^{65}Allard et al. (1995), n = 500

In bivariate analysis, drug use negatively correlated with morale (r = 0.32, p < 0.001)

^{6} Blazer et al. (2000), n = 4,162

In multivariate logistic regression, benzodiazepine use significantly associated with depressive symptoms at 10year followup (OR: 2.13, p < 0.05)

^{68} Dealberto et al. (1997), n = 2,812

In multivariate logistic regression, drug use significantly associated with depressive symptoms at 6year followup (OR: 2.78, p < 0.001)

^{69} Gustafsson et al. (1996), n = 421

In LISREL model, drug use positively correlated (0.63) with poor mental health (chi^{2} = 3.38, p = 0.33)

^{94} Paterniti et al. (1998), n = 1,389

In multivariate logistic regression, drug users 4 times more likely to report anxious and depressive symptoms than nonusers (OR: 4.0, CI: 2.5–6.5)

^{63} Gleason et al. (1998), n = 5,181

In multivariate logistic regression, benzodiazepine users 7 times more likely to report emotional or nervous problems than nonusers (OR: 6.66, CI 5.1, 8.7)

^{74} Taylor et al. (1998), n = 5,222

Depressive symptoms multiply by 3 the likelihood of using hypnotics and by 5.1 the likelihood of using anxiolytics. Anxious symptoms multiply by 4.2 the likelihood of using hypnotics and by 3 of using anxiolytics.
