Association of the T-G Polymorphism in Adiponectin (Exon 2) With Obesity and Insulin Sensitivity: Interaction With Family History of Type 2 Diabetes

In the Tübingen Family Study for type 2 diabetes, primarily normal glucose-tolerant individuals with (and without) a family history of type 2 diabetes are recruited and metabolically characterized. Recruitment mechanisms include word-of-mouth, fliers, and newspaper advertisements. In the present analysis, we report data from OGTT (n = 371) and euglycemic-hyperinsulinemic clamps (n = 210; characteristics shown in Tables 1 and 2). In our population, the prevalence of the G allele in adiponectin (exon 2) is ∼28%. The OGTT group includes 29 individuals with impaired glucose tolerance (IGT). This proportion represents the population’s prevalence of IGT (13). To ensure that we did not miss an effect of the polymorphism on glucose tolerance, we left this subgroup in the analysis. The characteristics of the subgroups that underwent the different protocols are shown in Tables 1 and 2. We additionally stratified the groups by presence or absence of family history of diabetes, which was known in 360 individuals (characteristics shown in Tables 3 and 4). Assessment of family history status depended solely on self-reports using a standardized questionnaire.

The genotype distribution of the study population (72.2% TT, 23.7% GT, 4% GG) was essentially in Hardy-Weinberg equilibrium (P = 0.09, χ² test). The statistical trend is due to an overrepresentation of the homozygotes by one or two individuals, who were not related. The protocols were approved by the local ethical committee, and all individuals gave informed written consent.

After a 10-h overnight fast, individuals ingested a solution that contained 75 g dextrose, and venous blood samples were obtained at 0, 30, 60, 90, and 120 min for determination of plasma glucose and plasma insulin. Stages of glucose tolerance were classified according to World Health Organization criteria (14).

After the baseline period, individuals received a primed insulin infusion at a rate of 1.0 mU · kg⁻¹ · min⁻¹ for 2 h as previously described (15). Blood was drawn every 5–10 min for determination of blood glucose, and the infusion rate of exogenous glucose was adjusted appropriately to maintain the baseline glucose level.

Blood glucose was determined using a bedside glucose analyser (glucose-oxidase method; Yellow Springs Instruments, Yellow Springs, CO). Plasma insulin was determined by microparticle enzyme immunoassay (Abbott Laboratories, Tokyo, Japan), and plasma C-peptide was determined by radioimmunoassay (Byk-Sangtec, Dietzenbach, Germany). The T-G polymorphism in exon 2 of the adiponectin gene was determined by sequencing the polymerase chain reaction product (primer sequence as previously described [4]) using an automated sequencer (ABI prism 310, Perkin-Elmer, Foster City, CA).

The insulin sensitivity index (in μmol · kg⁻¹ · min⁻¹ · [pmol/l]⁻¹) for systemic glucose uptake was calculated as mean infusion rate of exogenous glucose necessary to maintain euglycemia during the last 60 min of the standard clamp divided by the steady-state serum insulin concentration.

Data are given as mean ± SE unless otherwise stated. Differences were assessed by analysis of variance for three groups in the OGTT database. In addition, comparison between two groups (individuals heterozygous and homozygous for the polymorphism were pooled, GT + GG) was analyzed using an unpaired Student’s t test on log-transformed data. For analyses in the substantially smaller clamp group only, the pooled (GT + GG) genotypes were used. The nonparametric Wilcoxon test was used for non-normally distributed parameters. For the comparison of subgroups stratified by presence or absence of family history of type 2 diabetes, analysis of variance followed by a Tukey-Kramer test was used. The interaction effect of family history and genotype was assessed using factorial multivariate analysis of variance. P < 0.05 was considered to be statistically significant. The statistical software package JMP (SAS Institute, Cary, NC) was used for the statistical analyses.

In the 371 nondiabetic individuals who underwent the OGTT, we found a significantly greater BMI in GG + GT (25.5 ± 0.7 kg/m²) compared with TT (24.1 ± 0.3 kg/m²; P = 0.02; Table 1). A parallel difference for percentage body fat was observed. Stratification by FHD revealed that the difference observed in the whole population was entirely attributable to even more pronounced differences in the subgroup without FHD, whereas in the subgroup with FHD, the G allele had absolutely no effect (Table 3). Regardless of the genotype, individuals without FHD had a significantly lower BMI than individuals with FHD (25.2 ± 0.4 vs. 26.2 ± 0.5 kg/m²; P = 0.01). It is interesting that the BMI in the GG + GT genotype without FHD (27.0 ± 0.9 kg/m²) was not significantly different than in the whole group with FHD (Fig. 1). The significantly greater basal and postload insulin concentrations in the GT + GG group suggests that the BMI difference, not unexpectedly, leads to differences in insulin sensitivity.

In the subgroup that underwent the hyperinsulinemic-euglycemic clamp (n = 209), the findings regarding obesity parameters were reproducible (Tables 2 and 4). Insulin sensitivity was significantly lower in GG + GT (0.089 ± 0.007 units) compared with TT (0.112 ± 0.005 units; P = 0.02). This statistical difference disappeared completely on adjustment for BMI and percentage body fat. The relationships among family history of diabetes, BMI, insulin sensitivity, and genotype are shown in Fig. 2.

The purpose of the present studies was to examine the effect of the common T-G polymorphism in the adiponectin gene on measures of obesity and insulin sensitivity. The most interesting finding was the difference in BMI, percentage body fat, and insulin sensitivity, especially in the group without FHD. The G allele was significantly associated with obesity. The decreased insulin sensitivity associated with the G allele was entirely accounted for by the difference in BMI. These associations with the G allele were entirely absent in the subgroup with FHD.

Our findings suggest that in individuals without familial predisposition for type 2 diabetes, the adiponectin polymorphism may mildly increase obesity risk and secondarily cause insulin resistance. A possible interpretation for the absence of such an association in the subgroup with FHD could be as follows: the familial predisposition represents a much stronger genetic load than the presence of single polymorphism, with a weak effect on the individual’s phenotype. The latter effect may thus be buried by sum effects of other, unidentified genetic factors. Our observation of a significantly greater BMI in the subgroup with FHD compared with the subgroup without FHD but not compared with the GT + GG (without FHD) cohort supports this hypothesis.

It is unclear how a silent polymorphism (i.e., one not resulting in an amino acid exchange) can affect metabolism and/or alter the risk of obesity. Generally, a polymorphism need not be functionally relevant itself but can be in complete or near-complete linkage disequilibrium with a yet unidentified second polymorphism that has functional relevance, for example in the promoter region. However, sequencing of the promoter region of the human adiponectin gene in reasonably sized populations did not reveal common polymorphisms, let alone one in linkage disequilibrium with the T-G polymorphism in exon 2 (4,10). Alternatively, the polymorphism can be in linkage disequilibrium with polymorphism in introns that, in analogy to the UCSNP-43 polymorphism in calpain-10 (16), would destabilize pre-mRNA and result in reduced mRNA levels. Because intronic regions of the adiponectin gene have not yet been systematically screened for single nucleotide polymorphisms, this scenario remains a possibility. Providing a final answer to this question is beyond the capabilities of our laboratory. The strong linkage disequilibrium with type 2 diabetes, measures of adiposity, and insulin levels found in the chromosomal region where the adiponectin gene is located (11,12) suggests that somewhere on this locus a common genetic variant, such as the T-G polymorphism in the adiponectin gene, must have a measurable effect on obesity, glucose homoeostasis, or both.

It may be of interest that circulating adiponectin concentrations were significantly lower in Japanese carriers of the G allele (4). Although this did not quite reach statistical significance, it may point to a possible link among the GG genotype, reduced circulating adiponectin levels, and reduced insulin sensitivity. A conclusive assessment of these relationships probably requires larger sample sizes and the advent of a commercially available adiponectin assay.

Our findings are clearly at variance with reports from a Japanese (4) and a German population (10) in which no effect of G allele on measures of obesity were observed. Possible explanations for this divergence include differences in family history status by which we had stratified and a selection bias due to recruitment of individuals through an endocrinological outpatient clinic (10).

In conclusion, we found a significantly greater BMI in GG + GT accompanied by reduced insulin sensitivity compared with TT. This difference was attributable entirely to differences in the subgroup without FHD, whereas in the subgroup with FHD, the G allele had absolutely no effect. This suggests that in individuals without FHD, the adiponectin polymorphism may mildly increase the obesity risk (and secondarily insulin resistance). In contrast, in individuals who are already burdened by other genetic factors, this small effect may be very hard or impossible to detect.

This study was supported in part by the Deutsche Forschungsgemeinschaft DFG (to M.S.) and the European Community (QLRT-1999-00674).

We thank all of the research volunteers for their participation. We gratefully acknowledge the excellent technical assistance of our laboratory staff.