Relationships between plasma leptin levels and body composition parameters measured by different methods in postmenopausal women

Abstract

The aim of this study was to determine the effects of body composition measured by different methods with different measurement errors on fasting plasma leptin level in normal body mass and obese postmenopausal women. It was hypothesized that the relationship between plasma leptin concentration and body fat is higher using more sophisticated laboratory methods (dual energy X-ray absorptiometry, DXA) in comparison with field methods (bioelectrical impedance analysis, BIA, or skinfold thickness) for body fat measurement because of the greater precision of DXA measurements. Thirty-five postmenopausal (55-83 years of age) healthy Estonian women were divided into two groups: BMI < 27 kg/m(2) as non obese (n = 18) and BMI> 27 kg/m(2) as obese (n = 17). Body composition was determined using DXA (total body, arms, legs, and trunk fat percent, fat mass, and LBM) and BIA methods. Body fat percent was significantly higher using the DXA method. Subcutaneous adipose tissue distribution was determined by measuring nine skinfold thicknesses. Body fat distribution was defined as the ratio of waist-to-hip (WHR) and waist-to-thigh (WTR) circumferences. Leptin was determined by means of radioimmunoassays. Leptin concentration was not significantly different between groups (19.0 +/- 13.3 and 21.5 +/- 21.5 ng/ml in non obese and obese groups, respectively). Body fat percent and fat weight measured by DXA or BIA methods and all measured skinfold thickness values, except biceps and abdominal, were higher in obese women. Body height did not correlate significantly with leptin concentrations. The relationships between leptin concentration were highest with body weight (r = 0.67) and BMI (r = 0.73) values in the obese group. All measured body fat parameters using DXA or BIA methods correlated significantly with plasma leptin concentration in the obese group. LBM did not influence the leptin concentration in postmenopausal women. Stepwise multiple regression analysis indicated that the body fat percent measured using the DXA method was highly related to plasma leptin concentration in the obese group (63.2%; R(2) x 100). When absolute fat mass parameters were considered, leptin concentration was related to the mass of arms fat tissue in the obese group of women (62.3%). Body fat percent measured by BIA was highly related to plasma leptin concentration in the obese group (63.3%). Only biceps skinfold thickness was related to leptin concentration (22.5% and 58.9%, in the nonobese and obese groups, respectively) from the nine measured skinfold thicknesses. WHR and WTR did not reflect leptin concentration in different groups of postmenopausal women. It was concluded that different methods of body composition estimation generate different correlations with plasma leptin concentration. Body fat percent and especially fat mass measured by DXA are the main predictors relating to plasma leptin concentration in obese, but not in nonobese, postmenopausal women. In addition, fat mass in arms measured by DXA and biceps skinfold thickness were also highly related to leptin concentration.