Ethnic variations in risk trajectories for type 2 diabetes: an observational cohort study

Study type
Protocol
Date of Approval
Study reference ID
23_003562
Lay Summary

The number of people living with diabetes is increasing worldwide. Most of the scientific evidence about mechanisms of diabetes development, early prevention and diagnosis was based on studies which included predominantly older and white populations. However, some ethnic minority groups, including people of South Asian and African Caribbean ethnicities, are at higher risk of developing diabetes at a younger age and within normal body weight ranges compared to people with a white ethnic background. This could lead to delayed diagnosis and diabetes-related complications, which means current evidence may not be generalisable to the diverse world population. Therefore, this study aims to describe the ethnic differences in trends of diabetes onset and understand the pathways to diabetes utilising representative multi-ethnic UK/England population from electronic health records data to elucidate the role of different risk factors (demographic, socio-economic, clinical and behavioural) and potential interactions among them. The outputs of this research will contribute a body of evidence that can inform diabetes prevention guidelines and reduce ethnic inequalities and the burden of diabetes in the UK.

Technical Summary

This observational cohort study aims to investigate the ethnic differences in risk trajectories to T2D using the CPRD Aurum database.

The study population will be all UK-registered individuals between 2004 and 2023 in CPRD Aurum whose records are eligible for linkage to HES and IMD data. The follow-up period will start from the latest date of the individual’s registration with a general practitioner and end at the earliest of T2D diagnosis, death, de-registration, or last data collection. The primary exposure will be ethnicity. Outcomes of interest will be T2D risk and the age of its diagnosis. The covariates of interest will include patients’ demographic, socio-economic, clinical and behavioural characteristics.

Firstly, we will describe trends in the age-standardised incidence of early-onset T2D and the median age of diabetes onset across ethnicities by demographic, socio-economic and clinical risk factors using joinpoint regression analysis to estimate annual percentage change. Secondly, we will model ethnic- and sex-stratified trajectories to T2D by estimating the direct and indirect effects of demographic, socioeconomic, clinical and behavioural risk factors to capture potential differences in the mechanism of T2D development and its age of onset utilising the mediation analysis method. Thirdly, we will test for interactions in the pathways between risk factors across ethnic groups by incorporating the conditional process analysis method in the mediation analysis.

The findings of this study can potentially benefit people of ethnic minority groups by helping to understand better the mechanisms of early onset T2D and by facilitating earlier identification of high-risk populations and targeting interventions. The research can potentially reduce the high economic burden of healthcare costs associated with managing T2D patients by decreasing its incidence and enhancing better allocation of healthcare resources.

Health Outcomes to be Measured

Primary outcomes:
• Age at diagnosis of T2D across ethnicities (Age of diabetes onset)
• T2D risk across ethnicities

Secondary outcomes
• Age-standardised incidence and prevalence of early/late-onset T2D and the median age of T2D onset across ethnicities by demographic, socio-economic and clinical risk factors

Collaborators

Rohini Mathur - Chief Investigator - Queen Mary University of London
Binur Orazumbekova - Corresponding Applicant - Queen Mary University of London
Elizabeth Remfry - Collaborator - Queen Mary University of London
Moneeza Siddiqui - Collaborator - Queen Mary University of London

Linkages

HES Admitted Patient Care;Patient Level Index of Multiple Deprivation Domains;Practice Level Index of Multiple Deprivation Domains