Abstract
BACKGROUND: A considerable proportion of older adults exhibit multiple modifiable health risk behaviors. Identifying not only the co-occurrence of health risk behaviors but, more importantly, their specific clustering patterns is essential for developing targeted multiple health behavior interventions. METHODS: The study sample comprised 12,766 participants from the 2018 wave of the CLHLS. Data on their health risk behaviors (inadequate sleep, insufficient fruit intake, insufficient vegetable intake, salty diet, smoking, drinking, and irregular exercise), basic demographic information, and health outcomes (self-rated quality of life scores and self-rated health status) were obtained. Analytical methods included descriptive statistics, Spearman correlation, binary logistic regression, Odds/Expected (O/E) ratios, and network community detection. RESULTS: The median age of the participants was 84 (75, 93) years, with 5,891 (46.15%) males. A high prevalence (64.36%) of co-occurring health risk behaviors (≥2) was observed. There was a negative correlation between the number of co-occurring health risk behaviors and self-assessed quality of life (r = -0.173, p < 0.01), as well as self-assessed health level (r = -0.141, p < 0.01). Gender, age, years of education, residence, and economic status significantly influenced the co-occurrence of behaviors (all p < 0.01). Smoking and drinking (O/E = 2.67), insufficient fruit and vegetable intake (O/E = 1.60), and a salty diet and smoking (O/E = 1.32) demonstrated the most significant clustering risks. Two distinct clustering patterns were identified, which can be termed an Addiction Behavior Pattern and an Unhealthy Activity-Eating Pattern. CONCLUSION: The co-occurrence of health risk behaviors is prevalent and associated with adverse health outcomes. Behaviors exhibit a clustering effect and demonstrate distinct clustering patterns. Public health interventions should move beyond single-behavior approaches to develop integrated strategies that target these specific clustering patterns for more effective management of multiple health risks.