Weight trajectories and access to weight management services in individuals with severe mental illness in the UK: a population-based, matched cohort study

Background: Excess weight is common in people with severe mental illness, including schizophrenia spectrum disorders, bipolar disorder, and other non-organic psychotic disorders. Rapid weight gain often follows use of antipsychotics, but long-term weight trajectories are unclear. We aimed to compare 15-year weight trajectories and assess the incidence of weight management advice and referrals among people with and without severe mental illness. Methods: In this retrospective, matched cohort study, we used the Clinical Practice Research Datalink Aurum to identify people aged 18–65 years registered at 1454 primary care practices in England (UK) between Jan 1, 1998, and Oct 31, 2020. 12 people with lived experience of severe mental illness informed the research objectives and study design. Registered individuals who had been diagnosed with severe mental illness (ICD-10 F20–39) were eligible. Each person with severe mental illness was frequency-matched with up to four people without severe mental illness on age, sex, practice, and calendar year. The first coprimary outcome was change in weight (kg), as per all recorded weight measures in the electronic health record. The second coprimary outcome was the incidence of advice and referral to weight management services. We estimated weight trajectories using hierarchical mixed-effects linear regression models and the incidence of advice and referrals using zero-inflated Poisson regression models. Findings: We included 113 904 individuals (mean age 39·17 years [SD 12·38]), of whom 51 062 (44·8%) were male and 62 842 (55·2%) were female; 90 620 (79·6%) self-identified as White, 7430 (6·5%) as Black, 12 288 (10·8%) as Asian, 1983 (1·7%) as mixed, and 1583 (1·4%) as other. 90 879 (79·8%) individuals did not have severe mental illness; 23 025 (20·2%) individuals had received a diagnosis of severe mental illness: 11 039 (47·9%) had a schizophrenia spectrum disorder, 11 942 (51·9%) had bipolar disorder, and 44 (0·2%) had other psychoses. Weight in people with severe mental illness increased by 2·10 kg (95% CI 1·98–2·22; p<0·0001) at year 1 and by 5·55 kg (5·24–5·86; p<0·0001) at year 15, compared with 0·58 kg (0·51–0·65; p<0·0001) at year 1 and 1·62 kg (1·42–1·82; p<0·0001) at year 15 in people without severe mental illness. After adjusting for age, sex, race and ethnicity, socioeconomic status, alcohol consumption status, smoking status, and BMI, people with severe mental illness and a BMI of 25 kg/m2 or higher were 10% more likely to receive weight advice than people without severe mental illness (incidence rate ratio 1·10 [95% CI 1·07–1·13]; 8·51 × 10−10). There were no differences in the rates of referral to weight management programmes between people with and without severe mental illness. Interpretation: People with severe mental illness are more likely to gain weight rapidly after diagnosis than the general population, with effects lasting up to 15 years. Despite frequent advice to lose weight, weight gain is not matched with rates of referrals to services for weight management. Early intervention is crucial to reduce excess weight and associated cardiometabolic risks in this underserved patient group. Funding: None.