WE reported a population association between a tryptophan hydroxylase (TPH) intron 7 polymorphism with suicidal behavior in Finnish alcoholic offenders and with cerebral spinal fluid 5-hydroxyindoleacetic acid concentration in impulsive offenders.1 Abbar et al2 found no association between a different TPH polymorphism and suicidal behavior in a clinically and ethnically divergent sample. Our TPH association was presented as a preliminary result requiring replication. However, the study by Abbar et al2 has little relevance to the validity of the association we observed. They used a different, unpublished TPH marker to genotype both women and men who were undergoing psychiatric care and who had a dissimilar and mixed diagnostic profile and a different ethnic background. Therefore, the assertion by Abbar et al of a "discrepancy" between their and our results has no basis.
The Ava II restriction fragment length poIymorphism used by Abbar et al2 is distinct from