Objective: To explore the relationship between ultra-processed foods (UPF) consumption and dietary, lifestyle and social determinants using pathway analysis in the baseline of the Cohort of Universities of Minas Gerais (CUME project).
Design: Cross-sectional study, in which path analysis was used to estimate direct and indirect effects of dietary practices, sleep, time on the computer and professional status on UPF consumption.
Setting: Data were collected in 2016, through an online questionnaire composed of sociodemographic, anthropometric, lifestyle and dietary practices questions, and a FFQ.
Participants: Baseline participants from the CUME Project (n 2826), adults who graduated from Universidade Federal de Viçosa or Universidade Federal de Minas Gerais, Brazil.
Results: Being employed (P = 0·024), the time spent on the computer (P = 0·031) and the frequency of fried food intake (P < 0·001) were positively and directly associated with UPF consumption, whereas the sleep duration (P = 0·007) and the number of meals per d (P < 0·001) were negatively and directly associated with UPF consumption. Indirect effects were observed between being employed, mediated by the sleep duration (P = 0·032) and fried food intake (P = 0·005), whereas being a student is mediated by the time on the computer (P = 0·048).
Conclusion: The time spent on the computer, sleep duration and fried food consumption showed direct effects on UPF consumption. They also acted as mediators on the relationship between professional status and UPF consumption. Besides, the number of meals eaten each day also was directly associated with UPF consumption.
Keywords: Eating behaviour; Industrialised foods; Lifestyle; NOVA classification; Nutritional epidemiology.