RISK FACTORS FOR HIGH DIALYSATE GLUCOSE USE IN PD … · Use of high concentrations of glucose for...

Peritoneal Dialysis International, Vol. 30, pp. 448-455doi: 10.3747/pdi.2009.00063

0896-8608/08 $3.00 + .00Copyright © 2010 International Society for Peritoneal Dialysis

448

RISK FACTORS FOR HIGH DIALYSATE GLUCOSE USE IN PD PATIENTS—A RETROSPECTIVE 5-YEAR COHORT STUDY

Hon-Yen Wu,1,2,3 Kuan-Yu Hung,3 Fu-Chang Hu,4 Yung-Ming Chen,3 Tzong-Shinn Chu,3 Jenq-Wen Huang,3 Kwan-Dun Wu,3 and Tun-Jun Tsai3

Department of Internal Medicine,1 Far Eastern Memorial Hospital, Pan-Chiao, Taipei; Department of Internal Medicine,2 Yun-Lin Branch, National Taiwan University Hospital, Yun-Lin; and Department

of Internal Medicine3 and National Center of Excellence for General Clinical Trial and Research,4 National Taiwan University Hospital, Taipei, Taiwan

♦ Background: Use of high concentrations of glucose for peritoneal dialysis (PD) may produce unfavorable results. Our previous study showed that high initial glucose load is associated with poor PD technique survival.♦ Objective: This retrospective cohort study at a medical center in Taiwan aimed to understand the factors associated with high glucose load in long-term PD patients.♦ Patients and Methods: We reviewed 90 newly started PD patients over 5 years. We determined glucose load by calculating annual glucose weight and dialysate volume administered. Multiple linear regression analyses with time-dependent covariates were used to determine fac-tors that influence the annual average dialysate glucose concentration.♦ Results: The study group included 47 men and 43 women with a mean age of 53.4 ± 13.9 years. Technique survival rates were 91.0%, 84.1%, and 77.6% at the beginning of the second, third, and fourth year of PD therapy respectively. The presence of diabetes mellitus (DM), high body mass index (BMI), and low weekly renal Kt/V were significantly correlated with high average dialysate glucose concentra-tion during the first, second, and third years. For patients undergoing PD for more than 3 years, residual renal func-tion (RRF) deteriorated, and only DM significantly affected higher dialysate glucose concentration in the fourth year.♦ Conclusions: Patients with DM, high BMI, and low RRF were more likely to require a high glucose load for PD therapy, especially during the first 3 years. After those 3 years of PD, DM was the only significant factor in the need for higher glucose load. To reduce the glucose load in chronic PD patients, alternative osmotic agents such as icodextrin or amino acids should be considered in the daily PD regimen.

Perit Dial Int 2010; 30:448-455 www.PDIConnect.comepub ahead of print: 12 Mar 2010 doi: 10.3747/pdi.2009.00063

KEY WORDS: Body mass index; diabetes mellitus; glu-cose load; residual renal function.

Since the early days of peritoneal dialysis (PD), glucose has been used in the dialysate, and glucose remains

the most widely used osmotic agent. However, glucose absorption from daily PD with a glucose-based dialysate can be as high as 100 – 300 g daily (1–3). This absorp-tion can worsen a patient’s metabolic profile and lead to weight gain, hyperglycemia, dyslipidemia, and cardiovas-cular diseases (CVDs). In addition, bioincompatibility of the PD solution has harmful effects on the peritoneum in the long term (4).

Our previous study (5) showed that higher initial glucose load, defined as the average dialysate glucose concentration administered in the first 6 months of PD therapy, predicted PD technique survival and was corre-lated with serum albumin, residual renal function (RRF), and diabetes mellitus (DM). The disadvantages of using a high glucose–containing PD regimen are well known. In this retrospective cohort analysis, we further analyzed the factors that influence glucose load in patients un-dergoing long-term PD.

PATIENTS AND METHODS

STUDY PARTICIPANTS: INCLUSION AND EXCLUSION CRITERIA

Between August 2001 and February 2004, 123 patients initiated PD therapy at the National Taiwan University Hospital; they were followed until June 2006. We ex-cluded patients who had undergone PD previously (n = 6), who were younger than 18 years of age (n = 2), who

Correspondence to: J.W. Huang, Department of Internal Medicine, National Taiwan University Hospital, No. 7 Chung-Shan South Road, Taipei 100 Taiwan.

[email protected] 14 February 2009; accepted 4 August 2009

This single copy is for your personal, non-commercial use only. For permission to reprint multiple copies or to order presentation-ready copies

for distribution, contact Multimed Inc. at [email protected]

by guest on September 2, 2020

http://ww

w.pdiconnect.com

/D

ownloaded from

http://www.pdiconnect.com/

449

WU et al. JULY 2010 - VOL. 30, NO. 4 PDI

had active malignancy (n = 1), or who had a total PD du-ration of less than 6 months (n = 9). Patients who used Extraneal (Baxter Healthcare Corporation, Deerfield, IL, U.S.A.), Nutrineal (Baxter Healthcare), or Physioneal (Baxter Healthcare) during the study period were also excluded (n = 15), leaving 90 end-stage renal disease patients enrolled in the study.

We reviewed the clinical data for each patient (in-cluding age, sex, and comorbid diseases) and the data for body height, body weight, body mass index (BMI), cardiothoracic ratio (CTR), peritoneal transport charac-teristics, solute clearance, RRF, and blood biochemistry exams. Comorbid diseases included DM, CVD, and chronic hepatic diseases. History of myocardial infarction or stroke, left ventricular ejection fraction below 30% by echocardiogram, and coronary or peripheral artery diseases revealed by angiography were considered to be CVDs (6). Chronic hepatic diseases included chronic viral hepatitis and liver cirrhosis (7).

CALCULATION OF GLUCOSE LOAD FROM AVERAGE GLUCOSE CONCENTRATION

The determinants of high glucose load after long-term PD therapy, based on the detailed regimen administered to each patient, were the primary outcome. Glucose weight and glucose load are two measures of glucose exposure. We calculated glucose weight as the sum of the products of the volume and glucose concentration for each exchange (8). We calculated glucose load as the average dialysate glucose concentration: total glucose weight/total volume of PD solution prescribed (5,9,10). We obtained the annual glucose load by calculating the annual average dialysate glucose concentration. To describe the clinical characteristics of patients with different glucose loads, we divided all patients into two equal-sized groups for further analysis, according to the median average dialysate glucose concentration in the first year: low glucose load, <1.78%, and high glucose load, ≥1.78%. The PD regimen and modality for each patient were evaluated and prescribed monthly in our PD unit. The PD solutions prescribed in our hospital include Extraneal, Nutrineal, Physioneal, and Dianeal.

LABORATORY EXAMINATIONS AND PD ASSESSMENTS

The blood for biochemistry exams was drawn after patients had fasted for at least 8 hours. We evaluated peritoneal solute transport, dialysate-to-plasma ratio of creatinine (D/P Cr) and ratio of dialysate glucose at 4 and 0 hours of the dwell (D4/D0 Glu), annually by the perito-neal equilibration test (PET) (11), measured peritoneal

solute clearance by weekly peritoneal Kt/V, and evaluated RRF based on weekly renal Kt/V (12). Ultrafiltration (UF) volume was recorded when performing the PET. Anuria was defined as daily urine <100 mL. All biochemical tests were performed using automatic analyzers (TBA-120FR and TBA-200FR: Toshiba Corporation, Tokyo, Japan).

STATISTICAL ANALYSES

Data are expressed as mean ± standard deviation or percentage unless otherwise indicated. We performed comparisons of clinical data using the unpaired t-test, one-way ANOVA, or Pearson chi-square test, as appropri-ate. For descriptive analysis, we performed univariate analysis.

Because missing data in the repeated measurements of glucose concentration at later times was most likely attributable to technique failure, we assessed the predic-tors of each year’s annual average glucose concentration over the 4-year follow-up period sequentially, by fitting 4 separate multiple linear regression models, which used the observations still at risk at the beginning of each year. To assure the quality of our analyses, we applied basic model-fitting techniques for linear regression models, including

• variableselection,• goodness-of-fitassessment,and• egressiondiagnostics(forexample,residualanalysis,

detection of influential cases, and check for multicol-linearity).

We used the variance inflation factor (VIF) to detect the potential multicollinearity problem (VIF ≥ 10). A two-sided p value of 0.05 or less was considered statistically significant. The statistical analyses were conducted us-ing the SAS software (version 9.1.3: SAS Institute, Cary, NC, U.S.A.).

RESULTS

PATIENT CHARACTERISTICS

The mean follow-up period was 37.7 ± 11.1 months (range: 6 – 58 months). The annual technique survival rates by the Kaplan–Meier method were 91.0%, 84.1%, and 77.6% at the beginning of year 2, year 3, and year 4 respectively. During the study period, we diagnosed onset of DM in 4 patients (4.4%), and we considered this change in DM status in the multiple linear regression analysis of each year’s result. Table 1 shows the clinical characteristics of our PD patients at the beginning of each year. Patients with longer technique survival were




http://ww

w.pdiconnect.com

/D

ownloaded from


450

PDI JULY 2010 - VOL. 30, NO. 4 RISK FACTORS FOR HIGH GLUCOSE LOAD

younger, had a lower prevalence of DM, lower D/P Cr, and higher D4/D0 Glu. We noted a decrease in RRF with increasing UF volume and peritoneal Kt/V among patients who had more prolonged PD therapy.

Table 2 shows the annual baseline clinical characteris-tics of patients with different glucose loads. Patients who needed high glucose loads were significantly older, had a higher BMI, a higher CTR, higher serum triglycerides, lower serum albumin, and a higher prevalence of DM or CVD. Patients with high glucose loads had significantly lower RRF and hence higher weekly peritoneal Kt/V to attain adequate total solute clearance. Figure 1 shows that more than half our patients became anuric within 36.3 months.

FACTORS DETERMINING LONG-TERM GLUCOSE LOAD

To consider the gradual increase in the number of dropouts because of technique failure, we conducted multiple linear regression analysis to discover the determinants for glucose load within each year among patients who remained on PD at the beginning of that year. We used multiple linear regression models specif ic to the PD year, with the stepwise variable selection method, to analyze the effect of sex, age, body height, body weight, BMI, comorbid diseases,

peritoneal solute transport, peritoneal solute clearance, RRF, CTR, and results of blood biochemistry exams.

As shown in Table 3, the presence of DM, low weekly renal Kt/V, and high BMI were significantly correlated with higher average dialysate glucose concentration dur-ing the first year. We observed the same correlations for

TABLE 1 Demographic and Clinical Dataa of Peritoneal Dialysis (PD) Patients by Treatment Year

Variable Treatment year First Second Third Fourth

Patients (n) 90 81 73 43Baseline age (years) 53.4±13.9 52.4±13.0 52.3±12.0 49.7±10.6Sex (male:female) 47:43 42:39 37:36 21:22Body mass index (kg/m2) 21.5±2.7 22.9±3.1b 23.2±2.8b 23.1±2.7b

Weekly renal Kt/V 0.55±0.46 0.35±0.38b 0.23±0.32b,c 0.15±0.25b,c

Glucose concentration (%)d 1.83±0.28 1.80±0.32 1.81±0.35 1.81±0.40Glucose weight (g/month)e 5061±1472 5029±1739 5481±3198 4546±2102D/P creatinine 0.626±0.098 0.617±0.111 0.598±0.110 0.590±0.091PD modality (CAPD:APD) 73:17 61:20 52:21 32:11Total cholesterol (mg/dL)f 195.8±55.4 233.5±49.9b 226.9±54.4b 218.7±48.0b

Triglycerides (mg/dL)f 159.6±105.5 241.2±174.2b 228.8±176.3b 214.5±168.8b

Albumin (g/dL)f 3.72±0.58 3.99±0.42b 3.97±0.42b 4.00±0.32b

D/P = dialysate-to-plasma; CAPD = continuous ambulatory peritoneal dialysis; APD = automated peritoneal dialysis.a Number or mean ± standard deviation.b p < 0.05 compared with the first year.c p < 0.05 compared with the second year.d The annual average dialysate glucose concentration prescribed during the year.e Obtained from all the PD solution prescribed within each year.f To convert cholesterol in mg/dL to mmol/L, multiply by 0.02586; triglycerides in mg/dL to mmol/L, multiply by 0.0113; albumin

in g/dL to g/L, multiply by 10.

Figure 1 — Nelson–Aalen estimate of the cumulative risk for development of anuria. Anuria was defined as daily urine <100 mL. More than 50% of our patients become anuric within 36.3 months after initiation of peritoneal dialysis therapy. Dotted lines indicate 95% confidence intervals.




http://ww

w.pdiconnect.com

/D

ownloaded from


451


TABL

E 2

Long

itud

inal

Cha

nges

in C

linic

al V

aria

bles

Acc

ordi

ng to

Glu

cose

Loa

d—Lo

w o

r Hig

h—at

Bas

elin

ea

Va

riab

le

Trea

tmen

t yea

r

Firs

t Se

cond

Th

ird

Four

th

Lo

w

Hig

h Lo

w

Hig

h Lo

w

Hig

h Lo

w

Hig

h

Pati

ents

(n)

45

45

43

38

41

32

24

19

Base

line

age

(yea

rs)

49.8

±13.

7 57

.1±1

3.4b

49.0

±12.

5 56

.2±1

2.6b

49.9

±11.

6 55

.3±1

2.0

48.0

±9.9

51

.8±1

1.3

Sex

(mal

e:fe

mal

e)

26:1

9 21

:24

24:1

9 18

:20

22:1

9 15

:17

11:1

3 10

:9

Body

mas

s in

dex

(kg/

m2 )

20.7

±2.2

22

.3±3

b 21

.7±2

.5

24.1

±3.3

b 22

.3±2

.2

24.4

±3.1

b 22

.2±2

.1

24.3

±2.9

b

Diab

etes

mel

litus

3

(6.7

) 14

(31

.1)b

4 (9

.3)

9 (2

3.7)

4

(9.8

) 7

(21.

9)

2 (8

.3)

3 (1

5.8)

Card

iova

scul

ar d

isea

ses

8 (1

7.8)

18

(40

.0)b

7 (1

6.3)

15

(39

.5)b

7 (1

7.1)

11

(34

.4)

5 (2

0.8)

9

(47.

4)Gl

ucos

e co

ncen

trat

ion

(%)c

1.61

±0.0

8 2.

05±0

.22b

1.57

±0.2

9 2.

02±0

.32b

1.65

±0.2

0 2.

02±0

.39b

1.66

±0.4

1 2.

00±0

.3b

Gluc

ose

wei

ght (

g/m

onth

)d 43

51±8

07

5771

±164

3b 42

45±1

186

5896

±185

2b 45

79±1

165

6637

±442

1b 39

63±1

803

5251

±226

5b

Wee

kly

rena

l Kt/

V 0.

65±0

.49

0.45

±0.4

b 0.

50±0

.45

0.19

±0.1

8b 0.

34±0

.38

0.09

±0.1

6b 0.

19±0

.27

0.10

±0.2

2D/

P cr

eati

nine

0.

623±

0.10

6 0.

629±

0.09

0 0.

619±

0.10

8 0.

615±

0.11

7 0.

589±

0.11

2 0.

610±

0.10

8 0.

573±

0.08

7 0.

611±

0.09

3U

ltra

filt

rati

on v

olum

e (m

L)

311±

147

298±

134

367±

133

316±

183

369±

157

307±

203

374±

185

364±

127

Trig

lyce

ride

s (m

g/dL

)e 14

5.8±

63.4

17

3.4±

134.

6 19

8.5±

131.

7 28

9.6±

203.

5b 22

3.9±

177.

6 23

5.1±

177.

2 20

0.7±

141.

0 23

1.8±

201.

2LD

L (m

g/dL

)e 10

5.1±

35.4

10

7.0±

44.9

10

3.9±

26.9

10

1.0±

36.0

99

.1±3

2.5

99.8

±40.

8 95

.6±3

2.1

93.8

±36.

5

D/P

= di

alys

ate-

to-p

lasm

a; L

DL =

low

-den

sity

lipo

prot

ein

chol

este

rol.

a Su

bjec

ts w

ere

divi

ded

into

two

grou

ps a

ccor

ding

to th

e av

erag

e di

alys

ate

gluc

ose

load

pre

scri

bed

in th

e fi

rst t

reat

men

t yea

r: lo

w <

1.7

8%; h

igh ≥

1.78

%. D

ata

are

expr

esse

d as

num

ber (

perc

enta

ge),

or m

ean

± st

anda

rd d

evia

tion

.b

p <

0.05

com

pare

d w

ith

low

glu

cose

load

gro

up in

the

sam

e tr

eatm

ent y

ear.

c Th

e an

nual

ave

rage

dia

lysa

te g

luco

se c

once

ntra

tion

pre

scri

bed

duri

ng th

e tr

eatm

ent y

ear.

d Ob

tain

ed fr

om a

ll th

e pe

rito

neal

dia

lysi

s so

luti

on p

resc

ribe

d du

ring

the

trea

tmen

t yea

r.e

To c

onve

rt tr

igly

ceri

des

in m

g/dL

to m

mol

/L, m

ulti

ply

by 0

.011

3; L

DL in

mg/

dL to

mm

ol/L

, mul

tipl

y by

0.0

2586

.




http://ww

w.pdiconnect.com

/D

ownloaded from


452


the second and third years. However, with the decline of RRF, only DM still significantly predicted higher average dialysate glucose concentration during the fourth year. Within each year, DM has the strongest influence on glu-cose load; BMI had a relatively minor role in predicting glucose load. Because of the extremely low VIF for each covariate, we observed very small correlations between DM, RRF, and BMI within each year (Table 3).

LONG-TERM GLUCOSE LOAD OF PATIENTS WITH AND WITHOUT DM

We further analyzed the differences of annual average dialysate glucose concentration in patients with and without DM [Figure 2(A)]. During the entire study period, patients with DM had significantly higher glucose loads than did patients without DM, and this difference did not decrease over time [Figure 2(A)]. The differences in glucose load between years were not significant in patients with or without DM. We obtained the increasing rate of average dialysate glucose concentration for the second, third, and

fourth years compared with the first year [Figure 2(B)]. We observed no significant differences between patients with and without DM, nor between different years. However, when considering the entire study period, there was a trend for a more rapid rate of glucose concentration increase in patients with DM. Furthermore, in the DM and non-DM groups, the rate of glucose concentration increase is more prominent in later years [Figure 2(B)].

TABLE 3 Multiple Linear Regression Analyses of Predictors Associated

with Annual Average Dialysate Glucose Concentration Administered Each Year

Covariate Parameter Standard p Variance estimate error Value inflation (β)

First yeara Diabetes mellitus 0.2562 0.0590 <0.0001 1.0436 Weekly renal Kt/V –0.2514 0.0620 0.0001 1.0059 Body mass index 0.0248 0.0077 0.0017 1.0432

Second yearb Diabetes mellitus 0.2080 0.0848 0.0167 1.0742 Weekly renal Kt/V –0.3595 0.0923 0.0002 1.0188 Body mass index 0.0441 0.0101 <0.0001 1.0665

Third yearc Diabetes mellitus 0.3155 0.1069 0.0045 1.0610 Weekly renal Kt/V –0.4219 0.1276 0.0016 1.0086 Body mass index 0.0365 0.0117 0.0027 1.0535

Fourth yeard Diabetes mellitus 0.3170 0.1283 0.0184 1

a Linear regression model: n = 90, adjusted generalized R2 = 0.3768.

b Linear regression model: n = 74, adjusted generalized R2 = 0.4079.

c Linear regression model: n = 65, adjusted generalized R2 = 0.3505.

d Linear regression model: n = 38, adjusted generalized R2 = 0.1212.

Figure 2 — (A) Longitudinal comparison of annual average dialysate glucose concentration in patients with and without diabetes mellitus (DM). Data are expressed as means and standard error. * P < 0.001; ** P < 0.01. (B) Longitudinal comparison of the increasing rate of annual average dialysate glucose concentration in patients with and without DM. Data are expressed as means and standard error.




http://ww

w.pdiconnect.com

/D

ownloaded from


453


DISCUSSION

Ours is the first detailed longitudinal analysis of the effect of glucose load on patients undergoing long-term PD. Our previous study reported that high initial glucose load was associated with poor PD technique survival and that high initial glucose load was determined by the presence of DM, poor RRF, and low serum albumin (5). The present study demonstrates that PD patients with DM, low RRF, and high BMI required higher glucose loads for at least 3 years; after 3 years, DM was the only significant factor associated with a need for high glucose load (Table 3).

To analyze the long-term glucose load of PD patients, we performed longitudinal analyses of dialysate glucose concentration. Because of dropouts occurring from PD technique failure, and because of the possible influ-ence of the same covariates at different time points, we used a cross-sectional analysis for each year. Similar statistical analyses have been used in previous studies. For example, Davies et al. and Brown et al. (9,13) per-formed a longitudinal analysis of peritoneal membrane function of anuric PD patients from the cohort of the European Automated Peritoneal Dialysis Outcomes Study (EAPOS). They demonstrated that anuric PD patients can experience significant declines in peritoneal membrane function in 12 – 24 months, and that glucose appeared to enhance those declines (9,14).

We think that glucose concentration is preferable to glucose weight as an indicator for estimation of glucose load. A high glucose concentration is more highly cor-related with consumption of hypertonic PD solution, implying poor fluid control and more peritoneal dam-age. Glucose weight might be a less specific indicator of glucose load because high glucose weight may reflect the more frequent use of hypertonic solutions, or higher daily dialysate volume, or both. In addition, PD modality and dwell time also influence the glucose concentration and glucose weight. Compared with patients on continuous ambulatory PD, those on automated PD may contact and absorb more glucose because fresh dialysate is more fre-quently replenished. Calculation of the actual peritoneal glucose absorption requires precise measurement of glu-cose concentration and dialysate volume at the start and end of the dwell (1–3). Glucose absorption might influ-ence a number of metabolic imbalances including obesity, worse glycemic control, hyperinsulinemia, dyslipidemia, oxidative stress, and even altered adipokine levels, which are all involved in the determination of patient survival. Because of the retrospective nature of our study, we were not able to obtain the actual glucose absorption and to evaluate peritoneal changes for each patient precisely.

However, we choose to apply the average dialysate glu-cose concentration as a convenient and practical marker for evaluation of metabolic changes or patient outcomes in the PD population. For further studies concerning the issue of glucose content in PD dialysate, actual glucose absorption should be considered.

Our study has certain limitations. First, the number of patients, especially in the later years of the study, was somewhat small. Second, a retrospective cohort study places constraints on the analyses that can be used to in-terpret the mechanisms of the clinical factors associated with patient outcomes. Third, the results of retrospective cohort studies can be influenced by censoring. We be-lieve that this latter effect is unlikely to bias the results of this study because patients with DM (who often have high BMI and poor RRF) usually have worse technique survival and therefore may leave the study cohort earlier. These departures could lead only to an underestimate of the influence of DM, BMI, and RRF on glucose load, not an overestimate. A prospective, multicenter study with a large number of patients would be able to resolve the limitations of the present study.

Among patients undergoing PD, DM is known to be associated with technique failure and death. Patients with DM on PD are more likely to develop fluid overload, possibly because of their higher salt and water intake, higher prevalence of coronary artery disease, greater peritoneal transport of small solutes, and poor UF (15). In addition, patients with DM more frequently present with hyperglycemia and hence require more hypertonic PD fluids to achieve adequate UF (16). All of these fac-tors can result in higher glucose loads for patients with DM and can presumably explain why DM had the most significant role in determining the average glucose concentration administered to our patients.

According to previous studies, elevated BMI usually showed a neutral or deleterious effect on PD outcome (17). Reverse epidemiology about the influence of BMI on hemodialysis patients cannot be applied to PD patients (17,18). This discordance may be a result of high cata-bolic status or inability of BMI to differentiate adipose tissue from muscle mass in PD patients (19). Obesity in PD patients may result in faster decline of RRF, higher incidence of peritonitis, or proinflammatory status (20–22). Johnson et al. reported that each 1 m2 increase of body surface area was associated with an adjusted hazard ratio of 6.23 for the development of anuria (22). The results presented here show that high BMI and low RRF predicted a higher glucose load needed for PD until the end of year 3 (Table 3). Again, our study emphasizes the importance of RRF and weight control for patients undergoing long-term PD.




http://ww

w.pdiconnect.com

/D

ownloaded from


454


To reduce the need for peritoneal UF in PD patients, the use of diuretics in non-oliguric patients and reduc-tion in dietary salt and water intake are suggested (23). To achieve peritoneal UF with minimization of glucose use, PD solutions containing alternative osmotic agents such as icodextrin or amino acids have been developed (23,24). Using icodextrin in the long dwell has been shown to lower the glucose load and preserve RRF in PD patients (25,26). Newer glucose-based PD solutions with a combination buffer of bicarbonate and lactate have neutral pH and lower glucose degradation products (GDPs), and hence are thought to be more biocompatible (24). A glucose-sparing regimen combining icodextrin, amino acids, and low-GDP glucose-based PD solutions has been shown to be as safe and efficacious as the conventional regimens (27).

CONCLUSIONS

Our study suggests that the presence of DM, high BMI, and low RRF predicts high dialysate glucose load during long-term PD therapy. After 3 years of PD therapy, RRF had deteriorated, and DM was the only significant fac-tor associated with a need for high glucose load during PD. To decrease the glucose load in chronic PD patients, alternative osmotic agents such as icodextrin or amino acids should be considered in the daily PD regimen.

DISCLOSURES

The present work was supported by grants from the National Science Council of Taiwan (to KYH, NSC-95-2314-B-002-238-MY2) and by the Ta-Tung Kidney Foundation and the Mrs. Hsiu-Chin Lee Kidney Research Foundation (Taipei, Taiwan). The authors have no conflicts of inter-est to declare.

ACKNOWLEDGMENTS

The authors thank Ms. Ling-Chu Wu and Ms. Chia-Chi Cheng for their assistance in statistical computing.

REFERENCES

1. Lameire N, Matthys D, Matthys E, Beheydt R. Effects of long-term CAPD on carbohydrate and lipid metabolism. Clin Nephrol 1988; 30(Suppl 1):S53–8.

2. Grodstein GP, Blumenkrantz MJ, Kopple JD, Moran JK, Coburn JW. Glucose absorption during continuous ambula-tory peritoneal dialysis. Kidney Int 1981; 19:564–7.

3. De Santo NG, Capodicasa G, Senatore R, Cicchetti T, Cirillo D, Damiano M, et al. Glucose utilization from dialysate in

patients on continuous ambulatory peritoneal dialysis (CAPD). Int J Artif Organs 1979; 2:119–24.

4. Sitter T, Sauter M. Impact of glucose in peritoneal dialysis: saint or sinner? Perit Dial Int 2005; 25:415–25.

5. Wu HY, Hung KY, Huang JW, Chen YM, Tsai TJ, Wu KD. Initial glucose load predicts technique survival in pa-tients on chronic peritoneal dialysis. Am J Nephrol 2008; 28:765–71.

6. Peng YS, Chiang CK, Hung KY, Chiang SS, Lu CS, Yang CS, et al. The association of higher depressive symptoms and sexual dysfunction in male haemodialysis patients. Nephrol Dial Transplant 2007; 22:857–61.

7. Chaudhary K, Khanna R. Renal replacement therapy in end-stage renal disease patients with chronic liver disease and ascites: role of peritoneal dialysis. Perit Dial Int 2008; 28:113–17.

8. Davies SJ, Phillips L, Naish PF, Russell GI. Peritoneal glu-cose exposure and changes in membrane solute transport with time on peritoneal dialysis. J Am Soc Nephrol 2001; 12:1046–51.

9. Davies SJ, Brown EA, Frandsen NE, Rodrigues AS, Rodri-guez–Carmona A, Vychytil A, et al. Longitudinal membrane function in functionally anuric patients treated with APD: data from EAPOS on the effects of glucose and icodextrin prescription. Kidney Int 2005; 67:1609–15.

10. Davies SJ, Brown EA, Reigel W, Clutterbuck E, Heimbürger O, Diaz NV, et al. What is the link between poor ultrafiltra-tion and increased mortality in anuric patients on auto-mated peritoneal dialysis? Analysis of data from EAPOS. Perit Dial Int 2006; 26:458–65.

11. Twardowski ZJ, Nolph KD, Khanna R, Prowant BF, Ryan LP, Moore HL, et al. Peritoneal equilibration test. Perit Dial Bull 1987; 7:138–47.

12. Watson PE, Watson ID, Batt RD. Total body water vol-umes for adult males and females estimated from simple anthropometric measurements. Am J Clin Nutr 1980; 33:27–39.

13. Brown EA, Davies SJ, Rutherford P, Meeus F, Borras M, Riegel W, et al. Survival of functionally anuric patients on automated peritoneal dialysis: the European APD Outcome Study. J Am Soc Nephrol 2003; 14:2948–57.

14. Davies SJ, Brown EA. EAPOS: what have we learned? Perit Dial Int 2007; 27:131–5.

15. Yao Q, Lindholm B, Heimbürger O. Peritoneal dialysis prescription for diabetic patients. Perit Dial Int 2005; 25(Suppl 3):S76–9.

16. Gan HB, Chen MH, Lindholm B, Wang T. Volume control in diabetic and nondiabetic peritoneal dialysis patients. Int Urol Nephrol 2005; 37:575–9.

17. Abbott KC, Oliver DK, Hurst FP, Das NP, Gao SW, Perkins RM. Body mass index and peritoneal dialysis: “exceptions to the exception” in reverse epidemiology? Semin Dial 2007; 20:561–5.

18. Aslam N, Bernardini J, Fried L, Piraino B. Large body mass index does not predict short-term survival in peritoneal dialysis patients. Perit Dial Int 2002; 22:191–6.




http://ww

w.pdiconnect.com

/D

ownloaded from


455


19. Ramkumar N, Pappas LM, Beddhu S. Effect of body size and body composition on survival in peritoneal dialysis patients. Perit Dial Int 2005; 25:461–9.

20. McDonald SP, Collins JF, Rumpsfeld M, Johnson DW. Obesity is a risk factor for peritonitis in the Australian and New Zealand peritoneal dialysis patient populations. Perit Dial Int 2004; 24:340–6.

21. Stompór T, Sulowicz W, Dembinska–Kiec A, Janda K, Wójcik K, Zdzienicka A. An association between body mass index and markers of inflammation: is obesity the proinflamma-tory state in patients on peritoneal dialysis? Perit Dial Int 2003; 23:79–83.

22. Johnson DW, Mudge DW, Sturtevant JM, Hawley CM, Camp-bell SB, Isbel NM, et al. Predictors of decline of residual renal function in new peritoneal dialysis patients. Perit Dial Int 2003; 23:276–83.

23. Holmes C, Mujais S. Glucose sparing in peritoneal di-alysis: implications and metrics. Kidney Int Suppl 2006; (103):S104–9.

24. Chung SH, Stenvinkel P, Bergström J, Lindholm B. Bio-compatibility of new peritoneal dialysis solutions: what can we hope to achieve? Perit Dial Int 2000; 20(Suppl 5):S57–67.

25. Adachi Y, Nakagawa Y, Nishio A. Icodextrin preserves re-sidual renal function in patients treated with automated peritoneal dialysis. Perit Dial Int 2006; 26:405–7.

26. Rodríguez–Carmona A, Pérez Fontán M, García López E, García Falcón T, Díaz Cambre H. Use of icodextrin during nocturnal automated peritoneal dialysis allows sustained ultrafiltration while reducing the peritoneal glucose load: a randomized crossover study. Perit Dial Int 2007; 27:260–6.

27. le Poole CY, Welten AG, Weijmer MC, Valentijn RM, van Ittersum FJ, ter Wee PM. Initiating CAPD with a regimen low in glucose and glucose degradation products, with icodextrin and amino acids (NEPP) is safe and eff icacious. Perit Dial Int 2005; 25(Suppl 3):S64–8.




http://ww

w.pdiconnect.com

/D

ownloaded from


RISK FACTORS FOR HIGH DIALYSATE GLUCOSE USE IN PD … · Use of high concentrations of glucose for...

Documents

Transcript of RISK FACTORS FOR HIGH DIALYSATE GLUCOSE USE IN PD … · Use of high concentrations of glucose for...