Objective: In earlier analyses of nonHispanic White women we found a stronger relation between abuse history and midpregnancy elevated depressive symptoms in women with the serotonin transporter (5-HTTLPR) S/S genotype. Here, we focus on African-American women (N=698). Our inquiry is motivated by racial differences in depression diagnosis/treatment, stressors, and frequency of major 5-HTTLPR alleles (S, LA, LG).
Materials and methods: Stressful life events (lifetime) and depressive symptoms (current) were ascertained at 15-27 weeks gestation. A Center for Epidemiological Studies Depression Score of more than or equal to 18 was considered 'elevated'. Life events were scored together and separated into six subconstructs. 5-HTTLPR genotypes were grouped as follows: (i) L and S alleles, (ii) S-LG equivalence ('triallelic to biallelic'), and (iii) LA/LA, all others, S/S ('high/intermediate/low'). Odds ratios (OR) for 'elevated' depressive symptoms-life events (total and subconstructs) relations were calculated for each genotype grouping.
Results: The prevalence of 'elevated' depressive symptoms did not vary by genotype. The relation between stressful life events and 'elevated' depressive symptoms was stronger in S/S compared with LA/LA genotype (interaction P=0.11). Of the six subconstructs, only abuse showed a statistically significant gene-environment interaction. The OR for the abuse-'elevated' depressive symptoms association was greater for S/S vs. LA/LA genotype (interaction P=0.03) and in the 'triallelic to biallelic' grouping (interaction P=0.04). In the 'high/intermediate/low' grouping, 'low' (S/S) had a higher OR (5.5) than both 'intermediate' and 'high' (ORs≤2.3) (interaction P=0.10).
Conclusions: These results show the importance of examining racial groups, specific stressful events, and different 5-HTTLPR genotype groupings when exploring gene-environment interactions in depression.