MACRONUTRIENT INTAKE IN RELATION TO BODY MASS INDEX: TEHRAN LIPID AND GLUCOSE STUDY (TLGS)

Human studies investigating the relationship between obesity and diet composition have failed to achieve consistent findings. This study was undertaken to assess the relationship between macronutrient intake and body mass index (BMI) in a group from Tehran, Iran. A total number of 1,290 subjects over 10 years of age (565 males and 725 females) were selected randomly for dietary survey from the participants of Tehran Lipid and Glucose Study (TLGS). Anthropometric indices were measured according to standard protocols and BMI was calculated. Dietary data were collected by trained interviewers using two nonconsecutive, 24- hour dietary recalls. Under- and over-reporting of energy intake were defined as energy intake to BMI < 1.35 and ≥ 2.4, respectively. Calorie-adjusted amounts of macronutrients were calculated by the residual method, following which energy intakes from all calorie-adjusted macronutrients, were simultaneously included in the multiple regression model controlling for confounding variables and mutual effects of macronutrients. Total energy intake was not included to avoid co-linearity. After excluding under- and over-reporters, BMI did not change in males but decreased in females. Misreporting of energy intake increased with age and BMI. Excluding misreporters caused significant changes in actual amounts of macronutrient intakes of women, but not of men. In all cases, adjusting for energy caused a significant increment in the amounts of protein. Controlling for confounding variables, energy intake from fat was positively associated with BMI in males and females (β = 0.61, p < 0.01 and β = 0.75, p < 0.01, respectively). This relationship remained after excluding misreporters (β = 0.66, p < 0.01 and β = 0.78, p < 0.01, respectively). No association was seen between energy intake from protein and carbohydrate with BMI in subjects before and after excluding misreporters. We have found evidence indicating that energy from fat is positively associated with obesity. No other association was observed between energy from protein and carbohydrates with BMI.