Longitudinal patterns of lifestyle risk behaviours among UK adults with established cardiovascular disease: a latent transition analysis.

Abstract

People with cardiovascular disease (CVD) need to engage in healthy lifestyle behaviours. However, there is a gap in identifying longitudinal patterns of change in lifestyle behaviours among people with CVD. This study aimed to identify clustering of lifestyle risk behaviours and their 4 ± year changes among UK adults with CVD, and to determine the associated factors. We used the UK Biobank data collected at two time points (2006-2010/baseline data = T0 and 2014+/third visit data = T4). Six key lifestyle risk behaviours were assessed: smoking, high alcohol intake, poor fruit and vegetable consumption, physical inactivity, poor sleep balance (<7 or >8 h/night) and prolonged sitting. A random intercept latent transition analysis was performed to identify patterns of lifestyle risk behaviours at T0 and their changes from T0 to T4. We included 5,304 participants with CVD whose data on lifestyle risk behaviours were collected at two-time points. Alcohol intake and current smoking were 75.7% and 5.4% at baseline, respectively, and 67.4% and 3.0% at follow-up. Three latent classes emerged: Latent class (LC) 1-"high alcohol intake, poor sleep balance and poor fruit and vegetable intake", LC2-"high alcohol intake and poor fruit and vegetable intake", and LC3-"high alcohol intake". Most adults remained in the same LC over the 4 + years (range: 83.9%-100.0%). After 4 + years, 3.5% from LC3 and 10.4% from LC2 at baseline moved into LC1. The odds of transitioning to LC2 relative to staying in LC1 and LC3 were 2.22 and 4.13 times higher for males than for females, respectively. A single-year increase in participants' age was associated with a 1.16 times increase in the odds of moving to LC1 relative to staying in LC2. People with CVD did not show improvement in lifestyle risk behaviours, and interventions targeting multiple lifestyle risk behaviours are needed to improve CVD.