Effects of lifestyle risk behaviour clustering on cardiovascular disease among UK adults: latent class analysis with distal outcomes

Abstract Background: Lifestyle risk behaviours such as smoking, physical inactivity, and unhealthy diet account for a considerable disease burden globally. These risk behaviours tend to cluster within an individual, which could have detrimental health effects. In this study, we aimed to examine the clusteringeffect of lifestyle risk behaviours on cardiovascular disease (CVD) and CVD risk among adults in the United Kingdom (UK). Methods: We performed a latent class (LC) analysis with distal outcomes using the UK Biobank baseline (2006-2010) data. First, we estimated LC measurement models, followed by an auxiliary model conditional on LC variables. We reported continuous (mean difference - MD) and binary (odds ratio - OR) outcomes with 95% confidence intervals. Results: We included 283,172 and 174,030 UK adults who had data on CVD and CVD risk, respectively. Multiple lifestyle risk behaviour clustering (physically inactive, poor fruit & vegetable intake, high alcohol intake, and prolonged sitting) had a 3.29 mean increase in CVD risk compared to high alcohol intake. In addition, adults with three risk behaviours (physically inactive, poor fruit & vegetable intake, and high alcohol intake) had 25.18 higher odds of having CVD than those with two risk behaviours (physically inactive, and poor fruit & vegetable intake). Social deprivation, gender and age were also associated with CVD. Conclusion: Individuals' LC membership with two or more lifestyle risk behaviours negatively affects CVD. Interventions targeting multiple lifestyle behaviours and social circumstances should be prioritized to reduce the CVD burden.