A partial correlation was executed between 5HTTLPR risk status (one or two copies of the s allele versus two copies of the l allele) and the risk behavior index at long-term follow-up to determine whether genetic risk status was associated with increases in risk behavior across the 29 months from pretest to long-term follow-up. The pretest risk behavior index scores and group assignment were controlled when genetic risk status was correlated with the risk behavior index at long-term follow-up. Controlling for pretest levels of risk behaviors allows an analysis to determine whether risk status is associated with increases over time in risk behaviors. Genetic risk status was correlated with increases in youth risk behavior across the 29 months from pretest to long-term follow-up, r(419) = .11, p < .03.
