05). A chi-square test indicated that at baseline, the groups differed significantly in click here gender (χ2(3) = 7.9; p < .05), education (χ2(6) = 63.0; p < .001), and physical activity (χ2(6) = 30.7; p < .001) with small to medium effect sizes ( Table 1). The groups also differed significantly in the prevalence of current diagnoses of depression (χ2(3) = 14.6; p < .01), generalized anxiety disorder (χ2(3) = 29.9; p < .001), and panic with agoraphobia (χ2(3) = 25.2; p < .001). However, no significant group differences were found (ps > .05) in the current diagnoses of anxiety, social anxiety, agoraphobia,

and panic without agoraphobia ( Table 1). 1 A multivariate ANOVA indicated a significant difference among groups on a linear combination of the dependent variables (F(12,5076) = 7.45; p < .001; see more Pillai’s trace = 0.05; partial η2 = 0.02). All four dependent variables reached statistical significance: severity of depression (F(3,1693) = 18.4; p < .001; partial η2 = 0.03); anxiety (F(3,1693) = 20.9; p < .001; partial η2 = 0.04); social anxiety (F(3,1693) = 4.2; p < .01; partial η2 = 0.01); agoraphobia (F(3,1693) = 13.2; p < .001; partial η2 = 0.02). Tukey HSD revealed that on three of the dependent variables (severity of depression,

anxiety and agoraphobia) nicotine-dependent smokers had higher scores than non-dependent smokers, former smokers and never-smokers (ps < .001). The latter three groups were not different from each other on these variables (ps > .05). For the severity of social anxiety, results were slightly different. Nicotine-dependent smokers were more socially anxious than former smokers (p < .05) and non-dependent smokers, but they were not different from never-smokers (p > .05). The mean scores are presented in Table 2. We also repeated similar analyses by combining the two groups of current smokers and found that current smokers had significantly more severe

depressive and anxiety symptoms than former and never-smokers (p < .001), except for social anxiety symptoms. 1 Finally, four regression analyses were run. In the regression analysis with symptoms of depression as the dependent variable, the overall variance explained was 8.4% (p < .001). The regression analysis with symptoms also of anxiety as the dependent variable explained 8% of the significant overall variance (p < .001). Similarly, for the symptoms of social anxiety and agoraphobia, the overall variance explained was 2.3% (p < .05) and 7.4% (p < .001), respectively. For individual contribution of each variable in predicting symptom severity, see Table 3. 2 We carried out similar regression analyses by including baseline FTND score as continuous covariate. A significant positive linear relationship between FTND and severity of symptoms on all four measures were found, thus confirming our initially reported analyses (Table 3S).