Background There are few studies on the moderating role of physical activity (PA) between occupational stress and depressive symptoms, and even fewer studies evaluating the roles of different types of PA. Female nurses are a high-risk group with occupational stress and depressive symptoms. Studies of this occupational group will provide a basis for making intervention programs to reduce their occupational stress and depressive symptoms.

Objective This study investigates the relationship among occupational stress, PA, and depressive symptoms, and explores the potential role of PA moderating the relationship between occupational stress and depressive symptoms.

Methods A three-stage stratified random sampling method was applied to select 716 female nurses from 3 tertiary class-A hospitals, 5 secondary class-A hospitals, and 11 community health service centers from April to June 2018. They were asked to complete the Occupational Stress Questionnaire, the International Physical Activity Questionnaire (IPAQ), and the Selfrating Depressive Symptoms Scale (SDS) to assess occupational stress, PA, and depressive symptoms; a total of 649 valid questionnaires were returned, and the valid return rate was 90.6%. A linear stratified regression model was used to analyze the moderating effect of PA on the relationship between occupational stress and depressive symptoms.

Results The positive rates of occupational stress and depressive symptoms of selected female nurses were 80.4% and 50.8% respectively. Occupational stress was positively correlated with depressive symptoms (r_s=0.325, P < 0.01); occupation-related PA was positively correlated with occupational stress and depressive symptoms (r_s=0.225, 0.248, P < 0.01); however, leisure-related PA was negatively related to occupational stress and depressive symptoms (r_s=-0.241, -0.216, P < 0.01). Occupation-related and housework-related PA had positive impacts on depressive symptoms (b=0.251, 0.194, P < 0.05), while leisure-related PA had a negative impact (b=-0.234, P < 0.05). Occupation-related PA showed a positive regulating role between occupational stress and depressive symptoms (b=0.197, P < 0.01), while leisure-related PA showed a negative regulating role (b=-0.160, P < 0.01). The moderating effect slope showed that with increasing occupational stress level, the positive rate of depressive symptom showed an upward trend, and compared with the low-level group, the upward trend of the high-level occupation-related PA group was steeper (b=1.435, P < 0.05), while that of the high-level leisure-related PA group became flatter (b=0.875, P < 0.05).

Conclusion Different types of PA have varied moderating effects on the relationship between occupational stress and depressive symptoms. Occupation-related PA has a positive moderating effect, while leisure-related PA has a negative moderating effect.