Adiponectin, Metabolic Risk Factors, and Cardiovascular Events among Patients with End-Stage Renal Disease

Protocol

The protocol conformed to the ethical guidelines of our institutions, and informed consent was obtained from each participant. All studies were performed on nondialysis days, between 8 a.m. and 10 a.m.

Study Cohort

Two hundred twenty-seven hemodialysis patients (125 male and 102 female patients, all Caucasian) who had been receiving regular dialysis treatment (RDT) for at least 6 mo and did not exhibit clinical evidence of heart failure (defined as dyspnea in addition to two of the following conditions: increased jugular pressure, bibasilar crackles, pulmonary venous hypertension, or interstitial edema in chest x-rays, requiring hospitalization or extra ultrafiltration) were eligible for the study. The demographic, clinical, and biochemical characteristics of these patients are detailed in Table 1. These patients were recruited between January 1997 and May 1998 and represented approximately 70% of the entire hemodialysis population of four dialysis units. The remaining 30% of the patients were excluded because of the presence of circulatory congestion or major infections (20%) or because of hospitalization for treatment of intercurrent illnesses or logistic reasons/unwillingness to participate in the study (10%). Patients were being treated thrice-weekly with standard bicarbonate hemodialysis (Na⁺, 138 mM; HCO₃⁻, 35 mM; K⁺, 1.5 mM; Ca²⁺, 1.25 mM; Mg²⁺, 0.75 mM) or high-flux hemodialysis with either cellulose or semisynthetic membranes (dialysis filter surface area, 1.1 to 1.7 m²). Dry weight was targeted in each case to achieve a normotensive, edema-free state. Eighty-five patients were habitual smokers (21 ± 16 cigarettes/d). One hundred twenty-three patients were receiving erythropoietin treatment. Eighty-two patients were receiving antihypertensive drugs (58 received single therapy with angiotensin-converting enzyme inhibitors, AT₁ receptor antagonists, calcium channel blockers, or α- or β-blockers, and 24 received double or triple therapy with various combinations of these drugs).

Follow-Up Study

After the initial assessment, patients were monitored for 31 ± 13 mo (range, 1 to 45 mo). During the follow-up period, cardiovascular events (ECG documented anginal episodes or myocardial infarctions, heart failure, ECG documented arrhythmia, transient ischemic attacks, or strokes, peripheral vascular disease, or major arterial or venous thrombotic episodes, except arteriovenous fistula thrombosis) were accurately recorded. Each death was reviewed and assigned an underlying cause by a panel of five physicians. As part of the review process, all available medical information regarding each death was collected. This information always included study and hospitalization records. In the case of an out-of-hospital death, family members were interviewed by telephone, for better assessment of the circumstances surrounding the death.

Laboratory Measurements

Fasting blood samples for analysis of the levels of lipids, albumin (bromocresol green), hemoglobin, calcium, phosphate, insulin, glucose, C-reactive protein (CRP), von Willebrand factor, and homocysteine were obtained from all patients on a midweek nondialysis day. Plasma homocysteine levels were determined with a HPLC method based on ammonium-7-fluorobenzo-2-oxa-1,3-diazole-4-sulfonate fluorescence derivatization (10). CRP levels were measured by using a commercially available kit (Behring Scoppito, l’Aquila, Italy). Serum insulin levels were measured by using a RIA kit (Sorin Saluggia, Vercelli, Italy). Insulin sensitivity was estimated by using the homeostatic model assessment (HOMA-R) index [i.e., plasma glucose level × (plasma insulin level/22.5)], which was validated with the euglycemic-hyperinsulinemic clamp method (11). Plasma leptin measurements were made in a single assay, using a sensitive RIA developed at Linco Research Laboratories (St. Louis, MO). The assay was based on a polyclonal antibody raised in rabbits against highly purified recombinant human leptin. The recovery of leptin with the RIA ranged from 99 to 105% in the range of 0.5 to 100 μg/L. The intra-assay coefficient of variation ranged from 3.4 to 8.3%. Plasma ADPN concentrations were measured by using a sensitive enzyme-linked immunosorbent assay (7), in the laboratory of the Department of Internal Medicine and Molecular Science of the Medical School of Osaka (Osaka, Japan).

The reference group consisted of 15 healthy, normotensive, Caucasian subjects matched to dialysis patients with respect to gender (six male and nine female patients versus 125 male and 102 female patients) and body mass index (BMI) (24.9 ± 1.3 kg/m² versus 24.5 ± 4.3 kg/m²).

Echocardiography

All echocardiographic measurements were performed according to the recommendations of the American Society of Echocardiography (12). Left ventricular mass (LVM) was calculated according to the formula described by Devereux et al. (13) and was indexed to height^2.7 (LVM index) (14). The height-based indexing of LVM was specifically chosen to minimize potential distortions attributable to extracellular volume expansion (with surface area indexing being weight-sensitive). Left ventricular hypertrophy (LVH) was defined as a LVM index of >47 g/m^2.7 for women or >50 g/m^2.7 for men.

Statistical Analyses

Data are reported as mean ± SD or as median and interquartile range, as appropriate. Between groups, comparisons were made with the t test or the Mann-Whitney test. Relationships between plasma ADPN levels and other covariates were analyzed by using the least-squares method. Data that did not demonstrate a Gaussian distribution were logarithmically transformed.

Time-to-event analysis of cardiovascular outcomes was performed by using the Kaplan-Meier method. The independent prognostic power of plasma ADPN levels for death or fatal or nonfatal cardiovascular events was analyzed by using the Cox proportional-hazards model. Variables with independent effects on survival rates were identified by constructing hierarchical models, which included Framingham risk factors [age, gender, smoking (yes/no), diabetes mellitus, systolic pressure, serum cholesterol levels, LVH, and previous cardiovascular events], factors peculiar to ESRD (duration of RDT, serum albumin levels, hemoglobin levels, and calcium·phosphate product values), and emerging risk factors (serum CRP and plasma homocysteine levels). Hazard ratios (HR) and their 95% confidence intervals (CI) were calculated with the use of the estimated regression coefficients and their standard errors in the Cox regression analysis. All calculations were performed by using a standard statistical package (SPSS for Windows, version 9.0.1; SPSS, Chicago, IL).

Plasma ADPN Levels

The main clinical and biochemical characteristics of the study population are summarized in Table 1. Plasma ADPN levels among dialysis patients (15.0 ± 7.7 μg/ml) were approximately 2.5 times higher (P < 0.0001) than the average value for healthy subjects (6.3 ± 2.0 μg/ml) and exceeded the upper limit of the normal range (cut-off, >10.0 μg/ml) for the majority of dialysis patients (170 of 227 patients, i.e., 75%). Among dialysis patients, plasma ADPN levels were higher (P = 0.03) for women (15.2 ± 7.9 μg/ml) than for men (14.0 ± 7.4 μg/ml), and this difference remained significant after data adjustment for BMI (P = 0.01). The between-gender difference in plasma ADPN levels among dialysis patients was similar to that observed for healthy control subjects (men, 5.2 ± 1.8 μg/ml; women, 7.1 ± 1.9 μg/ml).

Biochemical and Clinical Correlates of Plasma ADPN Levels among Dialysis Patients

Plasma ADPN levels were strongly and inversely related to BMI and plasma leptin levels (Figure 1), as well as to insulin levels and HOMA-R indices among both men and women (Figure 2) and serum glucose levels among men only (r = −0.27, P = 0.002). Plasma ADPN levels were also directly related to plasma HDL cholesterol levels and were inversely related to serum triglyceride levels; such associations were again apparent for both genders (Figure 3). Slight inverse correlations were observed between ADPN and von Willebrand factor levels (men, r = −0.17, P = 0.098; women, r = −0.19, P = 0.04). Plasma ADPN levels were slightly related to Kt/V (r = 0.15, P = 0.02). Plasma ADPN levels were lower (P < 0.05) among diabetic women (11.9 ± 1.7 μg/ml) than among nondiabetic women (15.1 ± 1.6 μg/ml), whereas they were identical among diabetic (12.0 ± 1.8 μg/ml) and nondiabetic (12.3 ± 1.7 μg/ml) men. Overall, plasma ADPN levels were unrelated to age (P = 0.21), smoking (P = 0.13), duration of RDT (P = 0.15), mean arterial pressure (P = 0.75), heart rate (P = 0.30), serum albumin levels (P = 0.07), CRP levels (P = 0.17), or LDL cholesterol levels (P = 0.48).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Relationships between plasma adiponectin (ADPN) levels and body mass index (BMI) and plasma leptin levels (•, male patients; ○, female patients). The regression lines are presented separately for male patients (——) and female patients (– – –).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Relationships between plasma ADPN levels and plasma insulin levels and homeostatic model assessment (HOMA-R) indices (•, male patients; ○, female patients). The regression lines are presented separately for male patients (——) and female patients (– – –).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Relationships between plasma ADPN levels and serum HDL cholesterol and triglyceride levels (•, male patients; ○, female patients). The regression lines are presented separately for male patients (——) and female patients (– – –).

Plasma ADPN Levels and Drug Therapy

Plasma ADPN levels were higher (P = 0.006) among patients who were receiving treatment with erythropoietin (16.3 ± 8.2 μg/ml) than among the remaining patients (13.5 ± 6.9 μg/ml), and this difference remained significant after data adjustment for gender and BMI (P = 0.04). Antihypertensive treatment had no effect on plasma ADPN levels, either among women (treated, 16.0 ± 6.9 μg/ml; untreated, 16.4 ± 8.5 μg/ml; P = 0.77) or among men (treated, 13.9 ± 7.7 μg/ml; untreated, 14.1 ± 7.3 μg/ml; P = 0.84).

Plasma ADPN Levels, Survival Rates, and Cardiovascular Events

Overall, 78 patients died, 50 of them (64%) as a result of cardiovascular causes (Table 2). Plasma ADPN levels failed to predict overall mortality rates, which were explained by serum cholesterol levels (HR, 1.01; 95% CI, 1.00 to 1.01; P = 0.001), male gender (HR, 2.84; 95% CI, 1.47 to 5.46; P = 0.002), previous cardiovascular events (HR, 2.49; 95% CI, 1.41 to 4.42; P = 0.002), age (HR, 1.04; 95% CI, 1.01 to 1.06; P = 0.003), serum albumin levels (HR, 0.42; 95% CI, 0.24 to 0.75; P = 0.003), serum CRP levels (HR, 1.01; 95% CI, 1.00 to 1.02; P = 0.01), diabetes mellitus (HR, 2.28; 95% CI, 1.13 to 4.59; P = 0.02), and LVH (HR, 2.66; 95% CI, 1.02 to 6.94; P = 0.04).

Ninety-five cardiovascular events (fatal or nonfatal) occurred during the follow-up period (Table 2). Plasma ADPN levels were lower (P < 0.05) among patients who experienced new cardiovascular events (13.7 ± 7.3 μg/ml), compared with event-free patients (15.8 ± 7.8 μg/ml). In the unadjusted analysis (Table 3), there was a 3% risk reduction for each 1 μg/ml increase in plasma ADPN levels and the relative risk of adverse cardiovascular events was 1.56 (95% CI, 1.12 to 1.99) higher for patients in the first ADPN tertile than for those in the third tertile (Figure 4). The point estimate remained almost unchanged after adjustment for Framingham risk factors (including echocardiographically detected LVH and previous cardiovascular events), factors peculiar to ESRD, and emerging risk factors such as serum CRP levels and plasma homocysteine levels (Table 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4. Kaplan-Meier survival curves for cardiovascular events (fatal and nonfatal) in the study cohort. Patients were stratified into three tertiles according to plasma ADPN concentrations (first tertile, ADPN levels of <11.1 μg/ml; second tertile, >11.1 to <16.4 μg/ml; third tertile, ≥16.4 μg/ml).