bs_bs_banner
Therapeutic Apheresis and Dialysis 2015; 19(2):144–153 doi: 10.1111/1744-9987.12239 © 2014 The Authors Therapeutic Apheresis and Dialysis © 2014 International Society for Apheresis
Adipokines in Peritoneal Dialysis: Relevant Clinical Impact According to Body Composition Ana Paula Bernardo,1,2 Isabel Fonseca,1,2 José Carlos Oliveira,3 Olivia Santos,1 Maria João Carvalho,1 António Cabrita,1 and Anabela Rodrigues1,2 1
Nephrology Department, 3Clinical Pathology Department, Santo António Hospital, Oporto Hospital Center and 2UMIB/ICBAS/Porto University, Oporto, Portugal
Abstract: Adipokines impact on clinical outcomes is not adequately addressed in peritoneal dialysis (PD). We investigated the impact of leptin/adiponectin ratio (L/A) as a predictor of cardiovascular events (CVE) in PD, taking into consideration patient’s body composition and the potential role of glucose load.We prospectively followed 66 prevalent PD patients for 47.0 ± 28.2 months. New CVE were evaluated. Lean tissue index (LTI), relative fat mass (relFM) and relative overhydration (relOH) using multifrequency bioimpedance (BCM) were assessed; serum lipids, interleukin-6 (IL-6), leptin and adiponectin were measured. We established the determinants of L/A using multiple linear regression and the impact of L/A on CVE. Obesity was present in 47 (73.4%) patients according to relFM, and in seven (10.6%) according to body mass index (BMI). Leptin and L/A exhibited a stronger correlation with relFM (both r = 0.62, P < 0.0001) than with BMI (r = 0.46 and r = 0.51, respectively, both P < 0.0001). L/A showed a significant correlation with triglycerides (r = 0.41,
P = 0.001) and HDL-cholesterol (r = −0.358, P = 0.003), better than isolated leptin or adiponectin. RelFM (RR = 0.130, 95%confidence interval [CI]:0.086–0.174, P < 0.0001) and LTI (RR = 0.194, 95%CI:0.037–0.351, P = 0.016) were independent predictors of L/A (R2 = 0.67). Patients who suffered new CVE were older (59.12 ± 12.41 vs. 47.52 ± 13.84years, P = 0.003) and had a higher relOH (11.28 ± 7.29 vs. 6.60 ± 8.16%, P = 0.028). L/A was significantly higher in patients with CVE[2.29 (1.79) vs. 0.65 (1.73), P = 0.028] but this association was only put on evidence after excluding patients with wasting. BMI is an inaccurate method to classify obesity in PD since it underestimates its prevalence compared with body composition assessment using BCM. High adiponectin and low leptin are associated with a more favorable metabolic risk profile in PD. The L/A is determined by relFM and by LTI. A higher L/A is associated with CVE in PD patients without wasting. Key Words: Adipokines, Cardiovascular disease, Obesity, Peritoneal dialysis.
As in the general population (1,2), overweight is widely prevalent in peritoneal dialysis (PD) patients (3–5). Being a risk factor for cardiovascular disease (CVD) and death in general population (6), the clinical impacts of obesity and overweight are uncertain in PD patients (7–9). These conflicting results are partially explained by the use of body mass index (BMI) to characterize obesity. BMI is imprecise, since it does not distinguish lean body mass, fat mass and extracellular water. This characterization, being important in the general population, is even more crucial in chronic kidney disease (CKD) since those
patients can present simultaneous changes in nutritional status, body fat content and hydration status (10) that have different impacts on outcome. Adipose tissue is now recognized as a new endocrine organ, since biologically active molecules secreted by adipocytes, like leptin and adiponectin, are important determinants of inflammation, atherosclerosis and insulin-resistance (11,12). Although in the general population, high plasma leptin and low plasma adiponectin have been demonstrated to be risk factors for CVD (13–15), in CKD patients, similar evidence is less clear, as there are contradictory results about the impact of adiponectin and leptin on cardiovascular outcomes (16–21). These opposing results could be partially explained by differences in nutritional status and body composition of the studied populations that were not taken into consideration.
Received April 2014; revised May 2014. Address correspondence and reprint requests to Dr Ana Paula Bernardo, MD, Largo Prof.,Abel Salazar, 4099-001 Porto, Portugal. Email: [email protected]
144
Leptin/Adiponectin Ratio in Peritoneal Dialysis Leptin and adiponectin have opposite effects on the cardiovascular system and, in line with this view, two recent studies, in obese and type 2 diabetic patients without CKD, have established that leptin/ adiponectin ratio (L/A) correlates better with pulse wave velocity and carotid intima-media thickness than plasma leptin and adiponectin alone (22,23). This preliminary observation suggests that L/A is a new atherogenic index, and that patients with a higher ratio could be prone to developing cardiovascular events. To our knowledge, only one study demonstrated an association between L/A and mortality in a group of non-diabetic PD patients (24). However, this study was unable to assess the patient’s body composition and it did not address the potential role of glucose load and peritoneal membrane transport characteristics on adipokines profile. Low plasma leptin and high plasma adiponectin can both reflect a protein wasting condition (21,25), a diagnosis that cannot be ruled out with a simple BMI measurement, and which itself impacts patient outcome. In order to determine the impact of adipokines on CKD patient’s clinical outcomes, studies must give detailed information about patient’s body composition. With the purpose to overcome this knowledge limitation, we performed a prospective study to investigate the impact of L/A as a predictor of cardiovascular events in PD patients, taking into consideration the nutritional status and body composition. We characterized body composition using multifrequency bioimpedance (BCM, Fresenius Medical Care, Bad Homburg, Germany) and we further explored the potential role of glucose load and peritoneal membrane small solute transport status on adipokines and obesity profile. PATIENTS AND METHODS Patients This prospective observational study enrolled 66 PD patients attending the Centro Hospitalar do Porto—Peritoneal Dialysis Unit. All patients were treated with low glucose degradation product solutions. As a standard, only 1.36% and 2.27% glucose solutions were used. Hypertonic exchanges were used only by exception and for a short period of time. All patients provided informed consent and the study had been approved by Centro Hospitalar do Porto Ethics Committee. Study design Patients were prospectively followed from December 2008 until death, renal transplantation, transfer to © 2014 The Authors Therapeutic Apheresis and Dialysis © 2014 International Society for Apheresis
145
hemodialysis, or end of study (July 2013). Mean follow-up time was 47.0 ± 28.2 months. The primary endpoint was cardiovascular event. Transient ischemic attack and stroke were considered as cerebrovascular events. Coronary artery event was diagnosed when the patient had a myocardial infarction or an unstable angina episode. Peripheral artery disease was diagnosed in the presence of acute limb ischemia, major arterial thrombotic episode, or in the presence of intermittent claudication and significant arterial stenosis on angiography. Sudden death, in a patient with previously known coronary artery disease, was also considered as a cardiovascular event. Laboratory measurements Fasting venous blood samples were obtained at 9 am during a peritoneal equilibration test (PET). The preceding overnight dwell was standardized to 1.36% glucose solution. Serum was allowed to clot at room temperature and then was separated from cells within 60 min and stored at −70°C until analysis for adiponectin, leptin and interleukin-6 (IL-6). Serum total cholesterol, high-density lipoprotein (HDL) cholesterol and triglycerides were measured with an autoanalyzer (COBAS Integra 800, Roche Diagnostics GmbH, Berlin, Germany). Plasma leptin and adiponectin were measured using enzyme-linked immunosorbent assays (Mediagnost, Reutlingen, Germany). IL-6 was measured using enzyme-linked immunosorbent assay (Diasource, Louvain-La-Neuve, Belgium). Dialysis adequacy, urea kinetics, peritoneal glucose absorption and estimation of dietary protein intake Adequacy of dialysis was calculated from a 24-h urine and dialysate collection. Weekly Kt/V was determined using standard methods (26). Residual glomerular filtration rate was calculated as the average of 24-h urinary urea and creatinine as described elsewhere (27). Small solute peritoneal membrane transport was calculated from a 4-h, 3.86% glucose peritoneal equilibration test, according to the two-in-one protocol (28). Glucose drained during 24 h was measured. Peritoneal glucose absorption was calculated by subtracting the 24-h drained glucose from the total glucose influx by the dialysate. Dietary protein intake was estimated from protein catabolic rate and normalized to actual body weight (nPCR) using the PD Adequest software (Baxter Healthcare Corporation, Deerfield, IL, USA).
Ther Apher Dial, Vol. 19, No. 2, 2015
146
AP Bernardo et al.
Body composition assessment We performed multifrequency whole body bioimpedance assessment using Body Composition Monitor (BCM, Fresenius Medical Care, Bad Homburg, Germany). The bioimpedance method applied was validated by isotope dilution methods (29), by accepted reference body composition methods (30,31) and by extensive clinical assessment of the hydration state (32). Only measurements achieving simultaneously high quality (> 90%) and lower error percentage (18.5 kg/m
Therapeutic Apheresis and Dialysis 2015; 19(2):144–153 doi: 10.1111/1744-9987.12239 © 2014 The Authors Therapeutic Apheresis and Dialysis © 2014 International Society for Apheresis
Adipokines in Peritoneal Dialysis: Relevant Clinical Impact According to Body Composition Ana Paula Bernardo,1,2 Isabel Fonseca,1,2 José Carlos Oliveira,3 Olivia Santos,1 Maria João Carvalho,1 António Cabrita,1 and Anabela Rodrigues1,2 1
Nephrology Department, 3Clinical Pathology Department, Santo António Hospital, Oporto Hospital Center and 2UMIB/ICBAS/Porto University, Oporto, Portugal
Abstract: Adipokines impact on clinical outcomes is not adequately addressed in peritoneal dialysis (PD). We investigated the impact of leptin/adiponectin ratio (L/A) as a predictor of cardiovascular events (CVE) in PD, taking into consideration patient’s body composition and the potential role of glucose load.We prospectively followed 66 prevalent PD patients for 47.0 ± 28.2 months. New CVE were evaluated. Lean tissue index (LTI), relative fat mass (relFM) and relative overhydration (relOH) using multifrequency bioimpedance (BCM) were assessed; serum lipids, interleukin-6 (IL-6), leptin and adiponectin were measured. We established the determinants of L/A using multiple linear regression and the impact of L/A on CVE. Obesity was present in 47 (73.4%) patients according to relFM, and in seven (10.6%) according to body mass index (BMI). Leptin and L/A exhibited a stronger correlation with relFM (both r = 0.62, P < 0.0001) than with BMI (r = 0.46 and r = 0.51, respectively, both P < 0.0001). L/A showed a significant correlation with triglycerides (r = 0.41,
P = 0.001) and HDL-cholesterol (r = −0.358, P = 0.003), better than isolated leptin or adiponectin. RelFM (RR = 0.130, 95%confidence interval [CI]:0.086–0.174, P < 0.0001) and LTI (RR = 0.194, 95%CI:0.037–0.351, P = 0.016) were independent predictors of L/A (R2 = 0.67). Patients who suffered new CVE were older (59.12 ± 12.41 vs. 47.52 ± 13.84years, P = 0.003) and had a higher relOH (11.28 ± 7.29 vs. 6.60 ± 8.16%, P = 0.028). L/A was significantly higher in patients with CVE[2.29 (1.79) vs. 0.65 (1.73), P = 0.028] but this association was only put on evidence after excluding patients with wasting. BMI is an inaccurate method to classify obesity in PD since it underestimates its prevalence compared with body composition assessment using BCM. High adiponectin and low leptin are associated with a more favorable metabolic risk profile in PD. The L/A is determined by relFM and by LTI. A higher L/A is associated with CVE in PD patients without wasting. Key Words: Adipokines, Cardiovascular disease, Obesity, Peritoneal dialysis.
As in the general population (1,2), overweight is widely prevalent in peritoneal dialysis (PD) patients (3–5). Being a risk factor for cardiovascular disease (CVD) and death in general population (6), the clinical impacts of obesity and overweight are uncertain in PD patients (7–9). These conflicting results are partially explained by the use of body mass index (BMI) to characterize obesity. BMI is imprecise, since it does not distinguish lean body mass, fat mass and extracellular water. This characterization, being important in the general population, is even more crucial in chronic kidney disease (CKD) since those
patients can present simultaneous changes in nutritional status, body fat content and hydration status (10) that have different impacts on outcome. Adipose tissue is now recognized as a new endocrine organ, since biologically active molecules secreted by adipocytes, like leptin and adiponectin, are important determinants of inflammation, atherosclerosis and insulin-resistance (11,12). Although in the general population, high plasma leptin and low plasma adiponectin have been demonstrated to be risk factors for CVD (13–15), in CKD patients, similar evidence is less clear, as there are contradictory results about the impact of adiponectin and leptin on cardiovascular outcomes (16–21). These opposing results could be partially explained by differences in nutritional status and body composition of the studied populations that were not taken into consideration.
Received April 2014; revised May 2014. Address correspondence and reprint requests to Dr Ana Paula Bernardo, MD, Largo Prof.,Abel Salazar, 4099-001 Porto, Portugal. Email: [email protected]
144
Leptin/Adiponectin Ratio in Peritoneal Dialysis Leptin and adiponectin have opposite effects on the cardiovascular system and, in line with this view, two recent studies, in obese and type 2 diabetic patients without CKD, have established that leptin/ adiponectin ratio (L/A) correlates better with pulse wave velocity and carotid intima-media thickness than plasma leptin and adiponectin alone (22,23). This preliminary observation suggests that L/A is a new atherogenic index, and that patients with a higher ratio could be prone to developing cardiovascular events. To our knowledge, only one study demonstrated an association between L/A and mortality in a group of non-diabetic PD patients (24). However, this study was unable to assess the patient’s body composition and it did not address the potential role of glucose load and peritoneal membrane transport characteristics on adipokines profile. Low plasma leptin and high plasma adiponectin can both reflect a protein wasting condition (21,25), a diagnosis that cannot be ruled out with a simple BMI measurement, and which itself impacts patient outcome. In order to determine the impact of adipokines on CKD patient’s clinical outcomes, studies must give detailed information about patient’s body composition. With the purpose to overcome this knowledge limitation, we performed a prospective study to investigate the impact of L/A as a predictor of cardiovascular events in PD patients, taking into consideration the nutritional status and body composition. We characterized body composition using multifrequency bioimpedance (BCM, Fresenius Medical Care, Bad Homburg, Germany) and we further explored the potential role of glucose load and peritoneal membrane small solute transport status on adipokines and obesity profile. PATIENTS AND METHODS Patients This prospective observational study enrolled 66 PD patients attending the Centro Hospitalar do Porto—Peritoneal Dialysis Unit. All patients were treated with low glucose degradation product solutions. As a standard, only 1.36% and 2.27% glucose solutions were used. Hypertonic exchanges were used only by exception and for a short period of time. All patients provided informed consent and the study had been approved by Centro Hospitalar do Porto Ethics Committee. Study design Patients were prospectively followed from December 2008 until death, renal transplantation, transfer to © 2014 The Authors Therapeutic Apheresis and Dialysis © 2014 International Society for Apheresis
145
hemodialysis, or end of study (July 2013). Mean follow-up time was 47.0 ± 28.2 months. The primary endpoint was cardiovascular event. Transient ischemic attack and stroke were considered as cerebrovascular events. Coronary artery event was diagnosed when the patient had a myocardial infarction or an unstable angina episode. Peripheral artery disease was diagnosed in the presence of acute limb ischemia, major arterial thrombotic episode, or in the presence of intermittent claudication and significant arterial stenosis on angiography. Sudden death, in a patient with previously known coronary artery disease, was also considered as a cardiovascular event. Laboratory measurements Fasting venous blood samples were obtained at 9 am during a peritoneal equilibration test (PET). The preceding overnight dwell was standardized to 1.36% glucose solution. Serum was allowed to clot at room temperature and then was separated from cells within 60 min and stored at −70°C until analysis for adiponectin, leptin and interleukin-6 (IL-6). Serum total cholesterol, high-density lipoprotein (HDL) cholesterol and triglycerides were measured with an autoanalyzer (COBAS Integra 800, Roche Diagnostics GmbH, Berlin, Germany). Plasma leptin and adiponectin were measured using enzyme-linked immunosorbent assays (Mediagnost, Reutlingen, Germany). IL-6 was measured using enzyme-linked immunosorbent assay (Diasource, Louvain-La-Neuve, Belgium). Dialysis adequacy, urea kinetics, peritoneal glucose absorption and estimation of dietary protein intake Adequacy of dialysis was calculated from a 24-h urine and dialysate collection. Weekly Kt/V was determined using standard methods (26). Residual glomerular filtration rate was calculated as the average of 24-h urinary urea and creatinine as described elsewhere (27). Small solute peritoneal membrane transport was calculated from a 4-h, 3.86% glucose peritoneal equilibration test, according to the two-in-one protocol (28). Glucose drained during 24 h was measured. Peritoneal glucose absorption was calculated by subtracting the 24-h drained glucose from the total glucose influx by the dialysate. Dietary protein intake was estimated from protein catabolic rate and normalized to actual body weight (nPCR) using the PD Adequest software (Baxter Healthcare Corporation, Deerfield, IL, USA).
Ther Apher Dial, Vol. 19, No. 2, 2015
146
AP Bernardo et al.
Body composition assessment We performed multifrequency whole body bioimpedance assessment using Body Composition Monitor (BCM, Fresenius Medical Care, Bad Homburg, Germany). The bioimpedance method applied was validated by isotope dilution methods (29), by accepted reference body composition methods (30,31) and by extensive clinical assessment of the hydration state (32). Only measurements achieving simultaneously high quality (> 90%) and lower error percentage (18.5 kg/m
