*1.1. Health Behavior Clustering*

Albeit not specifically among Latinxs, several studies have explored simultaneously occurring clusters of behaviors, primarily focused on smoking, drinking, poor diet, and low physical activity, which comprise the main behavioral determinants of chronic diseases [13]. Past studies have uncovered various clusters among these behaviors. For instance, using data from the U.S. National Health and Nutrition Examination Survey (NHANES) III, Berrigan et al. [7] identified 32 patterns of health behavior among a representative sample of non-Hispanic Whites, non-Hispanic Blacks, and Mexican Americans. The most common pattern was adhering to recommendations on alcohol and tobacco consumption while not adhering to recommendations for physical activity, dietary fat intake, or fruit and vegetable consumption [7]. While this was also the most common pattern among Mexican Americans, it was less prevalent (13.8%) than among non-Hispanic Whites (14.7%) and non- Hispanic Blacks (18.4%). A second pattern that was particularly more prevalent among Mexican Americans than other groups was non-adherence to physical activity and to fruit consumption recommendations and adherence to tobacco, alcohol, and vegetable recommendations. Berrigan et al. [7] also found that men, regardless of their race or ethnicity, were 2.6 times more likely to not adhere to all five recommendations (physical activity, fruit and vegetable consumption, alcohol, tobacco use, and dietary fat intake) than women.

Using data from the Aerobics Center Longitudinal Study (ACLS), Héroux et al. [14] identified two clusters of unhealthy behaviors among a primarily non-Hispanic White (95%) sample of 13,621 participants in the U.S. The first group was composed of individuals more likely to engage in smoking, alcohol use, unhealthy diet, and low physical activity, while the second one was more likely not to engage in any of the four unhealthy behaviors [14]. They also found that all behaviors were significantly associated with each other, such that engaging in one behavior was related to increased odds of engaging in another one. For instance, individuals with an unhealthy diet (relative to those with a healthy diet) were 2.45 times more likely to engage in low physical activity, 2.02 times more likely to smoke, and 1.61 times more likely to drink heavily [14].

Research studies from other countries further support for the clustering of health behaviors and the importance of considering sociodemographic differences. In a populationbased study among Irish adults, for example, Conry et al. [15] found six different clusters of behaviors. Individuals in the healthy lifestyle cluster, characterized by non-smokers, high physical activity, healthy eating, and moderate alcohol use, tended to be women, older (65+ years), and of higher socioeconomic status (SES), while those in the mixed lifestyle cluster (non-smokers, moderate physical activity, and variable alcohol consumption) were more likely to be men, younger, and of low SES [15]. Similarly, using population-based data from 4238 German participants, Rabel et al. [16] identified three clusters of behaviors which varied by sex. The healthiest cluster (low to moderate drinking, favorable diet, moderate physical activity, and no smoking) was endorsed by women only, while the other two more heterogenous clusters were endorsed primarily by men (≥71%). Collectively, these studies suggest the co-occurrence of health behaviors is common, but patterns vary by sociodemographic factors. Men, in particular, seem to be more likely to engage in multiple risky behaviors than women; however, few studies have focused exclusively on men, and fewer on Latino men [17,18].
