Plasma androgens in autism.
Blood androgen levels do not differ between autistic and typical boys, so skip hormone tests and keep assessing behavior.
01Research in Context
What this study did
Doctors drew blood from 17 boys with autism and 17 boys without it.
They checked how much testosterone and DHEA-S each boy had.
The boys were 6-17 years old and lived near the hospital clinic.
What they found
Both groups had the same hormone levels.
High or low androgens did not line up with an autism label.
The team said hormone testing will not help spot autism.
How this fits with other research
Lyall et al. (2011) looked at moms, not kids.
They found that women who got their first period before age 10 or had high teen BMI were a bit more likely to later have a child with autism.
That study keeps the hormone idea alive, but it shifts the focus from the child’s blood to the mother’s early-life hormones.
Gandhi et al. (2022) and Erickson et al. (2016) also compared autistic and typical kids, yet they measured school executive functions and brain scans instead of blood.
All three papers, including Lalli et al. (1995), agree on one point: simple body or brain numbers rarely separate the groups cleanly.
Together they warn us not to chase single-cause biomarkers.
Why it matters
You can skip hormone labs when assessing autism.
Time is better spent on language, play, and social probes.
If parents ask about “too much testosterone,” show them this data and move the plan back to skill-based goals.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Drop any hormone-talk from intake spiel; spend that minute on a quick preference assessment instead.
02At a glance
03Original abstract
Plasma levels of testosterone and the adrenal androgen dehydroepiandrosterone sulfate (DHEA-S) were measured in male autistic subjects (31 prepubertal, 8 postpubertal), mentally retarded/cognitively impaired subjects (MR, 12 prepubertal), and normal control subjects (NC, 10 prepubertal, 11 postpubertal). Mean levels of plasma testosterone were similar in the postpubertal autistic (4.54 +/- 1.12 ng/ml) and postpubertal NC (5.02 +/- 1.87 ng/ml) groups. Plasma DHEA-S levels in postpubertal autistic (2170 +/- 1020 ng/ml) and postpubertal NC (1850 +/- 777 ng/ml) groups also were not significantly different. Similarly, no significant group differences were seen for testosterone or DHEA-S in the prepubertal autistic, MR, or NC individuals, although prepubertal MR individuals with cerebral palsy did have increased plasma DHEA-S levels compared to age-matched MR or NC individuals. Significant negative correlations were found between testosterone and whole blood serotonin (5-HT) levels in the combined (all subjects, all ages) groups and in the autistic group, suggesting that the effect of puberty on whole blood 5-HT may deserve further study. Data indicate that altered secretion of the androgens is not a common feature of autism. However, abnormalities of adrenal androgen secretion may be present in individuals with cerebral palsy.
Journal of autism and developmental disorders, 1995 · doi:10.1007/BF02179290