Cannabis smoking is highly prevalent among patients with psychotic disorders. Its use has been found to be related to clinical characteristics and the prognosis of the disorder. Recent evidence indicates a protective effect of cannabis on weight gain and related metabolic alterations. However, there are no previous studies on the long-term longitudinal effects of cannabis on first-episode drug-naïve patients, which would thereby avoid the confounding effects of chronicity and previous treatment exposure. We aimed to explore the effect of cannabis smoking on weight and lipid/glycaemic metabolic measures in a sample of first-episode non-affective psychosis patients.Anthropometric measurements and glycaemic and lipid parameters were obtained at baseline and three years after initiation of treatment. Patients self-reported their cannabis use at both time points. To explore the longitudinal effect of cannabis, patients were divided into three groups: continuers, discontinuers and non-users.Cannabis users at baseline presented a lower weight ( F=14.85, p<0.001), body mass index ( F=13.14, p<0.001), total cholesterol ( F=4.85, p=0.028) and low-density lipoprotein-cholesterol ( F=6.26, p=0.013) compared to non-users. These differences were also observed after three years: weight ( F=8.07, p=0.005), body mass index ( F=4.66, p=0.032) and low-density lipoprotein-cholesterol ( F=3.91, p=0.049). Moreover, those patients discontinuing cannabis use presented a higher increase in weight ( F=2.98, p=0.052), body mass index ( F=2.73, p=0.067) and triglyceride-high-density lipoprotein ratio ( F=2.72, p=0.067) than the 'non-users' and 'continuers'.The study suggests that cannabis use may produce a protective effect against weight gain and related metabolic alterations in psychosis. However, these results need to be replicated in a larger sample size.
