LAUSR

BACKGROUND AND OBJECTIVES The relationship between eating rate (ER) and increased risk of obesity in relation to body mass index (BMI, i.e., total body fat) and waist circumference (WC, i.e., abdominal fat) has not been fully examined. Considering gender differences, we identified unknown confounding factors (CFs) for each risk, and then assessed the two actual obesity risks, adjusting for the CFs. METHODS AND STUDY DESIGN Using a questionnaire, we collected data for ER (slow, normal as "reference," and fast) and related factors and measured BMI and WC for 3,393 men and 2,495 women. Using multiple logistic regression models, odds ratios (ORs) and their 95% confidence intervals (CIs) were estimated adjusting for both conventional and candidate CFs. RESULTS The following factors were identified as appropriate CFs, but were differed between the two obesity types: fast food consumption in both genders, sleep duration and restaurants/food service use in men, and family structure and packed lunch in women. In men, actual risks of BMI obesity and WC obesity were negatively associated with slow ER (ORs and 95% CIs; 0.70 and 0.52-0.96, and 0.69 and 0.50-0.96), but positively associated with fast ER (1.48 and 1.25-1.76, and 1.45 and 1.21-1.74). In women, those risks were positively related to fast ER (1.78 and 1.39-2.26, and 1.34 and 1.11-1.61). CONCLUSIONS For both BMI obesity risk and WC obesity risk, we conclude that slow and fast ER were related to decreased and increased risks when adjusted for appropriate CFs, which differed by gender and the obesity type.