A non-linear model of adherence to public health anticontagion measures during the COVID-19 pandemic is implemented by outlining distinct behavioral profiles based on five health measures. It also seeks to explain each profile based on subjective, psychosocial, and socioeconomic factors. The sample consisted of 2,387 university students from the state of Rio de Janeiro aged 18–35 years. A structured online questionnaire was applied with questions about their anticontagion attitudes. A three-class solution was identified based on lower BIC/AIC values and interpretability, with class 1 representing flexible adherence, class 2 strict adherence and class 3 risk adherence patterns. Logistic regression analyses indicated that psychosocial factors weighed more than objective factors in shaping adherence behaviors, especially anti-adherence beliefs (OR = 1.21 for Class 1; OR = 0.55 for Class 2; OR = 2.67 for Class 3), perceived family norms (OR = 0.76 for Class 1; OR = 1.42 for Class 2) and perceived friends norms (OR = 1.16 for Class 1; OR = 0.71 for Class 3). These results demonstrate that contrary to a linear high-low adherence pattern, this is a multifaceted phenomenon shaped by the interaction between personal and contextual factors, highlighting the need for tailored interventions according to behavioral profiles.
Bernardino et al. (Tue,) studied this question.