Although mental illness and smoking are known to be associated, the reasons why people with mental illness are more likely to smoke than the general population remain unclear. a study published July 10 in The Lancet shows that daily smoking is associated with an increased risk of mental illness and an earlier age of onset. The study tested four hypotheses, the first being that daily smoking was associated with an increased risk of mental illness in case-control and prospective studies; the second being that smoking was associated with an earlier onset of mental illness; the third being that the earlier one starts smoking, the higher the risk of mental illness; and the fourth assessing the prevalence of smoking at the time of first onset of mental illness in patients with mental illness. The study was conducted by searching Embase, Medline, and PsycINFO databases to screen observational studies in which patients with psychiatric disorders reported having a smoking rate compared with controls, to count the weighted mean difference between age at onset of psychiatric disorders and age at initiation of smoking, and to calculate a dominance ratio (OR) from cross-sectional studies and a risk ratio (RR) from prospective studies. Of the 3717 citations retrieved, a total of 61 studies including 72 samples met the inclusion criteria. The overall sample included 14555 smokers and 273162 nonsmokers. The proportion of patients with mental illness who smoked at first onset was 0.57 (95% CI 0.52-0.62, p<0.0001). For case-control studies, the overall OR of smokers versus non-smokers at first onset in patients with mental illness was 3.22 (95% CI 1.63-6.33), with some publication bias (Egger's test p=0.018, Begg's test p=0.007). For the prospective study, the overall RR for new-onset psychiatric disorders in daily smokers compared to non-smokers was 2.18 (95% CI 1.23-3.85). Daily smokers had an earlier age of onset of psychiatric disorders than non-smokers (weighted mean difference -1.04 years, 95% CI: -1.82 - -0.26). The age of initiation of smoking was not significantly earlier in patients with mental illness than in healthy controls (-0.44 years, 95% CI -1.21-0.34).