The Association Between Depressive Symptoms and the Weekly Duration of Physical Activity Subset by Intensity and Domain: Population-Based, Cross-Sectional Analysis of the National Health and Nutrition Examination Survey From 2007 to 2018.

Background: Prior literature suggests a dose-response relationship between physical activity (PA) and depressive symptoms. The intensity and domain of PA are suggested to be critical to its protective effect against depression; however, existing literature has shown mixed results.

Objective: The purpose of this population-based study is to examine the associations between depressive symptoms and weekly duration of (1) total PA and (2) PA subset by intensity, domain, or both.

Methods: A cross-sectional analysis of National Health and Nutrition Examination Survey data from 2007 to 2018 was conducted using multivariable logistic and linear regression models and survey weights. Participants (N=29,730) were 20 years and older and completed the Physical Activity Questionnaire and Depression Screener. The primary outcome was the presence of depressive symptoms, and the secondary outcomes were cognitive-affective and somatic symptoms of depression.

Results: Participants (N=29,730) had a weighted mean age of 47.62 (SD 16.99) years, and 15,133 (51.34%) were female. On average, participants without depressive symptoms engaged in 10.87 hours of total PA per week, whereas participants with depressive symptoms engaged in 8.82 hours (P<.001). No significant associations were seen between the weekly duration of total PA and depressive symptom odds, somatic, or cognitive-affective symptoms (all P>.05). Participants with an increased weekly duration of recreational PA had decreases in depressive symptom odds (adjusted odds ratio [aOR] 0.965, 95% CI 0.944-0.986) and in somatic (adjusted coefficient [aβ]=-0.016, 95% CI -0.022 to -0.009) and cognitive-affective (aβ=-0.015, 95% CI -0.023 to -0.007) symptoms. When recreational PA was subset by intensity, participants with an increased weekly duration of vigorous-intensity recreational PA had decreases in depressive symptom odds (aOR 0.926, 95% CI 0.883-0.972) and in somatic (aβ=-0.021, 95% CI -0.032 to -0.010) and cognitive-affective (aβ=-0.022, 95% CI -0.035 to -0.009) symptoms. However, significant associations were not seen for the weekly duration of work-related, moderate- or vigorous-intensity PAs (all P>.05).

Conclusions: Findings suggest that recreational, not work-related PA is associated with reduced symptoms of depression. Future studies should explore the impact of the different types and contexts of PA on depressive symptomatology.