The development of multiple long-term conditions (MLTCs, previously referred to as multimorbidity) is a priority for global health research. The complex relationship between lifestyle factors and MLTCs remains largely unexplored. Drug based interventions (e.g. statins) often focus on the risk reduction for single outcomes (e.g. prevention of cardiovascular disease). In contrary, many lifestyle factors (e.g. exercise) will have a preventative effect on more than one outcome (e.g. cancer, cardiovascular disease and diabetes). Unpicking the relationship between lifestyle factors and MLTCs is therefore of interest to take a more holistic view of patient care, in a time where individuals living with multiple long term-conditions is rising. The aim of this project is to explore the relationship between lifestyle factors and occurrence of multiple long-term conditions (MLTC) in the UK population.
Objective 1: Clustering Analysis to identify key relationships between lifestyle factors and multiple long term health conditions (MLTCs)
Apply clustering techniques: Apply clustering techniques to group the MLTCs and validate. Supervised algorithms will be trained on lifestyle factors (physical activity, sleep, smoking, diet, alcohol consumption and sun exposure).
Objective 2: Association Analysis for key relationships
Univariable and multivariable analyses, estimating association between lifestyle factor and multimorbid groups (and contributing conditions), quantified through risk difference, risk ratio and odds ratios.
Objective 3: Causal Analysis for key relationships
Directed Acyclic Graphs (DAGs): DAGs will be constructed to help identify confounding variables and motivate adjustment sets. We will use causal modelling techniques (standardisation, inverse probability weighting) to estimate the average treatment effects. Target trial emulation frameworks will be used where relevant. Direct effects will be estimated using mediation analysis.