To reevaluate platelet serotonin (5-HT) levels in autism, measuring and controlling for effects of race and puberty. The specificity of hyperserotonemia for autism versus cognitive impairment is also assessed.

Platelet 5-HT levels were measured in 77 individuals, aged 2 through 37 years, with autistic disorder; 65 normal controls; and 22 mentally retarded or otherwise cognitively impaired (MR/CI) prepubertal children. Effects of diagnosis, race, and pubertal status were evaluated by analysis of variance in separate pre- and postpubertal groups. 5-HT levels were expressed as ng/mL blood and ng/μL platelet volume.

Among prepubertal children, significant effects of diagnosis (ng/mL; F_2,109 = 5.9, p = .004) and race (F_2,109 = 14.7, p < .0005) were found. Autistic youngsters had significantly higher 5-HT concentrations than controls, although the elevation (25%) was less than typically reported; MR/CI children had levels very similar to those of controls. White children had significantly lower 5-HT levels than black or Latino youngsters, regardless of diagnosis. Diagnosis and race effects were nonsignificant in the postpubertal group. Postpubertal subjects had lower 5-HT concentrations than prepubertal subjects (ng/mL; F_1,114 = 28.5, p < .0005).

The data underscore the importance of matching for race and pubertal status in neuropsychiatric research and suggest that the prevalence of hyperserotonemia in autistic individuals may have been overestimated because of a failure to control for both variables. Hyperserotonemia was not found in MR/CI youngsters without autistic features. J. Am. Acad. Child Adolesc. Psychiatry, 1998, 37(7):767–776.