Low Plasma Adiponectin Concentrations Do Not Predict Weight Gain in Humans

RESEARCH DESIGN AND METHODS

Subjects.

A total of 219 Pima (or closely related Tohono O’odham) Indians (Table 1) were selected from two different ongoing longitudinal studies.

Ninety-eight subjects were selected from a longitudinal study of the pathogenesis of obesity and type 2 diabetes (study 1) performed at the Clinical Diabetes and Nutrition Section of the National Institutes of Health. After a baseline examination, which included weight and waist and thigh circumference measurements and a 75-g oral glucose tolerance test (OGTT), subjects were invited back at approximately annual intervals for repeated weight measurements and OGTTs. The group of subjects with adiponectin measurements was originally selected for a cross-sectional study designed to evaluate the relationships between plasma adiponectin concentrations, adiposity, and insulin sensitivity (3). All subjects were on weight-maintaining diets at least for 2 days, nonsmokers, and healthy according to a physical examination and routine laboratory tests. Subjects were nondiabetic (11) at baseline and follow-up examinations.

One hundred twenty-one subjects were selected from the longitudinal study of diabetes and its complications (study 2) conducted in the Gila River Indian Community. Residents of this community are invited to participate every 2 years, regardless of their health status. Each participant undergoes a physical examination, which includes height, weight, and waist and thigh circumference measurements and 75-g OGTTs. The group of subjects with adiponectin measurements was originally selected for a case-control series designed to evaluate the prospective relationship between plasma adiponectin concentrations and the risk of development of type 2 diabetes (9). The subjects had normal glucose tolerance at baseline (fasting plasma glucose <110 mg/dl and 2-h plasma glucose <140 mg/dl) and were nondiabetic at follow-up (11).

The protocols were approved by the Tribal Council of the Gila River Indian Community and by the Institutional Review Board of the National Institute of Diabetes and Digestive and Kidney Diseases. All subjects provided written informed consent before participation.

Methods.

After a 12-h overnight fast, all subjects underwent a 75-g OGTT for assessment of glucose tolerance according to World Health Organization diagnostic criteria (11). BMI was calculated as the ratio of weight (in kilograms) divided by height (in meters) squared. Waist and thigh circumferences were measured at the umbilicus in the supine position and at the gluteal fold in the standing position. The waist-to-thigh ratio (WTR) was calculated as an index of body fat distribution.

In study 1, plasma adiponectin concentrations were measured from fasting plasma samples drawn during OGTT with prechilled syringes, transferred into prechilled EDTA tubes, and immediately placed on ice. All tubes were cold-centrifuged (4°C) within several minutes of collection and stored at −70°C until assayed. Plasma glucose concentrations were determined by the glucose oxidase method (Beckman Instruments, Fullerton, CA). Body composition was estimated by underwater weighing, with simultaneous determination of residual lung volume by helium dilution (12) or by total-body dual-energy X-ray absorptiometry (DPX-L; Lunar Radiation, Madison, WI) with calculations of percent body fat, fat mass, and fat-free mass as previously described (13).

In study 2, plasma adiponectin concentrations were measured from 2-h samples during OGTT. Blood samples were drawn and stored at room temperature until centrifugation (within 3 h). Samples were chilled to −20°C and stored at −70°C. Plasma glucose concentrations were determined by potassium ferrocyanide (Technicon) or hexokinase method (Chiron).

All samples were assayed for adiponectin at the Department of Internal Medicine and Molecular Sciences, Osaka University, Japan. Plasma adiponectin concentrations were determined using a validated sandwich enzyme-linked immunosorbent assay employing an adiponectin-specific antibody (intra-assay and interassay coefficients of variation, 3.3 and 7.4%, respectively). Because plasma adiponectin concentrations show no diurnal change (5) and a pilot study indicated no significant differences in adiponectin concentration in plasma samples taken fasting or 2 h after an OGTT (2 h, 96% of fasting; absolute difference, 0.26 ± 0.3 μg/ml; paired t test, P = 0.44; unpublished data, R.S.L., J.K., W.C.K.), plasma adiponectin measurements from studies 1 and 2 were pooled together.

Statistical analyses.

Statistical analyses were performed using software of the SAS Institute (Cary, NC). Results are given as means ± SD. Fasting plasma adiponectin concentrations were logarithmically transformed to approximate a normal distribution.

The relationship between plasma adiponectin concentrations and body weight was examined by calculating Pearson correlation coefficients. The relationships between plasma adiponectin concentrations at baseline and changes in weight and/or other measures of adiposity (follow-up adjusted for baseline) were examined by multiple linear regression models to allow adjustment for covariables. Differences in anthropometric measurements between the visits were assessed by unpaired Student’s t test.

RESULTS

The anthropometric characteristics and plasma adiponectin concentrations of the subjects included in the study are summarized in Table 1.

Cross-sectionally, in the whole group, plasma adiponectin concentrations were negatively associated with body weight (r = −0.28, P = 0.0001), BMI (r = −0.38, P = 0.0001), and WTR (r = −0.24, P = 0.0001), but not waist circumference (r = −0.10, P = 0.2). In study 1, percent body fat (r² = 0.25, P = 0.0001) was the only independent determinant of plasma adiponectin concentrations. In study 2, BMI (r² = 0.16, P = 0.0001) and age (r² = 0.01, P = 0.03) were independent determinants of plasma adiponectin concentrations.

Prospectively, in the whole group, plasma adiponectin concentrations at baseline were not associated with change in weight after adjustment for time of follow-up (r = −0.72, P = 0.8) and additional adjustment for age at follow-up and sex (r = − 0.31, P = 0.9). When the term for participation in study 1 or 2 was included in the model, study was not a significant determinant of the weight change after adjustment for age at follow-up, time of follow-up, and sex (P = 0.7). The relationship between plasma adiponectin and weight change did not differ between subjects in studies 1 and 2 after adjustment for age at follow-up, time of follow-up, and sex (P = 0.4 for interaction). Plasma adiponectin concentrations at baseline were not associated with change in BMI after adjustment for age at follow-up, sex, and time of follow-up (P = 0.8). Similarly, plasma adiponectin concentrations at baseline were not associated with change in waist circumference (P = 0.8) or WTR (P = 0.2) after adjustment for age at follow-up, sex, time of follow-up, and baseline weight.

When analysis was performed separately in study 1 and study 2, plasma adiponectin concentrations were not associated with changes in body weight (Fig. 1), BMI, or WTR (data not shown).

In 98 Pima Indians (study 1) who had measurements of percent body fat at baseline and follow-up, plasma adiponectin concentrations at baseline were not associated with change in percent body fat before and after adjustment for age at follow-up, sex, and time of follow-up (both P = 0.9).

Interaction between adiponectin and follow-up time was not significant in any of these models (P > 0.05).

DISCUSSION

In the present study, we found that low plasma adiponectin concentrations do not predict future weight gain.

In mice, administration of adiponectin prevented the development of diet-induced obesity by increasing muscle lipid oxidation (7). We have previously shown that plasma adiponectin concentrations are not associated with fat oxidation in humans (14). In the present study, we showed that plasma adiponectin concentrations at baseline were not associated with subsequent change in weight and/or other measures of adiposity (BMI, WTR, or percent fat). This suggests that physiologic plasma concentrations of adiponectin do not seem to play an etiologic role in development of obesity in humans. Our finding is supported by two recent studies showing that body weight is not increased in adiponectin-deficient mice (15,16). It is still possible, however, that administration of adiponectin may decrease body weight in humans. This cannot be resolved in an observational study.

Other studies indicate that plasma adiponectin concentrations increase after weight reduction. Hotta et al. (5) showed that plasma adiponectin concentrations increased after weight reduction in both obese nondiabetic and diabetic subjects after 2 months on a calorie-restricted diet. Moreover, Yang et al. (6) showed that plasma adiponectin concentrations increased in severely obese subjects following bariatric surgery during an average follow-up of 8 months. These studies together with ours suggest that low plasma adiponectin concentrations are most likely the consequence rather than the cause of obesity.

Low plasma adiponectin concentrations predict decrease in insulin sensitivity and the development of type 2 diabetes in Pima Indians (8,9), a population with a high prevalence of type 2 diabetes (17). Because obesity is an important risk factor for the development of insulin resistance (17) and obesity is associated with low plasma adiponectin concentrations, adiponectin might influence insulin sensitivity via change in adiposity. The present study showed that low plasma adiponectin concentrations are not associated with weight gain in Pima Indians, suggesting that the relationship between plasma adiponectin concentrations and insulin sensitivity and/or type 2 diabetes is mediated by factors other than increased adiposity. Our observation is consistent with the association between low plasma adiponectin levels and insulin resistance in patients with generalized lipodystrophies (18).

Because information about adiponectin variation over time is scarce in humans, a limitation of our study is that our observation is based on a single fasting plasma adiponectin measurement. However, Hotta et al. (5) have shown no meal-related or circadian changes in plasma adiponectin concentrations.

In conclusion, low plasma adiponectin concentrations do not appear to have an etiologic role in development of obesity in Pima Indians.