J Epidemiol Community Health 59:377-379 doi:10.1136/jech.2004.031526
  • Evidence based public health policy and practice

Work related and non-work related stress in relation to low leisure time physical activity in a Swedish population

  1. Karl Magnus Wemme1,
  2. Maria Rosvall2
  1. 1Department of Social and Preventive Medicine, Malmö University Hospital, Malmö, Sweden
  2. 2Department of Community Medicine, Lund University, Malmö University Hospital
  1. Correspondence to:
 Dr K M Wemme
 Department of Social and Preventive Medicine, Malmö University Hospital, SE-205 02 Malmö, Sweden;
  • Accepted 23 November 2004

Physical activity is regarded as an important component of a healthy lifestyle. Several social and environmental factors have systematically emerged as barriers to low leisure time physical activity (LTPA), for example, lack of money, low social support due to lack of supportive family or friends, and living in high crime rate areas. Low LTPA has been found to be strongly associated with low socioeconomic status groups where psychosocial stressors have been suggested to play a mediating part.1,2

The overall aim of this study was to investigate the association between work and non-work related stressors, respectively, in relation to low LTPA in a general population sample. Our main hypothesis was that psychosocial stressors would act as barriers to physical activity thus generating sedentary behaviour as measured by LTPA.


The Scania health survey was conducted during the period of November 1999 through April 2000. After excluding refusals, non-contacts, and other non-responding groups, the final sample consisted of 13 715 persons, making the response rate 59% of the sample size. A self administered questionnaire provided information regarding health related behaviours, sociodemographic characteristic, and psychosocial factors. Our analysis was restricted to currently employed people (52% of the sample population) yielding a study population of 7169 subjects (3877 men and 3292 women). Odds ratios for low LTPA in relation to psychosocial variables were calculated using two logistic regression models, model one with adjustment for age, …