To the Editor: Dr. Brewerton raises an interesting point regarding a potential confound of seasonal variation in serotonin function and aggressive behavior. He suggests that our findings of lower CSF 5-HIAA concentrations in aggressive individuals may be accounted for by seasonal fluctuation. This seems unlikely. Our measure of aggression was a lifetime history of aggressive behavior, not a single incident that occurred at a specific time of the year. It was an accumulation of behaviors across seasons and over many years. Thus, our aggression measure was not seasonal.

Furthermore, although there is evidence of seasonality in aggression, the findings are not straightforward. Murder does not show a seasonal change (Michael and Zumpe, 1983); the rates of occurrence for rape (Michael and Zumpe, 1983) and the battering of women (Michael and Zumpe, 1986) are higher in the summer. When all forms of suicide are considered, seasonal variation is not always found (1). However, violent suicides are at their peak in the spring R1588BCFIEFDI.

According to Brewerton et al. (1988), CSF 5-HIAA concentrations are at their highest during the summer. When this finding is coupled with the increased rates of battering and rape found during the summer, the results are contrary to the inverse correlation we, as well as others R1588BCFDJFAC, R1588BCFBFEGI, have found between CSF 5-HIAA concentrations and aggression. Dr. Brewerton’s point would be more convincing if CSF 5-HIAA levels were lower, rather than higher, in the summer. Nevertheless, Dr. Brewerton’s point is well taken, and it seems worthwhile to remember to consider the effect of seasonality on both biological and behavioral measures.