The association of lifestyle and mortality risk in individuals with multimorbidity: a population-based cohort study

Abstract

Multimorbidity is highly prevalent and a major driver of premature mortality(1–5). While healthy lifestyle behaviours are known to reduce mortality risk in the general population(6,7), it remains unclear to what extent individuals with multimorbidity benefit from sustained lifestyle improvements. We aimed to evaluate the association between lifestyle and mortality risk in people with multimorbidity, and whether it varied across the specific multimorbidity patterns. This prospective study included 475,706 UK Biobank participants. A lifestyle score(8) according to nine risk factors (smoking, alcohol, physical activity, sleep, TV viewing, fruit and vegetable intake, oily fish, red meat, processed meat) was established and categorised as healthy and unhealthy lifestyle. Specific multimorbidity patterns were defined as metabolic (≥2 of diabetes, stroke, heart disease), cardio-renal (≥2 of diabetes, stroke, heart disease, kidney disease), and overall (≥2 of 43 conditions) multimorbidity(9). All-cause mortality was ascertained from death registries and hospital records. Cox proportional hazards models, with interaction terms, assessed associations between lifestyle and the outcome across multimorbidity types. Participants with missing covariates or implausible values of lifestyle risk factors were excluded. Models were adjusted for age, sex, ethnicity, deprivation, and BMI. All participants provided written informed consent(10). Among participants, 44.5% reported an unhealthy lifestyle, and 1.9%, 1.7% and 32.8% had cardiovascular, cardio-renal and overall multimorbidity, respectively. Compared with those without multimorbidity, hazard ratios (HRs) for all-cause mortality were 4.63 (95% CI, 4.30-4.98) for metabolic multimorbidity and 4.84 (95% CI,4.51-5.20) for cardio-renal multimorbidity. Compared with people without diseases and with healthy lifestyle, participants with cardio-renal multimorbidity and an unhealthy lifestyle had a markedly elevated risk of HR 6.65 (95% CI, 6.10–7.25) for all-cause mortality. Significant multiplicative and additive interactions between lifestyle and multimorbidity were observed for all-cause mortality in overall multimorbidity. Lifestyle behaviours substantially modify the mortality risk among people with multimorbidity. The greatest benefit is observed in cardio-renal multimorbidity. Targeted, cluster-specific lifestyle interventions could markedly reduce premature mortality and support precision prevention strategies in people with high multimorbidity burden.