bs_bs_banner
Nephrology 20 (2015) 617–624
Original Article
Hyperleptinaemia, insulin resistance and survival in peritoneal dialysis patients LIOU CAO, SHAN MOU, WEI FANG, LEYI GU, JIAYING HUANG, AIPING GU, JIAQI QIAN and ZHAOHUI NI Department of Nephrology, Molecular Cell Laboratory for Kidney Disease, Shanghai Center for Peritoneal Dialysis Research, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
KEY WORDS: hyperleptinaemia, insulin resistance, leptin, peritoneal dialysis, mortality. Correspondence: Dr Zhaohui Ni, Department of Nephrology, Molecular Cell Laboratory for Kidney Disease, Shanghai Center for Peritoneal Dialysis Research, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, No. 1630 Dongfang Road, Shanghai 200127, China. Email: [email protected] Conflict of interest: The authors have no conflict of interest to state.
Accepted for publication 16 April 2015. Accepted manuscript online 23 April 2015. doi:10.1111/nep.12491
SUMMARY AT A GLANCE This study of 157 peritoneal dialysis patients assessed the potential relationship of insulin resistance, and its related factors such as leptin, to glucose and lipid metabolism. They demonstrated that insulin resistance correlated with age, leptin and triglyceride levels, and was negatively correlated with BMI. In addition they showed that dialysis duration, leptin levels, nPCR and hsCRP may be important risk factors for mortality in PD patients.
ABSTRACT Aim: The aim of this study was to clarify the relationship between insulin resistance (IR) and glucose and lipid metabolism in peritoneal dialysis (PD) patients. The study also investigated the prognostic factors for survival in long-term peritoneal dialysis patients. Methods: Participants were recruited from July 1 to August 1, 2011, based on selection criteria. Patients were divided into two groups, high (H) and low (L) group according to the median value of homeostasis model assessment of IR (HOMA-IR). Type 2 diabetes mellitus (DM) patients were chosen as positive controls. Clinical, plasma biochemical and metabolic parameters were observed and recorded at the outset and follow-up of this study. Mortality related factors were also detected, and statistical analyses were performed. Results: A total of 157 cases with an average age of 55 ± 15 years were included. There were 26, 66 and 65 cases in the DM, H and L groups, respectively. Younger age, lower body mass index, high sensitive C-reactive protein (hsCRP) level, higher normalized protein catabolic rate (nPCR) were found in the L group compared with the other two groups. HOMA-IR positively correlated with age, leptin and triglyceride levels, and it negatively correlated with BMI. L group had better survival rate than H and DM groups. Dialysis duration, serum leptin level, nPCR and hsCRP were statistically associated with mortality. Conclusion: Insulin resistance may play an important role in the pathophysiology of glucose and lipid metabolism. Dialysis duration, leptin, nPCR and hsCRP may be risk factors for mortality in PD patients.
The dialysate solution generally used for peritoneal dialysis (PD) patients contains glucose concentrations between 1360 and 3860 mg/dL.1 Approximately 60%–80% of glucose, which is approximately equal to 100–300 g glucose, in dialysate is absorbed daily during peritoneal dialysis,2 which is closely associated with disturbances of carbohydrate metabolism in patients. A large amount of glucose absorption leads to increased blood glucose, hyperinsulinism and insulin © 2015 Asian Pacific Society of Nephrology
resistance (IR). IR in combination with impaired glucose tolerance is the major contributor to the pathogenesis of type 2 diabetes.3 The peritoneal glucose absorption rate is associated with the peritoneal solute transport rate (PSTR),4 which is also dependent on a high glucose level.5 Tuzcu et al.6 found that the HOMA [beta-cell function (%B)] values in non-diabetic PD patients was higher than hemodialysis patients and 617
L Cao et al.
healthy control subjects, and that newly diagnosed type 2 diabetes were more common in PD patients. Adipose cells are related to IR, and these cells are the source of leptin, adiponectin and other excreted factors. Serum leptin likely participates in the regulation of glucose, fat and energy metabolism, which are related to reduced feeding, increased energy liberation, and reduced weight.7 Leptin is primarily excreted and cleared by the kidneys, and higher plasma leptin levels are related to chronic kidney disease (CKD).8 Leptin is also involved in pancreatic β-cell proliferation and insulin secretion. Resistin is a peptide hormone that is an important, potential link between obesity and IR. Steppan et al.9 suggested that the neutralization of resistin using an anti-resistin antibody enhanced insulinstimulated glucose uptake. Other studies also found that resistin level is increased during adipocyte differentiation10 in mice, but resistin also plays a role in the inhibition of fat formation.11 Increased glycaemic load, protein loss in the peritoneum, dyslipidaemia and micro-inflammation are also involved in IR in PD patients. However, the interaction among these factors is controversial, especially the correlation between IR and leptin levels in PD patients. Whether IR and leptin are related to complications, prognosis, micro-inflammation, nutrition, lipid metabolism or dialysis adequacy is not clear. Therefore, in the current study, we observed these various parameters in non-diabetic and diabetic continuous ambulatory peritoneal dialysis (CAPD) patients in order to investigate the relationship between clinical and plasma biochemical parameters and IR. The mortality of the Low HOMA, High HOMA and DM groups was also compared.
Patients with various autoimmune diseases, signs of peritonitis or infection, or taking hormones or immunosuppressants, or who received a plasma transfusion or albumin during the 2 weeks prior to participant selection the study were excluded. Patients with serious cardiac insufficiency (New York Heart Association class III or greater), abnormal liver dysfunction and malignancies were also excluded.
Follow-up and endpoints Patients were followed up every month for 2 years (until August 1, 2013), unless one of the following incidents occurred: death, renal transplant, changes to haemodialysis, or treatment abandoning. Patients with the last three situations were deemed lost to follow-up, and they were statistically analyzed as censored data.
Groups Homeostasis model assessment of insulin resistance (HOMAIR) was calculated according to the following formula: HOMA-IR = fasting blood glucose (mmol/L) × fasting insulin (FINS, μmol/mL)/22.5. Patients were divided evenly into a high-level group (H group, HOMA-IR ≥ 2.05) and a lowlevel group (L group, HOMA-IR < 2.05) based on the median value of HOMA-IR. The H group was also defined as the insulin-resistance group. Type 2 diabetic patients (high blood glucose before PD and definitively diagnosed with type 2 diabetes) who were treated with PD were defined as the diabetes mellitus (DM) group simultaneously.
Clinical indices in detection
MATERIALS AND METHODS Participants The non-diabetic patients who receiving PD in peritoneal dialysis centers of Renji hospital were recruited between July 1 to August 1, 2011. Basic information was recorded, and peritoneal equilibrium tests and dialysis adequacy measurements were performed in three months. Type 2 diabetic patients who were treated with PD during the same period were selected as controls. Demographic and clinical data were recorded at study entry. Various indicators were detected and recorded at the start of the study. Ethics approval was granted from the Ethics Committee of Renji Hospital, Shanghai Jiao Tong University School of Medicine.
Inclusion and exclusion criteria Adult patients (>18 years old) who were on PD for more than 3 months were considered eligible, and regular follow-up was conducted every month using outpatient follow-up, home visits or telephone contacts. 618
Patients’ characteristics in general practice, including age, sex, height, weight and blood pressure, were recorded. Subjective global assessment (SGA) of nutrition analysis was conducted using the following conventional approaches. BMI, diet situations, gastrointestinal conditions, activities of daily living and disease response to stress were detected. Subcutaneous fat (skinfold thickness) over the left triceps and mid-arm muscle circumference12 (25.3 cm for males and 23.2 cm for females) were measured using conventional approaches. Edema of the ankles and sacral regions were noted. All of these indices were divided into A, B and C grades (ranked from good to worse). Previous investigations and human body measurements delineated that the presence of more than five of the above eight factors in grades B or C was defined as moderately malnourished (B) or severe malnutrition (C), and the other subjects were in good health. The National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III) Criteria for metabolic syndrome (MS) was adopted. The number of MS factors for each patient was recorded and compared among these groups. © 2015 Asian Pacific Society of Nephrology
Effects of peritoneal dialysis
All the patients were examined for 24-h urine and dialysate, dialysate glucose concentrations, exchange volume and frequency, dialysate calcium concentration, collections and dwell time were examined during follow-up studies. Concentrations of creatinine (Cr), blood urea nitrogen (BUN), urea, fasting blood glucose (FBG) and 2 hour postprandial blood glucose were detected. Total weekly urea clearance (Kt/V), renal Kt/V, total weekly creatinine clearance (WCcr) and renal Ccr were calculated using the urea kinetic model (UKM). The protein catabolic rate (PCR) and normalized protein catabolic rate (nPCR) was calculated. In addition, 2.5% glucose dialysis was conducted using a standard peritoneal equilibration test (PET). Peritoneal transport characteristics were detected, and the D/D0 (peritoneal glucose absorption) and D/Pcr were recorded at four hours. Glucose concentration and absorption of dialysis solutions were calculated as Bodnar DM et al. suggested13. Furthermore, drug use, including erythropoietin, iron supplements, hypertension drugs, vitamin D, statins, phosphorus-reducing drugs and others were recorded at the beginning of the study.
Biochemical parameter detection Serum creatinine (SCr), blood leukocytes, hemoglobin, albumin (Alb), calcium, phosphorus, intact parathyroid hormone (iPTH), serum potassium and sodium, chlorine iron, ferritin, total iron-binding capacity, high sensitive C-reactive protein (hsCRP), FINS, glycated albumin (GA), and CA125 (chemiluminescent immunoassay technique) were determined at the outset of this study in all patients.
Leptin, adiponectin and resistin detection Determination of leptin, adiponectin and resistin concentrations were performed according to the instruction manual of the Human Leptin (STH01090)/ Adiponectin (STH01002)/ Resistin (STH01145) ELISA Kit (Shanghai Sunteam Biotech, Shanghai, China) based on the double antibody sandwich ABC-ELISA method.
RESULTS The baseline characteristics The average age of 157 patients in the L group was younger than the DM and H groups (P < 0.001). There was a total of 79 males, including 17 in the DM (n = 26), 28 in the H (n = 66) and 34 in the L groups (n = 65). There was no significant difference in gender distribution among three groups. Dialysis duration of the DM, H and L groups were 33.69 ± 24.11, 36.29 ± 29.71 and 36.80 ± 30.35 months, respectively. There were no significant differences (P > 0.05) among these three groups. The BMI of the L group was lower than that of the DM and H groups (P = 0.007, P < 0.001). The mean arterial pressure of the L group was higher than that of the H group (P < 0.001). The nPCR of the L group was higher than that of the DM and H groups (P = 0.012, P = 0.019). The prevalence of MS and number of its factors in the DM and H groups were significantly higher than the L group (all P < 0.001). There were no significant differences in SGA, dialysis duration, various drugs (included statins) used, calcium density, glucose absorption, or glucose concentration in the dialysate among these three groups (Tables 1,2). Table 3 shows that there was no significant proportion difference in primary disease between the H and L groups.
Blood lipid and glucose level comparisons Blood glucose levels in the L group were lower than in the DM and H groups (P < 0.001, P = 0.017). Blood glucose levels in the DM group were much higher than in the H group (P < 0.001). The GA values of the DM group were higher than in the H and L groups (P < 0.001). The total cholesterol levels of the DM group were higher than those of the other two groups (P = 0.036, P = 0.037). Triglycerides levels in the H group were higher than those of the DM and L groups (P = 0.027, P < 0.001). HDL-C levels were lowest in the H group, and it was highest in the L group. There were no significant differences in low density lipoprotein (LDL) levels among groups (Table 1).
Comparison of other laboratory indexes Statistical approaches Normal distribution was confirmed before any statistical analyses were conducted, and a P > 0.05 was considered normally distributed. Simple linear regression and multiple linear regression analyses were performed for partial correlation analysis. Data analysis was performed using SPSS 17.0 statistical software (SPSS, Chicago, IL, USA). A two-sided P-value of 0.05 or less was considered statistically significant for all tests. © 2015 Asian Pacific Society of Nephrology
Blood WBCs, uric acid and CA125 values in the H group were much higher than in the L group (P < 0.001, P = 0.011, and P = 0.011, respectively) (Table 1).
Leptin, adiponectin and resistin comparisons Leptin values in the H group were significantly higher than in the DM and L groups (P = 0.005, P < 0.001). Adiponectin level was lower in the H group than in the L group (P < 0.001). Resistin levels in these three groups were similar. 619
L Cao et al.
Table 1 Baseline characteristics of all peritoneal dialysis patients Variable
Total (mean ± SD)
Diabetes mellitus group
High group IR > 2.05
Age (years) Amount (male %) Height (cm) Body mass index (kg/m2) Mean artery pressure (mmHg) Dialysis duration (month) Normalized protein catabolic rate (nPCR) Metabolic syndrome (%) Number of factors HOMA-IR Leptin (ng/mL) Adiponectin (ug/mL) Resistin (ng/mL)
55.24 ± 15.23 157 (79) 162.78 ± 8.05 23.28 ± 3.62 100.62 ± 15.00 36.07 ± 28.98 0.83 ± 0.19 76 (48.7) 2.44 ± 0.98
61.15 ± 12.97### 26 (17) 162.46 ± 8.14 24.14 ± 3.45## 100.05 ± 12.27 33.69 ± 24.11 0.77 ± 0.15# 19 (73.1) ### 2.88 ± 0.77###
19.40 ± 11.59 12.86 ± 6.18 33.77 ± 15.15
18.19 ± 11.62 12.38 ± 6.30 30.96 ± 16.13
58.44 ± 15.50### 66 (28) 161.30 ± 7.67# 24.26 ± 3.64### 96.54 ± 14.20### 36.29 ± 29.71 0.80 ± 0.18# 42 (63.6) ### 2.86 ± 0.84### 19.36 ± 27.24### 25.08 ± 9.18*# 11.01 ± 5.67# 36.05 ± 17.87
Subjective global assessment of nutrition (SGA) Normal (%) Moderate malnutrition (%) Severe malnutrition (%)
16 (61.5) 9 (34.6) 1 (3.8)
48 (72.7) 17 (25.8) 1 (1.5)
49 (75.4) 14 (21.5) 2 (3.1)
Drug Anti-hypertension drugs (%) Phosphorus reduction medicine (%) Vitamin D (%) Erythropoietin (%) Chalybeate (%) Statins (%)
26 (100) 22 (84.6) 2 (7.7) 26 (100) 23 (88.5) 7 (26.9)
57 (86.4) 53 (80.3) 7 (10.6) 64 (97) 58 (87.9) 13 (19.7)
60 (92.3) 57 (87.7) 10 (15.4) 63 (96.9) 58 (89.2) 9 (13.8)
8.00 ± 2.82 111.44 ± 19.39 208.81 ± 70.48 417.05 ± 79.42 2.30 ± 0.21 400.63 ± 376.07 5.34 ± 1.19 2.26 ± 1.21 3.31 ± 1.05 1.18 ± 0.35 7.46 ± 3.09
8.45 ± 2.82 114.96 ± 17.14 211.77 ± 74.87 421.85 ± 80.36 2.28 ± 0.25 341.03 ± 338.60 5.57 ± 1.08 2.17 ± 1.20 3.46 ± 0.99 1.21 ± 0.34 10.0 ± 3.19###
13.01 ± 4.22 6.70 ± 17.66 35.02 ± 17.82
15.75 ± 4.70 3.14 ± 5.11 36.40 ± 17.75
8.65 ± 3.26### 112.68 ± 21.53 217.72 ± 74.16 435.47 ± 90.51# 2.34 ± 0.23 401.87 ± 353.88 5.42 ± 1.06* 2.76 ± 1.34*### 3.26 ± 1.03 1.08 ± 0.33 7.53 ± 3.19#*** 5.24 ± 1.87### 12.68 ± 4.71* 10.78 ± 24.01# 40.59 ± 18.97#
7.19 ± 2.09 108.78 ± 17.85 198.71 ± 64.46 398.75 ± 69.09 2.27 ± 0.18 423.21 ± 413.28 5.17 ± 1.34* 1.79 ± 0.84 3.30 ± 1.05 1.28 ± 0.36 6.34 ± 2.21*** 1.28 ± 0.45 12.15 ± 2.79* 3.97 ± 11.6 28.82 ± 14.60
Laboratory index White blood cell (×109/L) Hemaglobin (g/L) Thrombocyte (×1012/L) Uric acid (umol/L) Calcium (mmol/L) Intact parathyroid hormone (pg/mL) Cholesterol (mmol/L) Triglyceride (mmol/L) Low density lipoprotein (mmol/L) High density lipoprotein (mmol/L) 2-hour-fasting blood-glucose (mmol/L) Insulin (IU/L) Glycated albumin (%) High sensitive C-reactive protein (mg/L) CA125 (U/mL)
Lower group IR < 2.05 49.63 ± 14.11 65 (34) 164.42 ± 8.21 21.95 ± 3.26 104.99 ± 15.76 36.80 ± 30.35 0.88 ± 0.21 15 (23.1) 1.83 ± 0.88 6.68 ± 9.80 14.12 ± 11.27 14.94 ± 6.08 32.59 ± 11.12
Compared to the DM group: *means P ≤ 0.05; **means P ≤ 0.01; ***means P ≤ 0.001. Compared to the L group: # means P ≤ 0.05; ## means P ≤ 0.01; ### means P ≤ 0.001.
HOMA-IR, FINS, and hsCRP values were higher in the H group than in the L group (P < 0.001, P < 0.001, and P = 0.027, respectively). There were no obvious differences in total Kt/V, renal Kt/V, renal Ccr, peritoneal transport type, D/D0, D/Pcr or urine volume at 4 h (Table 2) among three groups.
triglycerides (P = 0.039) and age (P = 0.012), but it was negatively correlated to BMI (P = 0.042). A simple linear regression illustrated the relationships between these four factors and HOMA-IR (Fig. 1), and the interactions among these factors were evaluated using multiple linear regression models (Table 4).
HOMA-IR-related factor analysis
Patient mortality
Simple and multiple linear regression analyses showed that HOMA-IR was positively correlated with leptin (P = 0.001),
The L group had better survival rate than the H and DM groups (P = 0.037). Four patients in the DM group died
620
© 2015 Asian Pacific Society of Nephrology
Effects of peritoneal dialysis
Table 2 Peritoneal dialysis related factors Variable
Glucose intake (g/day) Glucose absorption (g/day)
Total (mean ± SD)
Diabetes mellitus group
High group IR > 2.05
Lower group IR < 2.05
65.10 ± 17.67 38.38 ± 12.26
69.27 ± 10.87 40.18 ± 8.06
64.08 ± 15.94 37.84 ± 11.94
64.48 ± 21.17 38.20 ± 13.95
17 (65.4%) 9 (34.6%)
33 (50%) 33 (50%)
33 (50.8%) 32 (49.2%)
1.67 ± 0.28 0.25 ± 0.28 57.42 ± 12.96 13.01 ± 15.11 526.25 ± 674.16 0.61 ± 0.11 0.42 ± 0.08
1.75 ± 0.47 0.40 ± 0.55 61.17 ± 26.55 20.16 ± 27.94 484.92 ± 692.26 0.63 ± 0.10 0.41 ± 0.08
1.71 ± 0.38 0.32 ± 0.49 58.53 ± 22.49 15.62 ± 26.16 400.76 ± 552.16 0.62 ± 0.11 0.41 ± 0.09
2 (7.7%) 9 (34.6%) 4 (15.4%) 11 (42.3%)
4 (6.1%) 21 (31.8%) 6 (9.1%) 35 (53.0%)
3 (4.6%) 19 (29.2%) 5 (7.7%) 38 (58.5%)
Dialysate calcium concentration 1.25 (%) 1.75 (%) Dialysis related indicators Total Kt/V Residual kidney Kt/V Total creatinine clearance (Ccr) Renal Ccr 24-h urine (mL) Peritoneal creatinine clearance rate at 4 h Peritoneal glucose absorption at 4 h
1.72 ± 0.41 0.34 ± 0.49 59.42 ± 23.01 17.08 ± 25.49 455.80 ± 630.68 0.62 ± 0.11 0.41 ± 0.08
Peritoneal transport type High High-average Low- average Low
Table 3 The primary diseases of all peritoneal dialysis patients
Diabetic nephropathy (%) Hypertensive nephropathy (%) Chronic nephritis (%) Obstructive nephropathy (%) Renal tubular diseases (%) Polycystic kidney (%) Reason unknown (%)
Diabetes mellitus group
High group IR > 2.05
Lower group IR < 2.05
24 (92.4) 1 (3.8) 1 (3.8) 0 0 0 0
0 5 (7.6) 30 (45.5) 3 (4.5) 3 (4.5) 3 (4.5) 22 (33.3)
0 3 (4.6) 33 (50.8) 0 1 (1.5) 2 (3.1) 26 (40)
suffered sudden cardiac death, two patients suffered peritonitis infection and one patient suffered lung cancer metastasis). One patient abandoned treatment because of renal transplant, and one patient was transferred to haemodialysis because of peritonitis. Seven patients in the L group died (four patients suffered sudden cardiac death, two patients suffered peritonitis infection and one patient suffered inadequate dialysis). Four patients abandoned treatment because of renal transplantation, and one patient was transferred to haemodialysis because of peritonitis (Fig. 2).
Mortality-related factors Table 4 Multiple regression analysis for homeostasis model assessment of insulin resistance (HOMA-IR)in peritoneal dialysis (PD) patients Variable Age (years) Leptin (ng/mL) 2-hour postprandial blood glucose (mmol/L) Glycated albumin (%) Metabolic syndrome Triglyceride (mmol/L) High density lipoprotein (mmol/L) Normalized protein catabolic rate Body mass index (kg/m2) Log (high sensitive C-reactive protein)
β ± SE
P-value
0.07 ± 0.43 0.311 ± 0.086 0.302 ± 0.492 0.205 ± 0.202 1.539 ± 1.488 0.016 ± 0.008 0.343 ± 2.178 −1.709 ± 3.315 0.465 ± 0.227 −0.309 ± 0.908
0.012 0.001 0.540 0.312 0.303 0.039 0.875 0.607 0.042 0.735
(cardiovascular accident, cerebrovascular accident, peritonitis infection and upper gastrointestinal haemorrhage), and four patients were transferred to haemodialysis because of peritonitis. Eight patients in the H group died (five patients © 2015 Asian Pacific Society of Nephrology
Simple linear regression analysis demonstrated that leptin, dialysis duration, BMI, HOMA-IR, albumin, hsCRP, nPCR and triglycerides were related to mortality. Multiple linear regression analysis (Table 5) demonstrated that dialysis duration (P = 0.000), blood leptin level (P = 0.019), nPCR (P = 0.01) and hsCRP (P = 0.021) were potential prognostic factors for mortality.
DISCUSSION Cardiovascular disease (CVD) remains the main cause of death in patients with ESRD, and it is responsible for approximately 40% of the total mortality in CAPD patients.14 Identification of risk factors for CVD is important for CAPD patients. The gold standard is the insulin clamp technique, which is a time-consuming operation, and few studies use this technique in dialysis patients. However, insulin clamp is the most accurate determination of insulin release. HOMA-IR is the most commonly used method to detect IR in 621
L Cao et al.
Fig. 1 Homeostasis model assessment of insulin resistance (HOMA-IR) related factors. (A) HOMA-IR positively correlated with serum leptin levels. (B) HOMA-IR positively correlated with triglyceride levels. (C) HOMA-IR positively correlated with BMI. (D) HOMA-IR positively correlated with age.
Fig. 2 Kaplan–Meier survival curves in all peritoneal dialysis (PD) patients (P = 0.037). The blue, green and yellow lines represent the DM, , DM; , L; H and L groups, respectively. , H; , DM-censored; , L-censored; , H-censored.
622
© 2015 Asian Pacific Society of Nephrology
Effects of peritoneal dialysis
Table 5 Multiple regression analysis for mortality-related factors in peritoneal dialysis (PD) patients Variable Dialysis duration Leptin Normalized protein catabolic rate High sensitive C-reactive protein
β ± SE
P-value
−0.092 ± 0.024 −0.032 ± 0.014 −7.092 ± 2.762 0.03 ± 0.013
0.000 0.019 0.01 0.021
peritoneal dialysis patients. Our study demonstrated that HOMA-IR was positively correlated with leptin, PPG, GA, triglycerides, the presence of MS, the number of MS factors, age and log hsCRP, and it was negatively associated with HDL, BMI and nPCR. Multiple linear regression analysis revealed that HOMA-IR was positively correlated with age, leptin and triglyceride, and it was negatively correlated with BMI. These results suggested that HOMA-IR may play an important role in the pathophysiology of glucose and lipid metabolism. Our study also showed a good correlation between HOMA-IR and leptin. However, we failed to find any significant association between HOMA-IR and adiponectin or resistin, although these two factors work synergistically as adipokines. This finding suggest, leptin may play a more important role in the occurrence of IR than the other two parameters in PD patients. Independent correlations between BMI, age and IR were also found, which is similar to previous research.15 Although IR is multifactorial in origin, obesity has long been linked with IR. Once people develop obesity, leptin levels increase, which can suppress insulin secretion from the pancrease, and causes leptin resistance.16 Conversely, adiponectin has anti-inflammatory effects and could improve insulin sensitivity. Adiponectin levels were declined in obese people.17 Previous studies found that adiponectin exhibited a cardiovascular protective effect in haemodialysis patients, especially in patients with higher adiponectin.18 Briefly, inflammatory cytokines and adipokines regulate insulin levels under physiological conditions collaboratively and antagonistically. IR may occur when adiponectin and leptin homeostasis is disturbed. Johnson et al.19 found that the MS incidence in PD patients was much higher than pre-haemodialysis and haemodialysis patients, and was associated with a significantly increased risk of future cardiovascular events. A significantly lower incidence of MS in the L group than in the DM and H groups was found in our study. High cardiovascular morbidity and mortality in PD patients were discovered previously.7,20 Similarly, poorer survival in the DM and H groups compared to the L group was found in our follow-up visits. These results indicated that glucose and lipid metabolism in PD is prominent and require intensified attention. Therefore, it may be beneficial to discover the presence of MS and control blood pressure, blood glucose, and cholesterol level of PD patients to improve their survival rates. © 2015 Asian Pacific Society of Nephrology
Multiple regression analysis revealed that dialysis duration, leptin, nPCR and hsCRP were also involved in the mortality of PD patients. Serum hsCRP is a pro-inflammatory factor that is associated with CVD-related morbidity and mortality in PD and haemodialysis patients.21 CRP can also cause protein-energy malnutrition, which is induced by the inhibition of protein synthesis and catabolism, the high level of CRP was directly correlated with IR.22 Blood leptin levels of PD patients were significantly higher than those of haemodialysis patients even after correcting for differences in body fat, age and gender effects. These results may be related to dialysis and result in multiple parameter changes, including increased glucose load, glucose-stimulated insulin secretion, and aggravated hyperinsulinaemia, which finally promote blood leptin secretion.23 A previous study found that icodextin or amino acid dialysis solution could reduce blood leptin level and decrease the degree of IR.24 Pre-inflammatory indicators, including IL-6, IL-1 and metabolites of oxidative stress associated with IR, were significantly increased in PD patients.25 Downstream cascade inhibition of these inflammatory factors may reduce their reactivity to insulin. Insulin is an anti-inflammatory factor that regulates inflammation in vitro and in vivo.26 Therefore, the inflammatory response may aggravate the pathological process in PD patients, especially in people with high HOMAIR. Notably, hsCRP level in our DM group was reduced. This reduction may be the result of insulin therapy, which moderately reduces micro-inflammation. Our study also found that DM patients had higher fasting blood glucose, 2-hour postprandial blood-glucose and GA than the other two groups. A previous study found that GA accurately predicted the risk of death and hospitalizations in patients with diabetes mellitus and ESRD.27 This result suggests that GA may be a more accurate indicator of glucose metabolism in PD patients. Uncontrolled diabetes (especially type 2 diabetes) is an important risk factor for poor prognosis in PD patients,1,28 and the early monitoring and good control of GA may be one way to improve patient survival. Our study demonstrated that high leptin level might be involved in IR in patients undergoing PD. Serum leptin and triglyceride levels, BMI, and age may be risk factors for IR Low HOMA group had better survival rate than DM and H group patients. Leptin level, the duration of dialysis, leptin level, nPCR and hsCRP may also related to patient mortality in non-diabetic and diabetic PD patients.
ACKNOWLEDGEMENT This study was supported in part by the National Basic Research Program of China 973 Program No. 2012CB517600 (No. 2012CB517602). The study was also sponsored by the National Natural Science Foundation of China (81370794, 81102700, 81373865). 623
L Cao et al.
REFERENCES 1. Duong U, Mehrotra R, Molnar MZ et al. Glycemic control and survival in peritoneal dialysis patients with diabetes mellitus. Clin. J. Am. Soc. Nephrol. 2011; 6: 1041–8. 2. Cotovio P, Rocha A, Rodrigues A. Peritoneal dialysis in diabetics: There is room for more. Int. J. Nephrol. 2011; 2011. 3. Goldstein BJ. Insulin resistance as the core defect in type 2 diabetes mellitus. Am. J. Cardiol. 2002; 90: 3–10. 4. Wang T, Heimbürger O, Cheng H, Bergstroem J, Lindholm B. Amphotericin B does not increase peritoneal fluid removal. Adv. Perit. Dial. 1998; 14: 3–10. 5. Szeto C-C, Chow K-M, Leung C-B et al. Increased subcutaneous insulin requirements in diabetic patients recently commenced on peritoneal dialysis. Nephrol. Dial. Transplant. 2007; 22: 1697–702. 6. Tuzcu A, Bahceci M, Yilmaz ME, Turgut C, Kara IH. The determination of insulin sensitivity in hemodialysis and continuous ambulatory peritoneal dialysis in nondiabetic patients with end-stage renal disease. Saudi Med. J. 2005; 26: 786–91. 7. Ahima RS, Flier JS. Leptin. Annu. Rev. Physiol. 2000; 62: 413–37. 8. Shankar A, Syamala S, Xiao J, Muntner P. Relationship between plasma leptin level and chronic kidney disease. Int. J. Nephrol. 2012; 2012: 269532. 9. Steppan CM, Bailey ST, Bhat S et al. The hormone resistin links obesity to diabetes. Nature 2001; 409: 307–12. 10. Janke J, Engeli S, Gorzelniak K, Luft FC, Sharma AM. Resistin gene expression in human adipocytes is not related to insulin resistance. Obes. Res. 2002; 10: 1–5. 11. Pang S, Le Y. Role of resistin in inflammation and inflammation-related diseases. Cell. Mol. Immunol. 2006; 3: 29–34. 12. Gawkrodger D, Ferguson A, Barnetson R. Nutritional status in patients with dermatitis herpetiformis. Am. J. Clin. Nutr. 1988; 48: 355–60. 13. Bodnar DM, Busch S, Fuchs J et al. Estimating glucose absorption in peritoneal dialysis using peritoneal equilibratin tests. Adv. Perit. Dial. 1993; 9: 114–8. 14. Krishnasamy R, Badve SV, Hawley CM et al. Daily variation in death in patients treated by long-term dialysis: Comparison of in-center hemodialysis to peritoneal and home hemodialysis. Am. J. Kidney Dis. 2013; 61: 96–103. 15. Henriques VT, Martinez EZ, Divino-Filho JC, Pecoits-Filho R, Cardeal da Costa JA. Increase in BMI over time is associated with
624
16. 17. 18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
fluid overload and signs of wasting in incident peritoneal dialysis patients. J. Ren. Nutr. 2013; 23: e51–7. Enriori PJ, Evans AE, Sinnayah P, Cowley MA. Leptin resistance and obesity. Obesity 2006; 14: 254S–258S. Kadowaki T, Yamauchi T. Adiponectin and adiponectin receptors. Endocr. Rev. 2005; 26: 439–51. Amira OC, Naicker S, Manga P et al. Adiponectin and atherosclerosis risk factors in African hemodialysis patients: A population at low risk for atherosclerotic cardiovascular disease. Hemodial. Int. 2012; 16: 59–68. Johnson DW, Armstrong K, Campbell SB et al. Metabolic syndrome in severe chronic kidney disease: Prevalence, predictors, prognostic significance and effects of risk factor modification. Nephrology 2007; 12: 391–8. Rossum C, Hoebee B, Baak MA, Mars M, Saris WH, Seidell JC. Genetic variation in the leptin receptor gene, leptin, and weight gain in young Dutch adults. Obes. Res. 2003; 11: 377–86. Wang AY-M, Woo J, Lam CW-K et al. Is a single time point C-reactive protein predictive of outcome in peritoneal dialysis patients? J. Am. Soc. Nephrol. 2003; 14: 1871–9. McLaughlin T, Abbasi F, Lamendola C et al. Differentiation between obesity and insulin resistance in the association with C-reactive protein. Circulation 2002; 106: 2908–12. Johansen KL, Mulligan K, Tai V, Schambelan M. Leptin, body composition, and indices of malnutrition in patients on dialysis. J. Am. Soc. Nephrol. 1998; 9: 1080–84. Furuya R, Odamaki M, Kumagai H, Hishida A. Beneficial effects of icodextrin on plasma level of adipocytokines in peritoneal dialysis patients. Nephrol. Dial. Transplant. 2006; 21: 494–8. Nascimento MM, Suliman ME, Silva M et al. Effect of oral N-acetylcysteine treatment on plasma inflammatory and oxidative stress markers in peritoneal dialysis patients: A placebo-controlled study. Perit. Dial. Int. 2010; 30: 336–42. Hyun E, Ramachandran R, Hollenberg MD, Vergnolle N. Mechanisms behind the anti-inflammatory actions of insulin. Crit. Rev. Immunol. 2011; 31: 307–40. Freedman BI, Andries L, Shihabi ZK et al. Glycated albumin and risk of death and hospitalizations in diabetic dialysis patients. Clin. J. Am. Soc. Nephrol. 2011; 6: 1635–43. Fang W, Yang X, Kothari J et al. Patient and technique survival of diabetics on peritoneal dialysis: One-center’s experience and review of the literature. Clin. Nephrol. 2008; 69: 193–200.
© 2015 Asian Pacific Society of Nephrology
Nephrology 20 (2015) 617–624
Original Article
Hyperleptinaemia, insulin resistance and survival in peritoneal dialysis patients LIOU CAO, SHAN MOU, WEI FANG, LEYI GU, JIAYING HUANG, AIPING GU, JIAQI QIAN and ZHAOHUI NI Department of Nephrology, Molecular Cell Laboratory for Kidney Disease, Shanghai Center for Peritoneal Dialysis Research, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
KEY WORDS: hyperleptinaemia, insulin resistance, leptin, peritoneal dialysis, mortality. Correspondence: Dr Zhaohui Ni, Department of Nephrology, Molecular Cell Laboratory for Kidney Disease, Shanghai Center for Peritoneal Dialysis Research, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, No. 1630 Dongfang Road, Shanghai 200127, China. Email: [email protected] Conflict of interest: The authors have no conflict of interest to state.
Accepted for publication 16 April 2015. Accepted manuscript online 23 April 2015. doi:10.1111/nep.12491
SUMMARY AT A GLANCE This study of 157 peritoneal dialysis patients assessed the potential relationship of insulin resistance, and its related factors such as leptin, to glucose and lipid metabolism. They demonstrated that insulin resistance correlated with age, leptin and triglyceride levels, and was negatively correlated with BMI. In addition they showed that dialysis duration, leptin levels, nPCR and hsCRP may be important risk factors for mortality in PD patients.
ABSTRACT Aim: The aim of this study was to clarify the relationship between insulin resistance (IR) and glucose and lipid metabolism in peritoneal dialysis (PD) patients. The study also investigated the prognostic factors for survival in long-term peritoneal dialysis patients. Methods: Participants were recruited from July 1 to August 1, 2011, based on selection criteria. Patients were divided into two groups, high (H) and low (L) group according to the median value of homeostasis model assessment of IR (HOMA-IR). Type 2 diabetes mellitus (DM) patients were chosen as positive controls. Clinical, plasma biochemical and metabolic parameters were observed and recorded at the outset and follow-up of this study. Mortality related factors were also detected, and statistical analyses were performed. Results: A total of 157 cases with an average age of 55 ± 15 years were included. There were 26, 66 and 65 cases in the DM, H and L groups, respectively. Younger age, lower body mass index, high sensitive C-reactive protein (hsCRP) level, higher normalized protein catabolic rate (nPCR) were found in the L group compared with the other two groups. HOMA-IR positively correlated with age, leptin and triglyceride levels, and it negatively correlated with BMI. L group had better survival rate than H and DM groups. Dialysis duration, serum leptin level, nPCR and hsCRP were statistically associated with mortality. Conclusion: Insulin resistance may play an important role in the pathophysiology of glucose and lipid metabolism. Dialysis duration, leptin, nPCR and hsCRP may be risk factors for mortality in PD patients.
The dialysate solution generally used for peritoneal dialysis (PD) patients contains glucose concentrations between 1360 and 3860 mg/dL.1 Approximately 60%–80% of glucose, which is approximately equal to 100–300 g glucose, in dialysate is absorbed daily during peritoneal dialysis,2 which is closely associated with disturbances of carbohydrate metabolism in patients. A large amount of glucose absorption leads to increased blood glucose, hyperinsulinism and insulin © 2015 Asian Pacific Society of Nephrology
resistance (IR). IR in combination with impaired glucose tolerance is the major contributor to the pathogenesis of type 2 diabetes.3 The peritoneal glucose absorption rate is associated with the peritoneal solute transport rate (PSTR),4 which is also dependent on a high glucose level.5 Tuzcu et al.6 found that the HOMA [beta-cell function (%B)] values in non-diabetic PD patients was higher than hemodialysis patients and 617
L Cao et al.
healthy control subjects, and that newly diagnosed type 2 diabetes were more common in PD patients. Adipose cells are related to IR, and these cells are the source of leptin, adiponectin and other excreted factors. Serum leptin likely participates in the regulation of glucose, fat and energy metabolism, which are related to reduced feeding, increased energy liberation, and reduced weight.7 Leptin is primarily excreted and cleared by the kidneys, and higher plasma leptin levels are related to chronic kidney disease (CKD).8 Leptin is also involved in pancreatic β-cell proliferation and insulin secretion. Resistin is a peptide hormone that is an important, potential link between obesity and IR. Steppan et al.9 suggested that the neutralization of resistin using an anti-resistin antibody enhanced insulinstimulated glucose uptake. Other studies also found that resistin level is increased during adipocyte differentiation10 in mice, but resistin also plays a role in the inhibition of fat formation.11 Increased glycaemic load, protein loss in the peritoneum, dyslipidaemia and micro-inflammation are also involved in IR in PD patients. However, the interaction among these factors is controversial, especially the correlation between IR and leptin levels in PD patients. Whether IR and leptin are related to complications, prognosis, micro-inflammation, nutrition, lipid metabolism or dialysis adequacy is not clear. Therefore, in the current study, we observed these various parameters in non-diabetic and diabetic continuous ambulatory peritoneal dialysis (CAPD) patients in order to investigate the relationship between clinical and plasma biochemical parameters and IR. The mortality of the Low HOMA, High HOMA and DM groups was also compared.
Patients with various autoimmune diseases, signs of peritonitis or infection, or taking hormones or immunosuppressants, or who received a plasma transfusion or albumin during the 2 weeks prior to participant selection the study were excluded. Patients with serious cardiac insufficiency (New York Heart Association class III or greater), abnormal liver dysfunction and malignancies were also excluded.
Follow-up and endpoints Patients were followed up every month for 2 years (until August 1, 2013), unless one of the following incidents occurred: death, renal transplant, changes to haemodialysis, or treatment abandoning. Patients with the last three situations were deemed lost to follow-up, and they were statistically analyzed as censored data.
Groups Homeostasis model assessment of insulin resistance (HOMAIR) was calculated according to the following formula: HOMA-IR = fasting blood glucose (mmol/L) × fasting insulin (FINS, μmol/mL)/22.5. Patients were divided evenly into a high-level group (H group, HOMA-IR ≥ 2.05) and a lowlevel group (L group, HOMA-IR < 2.05) based on the median value of HOMA-IR. The H group was also defined as the insulin-resistance group. Type 2 diabetic patients (high blood glucose before PD and definitively diagnosed with type 2 diabetes) who were treated with PD were defined as the diabetes mellitus (DM) group simultaneously.
Clinical indices in detection
MATERIALS AND METHODS Participants The non-diabetic patients who receiving PD in peritoneal dialysis centers of Renji hospital were recruited between July 1 to August 1, 2011. Basic information was recorded, and peritoneal equilibrium tests and dialysis adequacy measurements were performed in three months. Type 2 diabetic patients who were treated with PD during the same period were selected as controls. Demographic and clinical data were recorded at study entry. Various indicators were detected and recorded at the start of the study. Ethics approval was granted from the Ethics Committee of Renji Hospital, Shanghai Jiao Tong University School of Medicine.
Inclusion and exclusion criteria Adult patients (>18 years old) who were on PD for more than 3 months were considered eligible, and regular follow-up was conducted every month using outpatient follow-up, home visits or telephone contacts. 618
Patients’ characteristics in general practice, including age, sex, height, weight and blood pressure, were recorded. Subjective global assessment (SGA) of nutrition analysis was conducted using the following conventional approaches. BMI, diet situations, gastrointestinal conditions, activities of daily living and disease response to stress were detected. Subcutaneous fat (skinfold thickness) over the left triceps and mid-arm muscle circumference12 (25.3 cm for males and 23.2 cm for females) were measured using conventional approaches. Edema of the ankles and sacral regions were noted. All of these indices were divided into A, B and C grades (ranked from good to worse). Previous investigations and human body measurements delineated that the presence of more than five of the above eight factors in grades B or C was defined as moderately malnourished (B) or severe malnutrition (C), and the other subjects were in good health. The National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III) Criteria for metabolic syndrome (MS) was adopted. The number of MS factors for each patient was recorded and compared among these groups. © 2015 Asian Pacific Society of Nephrology
Effects of peritoneal dialysis
All the patients were examined for 24-h urine and dialysate, dialysate glucose concentrations, exchange volume and frequency, dialysate calcium concentration, collections and dwell time were examined during follow-up studies. Concentrations of creatinine (Cr), blood urea nitrogen (BUN), urea, fasting blood glucose (FBG) and 2 hour postprandial blood glucose were detected. Total weekly urea clearance (Kt/V), renal Kt/V, total weekly creatinine clearance (WCcr) and renal Ccr were calculated using the urea kinetic model (UKM). The protein catabolic rate (PCR) and normalized protein catabolic rate (nPCR) was calculated. In addition, 2.5% glucose dialysis was conducted using a standard peritoneal equilibration test (PET). Peritoneal transport characteristics were detected, and the D/D0 (peritoneal glucose absorption) and D/Pcr were recorded at four hours. Glucose concentration and absorption of dialysis solutions were calculated as Bodnar DM et al. suggested13. Furthermore, drug use, including erythropoietin, iron supplements, hypertension drugs, vitamin D, statins, phosphorus-reducing drugs and others were recorded at the beginning of the study.
Biochemical parameter detection Serum creatinine (SCr), blood leukocytes, hemoglobin, albumin (Alb), calcium, phosphorus, intact parathyroid hormone (iPTH), serum potassium and sodium, chlorine iron, ferritin, total iron-binding capacity, high sensitive C-reactive protein (hsCRP), FINS, glycated albumin (GA), and CA125 (chemiluminescent immunoassay technique) were determined at the outset of this study in all patients.
Leptin, adiponectin and resistin detection Determination of leptin, adiponectin and resistin concentrations were performed according to the instruction manual of the Human Leptin (STH01090)/ Adiponectin (STH01002)/ Resistin (STH01145) ELISA Kit (Shanghai Sunteam Biotech, Shanghai, China) based on the double antibody sandwich ABC-ELISA method.
RESULTS The baseline characteristics The average age of 157 patients in the L group was younger than the DM and H groups (P < 0.001). There was a total of 79 males, including 17 in the DM (n = 26), 28 in the H (n = 66) and 34 in the L groups (n = 65). There was no significant difference in gender distribution among three groups. Dialysis duration of the DM, H and L groups were 33.69 ± 24.11, 36.29 ± 29.71 and 36.80 ± 30.35 months, respectively. There were no significant differences (P > 0.05) among these three groups. The BMI of the L group was lower than that of the DM and H groups (P = 0.007, P < 0.001). The mean arterial pressure of the L group was higher than that of the H group (P < 0.001). The nPCR of the L group was higher than that of the DM and H groups (P = 0.012, P = 0.019). The prevalence of MS and number of its factors in the DM and H groups were significantly higher than the L group (all P < 0.001). There were no significant differences in SGA, dialysis duration, various drugs (included statins) used, calcium density, glucose absorption, or glucose concentration in the dialysate among these three groups (Tables 1,2). Table 3 shows that there was no significant proportion difference in primary disease between the H and L groups.
Blood lipid and glucose level comparisons Blood glucose levels in the L group were lower than in the DM and H groups (P < 0.001, P = 0.017). Blood glucose levels in the DM group were much higher than in the H group (P < 0.001). The GA values of the DM group were higher than in the H and L groups (P < 0.001). The total cholesterol levels of the DM group were higher than those of the other two groups (P = 0.036, P = 0.037). Triglycerides levels in the H group were higher than those of the DM and L groups (P = 0.027, P < 0.001). HDL-C levels were lowest in the H group, and it was highest in the L group. There were no significant differences in low density lipoprotein (LDL) levels among groups (Table 1).
Comparison of other laboratory indexes Statistical approaches Normal distribution was confirmed before any statistical analyses were conducted, and a P > 0.05 was considered normally distributed. Simple linear regression and multiple linear regression analyses were performed for partial correlation analysis. Data analysis was performed using SPSS 17.0 statistical software (SPSS, Chicago, IL, USA). A two-sided P-value of 0.05 or less was considered statistically significant for all tests. © 2015 Asian Pacific Society of Nephrology
Blood WBCs, uric acid and CA125 values in the H group were much higher than in the L group (P < 0.001, P = 0.011, and P = 0.011, respectively) (Table 1).
Leptin, adiponectin and resistin comparisons Leptin values in the H group were significantly higher than in the DM and L groups (P = 0.005, P < 0.001). Adiponectin level was lower in the H group than in the L group (P < 0.001). Resistin levels in these three groups were similar. 619
L Cao et al.
Table 1 Baseline characteristics of all peritoneal dialysis patients Variable
Total (mean ± SD)
Diabetes mellitus group
High group IR > 2.05
Age (years) Amount (male %) Height (cm) Body mass index (kg/m2) Mean artery pressure (mmHg) Dialysis duration (month) Normalized protein catabolic rate (nPCR) Metabolic syndrome (%) Number of factors HOMA-IR Leptin (ng/mL) Adiponectin (ug/mL) Resistin (ng/mL)
55.24 ± 15.23 157 (79) 162.78 ± 8.05 23.28 ± 3.62 100.62 ± 15.00 36.07 ± 28.98 0.83 ± 0.19 76 (48.7) 2.44 ± 0.98
61.15 ± 12.97### 26 (17) 162.46 ± 8.14 24.14 ± 3.45## 100.05 ± 12.27 33.69 ± 24.11 0.77 ± 0.15# 19 (73.1) ### 2.88 ± 0.77###
19.40 ± 11.59 12.86 ± 6.18 33.77 ± 15.15
18.19 ± 11.62 12.38 ± 6.30 30.96 ± 16.13
58.44 ± 15.50### 66 (28) 161.30 ± 7.67# 24.26 ± 3.64### 96.54 ± 14.20### 36.29 ± 29.71 0.80 ± 0.18# 42 (63.6) ### 2.86 ± 0.84### 19.36 ± 27.24### 25.08 ± 9.18*# 11.01 ± 5.67# 36.05 ± 17.87
Subjective global assessment of nutrition (SGA) Normal (%) Moderate malnutrition (%) Severe malnutrition (%)
16 (61.5) 9 (34.6) 1 (3.8)
48 (72.7) 17 (25.8) 1 (1.5)
49 (75.4) 14 (21.5) 2 (3.1)
Drug Anti-hypertension drugs (%) Phosphorus reduction medicine (%) Vitamin D (%) Erythropoietin (%) Chalybeate (%) Statins (%)
26 (100) 22 (84.6) 2 (7.7) 26 (100) 23 (88.5) 7 (26.9)
57 (86.4) 53 (80.3) 7 (10.6) 64 (97) 58 (87.9) 13 (19.7)
60 (92.3) 57 (87.7) 10 (15.4) 63 (96.9) 58 (89.2) 9 (13.8)
8.00 ± 2.82 111.44 ± 19.39 208.81 ± 70.48 417.05 ± 79.42 2.30 ± 0.21 400.63 ± 376.07 5.34 ± 1.19 2.26 ± 1.21 3.31 ± 1.05 1.18 ± 0.35 7.46 ± 3.09
8.45 ± 2.82 114.96 ± 17.14 211.77 ± 74.87 421.85 ± 80.36 2.28 ± 0.25 341.03 ± 338.60 5.57 ± 1.08 2.17 ± 1.20 3.46 ± 0.99 1.21 ± 0.34 10.0 ± 3.19###
13.01 ± 4.22 6.70 ± 17.66 35.02 ± 17.82
15.75 ± 4.70 3.14 ± 5.11 36.40 ± 17.75
8.65 ± 3.26### 112.68 ± 21.53 217.72 ± 74.16 435.47 ± 90.51# 2.34 ± 0.23 401.87 ± 353.88 5.42 ± 1.06* 2.76 ± 1.34*### 3.26 ± 1.03 1.08 ± 0.33 7.53 ± 3.19#*** 5.24 ± 1.87### 12.68 ± 4.71* 10.78 ± 24.01# 40.59 ± 18.97#
7.19 ± 2.09 108.78 ± 17.85 198.71 ± 64.46 398.75 ± 69.09 2.27 ± 0.18 423.21 ± 413.28 5.17 ± 1.34* 1.79 ± 0.84 3.30 ± 1.05 1.28 ± 0.36 6.34 ± 2.21*** 1.28 ± 0.45 12.15 ± 2.79* 3.97 ± 11.6 28.82 ± 14.60
Laboratory index White blood cell (×109/L) Hemaglobin (g/L) Thrombocyte (×1012/L) Uric acid (umol/L) Calcium (mmol/L) Intact parathyroid hormone (pg/mL) Cholesterol (mmol/L) Triglyceride (mmol/L) Low density lipoprotein (mmol/L) High density lipoprotein (mmol/L) 2-hour-fasting blood-glucose (mmol/L) Insulin (IU/L) Glycated albumin (%) High sensitive C-reactive protein (mg/L) CA125 (U/mL)
Lower group IR < 2.05 49.63 ± 14.11 65 (34) 164.42 ± 8.21 21.95 ± 3.26 104.99 ± 15.76 36.80 ± 30.35 0.88 ± 0.21 15 (23.1) 1.83 ± 0.88 6.68 ± 9.80 14.12 ± 11.27 14.94 ± 6.08 32.59 ± 11.12
Compared to the DM group: *means P ≤ 0.05; **means P ≤ 0.01; ***means P ≤ 0.001. Compared to the L group: # means P ≤ 0.05; ## means P ≤ 0.01; ### means P ≤ 0.001.
HOMA-IR, FINS, and hsCRP values were higher in the H group than in the L group (P < 0.001, P < 0.001, and P = 0.027, respectively). There were no obvious differences in total Kt/V, renal Kt/V, renal Ccr, peritoneal transport type, D/D0, D/Pcr or urine volume at 4 h (Table 2) among three groups.
triglycerides (P = 0.039) and age (P = 0.012), but it was negatively correlated to BMI (P = 0.042). A simple linear regression illustrated the relationships between these four factors and HOMA-IR (Fig. 1), and the interactions among these factors were evaluated using multiple linear regression models (Table 4).
HOMA-IR-related factor analysis
Patient mortality
Simple and multiple linear regression analyses showed that HOMA-IR was positively correlated with leptin (P = 0.001),
The L group had better survival rate than the H and DM groups (P = 0.037). Four patients in the DM group died
620
© 2015 Asian Pacific Society of Nephrology
Effects of peritoneal dialysis
Table 2 Peritoneal dialysis related factors Variable
Glucose intake (g/day) Glucose absorption (g/day)
Total (mean ± SD)
Diabetes mellitus group
High group IR > 2.05
Lower group IR < 2.05
65.10 ± 17.67 38.38 ± 12.26
69.27 ± 10.87 40.18 ± 8.06
64.08 ± 15.94 37.84 ± 11.94
64.48 ± 21.17 38.20 ± 13.95
17 (65.4%) 9 (34.6%)
33 (50%) 33 (50%)
33 (50.8%) 32 (49.2%)
1.67 ± 0.28 0.25 ± 0.28 57.42 ± 12.96 13.01 ± 15.11 526.25 ± 674.16 0.61 ± 0.11 0.42 ± 0.08
1.75 ± 0.47 0.40 ± 0.55 61.17 ± 26.55 20.16 ± 27.94 484.92 ± 692.26 0.63 ± 0.10 0.41 ± 0.08
1.71 ± 0.38 0.32 ± 0.49 58.53 ± 22.49 15.62 ± 26.16 400.76 ± 552.16 0.62 ± 0.11 0.41 ± 0.09
2 (7.7%) 9 (34.6%) 4 (15.4%) 11 (42.3%)
4 (6.1%) 21 (31.8%) 6 (9.1%) 35 (53.0%)
3 (4.6%) 19 (29.2%) 5 (7.7%) 38 (58.5%)
Dialysate calcium concentration 1.25 (%) 1.75 (%) Dialysis related indicators Total Kt/V Residual kidney Kt/V Total creatinine clearance (Ccr) Renal Ccr 24-h urine (mL) Peritoneal creatinine clearance rate at 4 h Peritoneal glucose absorption at 4 h
1.72 ± 0.41 0.34 ± 0.49 59.42 ± 23.01 17.08 ± 25.49 455.80 ± 630.68 0.62 ± 0.11 0.41 ± 0.08
Peritoneal transport type High High-average Low- average Low
Table 3 The primary diseases of all peritoneal dialysis patients
Diabetic nephropathy (%) Hypertensive nephropathy (%) Chronic nephritis (%) Obstructive nephropathy (%) Renal tubular diseases (%) Polycystic kidney (%) Reason unknown (%)
Diabetes mellitus group
High group IR > 2.05
Lower group IR < 2.05
24 (92.4) 1 (3.8) 1 (3.8) 0 0 0 0
0 5 (7.6) 30 (45.5) 3 (4.5) 3 (4.5) 3 (4.5) 22 (33.3)
0 3 (4.6) 33 (50.8) 0 1 (1.5) 2 (3.1) 26 (40)
suffered sudden cardiac death, two patients suffered peritonitis infection and one patient suffered lung cancer metastasis). One patient abandoned treatment because of renal transplant, and one patient was transferred to haemodialysis because of peritonitis. Seven patients in the L group died (four patients suffered sudden cardiac death, two patients suffered peritonitis infection and one patient suffered inadequate dialysis). Four patients abandoned treatment because of renal transplantation, and one patient was transferred to haemodialysis because of peritonitis (Fig. 2).
Mortality-related factors Table 4 Multiple regression analysis for homeostasis model assessment of insulin resistance (HOMA-IR)in peritoneal dialysis (PD) patients Variable Age (years) Leptin (ng/mL) 2-hour postprandial blood glucose (mmol/L) Glycated albumin (%) Metabolic syndrome Triglyceride (mmol/L) High density lipoprotein (mmol/L) Normalized protein catabolic rate Body mass index (kg/m2) Log (high sensitive C-reactive protein)
β ± SE
P-value
0.07 ± 0.43 0.311 ± 0.086 0.302 ± 0.492 0.205 ± 0.202 1.539 ± 1.488 0.016 ± 0.008 0.343 ± 2.178 −1.709 ± 3.315 0.465 ± 0.227 −0.309 ± 0.908
0.012 0.001 0.540 0.312 0.303 0.039 0.875 0.607 0.042 0.735
(cardiovascular accident, cerebrovascular accident, peritonitis infection and upper gastrointestinal haemorrhage), and four patients were transferred to haemodialysis because of peritonitis. Eight patients in the H group died (five patients © 2015 Asian Pacific Society of Nephrology
Simple linear regression analysis demonstrated that leptin, dialysis duration, BMI, HOMA-IR, albumin, hsCRP, nPCR and triglycerides were related to mortality. Multiple linear regression analysis (Table 5) demonstrated that dialysis duration (P = 0.000), blood leptin level (P = 0.019), nPCR (P = 0.01) and hsCRP (P = 0.021) were potential prognostic factors for mortality.
DISCUSSION Cardiovascular disease (CVD) remains the main cause of death in patients with ESRD, and it is responsible for approximately 40% of the total mortality in CAPD patients.14 Identification of risk factors for CVD is important for CAPD patients. The gold standard is the insulin clamp technique, which is a time-consuming operation, and few studies use this technique in dialysis patients. However, insulin clamp is the most accurate determination of insulin release. HOMA-IR is the most commonly used method to detect IR in 621
L Cao et al.
Fig. 1 Homeostasis model assessment of insulin resistance (HOMA-IR) related factors. (A) HOMA-IR positively correlated with serum leptin levels. (B) HOMA-IR positively correlated with triglyceride levels. (C) HOMA-IR positively correlated with BMI. (D) HOMA-IR positively correlated with age.
Fig. 2 Kaplan–Meier survival curves in all peritoneal dialysis (PD) patients (P = 0.037). The blue, green and yellow lines represent the DM, , DM; , L; H and L groups, respectively. , H; , DM-censored; , L-censored; , H-censored.
622
© 2015 Asian Pacific Society of Nephrology
Effects of peritoneal dialysis
Table 5 Multiple regression analysis for mortality-related factors in peritoneal dialysis (PD) patients Variable Dialysis duration Leptin Normalized protein catabolic rate High sensitive C-reactive protein
β ± SE
P-value
−0.092 ± 0.024 −0.032 ± 0.014 −7.092 ± 2.762 0.03 ± 0.013
0.000 0.019 0.01 0.021
peritoneal dialysis patients. Our study demonstrated that HOMA-IR was positively correlated with leptin, PPG, GA, triglycerides, the presence of MS, the number of MS factors, age and log hsCRP, and it was negatively associated with HDL, BMI and nPCR. Multiple linear regression analysis revealed that HOMA-IR was positively correlated with age, leptin and triglyceride, and it was negatively correlated with BMI. These results suggested that HOMA-IR may play an important role in the pathophysiology of glucose and lipid metabolism. Our study also showed a good correlation between HOMA-IR and leptin. However, we failed to find any significant association between HOMA-IR and adiponectin or resistin, although these two factors work synergistically as adipokines. This finding suggest, leptin may play a more important role in the occurrence of IR than the other two parameters in PD patients. Independent correlations between BMI, age and IR were also found, which is similar to previous research.15 Although IR is multifactorial in origin, obesity has long been linked with IR. Once people develop obesity, leptin levels increase, which can suppress insulin secretion from the pancrease, and causes leptin resistance.16 Conversely, adiponectin has anti-inflammatory effects and could improve insulin sensitivity. Adiponectin levels were declined in obese people.17 Previous studies found that adiponectin exhibited a cardiovascular protective effect in haemodialysis patients, especially in patients with higher adiponectin.18 Briefly, inflammatory cytokines and adipokines regulate insulin levels under physiological conditions collaboratively and antagonistically. IR may occur when adiponectin and leptin homeostasis is disturbed. Johnson et al.19 found that the MS incidence in PD patients was much higher than pre-haemodialysis and haemodialysis patients, and was associated with a significantly increased risk of future cardiovascular events. A significantly lower incidence of MS in the L group than in the DM and H groups was found in our study. High cardiovascular morbidity and mortality in PD patients were discovered previously.7,20 Similarly, poorer survival in the DM and H groups compared to the L group was found in our follow-up visits. These results indicated that glucose and lipid metabolism in PD is prominent and require intensified attention. Therefore, it may be beneficial to discover the presence of MS and control blood pressure, blood glucose, and cholesterol level of PD patients to improve their survival rates. © 2015 Asian Pacific Society of Nephrology
Multiple regression analysis revealed that dialysis duration, leptin, nPCR and hsCRP were also involved in the mortality of PD patients. Serum hsCRP is a pro-inflammatory factor that is associated with CVD-related morbidity and mortality in PD and haemodialysis patients.21 CRP can also cause protein-energy malnutrition, which is induced by the inhibition of protein synthesis and catabolism, the high level of CRP was directly correlated with IR.22 Blood leptin levels of PD patients were significantly higher than those of haemodialysis patients even after correcting for differences in body fat, age and gender effects. These results may be related to dialysis and result in multiple parameter changes, including increased glucose load, glucose-stimulated insulin secretion, and aggravated hyperinsulinaemia, which finally promote blood leptin secretion.23 A previous study found that icodextin or amino acid dialysis solution could reduce blood leptin level and decrease the degree of IR.24 Pre-inflammatory indicators, including IL-6, IL-1 and metabolites of oxidative stress associated with IR, were significantly increased in PD patients.25 Downstream cascade inhibition of these inflammatory factors may reduce their reactivity to insulin. Insulin is an anti-inflammatory factor that regulates inflammation in vitro and in vivo.26 Therefore, the inflammatory response may aggravate the pathological process in PD patients, especially in people with high HOMAIR. Notably, hsCRP level in our DM group was reduced. This reduction may be the result of insulin therapy, which moderately reduces micro-inflammation. Our study also found that DM patients had higher fasting blood glucose, 2-hour postprandial blood-glucose and GA than the other two groups. A previous study found that GA accurately predicted the risk of death and hospitalizations in patients with diabetes mellitus and ESRD.27 This result suggests that GA may be a more accurate indicator of glucose metabolism in PD patients. Uncontrolled diabetes (especially type 2 diabetes) is an important risk factor for poor prognosis in PD patients,1,28 and the early monitoring and good control of GA may be one way to improve patient survival. Our study demonstrated that high leptin level might be involved in IR in patients undergoing PD. Serum leptin and triglyceride levels, BMI, and age may be risk factors for IR Low HOMA group had better survival rate than DM and H group patients. Leptin level, the duration of dialysis, leptin level, nPCR and hsCRP may also related to patient mortality in non-diabetic and diabetic PD patients.
ACKNOWLEDGEMENT This study was supported in part by the National Basic Research Program of China 973 Program No. 2012CB517600 (No. 2012CB517602). The study was also sponsored by the National Natural Science Foundation of China (81370794, 81102700, 81373865). 623
L Cao et al.
REFERENCES 1. Duong U, Mehrotra R, Molnar MZ et al. Glycemic control and survival in peritoneal dialysis patients with diabetes mellitus. Clin. J. Am. Soc. Nephrol. 2011; 6: 1041–8. 2. Cotovio P, Rocha A, Rodrigues A. Peritoneal dialysis in diabetics: There is room for more. Int. J. Nephrol. 2011; 2011. 3. Goldstein BJ. Insulin resistance as the core defect in type 2 diabetes mellitus. Am. J. Cardiol. 2002; 90: 3–10. 4. Wang T, Heimbürger O, Cheng H, Bergstroem J, Lindholm B. Amphotericin B does not increase peritoneal fluid removal. Adv. Perit. Dial. 1998; 14: 3–10. 5. Szeto C-C, Chow K-M, Leung C-B et al. Increased subcutaneous insulin requirements in diabetic patients recently commenced on peritoneal dialysis. Nephrol. Dial. Transplant. 2007; 22: 1697–702. 6. Tuzcu A, Bahceci M, Yilmaz ME, Turgut C, Kara IH. The determination of insulin sensitivity in hemodialysis and continuous ambulatory peritoneal dialysis in nondiabetic patients with end-stage renal disease. Saudi Med. J. 2005; 26: 786–91. 7. Ahima RS, Flier JS. Leptin. Annu. Rev. Physiol. 2000; 62: 413–37. 8. Shankar A, Syamala S, Xiao J, Muntner P. Relationship between plasma leptin level and chronic kidney disease. Int. J. Nephrol. 2012; 2012: 269532. 9. Steppan CM, Bailey ST, Bhat S et al. The hormone resistin links obesity to diabetes. Nature 2001; 409: 307–12. 10. Janke J, Engeli S, Gorzelniak K, Luft FC, Sharma AM. Resistin gene expression in human adipocytes is not related to insulin resistance. Obes. Res. 2002; 10: 1–5. 11. Pang S, Le Y. Role of resistin in inflammation and inflammation-related diseases. Cell. Mol. Immunol. 2006; 3: 29–34. 12. Gawkrodger D, Ferguson A, Barnetson R. Nutritional status in patients with dermatitis herpetiformis. Am. J. Clin. Nutr. 1988; 48: 355–60. 13. Bodnar DM, Busch S, Fuchs J et al. Estimating glucose absorption in peritoneal dialysis using peritoneal equilibratin tests. Adv. Perit. Dial. 1993; 9: 114–8. 14. Krishnasamy R, Badve SV, Hawley CM et al. Daily variation in death in patients treated by long-term dialysis: Comparison of in-center hemodialysis to peritoneal and home hemodialysis. Am. J. Kidney Dis. 2013; 61: 96–103. 15. Henriques VT, Martinez EZ, Divino-Filho JC, Pecoits-Filho R, Cardeal da Costa JA. Increase in BMI over time is associated with
624
16. 17. 18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
fluid overload and signs of wasting in incident peritoneal dialysis patients. J. Ren. Nutr. 2013; 23: e51–7. Enriori PJ, Evans AE, Sinnayah P, Cowley MA. Leptin resistance and obesity. Obesity 2006; 14: 254S–258S. Kadowaki T, Yamauchi T. Adiponectin and adiponectin receptors. Endocr. Rev. 2005; 26: 439–51. Amira OC, Naicker S, Manga P et al. Adiponectin and atherosclerosis risk factors in African hemodialysis patients: A population at low risk for atherosclerotic cardiovascular disease. Hemodial. Int. 2012; 16: 59–68. Johnson DW, Armstrong K, Campbell SB et al. Metabolic syndrome in severe chronic kidney disease: Prevalence, predictors, prognostic significance and effects of risk factor modification. Nephrology 2007; 12: 391–8. Rossum C, Hoebee B, Baak MA, Mars M, Saris WH, Seidell JC. Genetic variation in the leptin receptor gene, leptin, and weight gain in young Dutch adults. Obes. Res. 2003; 11: 377–86. Wang AY-M, Woo J, Lam CW-K et al. Is a single time point C-reactive protein predictive of outcome in peritoneal dialysis patients? J. Am. Soc. Nephrol. 2003; 14: 1871–9. McLaughlin T, Abbasi F, Lamendola C et al. Differentiation between obesity and insulin resistance in the association with C-reactive protein. Circulation 2002; 106: 2908–12. Johansen KL, Mulligan K, Tai V, Schambelan M. Leptin, body composition, and indices of malnutrition in patients on dialysis. J. Am. Soc. Nephrol. 1998; 9: 1080–84. Furuya R, Odamaki M, Kumagai H, Hishida A. Beneficial effects of icodextrin on plasma level of adipocytokines in peritoneal dialysis patients. Nephrol. Dial. Transplant. 2006; 21: 494–8. Nascimento MM, Suliman ME, Silva M et al. Effect of oral N-acetylcysteine treatment on plasma inflammatory and oxidative stress markers in peritoneal dialysis patients: A placebo-controlled study. Perit. Dial. Int. 2010; 30: 336–42. Hyun E, Ramachandran R, Hollenberg MD, Vergnolle N. Mechanisms behind the anti-inflammatory actions of insulin. Crit. Rev. Immunol. 2011; 31: 307–40. Freedman BI, Andries L, Shihabi ZK et al. Glycated albumin and risk of death and hospitalizations in diabetic dialysis patients. Clin. J. Am. Soc. Nephrol. 2011; 6: 1635–43. Fang W, Yang X, Kothari J et al. Patient and technique survival of diabetics on peritoneal dialysis: One-center’s experience and review of the literature. Clin. Nephrol. 2008; 69: 193–200.
© 2015 Asian Pacific Society of Nephrology
