The results for the combined sample in the model including race, deprivation, valence, spray, the 5-HTTLPR genotype, and their interactions showed a significant main effect of race (F (1,80) = 5.49, p = 0.02) indicating a higher overall startle for Caucasian vs. African American smokers. Additionally, the main effect of valence was significant (F (2,80) = 10.01, p < 0.0001) and revealed the expected ordering of the startle responses to each of the stimuli: negative (M = 86.7 (6.6)) >neutral (M = 77.4 (6.2)) >positive (M = 71.5 (6.0)). A significant SLC6A4 genotype by Spray interaction was replicated in this analysis of the entire sample (F (2,80) = 4.17, p = 0.02; see Figure 1). No other significant interactions with race or SLC6A4 genotype and other factors (e.g., valence, deprivation, sex) were detected. Both the separate race models and the combined racial analysis were repeated in an analysis that included startle responses that occurred during inter-slide intervals, as no significant interactions with valence were observed. The results of these analyses were identical to those reported above and will not be expanded upon here.
