directly assayed FT levels at 12–24 weeks of gestation in relation to subsequent ASD symptoms, ASD-related cognitions, social outcomes and playstyles prior to adolescence.

Method

A systematic term to subject heading search was conducted on Web of Science, Embase, PsycINFO, ‘Ovid Medline Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE Daily and Ovid Medline’, PsycARTICLES Full Text, and ProQuest up to December 2019. Studies that included the extraction of foetal fluid and children of both sexes were assessed in compliance with STROBE guidelines. Additional articles were obtained by reference list screening.

Results

22 FT-assay studies ($N=2284$) containing EMB-associated traits as dependent variables were identified, including ASD symptoms, ASD-related cognition, sociality and playstyles. Their STROBE ratings ranged from 50% to 86.4%. FT significantly accounted for ASD-related traits beyond the child’s sex in 3 of 4 studies. 4 out of 9 papers looking at sexed ASD-related cognitive-styles and 2 of 3 examining social outcomes showed significant FT effect. 2 of 6 found that FT accounted for significant variance in behavioral indices that differ on average between the sexes. Chi-square tests $({\chi }^2\left(2,N=22\right)=4.46,P<.05$) demonstrated that researchers affiliated with Baron-Cohen are significantly more likely to generate results fully supportive of EMB, with 25% ($N=3,P<.05$) of positive findings produced by independent authors. Homogeneity of data did not account for this.

Conclusion

The certainty with which FT was established as an agent in sexual differentiation varies by the psychological variable in question, but none of the conclusions were supported by an adequate number of studies. Nevertheless, this review yields the following preliminary conclusions, which can be tested in future research. FT plays a plausible role in driving social and non-social ASD-related cognition as well as ASD symptoms across the sexes. FT accounts for gender differences on eye contact frequency and value-laden proposition use and mediates the narrowing of interest toward systems and exerts sex-specific effects on numerical and language abilities, though these studies require independent replication. The role of FT on the differentiation of play is consistently non-significant. Where an effect exists, it is largely dwarfed by the effect of sex and hence it is equivocal that second trimester FT affects play. Biological implications for sex differences are considered and more lifespan longitudinal amniocentesis studies are suggested to pursue greater clarity in the empirical bases of EMB.