Weight change and cardio-metabolic outcomes in people with a serious mental illness in the UK from 1997 to 2020: a cohort study using the CPRD database

Date of Approval: 
2020-12-18 00:00:00
Lay Summary: 
Serious mental illnesses (SMIs) have a major and enduring effect on people’s lives. SMIs include schizophrenia, bipolar disorder, and psychotic depression. About 1 in 100 people develop an SMI and on average die 15 to 20 years earlier than the rest of the population. Heart disease is the leading cause of death for people with SMIs. Excess weight increases risk for developing heart disease, and people with SMIs are more likely to have excess weight compared with the general population. Studies show medications used to help manage the symptoms of SMIs may cause rapid weight gain. Other factors like diet, exercise and smoking may also influence weight. But these studies have usually studied small numbers of people for a short amount of time, and focused only on people with schizophrenia. There is little information about what happens to weight in the long-term or how important this is in explaining the risk of heart disease for everyone with an SMI. This study will examine the long-term weight change of people with SMIs and their risk for developing of heart disease. We also examine if excess weight is linked with referral for weight management services.
Technical Summary: 
Meta-analyses of observational studies suggest people with serious mental illnesses (SMIs) have a significant increased incidence of cardiovascular disease (CVD) and CVD-related mortality compared with the general population. Overweight and obesity cause hypertension, adverse lipid profiles, and type II diabetes, which are well-known risk factors for CVD. Evidence suggests that people with SMIs are 2 to 3 times more likely to have excess weight, perhaps due to medications and modifiable behaviours. However, long-term data is scarce. Furthermore, guidelines recommend people with SMIs are offered weight management services. However, no study has examined the proportion of people with both an SMI and excess weight who are offered and referred to these services by their general practitioner (GP). We will establish a cohort (Cohort 1) of patients from the Clinical Practice Research Datalink (CPRD) Aurum database with an SMI diagnosis between 1997 and present matched with four healthy controls on age, sex, GP practice and year by using the same index date. For objective 4, a second cohort (Cohort 2) will be selected from Cohort 1 to include people with obesity. Our objectives are to: 1. Examine the prospective association between SMIs and mean weight change; 2. Examine the prospective association between SMIs and biomarkers of CVD risk; 3. Examine the prospective association between SMIs and incidence of CVD and CVD fatalities; 4. Compare incidence of offer and actual referral to weight management services in people with both SMIs and obesity (BMI ≥30 kg/m2) with controls. To our knowledge, this cohort study will use a representative population sample to provide the first examination of the long-term association between SMIs and weight, CVD risk factors and CVD outcome. We aim to provide the evidence necessary to allow healthcare providers and policy makers to more effectively manage excess weight in this underserved group.
Health Outcomes to be Measured: 
Objectives 1, 2 and 4 are descriptive and we will run separate analyses for each outcome. Objective 3 is hypothesis testing of our primary outcome. We will interpret the data based on the associations seen for all outcomes by looking for consistency and interpretable patterns. The outcomes for each objective are as follows: Objective 1: Examine the prospective association between SMI status and mean weight change from index date* until censored**. Outcome(s): Absolute mean weight (kg), percentage weight change (kg) and body mass index (BMI) change. Objective 2: Examine the prospective association between SMI status and biomarkers of CVD risk from index date until censored. Outcome(s): Change in biomarkers for CVD risk: mean total-cholesterol, low-density lipoprotein (LDL), triglycerides, fasting plasma glucose (FPG), glycated haemoglobin (HbA1c,) and blood pressure (systolic and diastolic). Objective 3: Examine the prospective association between SMI status and incidence of CVD and CVD fatalities from index date until censored. Outcome: Total incident CVD: angina, coronary heart disease, congestive heart failure or cardiomyopathy, total stroke, and total fatal CVD events. Objective 4: Compare incidence of offer and actual referral to weight management services in people with SMIs and obesity (BMI ≥30 kg/m2) with controls without SMIs but with obesity from index date until censored. Outcome(s): Proportion (%) of exposed and unexposed who are recorded as offered and actually referred to weight management services. *Further details on index date are provided in section K **Further details on censored date are provided in section K
Application Number: 
20_000186
Collaborators: 

Paul Aveyard - Chief Investigator - University of Oxford
Charlotte Lee - Corresponding Applicant - University of Oxford
Carmen Piernas - Collaborator - University of Oxford
Clare Bankhead - Collaborator - University of Oxford
Felicity Waite - Collaborator - University of Oxford
Margaret Smith - Collaborator - University of Oxford
Min Gao - Collaborator - University of Oxford

Linkages: 
Practice Level Index of Multiple Deprivation