Leisure-time physical activity dose-response effects on obesity among US adults: results from the 1999-2006 National Health and Nutrition Examination Survey

Background It is not well established whether total volume of leisure-time physical activity (LTPA) has dose-response effects on obesity. Methods The dose-response relationship was examined using 12 227 non-institutionalised individuals, aged 20-64 years, drawn from the 8 years (1999-2006) of the continuous National Health and Nutrition Examination Survey (NHANES), a nationally representative sample of the US population. Results The age-adjusted prevalence of women's obesity was 41.4% for those with no LTPA in the past month; 39.1% for those who engaged in LTPA but fell short of the recommended minimum amount of LTPA (ie, <450 metabolic equivalent minutes per week (MET min/week)); 31.0% for those who met the recommended minimum guideline (ie, 450 to <750); 28.0% for those whose LTPA exceeded the minimum guideline but less than the first quartile among the overachievers (ie, 750 to <1260); 23.4% for the overachievers between the first and third quartile (ie, 1260 to <3556); and 19.5% for the overachievers at or above the third quartile (ie, 3556 MET min/week or above). This association was maintained even after occupational physical activity (OPA) was controlled. However, this pattern was not observed for Mexican and black adults and showed a floor effect as LTPA increased. Conclusions There is a crude graded inverse dose-response relationship between total volume of LTPA and obesity in US adult women, but not in men. Gender and racial/ethnic differences exist in the relationship of accumulated LTPA with obesity due, in part, to differential ratios of LTPA to OPA.