Although the correlations remained significant after controlling for behavioral risk factors known to be associated with IL-6, we explored the possibility that these variables may partially mediate the association between personality and IL-6. We simultaneously tested smoking, BMI, aspirin use, and disease burden as mediators, controlling for sex and age, using bootstrapping techniques (Preacher & Hayes, 2008). BMI and smoking both accounted for part of the association between N5: Impulsivity and IL-6: Impulsive participants had high circulating levels of IL-6, in part, because of their smoking habits and weight (point estimates = .0030 and .0011, 95% CIs = .0021-.0040 and .0006-.0017 for smoking and BMI, respectively, ps < .05). In addition, smoking mediated four facets of Conscientiousness (C1: Competence, C3: Dutifulness, C4: Achievement Striving, and C6: Deliberation; median point estimate = -.0005, 95% CI = -.001 to -.0001, p < .05) and BMI mediated C2: Order (point estimate = -.0013, 95% CI = -.002 to -.0006, p < .05), indicating participants low in Conscientiousness had higher IL-6, in part, because they smoked and were overweight (see Table 2).
