Potential influence of physical, psychological and lifestyle factors on the association between television viewing and depressive symptoms: A cross-sectional study.

To investigate the potential influence of physical, psychological, and lifestyle factors on the association between TV-viewing and depressive symptoms among Brazilian adults. We used cross-sectional data from the Brazilian National Survey, conducted in 2013 with 60,202 adults (≥18 years). Information regarding exposure (TV-viewing), potential influencing factors (multimorbidity, mobility, self-rated health, tobacco use, alcohol consumption, sugar consumption, and physical activity) as well as elevated depressive symptoms (through PHQ-9 – score > 9) (outcome) was collected via interview-administered questionnaires. Data on covariates were self-reported. Body mass index was estimated through the assessment of body mass and stature. Mediation models were estimated through the Karlson-Holm-Breen method. Individuals who reported >5 h/d of TV viewing showed a higher prevalence of depressive symptoms than those with <5 h/d of TV viewing [8.1%(99%CI:7.6%–8.6%) vs 14.2%(99%CI:12.2%–16.6%)]. The association between TV-viewing and depressive symptoms was influenced by tobacco use (Overall: 7.22%; men: 4.46%, women: 8.59%), physical activity (men: 3.99%, women: 2.28%), mobility (overall: 11.31%, men: 10.85%, women: 11.03%), and multimorbidity (overall: 9.11%, men: 11.6%, women: 6.03%). Poor self-rated health influenced the association between TV-viewing and elevated depressive symptoms only among men (15.55%). Similarly, the association between >4 h/d of TV viewing and depressive symptoms was influenced by tobacco use (men: 6.8%, women: 11.7%), physical activity for women (5.5%), self-rated health for men (14.7%), mobility (men: 8.7%, women: 17.0%), and multimorbidity (men: 9.6%, women: 12.3%). Tobacco use, physical activity, mobility, multimorbidity, and self-rated health (men) mediate the relationship between high TV-viewing and elevated depressive symptoms. Longitudinal research is required to confirm/refute our data which may also be useful to contribute to public health interventions. [ABSTRACT FROM AUTHOR]