Variables associated with unhealthy diet among Portuguese adults: a population-based logistic regression study.

Background: Noncommunicable diseases, responsible for most global deaths, are closely linked to unhealthy lifestyles, particularly a poor diet. There has been an important shift in dietary habits, deviating from national and international recommendations. In contrast, the adult Portuguese population appears to overestimate the healthfulness of their diet, highlighting the need to better understand these behaviors and objective associations. Methods: This study used weighted multiple logistic regression to identify associated variables with unhealthy diet among 891 Portuguese adults from a cross-sectional, population-based survey conducted in 2019. A healthy diet was defined by regular meals, adequate fruit and vegetables intake, and moderate alcohol consumption, according to established public health guidelines. Results: Non-smokers were less likely to have an unhealthy diet [OR = 0.21 (0.07-0.53)], while screen users were more likely [OR = 2.36 (1.12-5.11)]. The model showed moderate discrimination, AUC = 0.62. Conclusion: Smoking and screen-related behaviors independently were associated with unhealthy diet, underscoring the need for integrated public health strategies targeting multiple lifestyle risks to foster healthier habits.