Background
: The purpose of this study was to assess the lifestyle associated with weight gain and persistent overweight for 2 years.

Methods : The subjects were 580 workers (109 men, 471 women) at one university affiliated hospital. Their height and weight were measured twice in 2000 and 2002. A questionnaire, about lifestyle was filled out in 2000. The questionnaire included the average amount of alcohol intake and the number of days per week in which alcohol was consumed, smoking status, and the number of days per week in which exercise was done. Their weight change in 2 years was calculated and their weight status was classified into two groups: persistent normal weight (BMI＜23 kg/m²) and persistent overweight (BMI≥23 kg/m²).

Results : Among the total, 54.3% of men and 10.0% of women had persistent overweight over 2 years. Overall 93.4% of overweight men and 68.7% of overweight women remained overweight after 2 years. Compared with male non-smokers and male ex- smokers, current male smokers gained weight as much as 1.17 kg over 2 years after adjusting for age and initial BMI (P=0.033). Compared with females who exercised less than 3 per week, those who exercised 3 or more per week lost weight as much as 1.21 kg over 2 years after adjusting for age and initial BMI (P=0.005). Compared with males who consumed alcohol an average of ≤30 g per consumption, those who consumed an average of ＞30 g were 4.1 times (95% C.I. 1.04∼16.21) more likely to maintain persistent overweight over 2 years after adjustment for age and frequency of alcohol use. However no particular lifestyle was shown to predict the risk of persistent overweight for women.

Conclusion : Regular exercise of moderate intensity and modest alcohol consumption seemed to prevent weight gain and persistent overweight. Such lifestyle may be encouraged to maintain healthy weight.

Background
: The risk of overweight and obesity to health is well know. Physical activity is important in the treatment of obesity in terms of healthy body composition and maintenance of body function. This study is a follow-up study to identify the impact of changes of leisure time physical activities on body weight.

Methods : The study subjects were people who visited a health examination center in Seoul at least twice between 1995 and 2000. Seven hundred people who completed physical activity questionnaires each time they visited were included. Three hundred eighty two of them were men, and 318 were women. The questionnaire was Baecke physical activity questionnaire, consisted of items for work activities, sports activities, and non-sports leisure activities. Each index of activities was calculated.

Results : Ninety seven people (13.9%) reduced their weights more than 5% of initial body weights. Twenty eight people (4.0%) gained their weights more than 5% of initial body weights. In five hundred and seventy five people (82.1%) their body weights were changed within 5% of initial body weights. Non-sports leisure time indices increased greatly and were significantly more than sports indices in reduced weight group. Non-sports leisure time indices were not changed significantly in gained weight group. The dose-response relationship existed between changed body weight (kg) and difference of leisure time indices. The multivariate analysis showed that female was a significant factor related to reducing weight (odds ratio=4.5, 95% confidence interval 2.7~7.5). Increase of non-sports leisure indices showed borderline significance (odds ratio=1.4, 95% confidence interval 0.9~2.0).

Conclusion : To increase non-sports leisure activities for the purpose of reducing body weight is actively recommendable. Such activities performed on a daily basis may have impact on reducing weight through repeatability and ease of access.