Researchers report that children later diagnosed with autism were exposed to higher levels of several steroid hormones in the womb, including testosterone, progesterone and cortisol. The study, conducted by teams at the University of Cambridge and the Statens Serum Institute in Copenhagen, suggests prenatal steroidogenic activity may contribute to why autism is diagnosed more often in males, but the authors emphasize the findings are not suitable for screening or clinical use.
The research team, led by Professor Simon Baron‑Cohen and Dr Michael Lombardo from Cambridge together with Professor Bent Nørgaard‑Pedersen in Denmark, analyzed amniotic fluid samples held in a Danish biobank. The bank contains roughly 19,500 samples collected from pregnancies between 1993 and 1999. Amniotic fluid, sampled around 15–16 weeks of gestation when amniocentesis is performed, offers insight into a critical window of fetal brain development and sexual differentiation. From this collection the investigators identified 128 male individuals whose amniotic samples corresponded to later clinical diagnoses on the autism spectrum, and they compared those samples with matched typically developing male controls using national psychiatric registry data.
The team measured concentrations of four key Δ4 sex steroid hormones—progesterone, 17α‑hydroxy‑progesterone, androstenedione and testosterone—alongside the stress hormone cortisol. All measured hormones were strongly correlated with one another. Statistical analyses, including principal component analysis, revealed a single latent steroidogenic factor that captured much of the variation in hormone concentrations. On average, the amniotic fluid from males who later received an autism spectrum diagnosis showed higher levels across this generalized steroidogenic factor than the control group. Importantly, this elevation was observed consistently across different diagnostic subgroups on the autism spectrum, including Asperger syndrome, classic autism and PDD‑NOS.
Professor Baron‑Cohen said the finding represents one of the earliest non‑genetic biomarkers linked to later autism diagnoses. Previous work had associated higher prenatal testosterone with particular developmental patterns—such as slower social and language progress and increased attention to detail—but this study shows a broader prenatal steroid signature associated with clinical autism. Because several of these steroid hormones are produced in greater quantities in males, elevated fetal steroidogenic activity is a plausible factor contributing to the higher prevalence of autism diagnoses among males.
The researchers stress several important caveats. The hormone elevations were observed on average across groups and there is substantial overlap between individuals; the study does not provide a hormone threshold that could reliably predict autism for any given fetus. The authors explicitly advise against interpreting these results as justification for hormonal treatments or for prenatal screening, noting potential risks and the likelihood that fetal steroid effects may occur during an early, foundational period of brain development and therefore resist later intervention.
Dr Michael Lombardo noted that the results point to a specific developmental window—around 15 weeks of gestation—when steroidogenic activity may influence brain development and interact with genetic and other biological factors that together shape autism risk. The investigators propose that steroid hormones can modulate how genetic instructions are translated into proteins (gene expression), and that altering this regulation during early neurodevelopment could help explain how genetic risk is expressed.
Within amniotic fluid the study measured four key Δ4 sex steroid hormones synthesized in a stepwise pathway: progesterone, 17α‑hydroxy‑progesterone, androstenedione and testosterone. In addition, cortisol was measured as a steroid hormone that lies outside this Δ4 pathway.
Contact: Simon Baron‑Cohen – University of Cambridge
Source: University of Cambridge press release
Image Source: Image credited to John Jay Glenn and adapted from the University of Cambridge press release.
Original Research: The study “Elevated fetal steroidogenic activity in autism” by S. Baron‑Cohen, B. Auyeung, B. Nørgaard‑Pedersen, D. M. Hougaard, M. W. Abdallah, L. Melgaard, A. S. Cohen, B. Chakrabarti, L. Ruta and M. V. Lombardo was published in Molecular Psychiatry (published online June 3, 2014; doi:10.1038/mp.2014.48).
Elevated fetal steroidogenic activity in autism
Autism is more commonly diagnosed in males than in females, prompting investigation into prenatal steroid hormones as early biological risk factors. Using the Danish Historic Birth Cohort and the Danish Psychiatric Central Register, the investigators identified amniotic fluid samples from males born 1993–1999 who later received ICD‑10 diagnoses of autism, Asperger syndrome or PDD‑NOS (n = 128) and compared them with matched typically developing controls. Concentrations of Δ4 sex steroids and cortisol were measured by liquid chromatography‑tandem mass spectrometry. The hormones were positively correlated and a single generalized steroidogenic factor explained much of the variation. The autism group showed a significant elevation across this factor (Cohen’s d = 0.37, P = 0.0009) uniformly across diagnostic labels. These results provide direct evidence of elevated fetal steroidogenic activity associated with autism and suggest a potential role for prenatal steroid signaling as an epigenetic programming mechanism that may interact with other pathophysiological processes in autism.