Association of physical activity domains and television time with mental health profiles in Brazilian adolescents: a latent class analysis

Authors:
Open Access
Publisher website
Google Scholar
Abstract
This cross-sectional study identified mental health profiles among Brazilian adolescents and examined their associations with domain-specific physical activity (leisure time, school, and active commuting) and television (TV) viewing time. Data were obtained from 144,046 adolescents in the 2019 Brazilian National Survey of School Health (PeNSE, acronym in Portuguese). Self-reported indicators included the frequency of five emotional states, weekly minutes of physical activity in leisure time, physical education, and commuting, and daily TV hours. Latent class analysis classified adolescents based on their mental health profiles, and multinomial logistic regression estimated associations between physical activity domains and TV viewing time with mental health classes, adjusting for covariates. Six mental health profiles were identified: (1) school-related concern; (2) anxious and stressed; (3) subthreshold internalizing; (4) anxious, sad, and angry; (5) low symptom; (6) severe internalizing. Adolescents who engaged in physical activity in their leisure time and at school were more likely to belong to the low symptom class, while those who engaged in physical activity during commuting were more likely to be in class 6 (severe internalizing). Adolescents who watched two to four hours of TV per day were more likely to belong to class 1 (school-related concern), class 4 (anxious, sad, and angry), and class 5 (low symptom), whereas those who watched TV more than four hours/day were less likely to belong to class 3 (subthreshold internalizing) and class 6 (severe internalizing). These findings underscore that both the context and quantity of physical activity and TV time are differentially related to adolescent mental health profiles, implying the need for targeted surveillance and interventions.

This publication has 38 references indexed in Scilit: