Patterns of health lifestyle behaviours: findings from a representative sample of Israel.

backgroundResearchers are increasingly acknowledging the importance of understanding patterns of engagement in multiple, as opposed to isolated, health behaviours. Accordingly, several studies, targeting various geographical regions, have begun to identify and characterize unique classes of individuals in terms of their engagement in health behaviours, towards gaining insights that might guide tailored health interventions. Our study extends this stream of research to the Israeli context, as well as examines whether certain sociodemographic characteristics tend to be associated with affiliation to a specific group of individuals, i.e., class membership.

methodsTwo large representative samples were obtained from the 2010 and 2017 National Social Surveys of the Israel Central Bureau of Statistics. Latent Class Analysis was used to identify different classes, based on levels of engagement in five health behaviours: fruit-and-vegetable consumption, physical activity, smoking avoidance, sleep, and influenza vaccination. Multinomial logistic regression was applied to test the associations between sociodemographic characteristics (age, gender, religion, education level) and class membership.

resultsWe identified three distinct classes (denoted "healthy", "unhealthy", and "mixed"). Overall, the "healthy" class was characterized by healthy levels of fruit-and-vegetable consumption and physical activity, a low rate of currently-smoking individuals, and a high proportion of influenza vaccination. The "mixed" class was characterized by healthy levels of fruit-and-vegetable consumption and sleep duration, moderate levels of physical exercise, a high proportion of currently light smokers, and a low rate of vaccination. The "unhealthy" class was marked by relatively low levels of engagement in health behaviours. Generally, older, Jewish, and more-educated individuals were more likely to belong to the "healthy" class, while young, non-Jewish, and less-educated individuals were more likely to belong to the "unhealthy" class. We further identified differences between the 2010 and 2017 samples (e.g., differences in men's likelihood of being in the "unhealthy" class), suggesting that some behavioural patterns might have changed over time.

conclusionThis research identified three classes of engagement in health behaviours across two large and representative samples. Moreover, the classes were associated with unique sociodemographic characteristics. Consequently, our findings can highlight health-behavioural patterns relevant to different sub-populations that should be considered in public health interventions.