Project overview
Type 2 diabetes affects 4 million people in the UK and is characterised as a lifelong progressive disease.
Our recent work has shown that biochemical remission or temporary cure is achievable in a population-based primary care sample. Remission is defined as a state in which glycaemic levels are below the diagnostic threshold (HbA1c < 6.5% or 48 mmol/mol) in the absence of pharmacological interventions or bariatric surgery. The frequency of remission in the UK population is unclear as individuals will flux between states of remission and relapse over the course of diabetes. Analysis of longitudinal data from multiple time points is lacking and there are no studies identifying patterns of remission over time or their trajectories towards long-term health outcomes.
In this study, we will describe the prevalence and incidence of remission in a large UK primary care population of adults diagnosed with type 2 diabetes, develop clusters based on longitudinal patterns of remission and quantify the association between remission and long-term health outcomes over extended follow-up.
For our methods, we will use a cohort study sampling 108 000 people with type 2 diabetes within the CHIA database of routinely collected GP records over seven years. Patterns of remission will be summarised using descriptive statistics. Longitudinal latent class modelling will be used to generate clusters representing distinct patterns of remission. We will describe participant characteristics by remission cluster. Regression models will then be constructed to quantify the association between remission clusters and long-term outcomes including myocardial infarct, stroke, amputation and all-cause mortality. Models will be adjusted for confounding on a priori reasoning.
Primary care in the UK is an important context for the delivery of diabetes care to prevent complications of the disease. Understanding patterns of remission over the course of the disease and their impact on long-term health could provide motivation for patients, and also inform future targeted and personalised interventions for managing diabetes in primary care.
Our recent work has shown that biochemical remission or temporary cure is achievable in a population-based primary care sample. Remission is defined as a state in which glycaemic levels are below the diagnostic threshold (HbA1c < 6.5% or 48 mmol/mol) in the absence of pharmacological interventions or bariatric surgery. The frequency of remission in the UK population is unclear as individuals will flux between states of remission and relapse over the course of diabetes. Analysis of longitudinal data from multiple time points is lacking and there are no studies identifying patterns of remission over time or their trajectories towards long-term health outcomes.
In this study, we will describe the prevalence and incidence of remission in a large UK primary care population of adults diagnosed with type 2 diabetes, develop clusters based on longitudinal patterns of remission and quantify the association between remission and long-term health outcomes over extended follow-up.
For our methods, we will use a cohort study sampling 108 000 people with type 2 diabetes within the CHIA database of routinely collected GP records over seven years. Patterns of remission will be summarised using descriptive statistics. Longitudinal latent class modelling will be used to generate clusters representing distinct patterns of remission. We will describe participant characteristics by remission cluster. Regression models will then be constructed to quantify the association between remission clusters and long-term outcomes including myocardial infarct, stroke, amputation and all-cause mortality. Models will be adjusted for confounding on a priori reasoning.
Primary care in the UK is an important context for the delivery of diabetes care to prevent complications of the disease. Understanding patterns of remission over the course of the disease and their impact on long-term health could provide motivation for patients, and also inform future targeted and personalised interventions for managing diabetes in primary care.
Staff
Lead researcher
Other researchers
