http://informahealthcare.com/rnf ISSN: 0886-022X (print), 1525-6049 (electronic) Ren Fail, 2015; 37(1): 66–72 ! 2015 Informa Healthcare USA, Inc. DOI: 10.3109/0886022X.2014.964147
CLINICAL STUDY
Body composition monitoring and nutrition in maintenance hemodialysis and CAPD patients—a multicenter longitudinal study Sharon Mathew1, Georgi Abraham1,2,3, Madhusudan Vijayan4, Thigarajan Thandavan3, Milly Mathew1, Ilangovan Veerappan1, Laxmi Revathy1, and Merina E. Alex2 1
Department of Nephrology and Nutrition, Pondicherry Institute of Medical Sciences, Puducherry, India, 2Madras Medical Mission, Chennai, India, Tanker Foundation, Chennai, India, and 4Government Kilpauk Medical College, Chennai, India
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
3
Abstract
Keywords
Hydration and nutritional status of end stage renal disease (ESRD) patients are linked to increased morbidity and mortality. Body composition monitoring (BCM) by multi-frequency bioimpedance spectroscopy (MFBS) is considered to be a superior modality of fluid assessment in chronic kidney disease (CKD) dialysis. We did a longitudinal prospective study in South India on maintenance hemodialysis (MHD) and continuous ambulatory peritoneal dialysis (CAPD) patients over 24 months and looked at impact of baseline nutritional parameters and body composition parameters on 24-month mortality. Ninety-nine patients stable on dialysis for at least 3 months were recruited (MHD 85, CAPD 14) at baseline and at 24 months, 41 were alive and 33 had expired, 12 had undergone renal transplant and 13 were lost to follow-up. BCM and nutritional assessment were done at baseline and at follow-up. Baseline overhydration (OH) differed significantly between surviving and dead patients (p50.05). Receiver operating characteristic (ROC) curve between OH and mortality showed that the best cut-off point to differentiate between survived and expired patients was 3.15 L. ROC curve for BMI showed lower than cut-off of 22.65 kg/m2 to predict mortality with sensitivity 41.30% and specificity 81.81%. At follow-up, triceps skin fold thickness (TSF), biceps skin fold thickness (BSF) and mid arm circumference (MAC) increased significantly from baseline (p50.001, p ¼ 0.001 and p50.001, respectively). Overhydration and BMI are important predictors of mortality in dialysis patients. Improvement in anthropometric markers TSF, BSF and MAC in MHD patients was associated with survival.
Hemodialysis, multi-frequency bioimpedance spectroscopy, peritoneal dialysis, overhydration, survival
Introduction Cardiovascular diseases represent the leading cause of death in dialysis patients, both in maintenance hemodialysis (MHD) and continuous ambulatory peritoneal dialysis (CAPD) patients.1–4 This is partly attributed to fluid disturbances with abnormalities in salt and water intake which are highly prevalent in end stage renal disease (ESRD) patients.4 As a result of lack of trained renal dieticians in India, the dialysis patients are inappropriately advised regarding calorie, protein, micronutrients, salt and water intake. Malnutrition has been widely reported in Indian patients on dialysis.5,6 Malnutrition in CKD is associated with increased morbidity, mortality and poor quality of life.7,8 However, there is a paucity of multicentric longitudinal data on hydration and nutritional status of Indian dialysis patients and its effect on all-cause mortality.
Address correspondence to Georgi Abraham, Madras Medical Mission, No 4-A, Dr. J. Jayalalitha Nagar, Mogappair, Chennai-600037, India. Tel.: +91-9841710992; E-mail: [email protected]
History Received 31 May 2014 Revised 6 August 2014 Accepted 30 August 2014 Published online 24 September 2014
Renal replacement therapy (RRT) modalities comprising MHD and CAPD often influence the fluid and salt content of the patients. Excessive removal of fluid during dialysis leads to hypotension, and inadequate removal often leads to hypertension with left ventricular hypertrophy (LVH) and cardiac failure.4 Therefore, assessment of nutritional and hydration status, especially hypervolemia, can play an important role in the management of fluid overload in dialysis patients.4 Clinical assessment of fluid status is often inaccurate. Body composition monitor (BCM; Fresenius Medical Care, Bad Homburg, Germany) is a unique device based on multifrequency bioimpedance spectroscopy (MFBS) which measures overhydration (OH), total body water (TBW), extracellular water (ECW), intracellular water (ICW), fat tissue mass (FTM), lean tissue mass (LTM) and body cell mass (BCM). BCM has been validated against the following gold standard reference methods: bromide dilution for ECW, total body potassium (TBK) for ICW, deuterium dilution for TBW, dual energy X-ray absorptiometry (DEXA) for LTM, four compartment modeling, air displacement plethysmography, under water weighing for adipose tissue mass, magnet resonance tomography for BCM and an expert clinical assessment for overhydration (OH).9 BCM has been assessed
Maintenance hemodialysis and CAPD patients
DOI: 10.3109/0886022X.2014.964147
previously in ESRD patients, but has never been utilized in Indian dialysis patients. This prospective longitudinal multicentric study was conducted to assess the nutritional and hydration status of prevalent MHD and CAPD patients and to observe the impact of baseline characteristics of the above in predicting 2-year mortality.
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Subjects and methods After getting approval from the ethics committee, the patients were enrolled from three centers: Madras Medical Mission Hospital, TANKER foundation (www.tankerfoundation.org) and Pondicherry Institute of Medical Sciences. Those patients with ESRD who were older than 18 years and survived for first 3 months after initiation of dialysis were included in the study. Excluded patients were those with permanent cardiac pacemaker, patients with liver cirrhosis, infections such as TB, HIV, malignancy and patients on irregular dialysis due to financial constraints. Informed written consent was obtained from all the patients included in the study. The study was done at baseline in 99 patients (78 males and 21 females) in February 2010, and the follow-up study was done at 24 months in the surviving patients. Out of the 99 patients included, 85 (85.86%) were on hemodialysis (MHD) and 14 (14.14%) were on CAPD. In the CAPD patients, nine (64.29%) were on four exchanges of 2 L per day and five (35.71%) were on three exchanges of 2 L per day. In the MHD patients, 63 (74.12%) were on thrice weekly dialysis and rest (25.88%) were on biweekly dialysis, with the duration of each dialysis session being 4 h. Clinical, biochemical, nutritional and body composition measurements were made to assess the nutritional and fluid status of the patients at baseline and at 2 years. In the MHD group, the measurements were made before the onset of the dialysis session. In the CAPD group, the measurements were made after emptying the peritoneal cavity of fluid, before the morning exchange. Clinical characteristics collected were mode of dialysis, diabetes mellitus status, systolic and diastolic BP, urine output, dry weight dietary pattern (vegetarian/non-vegetarian), and daily calorie and protein intake. Biochemical parameters measured were serum potassium, blood urea and serum creatinine. Nutritional parameters assessed included body mass index (BMI), triceps skin fold thickness (TSF), biceps skin fold thickness (BSF), mid arm circumference (MAC), serum albumin and hemoglobin concentration. Body composition monitor parameters estimated were TBW, ECW, ICW, lean tissue index (LTI), fat tissue index (FTI), LTM, fat in kg, relative fat in percentage, adipose tissue mass (ATM) and BCM. The mean of two measurements was taken each time. The baseline characteristics were compared between the patients who survived and died, and a receiver operating characteristic (ROC) curve was constructed to determine the cut-off for significant parameters. The baseline characteristics were compared between MHD and CAPD patients. For the surviving patients, the baseline and follow-up characteristics were compared. The patients who underwent transplant and patients lost to follow-up were excluded from the repeat measurement at 2 years.
67
Statistical analysis Continuous variables were described in the form of mean, standard deviation, median and interquartile range. Dichotomous variables (yes/no) were summarized in terms of percentages. To compare two related continuous data samples (like for comparison of baseline and follow-up in our study), paired t test for normally distributed data and Wilcoxon signed rank test for non-normally distributed data were used. To compare two independent samples means (comparison of MHD and CAPD patients), independent t test for normally distributed data and Mann–Whitney U test for non-normally distributed data were used. Chi square test was used to compare proportions. Pearson’s correlation and Spearman’s correlation methods were used to determine the correlation coefficient between variables for normally and non-normally distributed data, respectively. ROC curve to determine the sensitivity and specificity of overhydration and BMI for prediction of mortality was used.
Results The study had 99 patients enrolled, and at the end of 24 months, 41 patients (MHD 33/CAPD 8) had survived, 33 had died, 12 patients had undergone renal transplant and 13 patients were lost to follow-up. The outcomes of the recruited patients are shown in Figure 1. Table 1 shows the baseline characteristics of the patients. The patients who had expired during the period of 2 years had higher degree of OH and lower BMI at baseline, compared to patients who had survived at 2 years (p ¼ 0.02 and 0.017, respectively). By ROC curve analysis of OH (Figure 2), the area under the curve was greater than 50%. Sensitivity– specificity analysis showed the best cut-off point for baseline OH between the survived and dead patients was 3.15 L with a sensitivity of 39.47% and specificity of 69.70%. By ROC curve analysis (Figure 3), the best cut-off for baseline BMI was found to be 22.65 kg/m2, area under the curve450% with sensitivity of 41.30% and specificity of 81.81%. While comparing the baseline parameters of survived and expired patients for MHD and CAPD patients separately, BMI and OH were significant in MHD patients (p ¼ 0.033 and 0.018, respectively). Table 2 shows the comparison of the variables of survived patients at baseline and at follow-up. On analyzing the body composition of the surviving patients at 2 years, it was found that ICW and OH were significantly different at 2 years compared with their baseline values (p ¼ 0.022 and p50.001, respectively). Among the nutritional parameters, TSF, BSF and MAC were found to be significantly improved (p 0.001, p50.001 and p 0.001) among those who had survived. Tables 3 and 4 show the comparison between MHD and CAPD patients. Lean tissue index correlated with BMI (r ¼ 0.209, p ¼ 0.042). We also looked for correlation between baseline lean tissue index and serum albumin as both are important indicators of nutrition and we found a weak positive correlation between the two, though not significant ( ¼ 0.194, p ¼ 0.153). On analyzing the correlation between OH and clinically determined weight gain by Spearman’s correlation method, we observed a correlation of ¼ 0.159
68
S. Mathew et al.
Ren Fail, 2015; 37(1): 66–72
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Figure 1. Flowchart showing outcomes of the recruited patients at follow-up.
Table 1. Baseline characteristics of MHD and CAPD patients.
Variables* Age (years) Gender Males (%) Females (%) Weight (kg) Height (cm) BMI (kg/m2) Mode of dialysis MHD (%) CAPD (%) Diabetic (%) Systolic BP (mmHg) Diastolic BP (mmHg) Urine output (mL) Dry weight (kg) Diet Vegetarian (%) Non-vegetarian (%) TSF (cm) BSF (cm) MAC (cm) Serum albumin (g/dL) Haemoglobin (g/dL) S.Potassium (mEq/L) Urea (mg/dL) Creatinine (mg/dL) eGFR BCM parameters Overhydration (L) TBW (L) ECW (L) ICW (L) LTI (kg/m2) FTI (kg/m2) LTM (kg) Fat (kg) Relative fat (kg) ATM (kg) BCM (kg)
All patients (n ¼ 99)
Survived (n ¼ 41)
55.26 ± 12.5
52.20 ± 13.1
54.76 ± 7.6
0.32
78 (78.79%) 21 59.05 ± 12.6 152.83 ± 9.0 22.23 ± 4.2
31 (75.6%) 10 56.52 ± 12.6 160.61 ± 8.9 22.68 ± 4.2
26 (78.8%) 7 54.55 ± 9.6 162.82 ± 9.5 20.53 ± 3.1
0.75
85 (86.14%) 14 (13.86%) 39 (39.4%) 147.45 ± 21.8 83.55 ± 12.2 100 (0–500)* 56.00 ± 12.3
33 (80.5%) 8 (19.5%) 15(36.6%) 147.49 ± 23.4 82.76 ± 12.1 312.20 ± 386.6 55.77 ± 12.7
27 (81.8%) 6 (18.2%) 15 (45.5%) 150.21 ± 21.2 84.48 ± 11.4 180 (0–700) 51.78 ± 9.2
22 (22.22%) 77 6 (4–8)* 4 (2–8)* 16 (10–23)* 3.6 ± 0.6 8.88 ± 1.8 4.76 ± 1.04 83 (6–100)* 6.8 ± 3.2 12.1 ± 11.59
13 (31.7%) 28 6 (4–8)* 4 (2–6)* 16.38 ± 7 3.52 ± 0.5 8.9 ± 2 4.80 ± 1.1 81.80 ± 29.3 7.23 ± 3.0 11.78 ± 13.43
6 (18.2%) 27 6 (3–8) 4 (2–6) 17.29 ± 6.9 3.52 ± 0.5 8.66 ± 1.6 4.68 ± 0.9 81.39 ± 28.8 6.28 ± 3.0 12.49 ± 8.98
0.42 0.87 0.62 1 0.58 0.62 0.95 0.18 0.8
2.90 (1.50–4.05)* 28.23 ± 6.9 14.42 ± 3.3 13.85 ± 3.9 10.48 ± 3.0 10.87 ± 4.3 27.49 ± 9.7 20.31 ± 7.8 34.65 ± 11 27.64 ± 10.6 14.00 ± 6.5
3.9 (2.65–4.80)* 28.55 ± 6.5 15.03 ± 3.4 13.62 ± 3.5 10.01 ± 2.3 9.07 ± 4.1 27.01 ± 8.3 17.45 ± 7.3 31.50 ± 10.7 23.74 ± 9.9 13.43 ± 5.3
0.02 0.84 0.52 0.79 0.46 0.07 0.82 0.11 0.22 0.11 0.69
3.1 (1.85–4.30)* 29.41 ± 7.1 14.91 ± 3.1 14.40 ± 4.0 10.52 ± 2.8 10.45 ± 4.6 28.30 ± 9.2 20.09 ± 8.5 33.55 ± 10.8 27.33 ± 11.5 14.39 ± 6.1
Died (n ¼ 33)
p Value for comparison between survived and died
0.46 0.31 0.017 0.89 0.44 0.61 0.53 0.48 0.13 0.186
Notes: MHD: Hemodialysis, CAPD: continuous ambulatory peritoneal dialysis, BMI: body mass index, TBW: total body water, ECW: extracellular water, ICW: intracellular water, LTI: lean tissue index, FTI: fat tissue index, LTM: lean tissue mass, ATM: adipose tissue mass, BCM: body cell mass, TSF: triceps skin fold, BSF: biceps skin fold, MAC: mid arm circumference. *Median (IQR) for all continuous variables, comparison by Mann–Whitney U test; percentages for categorical variables, comparison by Chi square test.
Maintenance hemodialysis and CAPD patients
DOI: 10.3109/0886022X.2014.964147
which was just short of being statistically significant (p ¼ 0.063). Multiple regression analysis was used to study the effect of various predictors on mortality, and the results are shown
69
in Table 5. The predictors were chosen on the basis of p value less than 0.1 in the comparison between survived and expired patients. It was found that OH was a significant predictor of mortality (p ¼ 0.042), if BMI and fat tissue index (FTI) were kept constant. There was no association between frequency of dialysis (twice/thrice weekly) and survival in MHD patients (p ¼ 0.61). We also compared demographic, body composition and nutritional parameters at baseline and follow-up for MHD and CAPD patients and are shown in Tables 3 and 4. Hemoglobin concentration was statistically significant in the survived CAPD patients compared with MHD patients at follow-up (p ¼ 0.014).
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Discussion
Figure 2. ROC curve between overhydration and mortality. Note: Area under the curve 450%, estimated cut off ¼ 3.15 L, sensitivity ¼ 39.47%, specificity ¼ 69.70%.
Figure 3. ROC curve between body mass index and mortality. Note: Area under the curve 450%, estimated cut off ¼ 22.65 kg/m2, sensitivity ¼ 41.30%, specificity ¼ 81.81%.
BCM by MFBS is an upcoming tool in the measurement of overhydration (OH) in fluid overload conditions such as CKD and congestive cardiac failure (CCF). It has been validated as the gold standard in the measurement of overhydration in CKD patients on MHD.9 In our study, baseline OH (43.15 L) (p 0.05) was found to have a significant correlation with 2-year mortality. This is in accordance to the study done by Wizemann et al., in which it was found that out of 269 chronic MHD patients, 58 patients were overhydrated and 41% of these patients died when followed-up for a period of 3.5 years.4 In the study done by Wabel et al., 25% of MHD patients suffered from overhydration of 42.5 L.9 The ROC curve analysis between body composition and nutritional parameters with mortality revealed OH as the only BCM parameter with a statistically significant area under the curve greater than 50% (p 0.05). Sensitivity–specificity analysis showed the best cut-off point for baseline overhydration to predict 2-year mortality was 3.15 L. In the patients who had survived, there was a significant reduction in the level of OH (p50.001). OH is a risk factor for cardiac dysfunction (LVH, MI, CCF), increased incidence of intradialytic morbid events and sudden death.10,11 Excessive ultrafiltration rate can lead
Table 2. Comparison of body composition and nutritional parameters of survived MHD and CAPD patients between baseline and follow-up. At baseline (n ¼ 41) Variables
Mean (SD)
TBW (L) ECW (L) ICW (L) LTI (kg/m2) FTI (kg/m2) LTM (kg) Fat (kg) Relative fat (%) ATM (kg) BCM (kg) OH (L) TSF (cm) BSF (cm) MAC (cm) S.albumin (gm/dL) Hb (gm/dL)
28.23 ± 6.91 14.42 ± 3.33 13.86 ± 3.97 10.28 ± 3.07 11.63 ± 3.76 26.54 ± 9.71 21.60 ± 6.99 36.68 ± 9.40 29.40 ± 9.49 13.44 ± 6.48 2.73 ± 1.87 6.66 ± 3.81 5.23 ± 4.36 16.38 ± 7.03 3.52 ± 0.50 8.94 ± 2.03
Median (IQR) 28.3 14.3 13 10.1 11.8 27.0 22.2 36 30.3 12.8 2.9 6 4 12 3.65 8.3
(22.9–34.1) (11.8–17.0) (10.4–7.15) (8.2–12.3) (7.2–13.8) (19.4–37.4) (13.60–26.1) (29.9–44.4) (18.5–35.5) (9–19.3) (1.5–4.1) (4–8) (2–6) (11–23) (3.12–3.9) (7.6–10.6)
At follow-up (n ¼ 41) Mean (SD)
Median (IQR)
p Value
29.94 ± 8.1 13.88 ± 2.8 16.03 ± 6.4 10.95 ± 3.0 12.18 ± 5.8 27.97 ± 8.7 22.42 ± 11.8 35.96 ± 11.8 30.77 ± 14.2 14.48 ± 6.0 0.95 ± 2.61 10 (6–25) 10 (4–22) 22.15 ± 6.4 3.44 ± 0.5 8.88 ± 1.6
27.60 (19.9–53.4) 13.3 (9.7–20.8) 13.9 (10.1–37.4) 10.4 (7–19) 12.7 (3–26) 25.2 (17–55) 22.5 (6–43) 37.4 (11–54) 30.9 (7–58) 13.00 (7–33) 1.40 (0.2–2.2) 14.66 ± 9.48 12.54 ± 10.13 23.00 3.4 8.8
0.140 0.171 0.022* 0.188 0.730 0.302 0.839 0.544 0.726 0.277 50.001* 50.001* 50.001* 50.001* 0.905 0.877
Notes: MHD: hemodialysis, CAPD: continuous ambulatory peritoneal dialysis, TBW: total body water, ECW: extracellular water, ICW: intracellular water, LTI: lean tissue index, FTI: fat tissue index, LTM: lean tissue mass, ATM: adipose tissue mass, BCM: body cell mass, OH: overhydration, TSF: triceps skin fold, BSF: biceps skin fold, MAC: mid arm circumference. *50.05 significant.
70
S. Mathew et al.
Ren Fail, 2015; 37(1): 66–72
Table 3. Comparison of demography, body composition and nutrition parameters at baseline for MHD and CAPD patients.
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Variables* Age (years) Gender Males (%) Females (%) Weight (kg) Height (cm) BMI (kg/m2) Diabetic (%) Systolic BP (mmHg) Diastolic BP (mmHg) Urine output (mL) Dry weight (kg) Diet Vegetarian Non-vegetarian TSF (cm) BSF (cm) MAC (cm) S. albumin (gm/dL) Haemoglobin (gm/dL) S. Potassium (mEq/L) Urea (mg/dL) Creatinine (mg/dL) Overhydration (L) TBW (L) ECW (L) ICW (L) LTI (kg/m2) FTI (kg/m2) LTM (kg) Relative LTM (%) Fat (kg) Relative fat (%) ATM (kg) BCM (kg)
MHD (n ¼ 85)
CAPD (n ¼ 14)
p Value
50.65 ± 12.23
60.43 ± 12.15
0.007
68 (80.0%) 17 (20.0%) 58.75 ± 13.02 163.58 ± 8.90 21.90 ± 4.30 31 (36.5%) 147.73 ± 21.32 84.67 ± 12.23 100 (0–450) 55.71 ± 12.71
10 (73.3%) 4 (26.7%) 60.89 ± 10.48 158.29 ± 8.82 24.24 ± 3.18 8 (57.14) 145.79 ± 25.50 76.71 ± 10.70 200 (0–762) 57.75 ± 10.34
19 (22.4%) 66 (77.6%) 6 (3–8)* 4 (2–8)* 16.12 ± 6.75 3.67 ± 0.66 8.75 ± 1.81 4.86 ± 0.96 83 (64–102)* 6.77 ± 3.16 3.2 (2.0–4.4)* 29.75 ± 7.20) 15.16 ± 3.40 14.61 ± 4.13 10.66 ± 2.79 9.94 ± 4.60 28.92 ± 9.19 50.09 ± 13.85 19.32 ± 8.62 32.34 ± 10.64 26.28 ± 11.72 14.76 ± 6.08
4 (26.7%) 10 (73.3%) 5.71 ± 2.09 4 (2–5.25)* 19.39 ± 6.92 3.31 ± 0.4 9.67 ± 1.80 4.17 ± 1.33 62 (45.5–94.75)* 6.92 ± 3.45 2.75 (0.57–3.6)* 27.36 ± 6.92 14.22 ± 3.38 13.14 ± 3.67 9.71 ± 2.86 13.41 ± 3.68 24.74 ± 9.20 40.25 ± 11.28 24.74 ± 6.44 40.55 ± 9.19 33.36 ± 8.7 12.23 ± 6.13
0.51 0.56 0.04 0.05 0.26 0.76 0.02 0.32 0.57 0.74 0.92 0.48 0.10 0.09 0.09 0.02 0.09 0.87 0.16 0.25 0.34 0.22 0.24 0.009 0.12 0.014 0.034 0.008 0.034 0.16
Notes: MHD: hemodialysis, CAPD: continuous ambulatory peritoneal dialysis, BMI: body mass index, TBW: total body water, ECW: extracellular water, ICW: intracellular water, LTI: lean tissue index, FTI: fat tissue index, LTM: lean tissue mass, ATM: adipose tissue mass, BCM: body cell mass, TSF: triceps skin fold, BSF: biceps skin fold, MAC: mid arm circumference. *Expressed as median (interquartile range).
to intradialytic hypotension in hemodialysis patients which could have adverse effects on patients with underlying cardiac dysfunction. BCM is a sensitive tool to measure changes in body composition in patients with hypervolemia, euvolemia and hypovolemia, and could be used to monitor volume status in dialysis patients and detect the time to intervene to prevent great fluctuation in body composition. In our study, while looking at the nutritional parameters, the mean BMI of the survived and expired patients were 22.6 ± 4.2 and 20.53 kg/m2, respectively. We observed increased mortality in patients who had lower baseline BMI (p ¼ 0.017). By ROC curve analysis, the best cut-off for predicting 2-year mortality was found to be 22.65 kg/m2, area under the curve 450% with sensitivity of 41.30% and specificity of 81.81%. Similar observations had been made earlier about under nutrition and mortality in dialysis patients.4 Lower BMI is a strong risk factor for increased mortality in dialysis patients.12,13 This has been referred to as ‘‘reverse epidemiology’’ by Kalantar-Zadeh et al., where obesity though a risk factor for development of CKD is associated with increased survival in various stages of CKD when compared with non-obese individuals.14,15 On performing multiple regression analysis, the predictors chosen on the
basis of p50.1 were OH, BMI and fat tissue index, and it was found that OH was significant predictor of mortality (p ¼ 0.042), when the effect of BMI and fat tissue index were adjusted. However, the paradoxical relationship between OH and BMI in ESRD could be a reason why BMI was not a significant predictor of mortality in multiple regression analysis. Further longitudinal large sample studies are required to appreciate the relationship between OH, BMI and risk of mortality in ESRD. While analyzing MHD and CAPD patients separately, it was found that higher OH and lower BMI at baseline were found to be significantly associated with mortality in MHD patients, but not in CAPD patients. This is probably because of smaller sample size in CAPD patients. Among the 41 survived patients, 36.6% were diabetic compared with 45.5% in the non-survived group, which was not significant (p ¼ 0.44). Although Raffaitin et al. in his study found that BMI, LTM (p ¼ 0.046) and serum albumin (p ¼ 0.05) decreased in diabetic patients on hemodialysis over a period of 2 years,16 we did not observe similar results. The baseline mean systolic blood pressure in the survived group was 147.45 ± 21.8 mmHg although in the non-survived group it showed higher trend to 150 ± 21.2 mmHg (p ¼ 0.60)
Maintenance hemodialysis and CAPD patients
DOI: 10.3109/0886022X.2014.964147
71
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Table 4. Comparison of body composition and nutritional parameters at 24 months for MHD and CAPD patients. Variables*
MHD (n ¼ 33)
CAPD (n ¼ 8)
p Value
Weight (kg) Height (cm) BMI (kg/m2) Diabetic (%) Systolic BP (mmHg) Diastolic BP (mmHg) Urine output (ml) Dry weight (kg) TSF (cm) BSF (cm) MAC (cm) S. albumin (gm/dL) Haemoglobin (gm/dL) S. Potassium (mEq/L) Urea (mg/dL) Creatinine (mg/dL) BCM parameters Overhydration (L) TBW (L) ECW (L) ICW (L) LTI (kg/m2) FTI (kg/m2) LTM (kg) Fat (kg) Relative fat (%) ATM (kg) BCM (kg)
56.21 ± 11.87 161.42 ± 9.36 21.45 ± 4.16 12 (36.5%) 157.36 ± 22.22 87.67 ± 14.02 0 (0–400) 55.24 ± 13.42 14 (8.50–26) 10 (5.50–24) 22.42 ± 6.34 3.48 ± 0.44 8.58 ± 1.51 5.82 ± 0.80 108.27 ± 28.97 10.23 ± 2.79
61.36 ± 15.17 156.63 ± 7.09 25.28 ± 4.01 5 (62.5%) 150.63 ± 25.31 78.13 ± 16.64 100 (0–287.50) 59.62 ± 10.35 6 (6–8.75) 3.50 (2–5) 21.06 ± 7.29 3.27 ± 0.71 10.12 ± 1.47 4.20 (3.85–4.70)* 66.13 ± 25.50 7.93 ± 2.65
0.30 0.18 0.024 0.11 0.46 0.10 0.70 0.40 0.034 0.008 0.60 0.30 0.014 0.002 0.001 0.042
1.45 (0.6–1.9)* 27.51 ± 6.85 13.85 ± 3.36 13.67 ± 3.48 10.06 ± 2.27 15.04 ± 1.78 25.20 ± 7.34 27.51 ± 4.61 43.18 ± 3.28 37.45 ± 6.36 12.64 ± 4.80
0.78 0.35 0.97 0.25 0.36 0.028 0.32 0.06 0.051 0.09 0.34
1.40 (0.2–2.4)* 30.53 ± 8.39 13.89 ± 3.71 16.60 ± 6.90 11.30 ± 3.18 11.20 (5.90–15)* 28.73 ± 9.10 21.70 (11.65–28.2) 33.97 ± 12.61 29.80 (15.85–39.1) 14.99 ± 6.27
Notes: MHD: hemodialysis, CAPD: continuous ambulatory peritoneal dialysis, BMI: body mass index, TBW: total body water, ECW: extracellular water, ICW: intracellular water, LTI: lean tissue index, FTI: fat tissue index, LTM: lean tissue mass, ATM: adipose tissue mass, BCM: body cell mass, TSF: triceps skin fold, BSF: biceps skin fold, MAC: mid arm circumference. *Expressed as Median (Interquartile range).
Table 5. Multiple logistic regressions for analyzing various predictors of mortality. Expired Factors
n
Row%
95% for OR Adj OR
LL
UL
p Value
0.768
7.203
Ref 0.134
1.038
8.460
Ref 0.042
0.631
6.353
Ref 0.238
2
BMI [kg/m ]—at baseline 4Median (21.25) 11 29.7 1.00 Median (21.25) 22 59.5 2.352 Overhydration [L]—at baseline Median (3.2) 11 28.9 1.00 4Median (3.2) 22 61.1 2.963 Fat tissue index [kg/m2]—at baseline 4Median (10.2) 11 27.8 1.00 Median (10.2) 22 60.5 2.003
which is not significant and which could be a manifestation of OH. LTM did not show any significant changes on follow-up (p ¼ 0.32) compared to study done by Kato et al. In the study done by Kalantar-Zadeh et al., he found that low-baseline relative fat percentage and fat loss over time were independently associated with higher mortality in MHD patients.17 In our study, the non-survived patients had baseline mean fat levels of 17.45 ± 7.3 kg which is due to malnutrition. Baseline relative fat and adipose tissue mass have been less for nonsurvived patients at 31.50 ± 10.7 and 23.74 ± 9.9 kg, respectively, though not significantly different from the survived patients (p ¼ 0.11 and 0.22, respectively).
TSF (p ¼ 0.001), BSF (p ¼ 0.001) and MAC (p ¼ 0.001) increased in the survived group over the period of 2 years which suggest that improvement in anthropometric markers of nutrition is associated with survival. Noori et al. showed that MAC is an independent predictor of greater survival in MHD patients.18 Dry weight measured clinically is crude and often imprecise.19 We looked for correlation between clinically determined weight gain (i.e., weight–dry weight) of the patient and OH measured by BCM. There was a better correlation with BCM compared with clinical weight gain measurement, but was not statistically significant (p ¼ 0.063). OH measured by BCM is superior to weight gain measured clinically.9 BCM is also used to assess nutritional status by measuring the LTI. Those with higher BMI had higher LTI and hence better survival rate18 than malnourished individuals who have a lesser LTI. Serum albumin at baseline and at follow-up did not show significant change (p ¼ 0.78). The study performed by Glenn et al. showed that change in albumin over a period of 2.6 years in MHD patients was significant in predicting survival (p50.0001).20 Serum albumin levels were increased in our CAPD patients, suggesting better nutritional status. Similar findings were found in the study done by Jager et al., on 132 MHD patients and 118 CAPD patients whose follow-up showed increased serum albumin in CAPD patients after a period of 2 years.21
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
72
S. Mathew et al.
There was an increase in body fat to 3.2 kg (95% CI 1.6–4.9) in our CAPD patients which was similar to the study done by Jager et al.21. LTM did not change over time in both MHD and CAPD patients which was similar to the finding of Jager et al were LTM remained the same at baseline and follow-up for both MHD and CAPD patients.21 The strength of this study is that it is a multicentric study evaluating nutritional and hydration status by BCM of dialysis patients and its effect on 2-year mortality. It represents the usefulness of body composition parameters particularly OH as a predictor of mortality. BCM is a cheap and effective tool to measure fluid status of dialysis patients, which could be useful in developing countries. The major limitation of this study is that BCM could have been done at regular intervals in the patients, which would have been ideal. Less number of patients are included in CAPD (n ¼ 14), which could have affected the comparison between MHD and CAPD patients. Despite the paradoxical relationship between OH and nutritional indicators such as BMI, lower BMI was associated with mortality in this study.
Conclusion This study showed that higher baseline OH and lower BMI were observed among those who had died. BCM is a better tool for looking at OH compared to clinically measured dry weight. Fat and adipose tissue mass were comparatively lower at baseline in patients who did not survive, though not statistically significant. The patients who survived had a significant improvement of TSF, BSF and MAC, and significant reduction in OH over a period of 2 years. Further large cohort longitudinal study using body composition parameters should be done for identifying the risk factors for mortality in maintenance dialysis patients.
Declaration of interest The authors acknowledge the support given by Madras Medical Mission, Chennai, India.
References 1. Al-Dadah A, Omran J, Nusair MB, Dellsperger KC. Cardiovascular mortality in dialysis patients. Adv Perit Dial. 2012;28:56–59. 2. Roberts MA, Polkinghorne KR, McDonald SP, Ierino FL. Secular trends in cardiovascular mortality rates of patients receiving dialysis compared with the general population. Am J Kidney Dis. 2011;58(1):64–72. 3. Kalantar-Zadeh K, Regidor DL, Kovesdy CP, et al. Fluid retention is associated with cardiovascular mortality in patients undergoing long-term hemodialysis. Circulation. 2009;119(5):671–679.
Ren Fail, 2015; 37(1): 66–72
4. Wizemann V, Wabel P, Chamney P, et al. The mortality risk of overhydration in hemodialysis patients. Nephrol Dial Transplant. 2009;24(5):1574–1579. 5. Prakash J, Raja R, Mishra RN, et al. High prevalence of malnutrition and inflammation in undialyzed patients with chronic renal failure in developing countries: A single center experience from eastern India. Ren Fail. 2007;29(7):811–816. 6. Vijayan M, Abraham G, Alex ME, et al. Nutritional status in stage V dialyzed patient versus CKD patient on conservative therapy across different economic status. Ren Fail. 2014;36(3): 384–389. 7. Abraham G, Varsha P, Mathew M, Sairam VK, Gupta A. Malnutrition and nutritional therapy of chronic kidney disease in developing countries: The Asian perspective. Adv Ren Replace Ther. 2003;10(3):213–221. 8. Bonanni A, Mannucci I, Verzola D, et al. Protein-energy wasting and mortality in chronic kidney disease. Int J Environ Res Public Health. 2011;8(5):1631–1654. 9. Wabel P, Chamney P, Moissl U, Jirka T. Importance of whole-body bioimpedance spectroscopy for the management of fluid balance. Blood Purif. 2009;27(1):75–80. 10. Jaeger JQ, Mehta RL. Assessment of dry weight in hemodialysis: An overview. J Am Soc Nephrol. 1999;10(2):392–403. 11. Hung S-C, Kuo K-L, Peng C-H, et al. Volume overload correlates with cardiovascular risk factors in patients with chronic kidney disease. Kidney Int. 2014;85(3):703–709. 12. Abbott KC, Glanton CW, Trespalacios FC, et al. Body mass index, dialysis modality, and survival: Analysis of the United States Renal Data System Dialysis Morbidity and Mortality Wave II Study. Kidney Int. 2004;65(2):597–605. 13. Fouque D, Vennegoor M, ter Wee P, et al. EBPG guideline on nutrition. Nephrol Dial Transplant. 2007;22(Suppl 2):ii45–ii87. 14. Park J, Ahmadi S-F, Streja E, et al. Obesity paradox in endstage kidney disease patients. Prog Cardiovasc Dis. 2014;56(4): 415–425. 15. Kalantar-Zadeh K, Block G, Humphreys MH, Kopple JD. Reverse epidemiology of cardiovascular risk factors in maintenance dialysis patients. Kidney Int. 2003;63(3):793–808. 16. Raffaitin C, Lasseur C, Chauveau P, et al. Nutritional status in patients with diabetes and chronic kidney disease: A prospective study. Am J Clin Nutr. 2007;85:96–101. 17. Kalantar-Zadeh K, Kuwae N, Wu DY, et al. Associations of body fat and its changes over time with quality of life and prospective mortality in hemodialysis patients. Am J Clin Nutr. 2006;83(2): 202–210. 18. Noori N, Kopple JD, Kovesdy CP, et al. Mid-arm muscle circumference and quality of life and survival in maintenance hemodialysis patients. Clin J Am Soc Nephrol. 2010;5(12): 2258–2268. 19. Jha V, Jairam A, Sharma MC, Sakhuja V, Piccoli A, Parthasarathy S. Body composition analysis with bioelectric impedance in adult Indians with ESRD: Comparison with healthy population. Kidney Int. 2006;69(9):1649–1653. 20. Chertow GM, Johansen KL, Lew N, Lazarus JM, Lowrie EG. Vintage, nutritional status, and survival in hemodialysis patients. Kidney Int. 2000;57(3):1176–1181. 21. Jager KJ, Merkus MP, Huisman RM, et al. Nutritional status over time in hemodialysis and peritoneal dialysis. J Am Soc Nephrol. 2001;12(6):1272–1279.
Notice of correction: The version of this article published online ahead of print on 24 Sep 2014 contained an error in the author list. Author ‘‘Madhusudan Vijayan’’ was mistakenly listed as ‘‘Madhusudan Viayan’’. The error has been corrected for this version.
CLINICAL STUDY
Body composition monitoring and nutrition in maintenance hemodialysis and CAPD patients—a multicenter longitudinal study Sharon Mathew1, Georgi Abraham1,2,3, Madhusudan Vijayan4, Thigarajan Thandavan3, Milly Mathew1, Ilangovan Veerappan1, Laxmi Revathy1, and Merina E. Alex2 1
Department of Nephrology and Nutrition, Pondicherry Institute of Medical Sciences, Puducherry, India, 2Madras Medical Mission, Chennai, India, Tanker Foundation, Chennai, India, and 4Government Kilpauk Medical College, Chennai, India
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
3
Abstract
Keywords
Hydration and nutritional status of end stage renal disease (ESRD) patients are linked to increased morbidity and mortality. Body composition monitoring (BCM) by multi-frequency bioimpedance spectroscopy (MFBS) is considered to be a superior modality of fluid assessment in chronic kidney disease (CKD) dialysis. We did a longitudinal prospective study in South India on maintenance hemodialysis (MHD) and continuous ambulatory peritoneal dialysis (CAPD) patients over 24 months and looked at impact of baseline nutritional parameters and body composition parameters on 24-month mortality. Ninety-nine patients stable on dialysis for at least 3 months were recruited (MHD 85, CAPD 14) at baseline and at 24 months, 41 were alive and 33 had expired, 12 had undergone renal transplant and 13 were lost to follow-up. BCM and nutritional assessment were done at baseline and at follow-up. Baseline overhydration (OH) differed significantly between surviving and dead patients (p50.05). Receiver operating characteristic (ROC) curve between OH and mortality showed that the best cut-off point to differentiate between survived and expired patients was 3.15 L. ROC curve for BMI showed lower than cut-off of 22.65 kg/m2 to predict mortality with sensitivity 41.30% and specificity 81.81%. At follow-up, triceps skin fold thickness (TSF), biceps skin fold thickness (BSF) and mid arm circumference (MAC) increased significantly from baseline (p50.001, p ¼ 0.001 and p50.001, respectively). Overhydration and BMI are important predictors of mortality in dialysis patients. Improvement in anthropometric markers TSF, BSF and MAC in MHD patients was associated with survival.
Hemodialysis, multi-frequency bioimpedance spectroscopy, peritoneal dialysis, overhydration, survival
Introduction Cardiovascular diseases represent the leading cause of death in dialysis patients, both in maintenance hemodialysis (MHD) and continuous ambulatory peritoneal dialysis (CAPD) patients.1–4 This is partly attributed to fluid disturbances with abnormalities in salt and water intake which are highly prevalent in end stage renal disease (ESRD) patients.4 As a result of lack of trained renal dieticians in India, the dialysis patients are inappropriately advised regarding calorie, protein, micronutrients, salt and water intake. Malnutrition has been widely reported in Indian patients on dialysis.5,6 Malnutrition in CKD is associated with increased morbidity, mortality and poor quality of life.7,8 However, there is a paucity of multicentric longitudinal data on hydration and nutritional status of Indian dialysis patients and its effect on all-cause mortality.
Address correspondence to Georgi Abraham, Madras Medical Mission, No 4-A, Dr. J. Jayalalitha Nagar, Mogappair, Chennai-600037, India. Tel.: +91-9841710992; E-mail: [email protected]
History Received 31 May 2014 Revised 6 August 2014 Accepted 30 August 2014 Published online 24 September 2014
Renal replacement therapy (RRT) modalities comprising MHD and CAPD often influence the fluid and salt content of the patients. Excessive removal of fluid during dialysis leads to hypotension, and inadequate removal often leads to hypertension with left ventricular hypertrophy (LVH) and cardiac failure.4 Therefore, assessment of nutritional and hydration status, especially hypervolemia, can play an important role in the management of fluid overload in dialysis patients.4 Clinical assessment of fluid status is often inaccurate. Body composition monitor (BCM; Fresenius Medical Care, Bad Homburg, Germany) is a unique device based on multifrequency bioimpedance spectroscopy (MFBS) which measures overhydration (OH), total body water (TBW), extracellular water (ECW), intracellular water (ICW), fat tissue mass (FTM), lean tissue mass (LTM) and body cell mass (BCM). BCM has been validated against the following gold standard reference methods: bromide dilution for ECW, total body potassium (TBK) for ICW, deuterium dilution for TBW, dual energy X-ray absorptiometry (DEXA) for LTM, four compartment modeling, air displacement plethysmography, under water weighing for adipose tissue mass, magnet resonance tomography for BCM and an expert clinical assessment for overhydration (OH).9 BCM has been assessed
Maintenance hemodialysis and CAPD patients
DOI: 10.3109/0886022X.2014.964147
previously in ESRD patients, but has never been utilized in Indian dialysis patients. This prospective longitudinal multicentric study was conducted to assess the nutritional and hydration status of prevalent MHD and CAPD patients and to observe the impact of baseline characteristics of the above in predicting 2-year mortality.
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Subjects and methods After getting approval from the ethics committee, the patients were enrolled from three centers: Madras Medical Mission Hospital, TANKER foundation (www.tankerfoundation.org) and Pondicherry Institute of Medical Sciences. Those patients with ESRD who were older than 18 years and survived for first 3 months after initiation of dialysis were included in the study. Excluded patients were those with permanent cardiac pacemaker, patients with liver cirrhosis, infections such as TB, HIV, malignancy and patients on irregular dialysis due to financial constraints. Informed written consent was obtained from all the patients included in the study. The study was done at baseline in 99 patients (78 males and 21 females) in February 2010, and the follow-up study was done at 24 months in the surviving patients. Out of the 99 patients included, 85 (85.86%) were on hemodialysis (MHD) and 14 (14.14%) were on CAPD. In the CAPD patients, nine (64.29%) were on four exchanges of 2 L per day and five (35.71%) were on three exchanges of 2 L per day. In the MHD patients, 63 (74.12%) were on thrice weekly dialysis and rest (25.88%) were on biweekly dialysis, with the duration of each dialysis session being 4 h. Clinical, biochemical, nutritional and body composition measurements were made to assess the nutritional and fluid status of the patients at baseline and at 2 years. In the MHD group, the measurements were made before the onset of the dialysis session. In the CAPD group, the measurements were made after emptying the peritoneal cavity of fluid, before the morning exchange. Clinical characteristics collected were mode of dialysis, diabetes mellitus status, systolic and diastolic BP, urine output, dry weight dietary pattern (vegetarian/non-vegetarian), and daily calorie and protein intake. Biochemical parameters measured were serum potassium, blood urea and serum creatinine. Nutritional parameters assessed included body mass index (BMI), triceps skin fold thickness (TSF), biceps skin fold thickness (BSF), mid arm circumference (MAC), serum albumin and hemoglobin concentration. Body composition monitor parameters estimated were TBW, ECW, ICW, lean tissue index (LTI), fat tissue index (FTI), LTM, fat in kg, relative fat in percentage, adipose tissue mass (ATM) and BCM. The mean of two measurements was taken each time. The baseline characteristics were compared between the patients who survived and died, and a receiver operating characteristic (ROC) curve was constructed to determine the cut-off for significant parameters. The baseline characteristics were compared between MHD and CAPD patients. For the surviving patients, the baseline and follow-up characteristics were compared. The patients who underwent transplant and patients lost to follow-up were excluded from the repeat measurement at 2 years.
67
Statistical analysis Continuous variables were described in the form of mean, standard deviation, median and interquartile range. Dichotomous variables (yes/no) were summarized in terms of percentages. To compare two related continuous data samples (like for comparison of baseline and follow-up in our study), paired t test for normally distributed data and Wilcoxon signed rank test for non-normally distributed data were used. To compare two independent samples means (comparison of MHD and CAPD patients), independent t test for normally distributed data and Mann–Whitney U test for non-normally distributed data were used. Chi square test was used to compare proportions. Pearson’s correlation and Spearman’s correlation methods were used to determine the correlation coefficient between variables for normally and non-normally distributed data, respectively. ROC curve to determine the sensitivity and specificity of overhydration and BMI for prediction of mortality was used.
Results The study had 99 patients enrolled, and at the end of 24 months, 41 patients (MHD 33/CAPD 8) had survived, 33 had died, 12 patients had undergone renal transplant and 13 patients were lost to follow-up. The outcomes of the recruited patients are shown in Figure 1. Table 1 shows the baseline characteristics of the patients. The patients who had expired during the period of 2 years had higher degree of OH and lower BMI at baseline, compared to patients who had survived at 2 years (p ¼ 0.02 and 0.017, respectively). By ROC curve analysis of OH (Figure 2), the area under the curve was greater than 50%. Sensitivity– specificity analysis showed the best cut-off point for baseline OH between the survived and dead patients was 3.15 L with a sensitivity of 39.47% and specificity of 69.70%. By ROC curve analysis (Figure 3), the best cut-off for baseline BMI was found to be 22.65 kg/m2, area under the curve450% with sensitivity of 41.30% and specificity of 81.81%. While comparing the baseline parameters of survived and expired patients for MHD and CAPD patients separately, BMI and OH were significant in MHD patients (p ¼ 0.033 and 0.018, respectively). Table 2 shows the comparison of the variables of survived patients at baseline and at follow-up. On analyzing the body composition of the surviving patients at 2 years, it was found that ICW and OH were significantly different at 2 years compared with their baseline values (p ¼ 0.022 and p50.001, respectively). Among the nutritional parameters, TSF, BSF and MAC were found to be significantly improved (p 0.001, p50.001 and p 0.001) among those who had survived. Tables 3 and 4 show the comparison between MHD and CAPD patients. Lean tissue index correlated with BMI (r ¼ 0.209, p ¼ 0.042). We also looked for correlation between baseline lean tissue index and serum albumin as both are important indicators of nutrition and we found a weak positive correlation between the two, though not significant ( ¼ 0.194, p ¼ 0.153). On analyzing the correlation between OH and clinically determined weight gain by Spearman’s correlation method, we observed a correlation of ¼ 0.159
68
S. Mathew et al.
Ren Fail, 2015; 37(1): 66–72
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Figure 1. Flowchart showing outcomes of the recruited patients at follow-up.
Table 1. Baseline characteristics of MHD and CAPD patients.
Variables* Age (years) Gender Males (%) Females (%) Weight (kg) Height (cm) BMI (kg/m2) Mode of dialysis MHD (%) CAPD (%) Diabetic (%) Systolic BP (mmHg) Diastolic BP (mmHg) Urine output (mL) Dry weight (kg) Diet Vegetarian (%) Non-vegetarian (%) TSF (cm) BSF (cm) MAC (cm) Serum albumin (g/dL) Haemoglobin (g/dL) S.Potassium (mEq/L) Urea (mg/dL) Creatinine (mg/dL) eGFR BCM parameters Overhydration (L) TBW (L) ECW (L) ICW (L) LTI (kg/m2) FTI (kg/m2) LTM (kg) Fat (kg) Relative fat (kg) ATM (kg) BCM (kg)
All patients (n ¼ 99)
Survived (n ¼ 41)
55.26 ± 12.5
52.20 ± 13.1
54.76 ± 7.6
0.32
78 (78.79%) 21 59.05 ± 12.6 152.83 ± 9.0 22.23 ± 4.2
31 (75.6%) 10 56.52 ± 12.6 160.61 ± 8.9 22.68 ± 4.2
26 (78.8%) 7 54.55 ± 9.6 162.82 ± 9.5 20.53 ± 3.1
0.75
85 (86.14%) 14 (13.86%) 39 (39.4%) 147.45 ± 21.8 83.55 ± 12.2 100 (0–500)* 56.00 ± 12.3
33 (80.5%) 8 (19.5%) 15(36.6%) 147.49 ± 23.4 82.76 ± 12.1 312.20 ± 386.6 55.77 ± 12.7
27 (81.8%) 6 (18.2%) 15 (45.5%) 150.21 ± 21.2 84.48 ± 11.4 180 (0–700) 51.78 ± 9.2
22 (22.22%) 77 6 (4–8)* 4 (2–8)* 16 (10–23)* 3.6 ± 0.6 8.88 ± 1.8 4.76 ± 1.04 83 (6–100)* 6.8 ± 3.2 12.1 ± 11.59
13 (31.7%) 28 6 (4–8)* 4 (2–6)* 16.38 ± 7 3.52 ± 0.5 8.9 ± 2 4.80 ± 1.1 81.80 ± 29.3 7.23 ± 3.0 11.78 ± 13.43
6 (18.2%) 27 6 (3–8) 4 (2–6) 17.29 ± 6.9 3.52 ± 0.5 8.66 ± 1.6 4.68 ± 0.9 81.39 ± 28.8 6.28 ± 3.0 12.49 ± 8.98
0.42 0.87 0.62 1 0.58 0.62 0.95 0.18 0.8
2.90 (1.50–4.05)* 28.23 ± 6.9 14.42 ± 3.3 13.85 ± 3.9 10.48 ± 3.0 10.87 ± 4.3 27.49 ± 9.7 20.31 ± 7.8 34.65 ± 11 27.64 ± 10.6 14.00 ± 6.5
3.9 (2.65–4.80)* 28.55 ± 6.5 15.03 ± 3.4 13.62 ± 3.5 10.01 ± 2.3 9.07 ± 4.1 27.01 ± 8.3 17.45 ± 7.3 31.50 ± 10.7 23.74 ± 9.9 13.43 ± 5.3
0.02 0.84 0.52 0.79 0.46 0.07 0.82 0.11 0.22 0.11 0.69
3.1 (1.85–4.30)* 29.41 ± 7.1 14.91 ± 3.1 14.40 ± 4.0 10.52 ± 2.8 10.45 ± 4.6 28.30 ± 9.2 20.09 ± 8.5 33.55 ± 10.8 27.33 ± 11.5 14.39 ± 6.1
Died (n ¼ 33)
p Value for comparison between survived and died
0.46 0.31 0.017 0.89 0.44 0.61 0.53 0.48 0.13 0.186
Notes: MHD: Hemodialysis, CAPD: continuous ambulatory peritoneal dialysis, BMI: body mass index, TBW: total body water, ECW: extracellular water, ICW: intracellular water, LTI: lean tissue index, FTI: fat tissue index, LTM: lean tissue mass, ATM: adipose tissue mass, BCM: body cell mass, TSF: triceps skin fold, BSF: biceps skin fold, MAC: mid arm circumference. *Median (IQR) for all continuous variables, comparison by Mann–Whitney U test; percentages for categorical variables, comparison by Chi square test.
Maintenance hemodialysis and CAPD patients
DOI: 10.3109/0886022X.2014.964147
which was just short of being statistically significant (p ¼ 0.063). Multiple regression analysis was used to study the effect of various predictors on mortality, and the results are shown
69
in Table 5. The predictors were chosen on the basis of p value less than 0.1 in the comparison between survived and expired patients. It was found that OH was a significant predictor of mortality (p ¼ 0.042), if BMI and fat tissue index (FTI) were kept constant. There was no association between frequency of dialysis (twice/thrice weekly) and survival in MHD patients (p ¼ 0.61). We also compared demographic, body composition and nutritional parameters at baseline and follow-up for MHD and CAPD patients and are shown in Tables 3 and 4. Hemoglobin concentration was statistically significant in the survived CAPD patients compared with MHD patients at follow-up (p ¼ 0.014).
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Discussion
Figure 2. ROC curve between overhydration and mortality. Note: Area under the curve 450%, estimated cut off ¼ 3.15 L, sensitivity ¼ 39.47%, specificity ¼ 69.70%.
Figure 3. ROC curve between body mass index and mortality. Note: Area under the curve 450%, estimated cut off ¼ 22.65 kg/m2, sensitivity ¼ 41.30%, specificity ¼ 81.81%.
BCM by MFBS is an upcoming tool in the measurement of overhydration (OH) in fluid overload conditions such as CKD and congestive cardiac failure (CCF). It has been validated as the gold standard in the measurement of overhydration in CKD patients on MHD.9 In our study, baseline OH (43.15 L) (p 0.05) was found to have a significant correlation with 2-year mortality. This is in accordance to the study done by Wizemann et al., in which it was found that out of 269 chronic MHD patients, 58 patients were overhydrated and 41% of these patients died when followed-up for a period of 3.5 years.4 In the study done by Wabel et al., 25% of MHD patients suffered from overhydration of 42.5 L.9 The ROC curve analysis between body composition and nutritional parameters with mortality revealed OH as the only BCM parameter with a statistically significant area under the curve greater than 50% (p 0.05). Sensitivity–specificity analysis showed the best cut-off point for baseline overhydration to predict 2-year mortality was 3.15 L. In the patients who had survived, there was a significant reduction in the level of OH (p50.001). OH is a risk factor for cardiac dysfunction (LVH, MI, CCF), increased incidence of intradialytic morbid events and sudden death.10,11 Excessive ultrafiltration rate can lead
Table 2. Comparison of body composition and nutritional parameters of survived MHD and CAPD patients between baseline and follow-up. At baseline (n ¼ 41) Variables
Mean (SD)
TBW (L) ECW (L) ICW (L) LTI (kg/m2) FTI (kg/m2) LTM (kg) Fat (kg) Relative fat (%) ATM (kg) BCM (kg) OH (L) TSF (cm) BSF (cm) MAC (cm) S.albumin (gm/dL) Hb (gm/dL)
28.23 ± 6.91 14.42 ± 3.33 13.86 ± 3.97 10.28 ± 3.07 11.63 ± 3.76 26.54 ± 9.71 21.60 ± 6.99 36.68 ± 9.40 29.40 ± 9.49 13.44 ± 6.48 2.73 ± 1.87 6.66 ± 3.81 5.23 ± 4.36 16.38 ± 7.03 3.52 ± 0.50 8.94 ± 2.03
Median (IQR) 28.3 14.3 13 10.1 11.8 27.0 22.2 36 30.3 12.8 2.9 6 4 12 3.65 8.3
(22.9–34.1) (11.8–17.0) (10.4–7.15) (8.2–12.3) (7.2–13.8) (19.4–37.4) (13.60–26.1) (29.9–44.4) (18.5–35.5) (9–19.3) (1.5–4.1) (4–8) (2–6) (11–23) (3.12–3.9) (7.6–10.6)
At follow-up (n ¼ 41) Mean (SD)
Median (IQR)
p Value
29.94 ± 8.1 13.88 ± 2.8 16.03 ± 6.4 10.95 ± 3.0 12.18 ± 5.8 27.97 ± 8.7 22.42 ± 11.8 35.96 ± 11.8 30.77 ± 14.2 14.48 ± 6.0 0.95 ± 2.61 10 (6–25) 10 (4–22) 22.15 ± 6.4 3.44 ± 0.5 8.88 ± 1.6
27.60 (19.9–53.4) 13.3 (9.7–20.8) 13.9 (10.1–37.4) 10.4 (7–19) 12.7 (3–26) 25.2 (17–55) 22.5 (6–43) 37.4 (11–54) 30.9 (7–58) 13.00 (7–33) 1.40 (0.2–2.2) 14.66 ± 9.48 12.54 ± 10.13 23.00 3.4 8.8
0.140 0.171 0.022* 0.188 0.730 0.302 0.839 0.544 0.726 0.277 50.001* 50.001* 50.001* 50.001* 0.905 0.877
Notes: MHD: hemodialysis, CAPD: continuous ambulatory peritoneal dialysis, TBW: total body water, ECW: extracellular water, ICW: intracellular water, LTI: lean tissue index, FTI: fat tissue index, LTM: lean tissue mass, ATM: adipose tissue mass, BCM: body cell mass, OH: overhydration, TSF: triceps skin fold, BSF: biceps skin fold, MAC: mid arm circumference. *50.05 significant.
70
S. Mathew et al.
Ren Fail, 2015; 37(1): 66–72
Table 3. Comparison of demography, body composition and nutrition parameters at baseline for MHD and CAPD patients.
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Variables* Age (years) Gender Males (%) Females (%) Weight (kg) Height (cm) BMI (kg/m2) Diabetic (%) Systolic BP (mmHg) Diastolic BP (mmHg) Urine output (mL) Dry weight (kg) Diet Vegetarian Non-vegetarian TSF (cm) BSF (cm) MAC (cm) S. albumin (gm/dL) Haemoglobin (gm/dL) S. Potassium (mEq/L) Urea (mg/dL) Creatinine (mg/dL) Overhydration (L) TBW (L) ECW (L) ICW (L) LTI (kg/m2) FTI (kg/m2) LTM (kg) Relative LTM (%) Fat (kg) Relative fat (%) ATM (kg) BCM (kg)
MHD (n ¼ 85)
CAPD (n ¼ 14)
p Value
50.65 ± 12.23
60.43 ± 12.15
0.007
68 (80.0%) 17 (20.0%) 58.75 ± 13.02 163.58 ± 8.90 21.90 ± 4.30 31 (36.5%) 147.73 ± 21.32 84.67 ± 12.23 100 (0–450) 55.71 ± 12.71
10 (73.3%) 4 (26.7%) 60.89 ± 10.48 158.29 ± 8.82 24.24 ± 3.18 8 (57.14) 145.79 ± 25.50 76.71 ± 10.70 200 (0–762) 57.75 ± 10.34
19 (22.4%) 66 (77.6%) 6 (3–8)* 4 (2–8)* 16.12 ± 6.75 3.67 ± 0.66 8.75 ± 1.81 4.86 ± 0.96 83 (64–102)* 6.77 ± 3.16 3.2 (2.0–4.4)* 29.75 ± 7.20) 15.16 ± 3.40 14.61 ± 4.13 10.66 ± 2.79 9.94 ± 4.60 28.92 ± 9.19 50.09 ± 13.85 19.32 ± 8.62 32.34 ± 10.64 26.28 ± 11.72 14.76 ± 6.08
4 (26.7%) 10 (73.3%) 5.71 ± 2.09 4 (2–5.25)* 19.39 ± 6.92 3.31 ± 0.4 9.67 ± 1.80 4.17 ± 1.33 62 (45.5–94.75)* 6.92 ± 3.45 2.75 (0.57–3.6)* 27.36 ± 6.92 14.22 ± 3.38 13.14 ± 3.67 9.71 ± 2.86 13.41 ± 3.68 24.74 ± 9.20 40.25 ± 11.28 24.74 ± 6.44 40.55 ± 9.19 33.36 ± 8.7 12.23 ± 6.13
0.51 0.56 0.04 0.05 0.26 0.76 0.02 0.32 0.57 0.74 0.92 0.48 0.10 0.09 0.09 0.02 0.09 0.87 0.16 0.25 0.34 0.22 0.24 0.009 0.12 0.014 0.034 0.008 0.034 0.16
Notes: MHD: hemodialysis, CAPD: continuous ambulatory peritoneal dialysis, BMI: body mass index, TBW: total body water, ECW: extracellular water, ICW: intracellular water, LTI: lean tissue index, FTI: fat tissue index, LTM: lean tissue mass, ATM: adipose tissue mass, BCM: body cell mass, TSF: triceps skin fold, BSF: biceps skin fold, MAC: mid arm circumference. *Expressed as median (interquartile range).
to intradialytic hypotension in hemodialysis patients which could have adverse effects on patients with underlying cardiac dysfunction. BCM is a sensitive tool to measure changes in body composition in patients with hypervolemia, euvolemia and hypovolemia, and could be used to monitor volume status in dialysis patients and detect the time to intervene to prevent great fluctuation in body composition. In our study, while looking at the nutritional parameters, the mean BMI of the survived and expired patients were 22.6 ± 4.2 and 20.53 kg/m2, respectively. We observed increased mortality in patients who had lower baseline BMI (p ¼ 0.017). By ROC curve analysis, the best cut-off for predicting 2-year mortality was found to be 22.65 kg/m2, area under the curve 450% with sensitivity of 41.30% and specificity of 81.81%. Similar observations had been made earlier about under nutrition and mortality in dialysis patients.4 Lower BMI is a strong risk factor for increased mortality in dialysis patients.12,13 This has been referred to as ‘‘reverse epidemiology’’ by Kalantar-Zadeh et al., where obesity though a risk factor for development of CKD is associated with increased survival in various stages of CKD when compared with non-obese individuals.14,15 On performing multiple regression analysis, the predictors chosen on the
basis of p50.1 were OH, BMI and fat tissue index, and it was found that OH was significant predictor of mortality (p ¼ 0.042), when the effect of BMI and fat tissue index were adjusted. However, the paradoxical relationship between OH and BMI in ESRD could be a reason why BMI was not a significant predictor of mortality in multiple regression analysis. Further longitudinal large sample studies are required to appreciate the relationship between OH, BMI and risk of mortality in ESRD. While analyzing MHD and CAPD patients separately, it was found that higher OH and lower BMI at baseline were found to be significantly associated with mortality in MHD patients, but not in CAPD patients. This is probably because of smaller sample size in CAPD patients. Among the 41 survived patients, 36.6% were diabetic compared with 45.5% in the non-survived group, which was not significant (p ¼ 0.44). Although Raffaitin et al. in his study found that BMI, LTM (p ¼ 0.046) and serum albumin (p ¼ 0.05) decreased in diabetic patients on hemodialysis over a period of 2 years,16 we did not observe similar results. The baseline mean systolic blood pressure in the survived group was 147.45 ± 21.8 mmHg although in the non-survived group it showed higher trend to 150 ± 21.2 mmHg (p ¼ 0.60)
Maintenance hemodialysis and CAPD patients
DOI: 10.3109/0886022X.2014.964147
71
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
Table 4. Comparison of body composition and nutritional parameters at 24 months for MHD and CAPD patients. Variables*
MHD (n ¼ 33)
CAPD (n ¼ 8)
p Value
Weight (kg) Height (cm) BMI (kg/m2) Diabetic (%) Systolic BP (mmHg) Diastolic BP (mmHg) Urine output (ml) Dry weight (kg) TSF (cm) BSF (cm) MAC (cm) S. albumin (gm/dL) Haemoglobin (gm/dL) S. Potassium (mEq/L) Urea (mg/dL) Creatinine (mg/dL) BCM parameters Overhydration (L) TBW (L) ECW (L) ICW (L) LTI (kg/m2) FTI (kg/m2) LTM (kg) Fat (kg) Relative fat (%) ATM (kg) BCM (kg)
56.21 ± 11.87 161.42 ± 9.36 21.45 ± 4.16 12 (36.5%) 157.36 ± 22.22 87.67 ± 14.02 0 (0–400) 55.24 ± 13.42 14 (8.50–26) 10 (5.50–24) 22.42 ± 6.34 3.48 ± 0.44 8.58 ± 1.51 5.82 ± 0.80 108.27 ± 28.97 10.23 ± 2.79
61.36 ± 15.17 156.63 ± 7.09 25.28 ± 4.01 5 (62.5%) 150.63 ± 25.31 78.13 ± 16.64 100 (0–287.50) 59.62 ± 10.35 6 (6–8.75) 3.50 (2–5) 21.06 ± 7.29 3.27 ± 0.71 10.12 ± 1.47 4.20 (3.85–4.70)* 66.13 ± 25.50 7.93 ± 2.65
0.30 0.18 0.024 0.11 0.46 0.10 0.70 0.40 0.034 0.008 0.60 0.30 0.014 0.002 0.001 0.042
1.45 (0.6–1.9)* 27.51 ± 6.85 13.85 ± 3.36 13.67 ± 3.48 10.06 ± 2.27 15.04 ± 1.78 25.20 ± 7.34 27.51 ± 4.61 43.18 ± 3.28 37.45 ± 6.36 12.64 ± 4.80
0.78 0.35 0.97 0.25 0.36 0.028 0.32 0.06 0.051 0.09 0.34
1.40 (0.2–2.4)* 30.53 ± 8.39 13.89 ± 3.71 16.60 ± 6.90 11.30 ± 3.18 11.20 (5.90–15)* 28.73 ± 9.10 21.70 (11.65–28.2) 33.97 ± 12.61 29.80 (15.85–39.1) 14.99 ± 6.27
Notes: MHD: hemodialysis, CAPD: continuous ambulatory peritoneal dialysis, BMI: body mass index, TBW: total body water, ECW: extracellular water, ICW: intracellular water, LTI: lean tissue index, FTI: fat tissue index, LTM: lean tissue mass, ATM: adipose tissue mass, BCM: body cell mass, TSF: triceps skin fold, BSF: biceps skin fold, MAC: mid arm circumference. *Expressed as Median (Interquartile range).
Table 5. Multiple logistic regressions for analyzing various predictors of mortality. Expired Factors
n
Row%
95% for OR Adj OR
LL
UL
p Value
0.768
7.203
Ref 0.134
1.038
8.460
Ref 0.042
0.631
6.353
Ref 0.238
2
BMI [kg/m ]—at baseline 4Median (21.25) 11 29.7 1.00 Median (21.25) 22 59.5 2.352 Overhydration [L]—at baseline Median (3.2) 11 28.9 1.00 4Median (3.2) 22 61.1 2.963 Fat tissue index [kg/m2]—at baseline 4Median (10.2) 11 27.8 1.00 Median (10.2) 22 60.5 2.003
which is not significant and which could be a manifestation of OH. LTM did not show any significant changes on follow-up (p ¼ 0.32) compared to study done by Kato et al. In the study done by Kalantar-Zadeh et al., he found that low-baseline relative fat percentage and fat loss over time were independently associated with higher mortality in MHD patients.17 In our study, the non-survived patients had baseline mean fat levels of 17.45 ± 7.3 kg which is due to malnutrition. Baseline relative fat and adipose tissue mass have been less for nonsurvived patients at 31.50 ± 10.7 and 23.74 ± 9.9 kg, respectively, though not significantly different from the survived patients (p ¼ 0.11 and 0.22, respectively).
TSF (p ¼ 0.001), BSF (p ¼ 0.001) and MAC (p ¼ 0.001) increased in the survived group over the period of 2 years which suggest that improvement in anthropometric markers of nutrition is associated with survival. Noori et al. showed that MAC is an independent predictor of greater survival in MHD patients.18 Dry weight measured clinically is crude and often imprecise.19 We looked for correlation between clinically determined weight gain (i.e., weight–dry weight) of the patient and OH measured by BCM. There was a better correlation with BCM compared with clinical weight gain measurement, but was not statistically significant (p ¼ 0.063). OH measured by BCM is superior to weight gain measured clinically.9 BCM is also used to assess nutritional status by measuring the LTI. Those with higher BMI had higher LTI and hence better survival rate18 than malnourished individuals who have a lesser LTI. Serum albumin at baseline and at follow-up did not show significant change (p ¼ 0.78). The study performed by Glenn et al. showed that change in albumin over a period of 2.6 years in MHD patients was significant in predicting survival (p50.0001).20 Serum albumin levels were increased in our CAPD patients, suggesting better nutritional status. Similar findings were found in the study done by Jager et al., on 132 MHD patients and 118 CAPD patients whose follow-up showed increased serum albumin in CAPD patients after a period of 2 years.21
Ren Fail Downloaded from informahealthcare.com by Washington University Library on 01/04/15 For personal use only.
72
S. Mathew et al.
There was an increase in body fat to 3.2 kg (95% CI 1.6–4.9) in our CAPD patients which was similar to the study done by Jager et al.21. LTM did not change over time in both MHD and CAPD patients which was similar to the finding of Jager et al were LTM remained the same at baseline and follow-up for both MHD and CAPD patients.21 The strength of this study is that it is a multicentric study evaluating nutritional and hydration status by BCM of dialysis patients and its effect on 2-year mortality. It represents the usefulness of body composition parameters particularly OH as a predictor of mortality. BCM is a cheap and effective tool to measure fluid status of dialysis patients, which could be useful in developing countries. The major limitation of this study is that BCM could have been done at regular intervals in the patients, which would have been ideal. Less number of patients are included in CAPD (n ¼ 14), which could have affected the comparison between MHD and CAPD patients. Despite the paradoxical relationship between OH and nutritional indicators such as BMI, lower BMI was associated with mortality in this study.
Conclusion This study showed that higher baseline OH and lower BMI were observed among those who had died. BCM is a better tool for looking at OH compared to clinically measured dry weight. Fat and adipose tissue mass were comparatively lower at baseline in patients who did not survive, though not statistically significant. The patients who survived had a significant improvement of TSF, BSF and MAC, and significant reduction in OH over a period of 2 years. Further large cohort longitudinal study using body composition parameters should be done for identifying the risk factors for mortality in maintenance dialysis patients.
Declaration of interest The authors acknowledge the support given by Madras Medical Mission, Chennai, India.
References 1. Al-Dadah A, Omran J, Nusair MB, Dellsperger KC. Cardiovascular mortality in dialysis patients. Adv Perit Dial. 2012;28:56–59. 2. Roberts MA, Polkinghorne KR, McDonald SP, Ierino FL. Secular trends in cardiovascular mortality rates of patients receiving dialysis compared with the general population. Am J Kidney Dis. 2011;58(1):64–72. 3. Kalantar-Zadeh K, Regidor DL, Kovesdy CP, et al. Fluid retention is associated with cardiovascular mortality in patients undergoing long-term hemodialysis. Circulation. 2009;119(5):671–679.
Ren Fail, 2015; 37(1): 66–72
4. Wizemann V, Wabel P, Chamney P, et al. The mortality risk of overhydration in hemodialysis patients. Nephrol Dial Transplant. 2009;24(5):1574–1579. 5. Prakash J, Raja R, Mishra RN, et al. High prevalence of malnutrition and inflammation in undialyzed patients with chronic renal failure in developing countries: A single center experience from eastern India. Ren Fail. 2007;29(7):811–816. 6. Vijayan M, Abraham G, Alex ME, et al. Nutritional status in stage V dialyzed patient versus CKD patient on conservative therapy across different economic status. Ren Fail. 2014;36(3): 384–389. 7. Abraham G, Varsha P, Mathew M, Sairam VK, Gupta A. Malnutrition and nutritional therapy of chronic kidney disease in developing countries: The Asian perspective. Adv Ren Replace Ther. 2003;10(3):213–221. 8. Bonanni A, Mannucci I, Verzola D, et al. Protein-energy wasting and mortality in chronic kidney disease. Int J Environ Res Public Health. 2011;8(5):1631–1654. 9. Wabel P, Chamney P, Moissl U, Jirka T. Importance of whole-body bioimpedance spectroscopy for the management of fluid balance. Blood Purif. 2009;27(1):75–80. 10. Jaeger JQ, Mehta RL. Assessment of dry weight in hemodialysis: An overview. J Am Soc Nephrol. 1999;10(2):392–403. 11. Hung S-C, Kuo K-L, Peng C-H, et al. Volume overload correlates with cardiovascular risk factors in patients with chronic kidney disease. Kidney Int. 2014;85(3):703–709. 12. Abbott KC, Glanton CW, Trespalacios FC, et al. Body mass index, dialysis modality, and survival: Analysis of the United States Renal Data System Dialysis Morbidity and Mortality Wave II Study. Kidney Int. 2004;65(2):597–605. 13. Fouque D, Vennegoor M, ter Wee P, et al. EBPG guideline on nutrition. Nephrol Dial Transplant. 2007;22(Suppl 2):ii45–ii87. 14. Park J, Ahmadi S-F, Streja E, et al. Obesity paradox in endstage kidney disease patients. Prog Cardiovasc Dis. 2014;56(4): 415–425. 15. Kalantar-Zadeh K, Block G, Humphreys MH, Kopple JD. Reverse epidemiology of cardiovascular risk factors in maintenance dialysis patients. Kidney Int. 2003;63(3):793–808. 16. Raffaitin C, Lasseur C, Chauveau P, et al. Nutritional status in patients with diabetes and chronic kidney disease: A prospective study. Am J Clin Nutr. 2007;85:96–101. 17. Kalantar-Zadeh K, Kuwae N, Wu DY, et al. Associations of body fat and its changes over time with quality of life and prospective mortality in hemodialysis patients. Am J Clin Nutr. 2006;83(2): 202–210. 18. Noori N, Kopple JD, Kovesdy CP, et al. Mid-arm muscle circumference and quality of life and survival in maintenance hemodialysis patients. Clin J Am Soc Nephrol. 2010;5(12): 2258–2268. 19. Jha V, Jairam A, Sharma MC, Sakhuja V, Piccoli A, Parthasarathy S. Body composition analysis with bioelectric impedance in adult Indians with ESRD: Comparison with healthy population. Kidney Int. 2006;69(9):1649–1653. 20. Chertow GM, Johansen KL, Lew N, Lazarus JM, Lowrie EG. Vintage, nutritional status, and survival in hemodialysis patients. Kidney Int. 2000;57(3):1176–1181. 21. Jager KJ, Merkus MP, Huisman RM, et al. Nutritional status over time in hemodialysis and peritoneal dialysis. J Am Soc Nephrol. 2001;12(6):1272–1279.
Notice of correction: The version of this article published online ahead of print on 24 Sep 2014 contained an error in the author list. Author ‘‘Madhusudan Vijayan’’ was mistakenly listed as ‘‘Madhusudan Viayan’’. The error has been corrected for this version.
