|
Authors assessed the measurement properties of Baecke s habitual physical activity questionnaire and related sociodemographic and personal life-style variables Translated Baecke s questionnaire was administered to 526 subjects selected at authors convenience from those who visited to the Dongsan Medical Center July to August in 1991 for biannual periodic physical examination in Daegu. 315 men (mean age 31.1. standard deviation. SD 7.3) and 181 women (mean age 24. 6. SD 4.7) were used in final analysis excluding 28 having conditions limiting physical activities. Mean scores of physical activity at work and sport during leisure time were significantly higher in men than in women (p <0.01. p <0.05. respectively) while that _of physical activity during leisure time excluding sport was not different between them. Score of sport index showed a relatively large positive skewness. Cronbach´s alphas for work, sport and leisure-time index was 0.64. 0.75, 0.28 for men, 0.42, 0.76, 0.15 for women respectively. In both males and females, factor structure was not emerged explicitly as in Netherlands. Many of the items underlying 3 components of physical activities were related each other suggesting poor construct validity. Multiple regression analysis by forced entry method showed that educational attainment and monthly income, both of which are indicators of socioeconomic status were related to index at work. The educational attainment was inversely related to work index (regression coefficient. B = - 0.173. p < 0.01) and monthly income 50-99 million Won was inversely related to reference income less than 50 million Won(B = - 0.241, p < 0.01) and more than 100 million Won was inversely related to reference income(B =- 0.552. p < 0.01) suggesting a dose-response relationship. Monthly income was significantly related inversely to work index in females suggesting a similar dose-response relationship as in males. Drinkers in females showed more physical activity in work index (B =0.246. p < 0.05) than nondrinkers unlike in univariate analysis. Age was paradoxically related to physical activity. More than 40 years of age group were more physically active in sport index than reference 20-29 age group in males (B = 0.402. p ( 0.05)and more than 30 years of age group were more active in sport than reference 20-29 age group in females(B = 0.404. p ( 0.05). Lower educational attainment was related to leisure-time index excluding sport in males (B = - 0.153. p (0.05) and age showed a borderline significance (B = 0. 212) in males. The married was less active in leisure-time activity than others (B 0.337. p < 0.01). None of sociodemographic variables were significant in females for leisure-time index. Body mass index was related to none of the 3 component indices of Baecke´s questionnaire in both of males and females. These results suggest that Baecke´s Habitual Physical Activity Questionnaire may be not appropriate in the present study sample.
|