Adult patients with schizophrenia or schizoaffective disorder who had been treated with clozapine for 5 years or longer were eligible to participate in the study (index group). The end point of observation was death, loss to follow-up or 1 January 2010. The control group consisted of patients never treated with clozapine, matched on age, diagnosis and gender.

During the study period we identified 94 patients in each group. The mean follow-up in the index and control group was 12.3 and 13.5 years respectively. The vast majority of patients continued clozapine until the end of the follow-up period (mean duration of therapy: 10.7 years). The cumulative incidence of diabetes in the clozapine group was 22.3% compared to 16.0% in the control group (NS). The absolute risk difference was 6.3% (95%CI: −4.9% to 17.5%). The hazard ratio for time to new-onset diabetes was 1.54 (95%CI: 0.80–3.00) (see Fig 1 for Kaplan-Meier curve).

To the best of our knowledge, this study has the longest follow-up of a sizeable cohort of clozapine users in comparison to matched controls. The absolute risk difference was 6.3% and not significant despite an adequate power and extended follow-up of more than 13 years. Our result is in agreement with four randomized studies with a follow-up varying from 14 weeks up to 9 years. A case-control study matched 7,227 cases of newly treated diabetes by year of birth and gender to 6,780 controls and found no significant association between the incidence of new-onset diabetes and clozapine.

Fig. 1

[Kaplan-Meier curve: time to new-onset diabetes]