BANDINI, LINDA G. DUNG VU, AVIVA MUST, HELENE CYR, ALISON GOLDBERG, AND WILLIAM H. DIETZ. Comparison of high-calorie, low-nutrient-dense food consumption among obese and non-obese adolescents. Obes Res. 1999;7:438-443. Objective: The purpose of this study was to determine whether obese adolescents eat more high-calorie lownutrient-dense foods than non-obese adolescents. Research Methods and Procedures: Using a crosssectional design, 22 non-obese and 21 obese adolescents kept 14-day food records. Records provided estimates of total daily energy intake and caloric intake from five categories of high-calorie, low-nutrient-dense (HC) foods: candy, chips, soda, baked goods, and ice cream. Body composition was determined by "0 dilution and daily energy expenditure by doubly labeled water. Percentage of energy intake reported (%report) was calculated as the ratio of reported energy intake to measured energy expenditure (x 100%). Results: Both groups underreported energy intake, but the percentage reported was significantly greater in the nonobese group (78.2*20.5% non-obese vs. 55.5+21.8% obese, p

Objectives: To determine in non-obese pre-menarcheal girls if bioelectrical impedance (BIA) is a better predictor of body fatness than triceps skinfold (TSF) or body mass index (BMI) and to cross-validate published equations for determination of fat-free mass (FFM) from BIA in pre-menarcheal girls. Design: Cross-sectional analysis of data from 132 non-obese pre-menarcheal girls. The relationship of percent body fat (%BF), derived from isotopic dilution of H 2 18 O to TSF, BMI, and %BF by BIA, calculated from measures of height, weight and resistance was examined by correlation analysis. Setting: Massachusetts Institute of Technology (MIT) Clinical Research Center in Cambridge, MA, USA. Subjects: Pre-menarcheal girls aged 8±12 y were recruited from local schools, MIT summer day camp and by word of mouth. Results: TSF accounted for 68% of the explained variance (R 2 ) in the prediction of %BF measured by H 2 18 O, compared to 38% for BMI and 70% for BIA. Prediction of FFM by comparison of published equations was evaluated in this population. The predictive ability differed by Tanner stage. Kushner's equation (Kushner et al, 1992), based solely on height 2 /resistance was the only equation that provided estimates that did not differ signi®cantly from measured values among all Tanner stages.Conclusions: BIA appears to be a valid and reliable measure of FFM but is no better than TSF in predictions of body fat. Sponsorship: Studies were conducted at the General Clinical Research Center at MIT and were supported by NIH grants RR-00088, DK-HD50537, and DK46200.