Extracellular vesicle markers in relation to obesity and metabolic complications in patients with manifest cardiovascular disease

Abstract

Background: Alterations in extracellular vesicles (EVs), including exosomes and microparticles, contribute to cardiovascular disease. We hypothesized that obesity could favour enhanced release of EVs from adipose tissue, and thereby contribute to cardiovascular risk via obesity-induced metabolic complications. The objectives of this study were: 1) to investigate the relation between the quantity, distribution and (dys) function of adipose tissue and plasma concentrations of atherothrombotic EV-markers; 2) to determine the relation between these EV-markers and the prevalence of the metabolic syndrome; and 3) to assess the contribution of EV markers to the risk of incident type 2 diabetes.

Methods: In 1012 patients with clinically manifest vascular disease, subcutaneous and visceral fat thickness was measured ultrasonographically. Plasma EVs were isolated and levels of cystatin C, serpin G1, serpin F2 and CD14 were measured, as well as fasting metabolic parameters, hsCRP and adiponectin. The association between adiposity, EV-markers, and metabolic syndrome was tested by multivariable linear and logistic regression analyses. As sex influences body fat distribution, sex-stratified analyses between adipose tissue distribution and EV-markers were performed. The relation between EV-markers and type 2 diabetes was assessed with Cox regression analyses.

Results: Higher levels of hsCRP (β 5.59; 95% CI 3.00-8.18) and lower HDL-cholesterol levels (β-11.26; 95% CI -18.39 - -4.13) were related to increased EV-cystatin C levels, and EV-cystatin C levels were associated with a 57% higher odds of having the metabolic syndrome (OR 1.57; 95% CI 1.19-2.27). HDL-cholesterol levels were positively related to EV-CD14 levels (β 5.04; 95% CI 0.07-10.0), and EV-CD14 levels were associated with a relative risk reduction of 16% for development of type 2 diabetes (HR 0.84, 95% CI 0.75-0.94), during a median follow up of 6.5 years in which 42 patients developed type 2 diabetes.

Conclusions: In patients with clinically manifest vascular disease, EV-cystatin C levels were positively related, and EV-CD14 levels were negatively related to metabolic complications of obesity.

Figure 1

Relation between visceral fat, subcutaneous fat, waist circumference and BMI and EV-markers in males and females with manifest cardiovascular disease. Beta regression coefficients (β) with 95% confidence interval (CI) indicates the difference in log EV-cystatin C, log EV-serpin G1, square root EV-serpin F2 or log EV-CD14 levels per unit increase in AT parameter, adjusted for age, current smoking, eGFR, type 2 diabetes, blood pressure lowering medication, lipid lowering medication and year of inclusion in SMART.

Figure 2

Relation between metabolic parameters of adipose tissue (dys) function and EV-markers in patients with manifest cardiovascular disease.β with 95% CI indicates the difference in log EV-cystatin C, log EV-serpin G1, square root EV-serpin F2 or log EV-CD14 levels per unit increase in log hsCRP, log adiponectin, log HOMA-IR or HDL-cholesterol, adjusted for age, sex, current smoking, eGFR, type 2 diabetes, blood pressure lowering medication, lipid lowering medication, platelet aggregation inhibitors and year of inclusion in SMART. AdipoQ: adiponectin.

Figure 3

Relation between EV-markers and metabolic syndrome in patients with manifest cardiovascular disease. Odds ratios with 95% CI indicate the odds for metabolic syndrome per increase in log EV-cystatin C, log EV-serpin G1, square root EV-serpin F2 or log EV-CD14 concentration, adjusted for age, sex, current smoking, eGFR and year of inclusion in SMART.

Figure 4

EV-markers and the risk of new onset type 2 diabetes in patients with manifest cardiovascular disease. Hazard ratios with 95% CI indicate the relative risk for incident type 2 diabetes per increase in EV-cystatin C, EV-serpin G1, EV-serpin F2 or EV-CD14 concentration during 6.5 years (interquartile range 5.8–7.1 years) follow-up, adjusted for age, sex, current smoking, hsCRP, eGFR and year of inclusion in SMART.