Adiponectin, the metabolic syndrome and atherosclerosis. Observation and intervention studies

Abstract

Adiponectin (Adn) has beneficial effects on insulin- and glucose metabolism as well as anti-inflammatory properties. High Adn concentrations are associated with reduced risks for insulin resistance and future diabetes, and coronary heart disease. Adn is uniquely produced by adipocytes. Paradoxically, it is negatively correlated to measures of obesity.The main aim of the thesis was to test two hypotheses: (i) circulating levels of Adn are explained by the occurrence and degree of obesity, insulin resistance and inflammation; and (ii) low levels of serum Adn are related to subclinical atherosclerosis, independently of traditional risk factors for cardiovascular disease.A cohort of 64-year women had previously been screened. Women with diabetes mellitus (n=210) were included together with those with impaired glucose tolerance (IGT, n=201) and normal glucose tolerance consisting of two groups: one matched to IGT for BMI and WHR (NGTm, n=97), and one randomly selected and healthy group (NGTr, n=89). These women were examined with B-mode ultrasound for measurement of the intima-media thickness (IMT) in both carotid arteries. A population-based sample of 102 apparently healthy 58-year old men with varying degrees of obesity and insulin sensitivity were examined with euglycemic, hyperinsulinemic clamp. Forty obese men and women with or without the metabolic syndrome (MetS+, MetS-, n=20 in each group) were treated with very low calorie diet for 16 weeks, followed by a weight maintenance diet for 2 weeks. Abdominal subcutaneous adipose tissue biopsies were obtained initially, after 8, 16, and 18 weeks for DNA-microarray analyses of adiponectin and v-SNARE Vti1a mRNA occurrence, the latter related to Adn secretion from the adipocyte. An age- and sex matched reference group was obtained from a population study. Adn was measured with ELISA or radioimmunoassay technique. The results showed that NGTm women had higher Adn than the IGT group and that the obese MetS- group had similar Adn as the non-obese reference group. Further, BMI or WHR did not remain as independent covariates to Adn in multivariate analyses. During diet, weight loss did not correlate to Adn that continued to increase during refeeding far above the levels in the reference group. Hyperglycaemia was not associated with Adn. In all substudies Adn was related to measures of insulin resistance. This could be explained by genetic factors as family history of diabetes was associated with low Adn and indicators of insulin resistance. Inflammatory markers were also associated with low Adn and insulin resistance. However, in multivariate analyses measures of insulin resistance, C-peptide, family history of diabetes, and hs-CRP only explained one third of the variability in Adn. During weight loss changes in Adn and serum insulin did not correlate. The expression of the adiponectin gene or v-SNARE Vti1a did not differ between the groups or change during or after weight loss. Carotid IMT was related to Adn but not independently of waist circumference, serum insulin, or blood glucose.In conclusion, the variability in Adn was dissociated from obesity and could only partially be related to factors such as family history of diabetes, inflammation, and insulin resistance. The subcutaneous adipocyte expression or secretion of Adn were not affected by insulin resistant states or excessive weight reduction indicating post-translational effects. Adn showed only weak associations with subclinical atherosclerosis.