OBJECTIVE:

A link between serum testosterone and aggressive behavior, which has been demonstrated in numerous animal studies and suggested in several studies of adult men, has never been investigated in children before the time of puberty.

METHOD:

We measured serum testosterone, sex hormone binding globulin (SHBG), dehydroepiandrosterone (DHEA), and dehydroepiandrosterone sulfate (DHEAS) in 18 highly aggressive prepubertal boys, ages 4 to 10, hospitalized for violent or unmanageable behavior at a state children's psychiatric facility in New York City (the Bronx). We compared them with a group of age and race matched controls from the same demographic area, screened negative for aggressive behavior problems. All the aggressive subjects met DSM-III-R criteria for conduct disorder and scored higher than the 98th percentile on the aggression subscale of the Child Behavior Checklist (mean T = 80 for the group).

RESULTS:

There were no significant differences between aggressive and nonaggressive children for T, SHBG, DHEA, DHEAS, or ratios of combinations of these variables.

CONCLUSIONS:

These findings raise questions about inferences from adult studies that testosterone may play a causal role in the development of human aggression. Testosterone does not appear to be a useful biological marker for aggressivity in early childhood.