Adiponectin levels do not predict clinical onset of type 1 diabetes in antibody-positive relatives

Background and Aims: Insulin resistance has been proposed as risk factor for type 1 diabetes. We investigated whether decreased circulating levels of adiponectin - an abundant insulin sensitizer - can serve as additional predictive marker for type 1 diabetes in first degree relatives of known patients.
Materials and Methods: Adiponectin was measured yearly (ELISA, Linco) in 211 persistently islet antibody-positive (Ab+; positive for IAA, ICA, IA-2A and/or GADA at entry and in a repeat sample) first degree relatives (0-39 years) of type 1 diabetic patients and in 211 age- and sex-matched persistently antibody-negative (Ab-) relatives. Adiponectin levels were correlated with demographic and anthropometric data, antibody status, random proinsulin/C-peptide (PI/C) ratio and HLA-DQ genotype. During follow-up (median [interquartile range]: 81 [59-90] months), 37 Ab+ relatives developed type 1 diabetes.
Results: In the group of 422 relatives, baseline adiponectin correlated inversely with age and body mass index and was lower in males than in females especially after age 15 (p <0.001). There was no correlation with antibody status or later development of diabetes. In 24 Ab+ relatives sampled fasted, adiponectin levels correlated significantly with homeostasis model assessment of insulin sensitivity (HOMA%S; p = 0.006). For 28 prediabetic relatives with at least 2 blood samples before clinical onset, adiponectin levels adjusted for age and sex decreased prior to diagnosis (p = 0.001) whereas no decrease occurred in 28 age- and sex-matched Ab+ non-prediabetic relatives followed over a similar interval. Adiponectin increased again significantly after start of insulin therapy (p = 0.003) in those relatives for whom a sample after diagnosis was available (n = 13). In Ab+ relatives (n = 211), a marked decrease in age- and sex-adjusted adiponectin levels during the first year after study entry could not predict type 1 diabetes nor complement risk assessment based on islet antibodies, HLA-DQ genotype and pancreatic hormones in Cox regression analysis.
Conclusion: Adiponectin levels decrease shortly before clinical onset of type 1 diabetes, possibly reflecting insulinopenia, but do not contribute to disease prediction in Ab+ relatives.