Physical activity, exercise and self-rated health: a population-based study from Sweden

BACKGROUND In order to screen for the most inactive individuals in the population and target health-related interventions where they are most needed it is important to assess different forms of physical activity in population-based studies. The aims were (1) to identify the most inactive individuals in the population by assessing two dimensions of physical activity, (2) to investigate the correlation between exercise and total physical activity and (3) to investigate the association between exercise, total physical activity and good self-rated health. METHODS A simple random sample of the Swedish population aged 25-64 years were interviewed about their living conditions, health and lifestyle in a survey performed by Statitics Sweden. In total 1876 women and 1880 men completed the survey during 1999 (response rate 76.6%) when two different questions about physical activity assessed exercise and total physical activity in all domains (e.g. transportation, exercise, and at work). Logistic regression models were used to estimate odds ratios. RESULTS The most inactive individuals (no exercise and total physical activity </= 2 hours per week) constituted 4.3% of the sample. The correlation between exercise and total physical activity was low (gamma = 0.4, p = 0.02). There were significant associations between higher levels of exercise, total physical activity and good self-rated health after adjustment for age, gender, country of birth, education, employment, marital status, housing tenure, smoking and BMI. CONCLUSION Both exercise and total physical activity were independently associated with good self-rated health. It seems to be advantageous to use more than one question in population based surveys in order to evaluate several dimensions of physical activity and identify the most inactive individuals.