Background: Metformin is a pharmacological candidate to mitigate second-generation antipsychotic (SGA)-induced weight gain in patients with severe mental illnesses (SMI). Objective: To evaluate the incidence, prevalence, and demographic patterns of metformin co-prescription among patients diagnosed with SMI initiating SGAs. To estimate the impact of co-prescription on weight. Methods: A cohort study of patients diagnosed with SMI initiating aripiprazole, olanzapine, quetiapine, or risperidone in 2005-2019 using primary care data from Clinical Practice Research Datalink. We estimated cumulative incidence and period prevalences of co-prescription and explored prescribing differences by demographic and clinical factors. We compared weight change among patients prescribed an SGA only versus an SGA plus metformin, accounting for confounders using linear regression. Findings: Among 26,537 patients initiating SGAs, 4652 were ever prescribed metformin and 21,885 were not. Two-year incidence of first metformin prescription was 3.3%. The SGA plus metformin group were more ethnically diverse, had greater social deprivation, more comorbidities, and higher baseline weight (mean 90.4 vs. 76.8 kg). By two years post-SGA initiation, mean weight in the SGA-only group had changed by +4.16% (95% CI, -1.26 to +9.58) compared to -0.65% (95% CI, -4.26 to +2.96) in the SGA plus metformin group. After confounder adjustment, the two-year mean difference in weight with metformin co-prescription was -1.48 kg (95% CI, -4.03 to 1.07) among females and -1.84 kg (95% CI, -4.67 to 0.98) among males. Conclusion: Metformin is infrequently co-prescribed, despite established efficacy and guidelines. Clinical implications: Primary and secondary care collaboration should be strengthened and barriers to co-prescribing addressed.
