Adiposity and insulin resistance in nondiabetic hemodialysis patients:

effects of high energy supplementation1–3

Szu-Chun Hung and Der-Cherng Tarng

ABSTRACT hypothesized because of this obesity paradox (4). Nonetheless,

Background: In contrast to the general population, a higher body obesity is also associated with an increased risk of developing
mass index is associated with better survival among hemodialysis insulin resistance (5), which may contribute to accelerated CVD,
patients. Theoretically, high energy supplementation in these pa- which is the major cause of death among chronic hemodialysis
tients ought to lead to weight gain over time, but the benefits of patients (6). Hence, the protective effects of excess weight in
this strategy are unclear. hemodialysis patients remain debatable and need to be tested
Objective: The objective was to assess whether high energy sup- with an appropriately designed prospective study (7).
plementation in nondiabetic hemodialysis patients might adversely Increased body fat mass is associated with poorer insulin
affect insulin resistance—a known risk factor for cardiovascular

Downloaded from www.ajcn.org by on September 22, 2010

resistance in stage 3–4 nondiabetic chronic kidney disease (CKD)
disease. patients who are not receiving dialysis (8). However, potential
Design: We first investigated the association between body fat mass determinants of insulin resistance in stage 5 CKD patients un-
and insulin resistance (homeostasis model assessment of insulin dergoing chronic hemodialysis have not been studied in detail.
resistance; HOMA-IR) in nondiabetic hemodialysis patients in a We hypothesized that body fat mass may also be independently
cross-sectional analysis (study 1). Of the 106 individuals studied, 55 associated with insulin resistance among nondiabetic chronic
were randomly assigned to either high energy supplementation (an hemodialysis patients and an increase in energy intake will result
extra 475 kcal/d; n = 28) or not (n = 27) for 12 wk to assess pro- in increased insulin resistance due to the increase in body fat
spective changes in body fat mass and insulin resistance (study 2). mass. To test this hypothesis, we examined the relation between
Results: In study 1, body fat mass (P , 0.05) and C-reactive pro- body fat mass and insulin resistance as determined by homeo-
tein (CRP) (P , 0.05) each contributed independently to HOMA-
stasis model assessment of insulin resistance (HOMA-IR) among
IR. In study 2, 41 patients completed the study. The 20 patients who
nondiabetic chronic hemodialysis patients in a cross-sectional
received high energy supplementation had a significantly greater
observation (study 1). We further conducted a randomized
increase in body fat mass (P , 0.05), CRP (P , 0.05), and HOMA-
controlled trial to verify whether the level of insulin resistance is
IR (P , 0.001) than did the 21 controls.
associated with the increase in adiposity that occurs after high
Conclusions: Body fat mass and CRP are primary determinants of
energy supplementation (study 2).
insulin resistance in nondiabetic hemodialysis patients. High energy
supplementation, because it increases adiposity and inflammation,
exacerbates insulin resistance. A long-term study is needed to clar-
SUBJECTS AND METHODS
ify the metabolic effects of high energy supplementation on cardio-
vascular disease outcomes in hemodialysis patients. Am J Clin The study consisted of a cross-sectional analysis at baseline
Nutr 2009;90:64–9. (study 1), followed by a prospective randomized controlled in-
tervention over 12 wk (study 2). The protocol was approved by
the local medical ethics committee, and all patients gave their
INTRODUCTION
1
From the Department and Institute of Physiology, National Yang-Ming
Obesity is a well-established risk factor for cardiovascular
University, Taipei (D-CT); the Division of Nephrology, Buddhist Tzu Chi
disease (CVD) in the general population (1). Paradoxically, an General Hospital, Taipei Branch (S-CH); and the Division of Nephrology,
increasing number of epidemiologic studies have indicated that Department of Medicine and Immunology Research Center, Taipei Veterans
there is a survival advantage in having an elevated body mass General Hospital, Taipei, Taiwan (D-CT).
index (BMI; in kg/m2) among hemodialysis patients (2). In fact, 2
Supported by grants from the National Science Council (NSC 95-2314-
whereas traditional CVD risk factors are associated with an B-010-077 and 96-2628-B-010-001-MY3), Taipei Veterans General Hospital
unfavorable outcome in the general population, these factors (V97S5-004 and V97C1-093), and Buddhist Tzu Chi General Hospital, Tai-
would appear to be protective in chronic hemodialysis patients. pei Branch (TCRD-TPE-95-45).
3
Address correspondence to D-C Tarng, Department and Institute of
This phenomenon has been referred to as “reverse epidemiol-
Physiology, National Yang-Ming University and Division of Nephrology,
ogy” (3). Department of Medicine, Taipei Veterans General Hospital. No. 201, Section
Chronic hemodialysis patients frequently have a low energy 2, Shih-Pai Road, Taipei 112, Taiwan. E-mail: [email protected].
intake and are underweight. A greater caloric intake by hemo- Received January 5, 2009. Accepted for publication April 14, 2009.
dialysis patients to engender a better survival outcome has been First published online May 20, 2009; doi: 10.3945/ajcn.2009.27438.

64 Am J Clin Nutr 2009;90:64–9. Printed in USA. Ó 2009 American Society for Nutrition
 HIGH ENERGY SUPPLEMENTATION IN HEMODIALYSIS 65
written informed consent. The study commenced in 2006 and was The primary outcome measures were changes in body fat mass
concluded in 2007. and HOMA-IR over the 12-wk study period. The secondary out-
come measures were changes in metabolic variables, including
plasma leptin, serum cholesterol, triglyceride, albumin, and CRP.
Study 1 Patients in the control group were observed in parallel to estimate
Subjects were recruited among chronic hemodialysis patients any changes in the main outcome measures. Over the course of
at the dialysis centers of affiliated hospitals of National Yang- the study, the previous level of physical activity was kept constant
Ming University, Taipei. Inclusion criteria were age .20 y and for both groups. The methods of data collection were the same as
duration of prior dialysis .6 mo. The patients were being di- those in study 1.
alyzed for 424.5 h three times/wk by using a single-use dialyzer
equipped with a membrane surface area of 1.6–1.7 m2 at a blood Statistical analysis
flow of 3002350 mL/min and a dialysate flow of 500 mL/min.
Descriptive statistics included means 6 SDs for continuous
Exclusion criteria were diabetes mellitus, an inadequate dose of data and percentages for categorical data. The values for HOMA-
dialysis (single-pool Kt/V , 1.2), comorbidity with malignancy,
IR, plasma insulin and leptin, and serum CRP were not normally
liver disease, significant fluid overload, and an infectious disease
distributed and are reported as medians with interquartile ranges.
4 wk before enrollment.
In study 1 and study 2, for between-group comparison, Student’s
Baseline information was obtained from each patient, in-
t test was used for normally distributed data and the Mann-
cluding age, sex, anthropometric measures (eg, body weight,
Whitney rank-sum test was used for nonnormally distributed
BMI, and body fat mass), routine chemistry, insulin concentra-
data. Pearson’s chi-square test was used for frequency measures.
tion, leptin concentration, and C-reactive protein (CRP) con-
In study 1, Pearson’s correlation analysis was performed to
centration. All laboratory tests were conducted after the patients

Downloaded from www.ajcn.org by on September 22, 2010

examine the association between HOMA-IR and body fat mass
had fasted and before the start of midweek dialysis. Glucose in
as well as other potentially explanatory variables. Forward step-
the plasma and albumin together with total cholesterol and tri- wise multiple regression was performed to assess the independent
glyceride concentrations in the serum were measured by using an
effect of the 11 covariables, including age, sex, dialysis vintage,
autoanalyzer (model 736-60; Hitachi, Naka, Japan). Plasma in-
urea reduction rate, BMI, body fat mass, serum albumin, cho-
sulin (Diagnostic Products Corporation, Los Angeles, CA) and
lesterol, triglyceride, CRP, and plasma leptin on the HOMA-IR
leptin (Linco Research Inc, St Louis, MO) were measured by
(dependent variable) in study 1 and the changes in HOMA-IR
radioimmunoassay according to the procedures recommended
(dependent variable) in study 2. Statistical analysis was per-
by the manufacturers. Serum CRP was measured by immu-
formed by using the computer software Statistical Package of
noturbidimetric assay by using rate nephelometry (Beckman,
Social Science (SPSS 12.0, 2003; SPSS Inc, Chicago, IL). A
Galway, Ireland). Insulin resistance was quantified by HOMA-
P value ,0.05 was considered statistically significant.
IR. HOMA-IR was calculated with the following formula: fasting
On the basis of our previous data (12), we estimated that 40
plasma glucose (mg/dL) · fasting plasma insulin (lU/mL)/405,
patients (20 in each group) would be required for study 2 to have
with higher values representing greater degrees of insulin re-
a power of 90% to detect an absolute difference in the primary
sistance. The validity of HOMA-IR as a surrogate measure of
endpoint and a probability of a type I error of 0.05 with the use of
insulin resistance has been established in the CKD population a 2-sided test. We allowed for the possibility of a withdrawal rate
(9, 10). Body fat mass was measured with the multifrequency
of 25% by including a total of 55 patients for the study 2.
bioimpedance method (model 310; Biodynamics, Seattle, WA),
which was performed within 30 min after a dialysis session at
presumed dry weight and involved a single well-trained dietitian
to avoid interobserver variation. RESULTS

Study 1
Study 2 Patient characteristics
A subset of patients in study 1 who provided informed consent A total of 106 nondiabetic hemodialysis patients were en-
was randomly assigned to 1 of 2 groups: the high-energy group rolled. The etiologies of the end-stage renal disease (ESRD)
and the control group. Treatment order was block-randomized included glomerulonephritis (n = 40), interstitial nephritis (n =
with the use of computer-generated random numbers. Patients 19), hypertension (n = 18), various polycystic kidney diseases
in the high-energy group received one can of a commercially (n = 9), and shrunken kidneys of unknown etiology (n = 20). The
available oral nutritional supplement (Nepro; Abbott Laborato- patients were subdivided into either low (median: 1.16; inter-
ries, Ross Products Division, Chicago, IL) daily for 12 wk. Each quartile range: 0.86–1.53; n = 53) or high (median: 3.48; in-
can of supplement contained 16.6 g protein, 22.7 g fat, and 52.8 g terquartile range: 2.45–6.54; n = 53) HOMR-IR groups. The
carbohydrate and provided 475 kcal. These values for energy and baseline characteristics of the patients are shown in Table 1.
protein were added to the food intake to calculate total intake. Patients in the high HOMA-IR group were significantly older
Daily dietary records over 3 consecutive days were used to es- than those in the low HOMA-IR group. There were no signifi-
timate intakes of protein and calories at the beginning and the end cant differences between the 2 groups with regard to sex, he-
of the study (11). Supplements were dispensed weekly. Com- modialysis duration, BMI, and body weight. Plasma leptin,
pliance was verified by dietary interviews during dialysis ses- fasting plasma glucose, serum albumin, and total cholesterol
sions, as described previously (12). also did not differ significantly. Percentage body fat (P , 0.05),
66 HUNG AND TARNG
TABLE 1
Baseline characteristics of nondiabetic hemodialysis (HD) patients with low and high insulin resistance determined
by homeostasis model assessment (HOMA-IR) based on median values in study 1
Low HOMA-IR High HOMA-IR
Variable (n = 53) (n = 53) P1

HOMA-IR2 1.16 (0.8621.53)3 3.48 (2.4526.54) ,0.001

Age (y) 55 6 164 62 6 18 0.042
Sex (% male) 45 49 0.820
HD duration (y) 5.9 6 4.2 5.3 6 5.0 0.291
BMI (kg/m2) 20.9 6 3.3 21.6 6 3.1 0.230
Body weight (kg) 54.5 6 8.5 56.6 6 9.4 0.169
Body fat (%) 26.5 6 8.8 31.0 6 8.9 0.024
Body fat mass (kg) 14.6 6 5.6 17.5 6 5.8 0.001
Leptin (ng/mL) 11.8 (3.1222.5) 18.2 (8.6222.8) 0.215
Fasting plasma glucose (mg/dL) 75.7 6 13.8 95.5 6 36.5 0.063
Fasting plasma insulin (lU/mL) 6.6 (4.528.6) 15.8 (13.0220.8) ,0.001
Albumin (g/dL) 4.0 6 0.2 3.8 6 0.2 0.057
Total cholesterol (mg/dL) 169 6 36 172 6 31 0.690
Triglycerides (mg/dL) 101 6 72 175 6 81 0.013
C-reactive protein (mg/L) 3.2 (3.024.0) 3.5 (3.229.4) 0.016
1
Calculated by using Student’s t test for normally distributed data, the Mann-Whitney rank-sum test for
nonnormally distributed data, and Pearson’s chi-square test for frequency measures.

Downloaded from www.ajcn.org by on September 22, 2010

2
Median value of HOMA-IR was 1.9.
3
Median; interquartile range in parentheses (all such values).
4
Mean 6 SD (all such values).

body fat mass (P , 0.05, and fasting plasma insulin concen- regression analysis (Table 3), CRP (P , 0.01) and body fat
trations (P , 0.001) were significantly greater in the high mass (P , 0.05) were independent predictors of HOMA-IR.
HOMA-IR group than in the low HOMA-IR group. Moreover,
patients with high HOMA-IR had significantly greater serum
triglyceride (P , 0.05) and CRP (P , 0.05) concentrations than Study 2
did those with low HOMA-IR. Patient characteristics
Of the 106 individuals in study 1, 55 were randomly assigned
Predictors of insulin resistance to either the high-energy group (n = 28) or the control group (n =
27) (Figure 1). The high-energy and control groups were not
The potential predictors of HOMA-IR are depicted in Table 2.
significantly different at baseline with respect to age (58 6 14 y
In the unadjusted analysis, there was a strong positive correla-
compared with 57 6 14 y), sex distribution (63% male com-
tion between HOMA-IR and CRP (r = 0.343, P , 0.001), fol-
pared with 58% male), hemodialysis duration (6.9 6 6.2 y
lowed by body fat mass (r = 0.296, P , 0.01), BMI (r = 0.279,
compared with 6.6 6 6.1 y), BMI (20.9 6 2.7 compared with
P , 0.01), and plasma leptin (r = 0.268, P , 0.01). In multiple
21.4 6 2.8), body fat mass (16.1 6 3.5 kg compared with 15.6 6
5.1 kg), fasting plasma insulin concentration [10.4 (8.2–17.3)
TABLE 2 lU/mL compared with 13.3 (9.9–19) lU/mL], and HOMA-IR
Association between insulin resistance determined by homeostasis model [1.89 (1.29–6.33) compared with 2.80 (2.15–7.46)], respectively.
assessment (HOMA-IR) and potentially explanatory variables among
nondiabetic hemodialysis patients in study 11
Effects of high energy supplementation on body fat mass and
Variable r insulin resistance
Age 0.143 Forty-one patients completed the study (20 in the high-energy
Male sex (yes/no) 20.2662 group and 21 in the control group) (Figure 1). The high-energy
Hemodialysis duration 0.095 group had a significantly greater increase in body fat mass (P ,
BMI 0.2793 0.05), plasma leptin (P , 0.001), CRP (P , 0.05), and HOMA-
Body fat mass 0.2963
IR (P , 0.001) than did the control group (Table 4). Multi-
ln Leptin 0.2683
Cholesterol 0.204
variate regression analysis showed that the change in body fat
Triglyceride 0.2512 mass was an independent predictor of the increase in HOMA-IR
Albumin 0.157 (P , 0.001), even after adjustment for the change in serum CRP.
ln C-reactive protein 0.3434
1
n = 106. The dependent variable was ln HOMA-IR. Associations were DISCUSSION
calculated by using Pearson’s correlation coefficient.
2
P , 0.05. Insulin resistance and its associated metabolic abnormalities
3
P , 0.01. are known complications of advanced CKD. Previous studies
4
P , 0.001. that used the glucose clamp technique described by Smith and
 HIGH ENERGY SUPPLEMENTATION IN HEMODIALYSIS 67
TABLE 3
Forward stepwise multiple regression analysis of major determinants of insulin resistance determined by
homeostasis model assessment (HOMA-IR) among nondiabetic hemodialysis patients in study 11
Variable Unit of increase Regression coefficient 95% CI P

ln C-reactive protein ln 1 mg/L 0.626 0.330, 0.923 0.002

Body fat mass 1 kg 0.042 0.001, 0.083 0.041
1
n = 106. The dependent variable was ln HOMA-IR. The independent variables included were age, sex, dialysis
vintage, urea reduction rate, BMI, body fat mass, plasma leptin, and serum albumin, cholesterol, triglycerides, and C-
reactive protein. All variables that were significant (P , 0.05) were left in the model.

DeFronzo (13) showed that a postbinding defect in insulin action, The role of body fat mass in the development of uremic insulin
probably resulting from retained uremic toxins, was the primary resistance can be attributed to adipocyte-derived hormones such
cause of glucose intolerance in nondiabetic patients with ESRD. as leptin, adiponectin, and resistin, which regulate peripheral
Vitamin D deficiency, metabolic acidosis, and inflammation are insulin sensitivity (16). Leptin, a protein that is secreted exclu-
other potential contributors to the development of uremic insulin sively by adipocytes, plays an important role in insulin resistance
resistance (14). In the present study, we report 2 novel findings. and participates in the pathogenesis of arterial hypertension (17,
First, in the analysis of the cross-sectional observation (study 1), 18). As expected, plasma leptin concentrations correlated pos-
body fat mass and CRP, independently predicted insulin resis- itively with HOMA-IR in the present study. Proinflammatory
tance in nondiabetic ESRD patients receiving maintenance he- adipokines, such as tumor necrosis factor-a, have also been
modialysis. Second, in the randomized controlled trial (study 2), shown to mediate insulin resistance (19). Ramos et al (20) found

Downloaded from www.ajcn.org by on September 22, 2010

an increase in energy intake beyond current recommendations that percentage body fat independently predicts the concen-
(15), because it increases adiposity and inflammation, signif- trations of markers of oxidative stress and inflammation among
icantly augmented insulin resistance evaluated in terms of patients with moderate-to-severe CKD. CKD is a chronic in-
HOMA-IR. flammatory state. Therefore, increased adiposity can enhance
the inflammatory responses that accompany CKD and cause
a further increase in insulin resistance. Because of the cross-
sectional design of the study, however, the cause-effect relation
between body fat mass and HOMA-IR could not be completely
elucidated. Study 2 showed that an increase in calorie intake
appears to increase body fat mass and subsequently increase
HOMA-IR in hemodialysis patients. Further analysis showed an
independent metabolic effect of fat mass on HOMA-IR, even
after adjustment for the change in serum CRP. In fact, high
energy supplementation alone can induce immediate oxidative
stress and inflammation (21). Therefore, the direct effect of body
fat mass on insulin resistance should be evaluated after suppress-
ing inflammation in hemodialysis patients to eliminate confound-
ing factors, such as systemic proinflammatory cytokines.
The interrelation between body fat mass, insulin resistance,
and inflammation in the nondiabetic hemodialysis patients in the
present study was of interest because a deleterious combination
might explain the overwhelming CVD risk of ESRD patients.
Nonetheless, in contrast with the general population, overweight
and obese hemodialysis patients (BMI  27.5) have a signifi-
cantly higher survival rate than do persons who are normal in
weight (BMI = 20227.5) or underweight (BMI , 20) (2). One
recent study in chronic hemodialysis patients showed that those
with increased body weight over time had a better survival rate
than did those with stable or decreased body weight (22). This
“obesity paradox” in ESRD might be explained by statistical
fallacies such as age-related mortality patterns and time dis-
crepancies between competing risk factors. In a recent study, de
Mutsert et al (23) found no reverse epidemiology of BMI and
mortality in hemodialysis patients compared with a general
population of equal baseline age and duration of follow-up. It is
possible that, in the long term, overweight ESRD patients may
FIGURE 1. Numbers of patients who entered the study for a cross-
sectional analysis at baseline (study 1), who were later randomly assigned
experience more CVD events if they survive long enough (3). In
to receive either high energy supplementation or no supplementation (study the United States, Asian dialysis patients have a better survival
2), and who completed the study. rate than do white patients. Indeed, the relation between BMI
68 HUNG AND TARNG
TABLE 4
Effects of high energy supplementation for 12 wk on metabolic variables in nondiabetic patients undergoing chronic hemodialysis in study 21
High-energy group Control group
Variable (n = 20) (n = 21) P2

Total energy intake at baseline (kcal
kg21
d21) 29.9 6 6.73 29.2 6 5.7 0.705
DTotal energy intake, 12 wk 2 baseline (kcal
kg21
d21) 7.9 6 2.6 0.1 6 2.1 ,0.001
BMI at baseline (kg/m2) 20.8 6 2.5 21.7 6 2.7 0.410
DBMI, 12 wk 2 baseline (kg/m2) 0.6 6 1.0 0.3 6 1.5 0.327
Body fat mass at baseline (kg) 15.1 6 3.2 16.9 6 4.8 0.337
DBody fat mass, 12 wk 2 baseline (kg) 2.5 6 1.2 20.4 6 2.0 0.031
Fasting plasma glucose at baseline (mg/dL) 81 6 14 87 6 15 0.100
DFasting plasma glucose, 12 wk 2 baseline (mg/dL) 25 6 12 4 6 13 0.072
Fasting plasma insulin at baseline (lU/mL) 10.3 (8.0–14.9)4 13.2 (8.6–18.7) 0.877
DFasting plasma insulin, 12 wk 2 baseline (lU/mL) 7.7 (3.4–14.6) 0.9 (21.7–3.9) ,0.001
HOMA-IR at baseline 1.73 (1.25–4.88) 2.61 (1.42–5.32) 0.546
DHOMA-IR, 12 wk 2 baseline 2.31 (1.07–7.91) 0.04 (20.89 to 2.44) ,0.001
Plasma leptin at baseline (ng/mL) 11.8 (7.7–18.5) 12.5 (9.3–15.1) 0.623
DPlasma leptin, 12 wk 2 baseline (ng/mL) 14.7 (6.6–25.6) 3.3 (1.3–7.2) ,0.001
Total cholesterol at baseline (mg/dL) 171 6 37 174 6 32 0.877
DTotal cholesterol, 12 wk 2 baseline (mg/dL) 15 6 19 6 6 20 0.724
Triglycerides at baseline (mg/dL) 125 6 55 138 6 53 0.087
DTriglycerides, 12 wk 2 baseline (mg/dL) 16 6 24 1 6 36 0.493

Downloaded from www.ajcn.org by on September 22, 2010

CRP at baseline (mg/L) 3.0 (3.2–7.0) 3.4 (3.0–8.9) 0.785
DCRP, 12 wk 2 baseline (mg/L) 2.5 (0.7–5.2) 1.3 (0.5–2.3) 0.038
Serum albumin at baseline (g/dL) 3.8 6 0.3 3.9 6 0.3 0.301
DSerum albumin, 12 wk 2 baseline (g/dL) 0.2 6 0.1 0.0 6 0.1 0.864
1
HOMA-IR, insulin resistance determined by homeostasis model assessment; CRP, C-reactive protein.
2
Calculated by using Student’s t test for normally distributed data and the Mann-Whitney rank-sum test for nonnormally distributed data.
3
Mean 6 SD (all such values).
4
Median; interquartile range in parentheses (all such values).

and survival was U-shaped rather than reversed in the Asian risk factor for CVD in ESRD (6). Our present findings have
dialysis patients (24). important implications for chronic hemodialysis patients who are
A limitation of this study was the lack of a placebo control or willing to gain weight because of the “obesity paradox.” Simply
a supplement with protein in the control group, which may have increasing energy intakes above current recommendations might
affected the outcome of any treatment effect. Another limitation worsen insulin resistance and other associated metabolic dis-
was that body composition analysis with the bioimpedance orders. Strategies to attenuate adverse metabolic responses without
method may be confounded by excessive extracellular volume in negating the beneficial effects of increases in energy intake need
patients with uremia. Nevertheless, persons with potential sig- to be evaluated in future studies (26). Agents that enhance insulin
nificant fluid overload were excluded from this study, and bio- sensitivity, such as peroxisome proliferator–activated receptor c
impedance was performed after a dialysis session when patients agonists (27) and angiotensin II blockade (28), might hold prom-
had reached their dry weight. Furthermore, because of limitations ise as adjunctive therapies that improve insulin resistance among
associated with the use of bioimpedance analysis for assessing chronic hemodialysis patients receiving high energy supplemen-
body fat mass, the differentiation between subcutaneous and tation.
metabolically active visceral fat could not be assessed in the
The authors’ responsibilities were as follows—D-CT: protocol design and
present study (25). Finally, insulin resistance is a consequence of overall conduct of the study; and D-CT and S-CH: data analysis and manu-
high energy intake in nondiabetic hemodialysis patients, but it script preparation. Neither of the authors had a conflict of interest.
was not possible to infer that the increase in HOMA-IR was
causally related to an adverse CV outcome in a 12-wk in-
tervention. However, in a cohort study with a mean follow-up
REFERENCES
period of 67 mo, Shinohara et al (6) showed that HOMA-IR 1. Adams KF, Schatzkin A, Harris TB, et al. Overweight, obesity, and
independently predicted a higher risk of CV mortality among mortality in a large prospective cohort of persons 50 to 71 years old. N
chronic hemodialysis patients. Long-term, adequately sized Engl J Med 2006;355:763–78.
randomized controlled studies of CV outcomes among chronic 2. Fleischmann E, Teal N, Dudley J, May W, Bower JD, Salahudeen AK.
Influence of excess weight on mortality and hospital stay in 1346 he-
hemodialysis patients receiving high energy supplementation are modialysis patients. Kidney Int 1999;55:1560–7.
needed to better define the appropriate clinical role, if any, of this 3. Kalantar-Zadeh K, Block G, Humphreys MH, Kopple JD. Reverse ep-
therapy. idemiology of cardiovascular risk factors in maintenance dialysis pa-
In conclusion, the present study showed that body fat mass and tients. Kidney Int 2003;63:793–808.
4. Schmidt DS, Salahudeen AK. Obesity-survival paradox—still a contro-
CRP each closely correlate with insulin resistance in nondiabetic versy? Semin Dial 2007;20:486–92.
chronic hemodialysis patients. Moreover, high energy supple- 5. Kahn SE, Hull RL, Utzschneider KM. Mechanisms linking obesity to
mentation results in increased insulin resistance, which is a novel insulin resistance and type 2 diabetes. Nature 2006;444:840–6.
 HIGH ENERGY SUPPLEMENTATION IN HEMODIALYSIS 69
6. Shinohara K, Shoji T, Emoto M, et al. Insulin resistance as an in- 18. Stenvinkel P. Leptin and blood pressure-is there a link? Nephrol Dial
dependent predictor of cardiovascular mortality in patients with end- Transplant 2000;15:1115–7.
stage renal disease. J Am Soc Nephrol 2002;13:1894–900. 19. Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose expression of
7. Ikizler TA, Stenvinkel P, Lindholm B. Resolved: being fat is good for tumor necrosis factor-alpha: direct role in obesity-linked insulin re-
dialysis patients: the Godzilla effect. J Am Soc Nephrol 2008;19:1059– sistance. Science 1993;259:87–91.
64. 20. Ramos LF, Shintani A, Ikizler TA, Himmelfarb J. Oxidative stress and
8. Trirogoff ML, Shintani A, Himmelfarb J, Ikizler TA. Body mass index inflammation are associated with adiposity in moderate to severe CKD. J
and fat mass are the primary correlates of insulin resistance in non- Am Soc Nephrol 2008;19:593–9.
diabetic patients with stage 3–4 chronic kidney disease. Am J Clin Nutr 21. O’Keefe JH, Gheewala NM, O’Keefe JO. Dietary strategies for im-
2007;86:1642–8. proving post-prandial glucose, lipids, inflammation, and cardiovascular
9. Shoji T, Emoto M, Nishizawa Y. HOMA index to assess insulin re- health. J Am Coll Cardiol 2008;51:249–55.
sistance in renal failure patients. Nephron 2001;89:348–9. 22. Kalantar-Zadeh K, Kuwae N, Wu DY, et al. Associations of body fat and
10. Crutchlow MF, Robinson B, Pappachen B, et al. Validation of steady- its changes over time with quality of life and prospective mortality in
state insulin sensitivity indices in chronic kidney disease. Diabetes Care hemodialysis patients. Am J Clin Nutr 2006;83:202–10.
2007;30:1813–8. 23. de Mutsert R, Snijder MB, van der Sman-de Beer F, et al. Association
11. Lee MMS, Pan WH, Yu SL, Huang PC. Food predictive of nutrient between body mass index and mortality is similar in the hemodialysis
intake in Chinese diet in Taiwan: I. Total calories, protein, fat and fatty population and the general population at high age and equal duration of
acids. Int J Epidemiol 1992;21:922–8. follow-up. J Am Soc Nephrol 2007;18:967–74.
12. Hung SC, Tung TY, Yang CS, Tarng DC. High-calorie supplementation 24. Wong JS, Port FK, Hulbert-Shearon TE, et al. Survival advantage in
increases serum leptin levels and improves response to rHuEPO in Asian American end-stage renal disease patients. Kidney Int 1999;55:
long-term hemodialysis patients. Am J Kidney Dis 2005;45:1073–83. 2515–23.
13. Smith D, DeFronzo RA. Insulin resistance in uremia mediated by 25. Gohda T, Gotoh H, Tanimoto M, et al. Relationship between abdominal
postbinding defects. Kidney Int 1982;22:54–62. fat accumulation and insulin resistance in hemodialysis patients. Hy-
14. Siew ED, Ikizler TA. Determinants of insulin resistance and its effects pertens Res 2008;31:83–8.
on protein metabolism in patients with advanced chronic kidney disease. 26. Kwan BCH, Murtaugh MA, Beddhu S. Associations of body size with
Contrib Nephrol 2008;161:138–44. metabolic syndrome and mortality in moderate chronic kidney disease.

Downloaded from www.ajcn.org by on September 22, 2010

15. National Kidney Foundation. Clinical practice guidelines for nutrition in Clin J Am Soc Nephrol 2007;2:992–8.
chronic renal failure. Am J Kidney Dis 2000;35(suppl):S1–140. 27. Lin SH, Lin YF, Kuo SW, Hsu YJ, Hung YJ. Rosiglitazone improves
16. Axelsson J, Heimbürger O, Stenvinkel P. Adipose tissue and in- glucose metabolism in non-diabetic uremic patients on CAPD. Am J
flammation in chronic kidney disease. Contrib Nephrol 2006;151:165– Kidney Dis 2003;42:774–80.
74. 28. de Vinuesa SG, Goicoechea M, Kanter J, et al. Insulin resistance, in-
17. Wisse BE. The inflammatory syndrome: the role of adipose tissue cy- flammatory biomarkers, and adipokines in patients with chronic kidney
tokines in metabolic disorders linked to obesity. J Am Soc Nephrol disease: effects of angiotensin II blockade. J Am Soc Nephrol 2006;17:
2004;15:2792–800. S206–12.