2718) in 1993 were used for this study. In the first follow-up survey, information on physical activity was obtained using questionnaires developed by Baecke et al. Levels of physical activity were scored into four categories: physical activity at work (work index), leisure time exercises (exercise index), leisure time activity without exercise (leisure-time index) and total activity index. Sociodemographic characteristics, lifestyle and medical history etc, were also obtained from questionnaires. Overnight fasting blood samples were obt ained for lipoprotein analysis; other anthropometric and blood pressure measurements were also performed at the same time. We compared blood pressure and total cholesterol (TC) using means or proportional values, included with 95% confidence intervals, by the level of physical activity. Logistic regression analysis was also used for risk measurements. Results: Controlling for age, body mass index, smoking, drink, diabetes and family history of cardiovascular disease, significant inverse trends were found f or TC across groups with low, moderate and high levels of total physical activity. The average adjusted TC for individuals with high and those with low total activity levels were 190.0 vs. 197.8 mg/dl for men and 195.8 vs. 203.2 mg/dl for women. In a multiple logistic regression model, a negative association was observed between hypertension and total physical activity in both men (OR=0.7) and women (OR=0.66) after controlling for other co-factors. We also observed that TC was negatively associated with le isure time activities both in men and women. Conclusions: The effect of total physical activity is greater to total cholesterol than to blood pressure. No strong correlation among exercise activity and lipid level and blood pressure was found for the study population.
