Altered metabolic parameters in association with antipsychotic medication use in diabetes: A population based case-control study

• This study confirms higher diabetes rates in the schizophrenia population up to age 70 but not beyond, suggesting a survivor effect.
• This study utilises a national linked diabetes dataset to assess associations between antipsychotic use and diabetes parameters/outcomes.
• This study is novel, demonstrating slight improved diabetes-related parameters/outcomes in patients prescribed antipsychotics vs controls.

AimsThis study assess differences in clinical variables in diabetes patients prescribed antipsychotic medication and determines relative schizophrenia prevalence in the diabetes population.MethodsThis population-based case-control study utilizing Scotland’s national diabetes registry (SCI-diabetes) and linked psychiatric hospital discharge data (SMR04) established diabetes phenotypes in a patient cohort prescribed long term antipsychotic medication (n = 2362) (cases). Cases were matched 1:10 to diabetes patients not prescribed antipsychotic medication (controls) for BMI, gender; diabetes type; birth year; diagnosis date; smoking status. Sub-groups with defined schizophrenia (n = 196) or bipolar disorder (n = 190) were further examined. Schizophrenia prevalence in the diabetes versus general population was compared.ResultsDuring follow up, antipsychotic prescription was associated with lower HbA1c (55.1 (95% CI 54.5–55.8) or 7.2 (95% CI 7.1–7.3)% vs 58.2 (58.0–58.4) mmol or 7.5 (95% CI 7.5–7.5)% p < 0.001) lower serum total cholesterol, 4.2 (4.1–4.2) vs 4.3 (4.2–4.3) mmol/l, p < 0.001, lower blood pressure (systolic 130 (130.17–131.29) vs 134 (134.3–134.7) mmHg, p < 0.001), higher prescription of oral hypoglycaemic medication (42% (40–45) vs 38% (37–39) p < 0.001), similar statin prescriptions (85% (81–89) vs 85% (84–86), p = 0.55), and lower retinopathy rates (28% (25.6–30.5) vs 32% (31.5–33.1), p < 0.001). HbA1c at diagnosis was similar (p = 0.27). Schizophrenia prevalence was higher in the diabetes versus general population with differences across age groups (Scottish population versus diabetic population rate of 522.2 (522.1–522.3) versus 717.4 (703.4–731.9) per 100,000).ConclusionsWe confirm higher diabetes rates in schizophrenia up to age 70, similar attendance rates and clinical measurements that are not worse in a large well-matched population-based Scottish sample prescribed antipsychotic medication versus matched general diabetes patients.