Obesity and health: what can primary care data tell us?

Supervisor

• Professor Tom Yates

Project description

Obesity, measured by body mass index (BMI), is a fundamental determinant of health and a leading cause of morbidity and premature mortality globally. The importance of obesity to health has been further unmasked by the COVID-19 pandemic, where it has emerged as one of the most consistent and strongest predictors of disease severity and survival. Understanding the impact of obesity on health has largely been driven by relatively small national surveys or research cohorts. However, recent developments with data governance and linkage have provided a step-change in the type and amount of data available. Specifically, linkage between primary and secondary care clinical records with mortality records and other data assets, such as UK Census data, enable common clinical risk factors to be examined on a population level. Early examination of this data by the supervision team reported marked ethnic differences in the association of primary care coded BMI with health outcomes [1], suggesting BMI may act as an important determinant of health inequality within ethnic minority groups. 

The aim of this PhD is to undertake a comprehensive programme of research examining the pattern of routine measures of BMI within primary care in England over time and associations with survival and health outcomes. 

The overarching aim of the PhD will be achieved by the following objectives:

1. Investigate the pattern of missingness of BMI in primary care records and develop optimal methods for imputation.
2. Define the pattern and stability of BMI over time and across age, sex, deprivation and ethnicity
3. Undertake a series of survival models investigating the association between BMI and different morbidity/mortality outcomes, exploring interactions by age, sex, ethnicity and deprivation.

The student will work within the Leicester Real World Evidence Unit with access to linked primary care data through the Clinical Practice Research Datalink (CPRD) dataset, as well as national datasets accessed through different Trusted Research Environments (TREs). The project objectives will be addressed by using a diversity of statistical methods, including imputation, time series regression modelling and flexible parametric survival modelling. 

The project has been designed to add value to existing programmes of research funded by Health Data Research UK and our industry partners. It also aligns to the aspirations of the NIHR Leicester BRC, consolidating collaboration across the Lifestyle theme with data innovation and ethnic health. The project will help support the continuation or expansion of this funding in the future.

References

Yates T, Summerfield A, Razieh C, Banerjee A, Chudasama Y, Davies MJ, Gillies C, Islam N, Lawson C, Mirkes E, Zaccardi F. A population-based cohort study of obesity, ethnicity and COVID-19 mortality in 12.6 million adults in England. Nature communications. 2022 Feb 2;13(1):1-9.