Factors predicting body composition in postmenopausal women

Abstract

Factors predicting body composition were determined in 52 postmenopausal caucasian women. Body composition was measured by performing a total body scan using dual energy x-ray absorptiometry (DXA). Height and weight were measured, and body mass index (BMI) was calculated. Body fat distribution was assessed by waist-to-hip ratio (WHR) and abdominal fat weight (between the iliac crest and L1) as determined by DXA. Hormones including total estradiol and testosterone and sex hormone binding globulin (SHBG) were determined by radioimmunoassay. Free androgen index (FAI) was calculated from total testosterone and SHBG. A health history questionnaire and 3 day diet record were completed by each subject. Stepwise multiple regression was performed to determine factors significantly associated with the dependent variables BMI, body weight, total fat weight, lean tissue weight, WHR, and abdominal fat weight. Age, number of children, total months breastfeeding, number of years estrogen deficient; caloric and macronutrient intakes, number of hours per week in aerobic exercise, and hormone concentrations were used as independent variables. SHBG was the only significant (negative) predictor of BMI and body weight, SHBG and caloric intake were significant (negative) predictors of total fat weight, SHBG and estradiol concentrations were significant (negative and positive, respectively) predictors of WHR, SHBG and carbohydrate intake were significant (negative) predictors of abdominal fat weight, and years estrogen deficient and SHBG were significant (negative) predictors of lean tissue weight, all at p < .05. In conclusion, SHBG seemed to be the most consistent predictor of body composition, and macronutrient intake was not a significant predictor of body composition except for caloric intake for total fat weight, and carbohydrate intake for abdominal fat weight in a negative manner in both cases.