Temporal Trends in Mortality Associated with Comorbid Type 2 Diabetes and Schizophrenia: The Fremantle Diabetes Study

Abstract

Background: In Phase I of the community-based Fremantle Diabetes Study (FDS1), there was evidence of a deleterious interactive effect of schizophrenia and type 2 diabetes on mortality. Our aim was to investigate whether the mortality gap had improved in FDS Phase II (FDS2) conducted 15 years later. Methods: Participants with type 2 diabetes from FDS1 (n = 1291 recruited 1993–1996) and FDS2 (n = 1509 recruited 2008–2011) were age-, sex- and postcode-matched 1:4 to people without diabetes. Schizophrenia at entry and incident deaths were ascertained from validated administrative data. Results: Schizophrenia affected 50/11,195 (0.45%) of participants without diabetes and 17/2800 (0.61%) of those with type 2 diabetes (p = 0.284). During 142,304 person-years of follow-up, the mortality rate (95% CI) was lowest for the FDS2 subgroup without diabetes/schizophrenia (18.2 (16.9, 19.6)/1000 person-years) and highest in FDS2 and FDS1 subgroups with type 2 diabetes/schizophrenia (53.3 (14.5, 136.6) and 98.0 (31.8, 228.8)/1000 person-years, respectively). Compared to the respective FDS subgroup without diabetes/schizophrenia, the mortality rate ratio was approximately 50% higher in the type 2 diabetes subgroup, and three times higher in those with type 2 diabetes/schizophrenia. In Cox regression, unadjusted hazard ratios were highest in those with type 2 diabetes/schizophrenia in FDS1 (HR (95% CI): 3.71 (1.54, 8.93) and FDS2 (2.96 (1.11, 7.91)), increasing to 5.61 (2.33, 13.5) and 26.9 (9.94, 72.6), respectively, after adjustment for age. Conclusions: Although limited by small numbers of schizophrenia cases, these data suggest that comorbid type 2 diabetes and schizophrenia remains associated with a substantial and possibly increasing mortality gap.