Assessing the prevalence of protein-energy wasting in haemodialysis patients: A cross-sectional monocentric study

(1)

Original article

Assessing the prevalence of protein-energy wasting in haemodialysis patients: A cross-sectional monocentric study

Rajaa Essadik

^a

, Rajaa Msaad

^a

, Halima Lebrazi

^a

, Hassan Taki

^a

, El Hassane Tahri

^b

, Anass Kettani

^a

, Ghizlane Madkouri

^c

, Benyounes Ramdani

^c

, Rachid Saı¨le

^a,

*

aLaboratoryofbiologyandhealth,URAC34,HassanIIuniversity-Casablanca,facultyofsciencesBenM’Sik,avenueCdtDrissElHarti,BP7955,SidiOthmane, Casablanca,Morocco

bLaboratoryofmoleculargeneticsandpathophysiology,HassanIIuniversity-Casablanca,facultyofsciencesBenM’sik,avenueCdtDrissElHarti,BP7955, SidiOthmane,Casablanca,Morocco

cDepartmentofnephrology-transplantationandhemodialysis,universityhospitalcenterIbnRochd,Casablanca,Morocco

1. Abbreviations

BIA bioelectrical impedance analysis BMI body mass index

CKD chronic kidney disease

CRP C-reactive protein

DEXA dual-energy X-ray absorptiometry ESRD end-stage renal disease

HD haemodialysis

HDL-C HDL-cholesterol

ISRNM International Society of Renal Nutrition and Metabolism

ARTICLE INFO

Articlehistory:

Received9August2016 Accepted23February2017

Keywords:

Albumin

Chronickidneydisease Dietaryintake Nutritionalassessment Prealbumin

Protein-energyintake

ABSTRACT

Introduction. –

Protein-energy wasting (PEW) is a strong predictive factor for morbidity and mortality in haemodialysis (HD) patients. However, there is no consensus for its assessment. The present study aimed to assess the nutritional status of patients on chronic HD by use of different nutritional assessment parameters, and at verifying which can identify the greatest number of HD patients with PEW. Also, to investigate predictors of nutritional status in a haemodialysis center in Morocco.

Patients and methods.–

This is a cross-sectional analysis performed on 126 patients aged 44.82 14.01 years, undergoing maintenance HD in the Department of nephrology of the university hospital centre of Casablanca, Morocco. Energy and nutrients intake assessment was obtained by a three-day period food recall. Biochemical parameters, bioelectric impedance analysis, and subjective global assessment (SGA), have been performed to assess nutritional status.

Results. –

According to SGA the prevalence of PEW was 74.62%. However, when using the ISRMN malnutrition criteria only 36.50% of the patients were diagnosed with PEW. Pearson correlation showed a negative association between the degree of malnutrition evaluated by SGA and serum prealbumin (r = 0.54;

P

= 0.0001), serum albumin (r = 0.50;

P

= 0.001), energy (r = 0.34;

P

= 0.002), protein intake (r = 0.41;

P

= 0.0001), and a signiﬁcant positive correlation with CRP (r = 0.65;

P

= 0.0001) was determined, but not with anthropometric measurements nor lipids proﬁle. The areas under the receiver operating characteristic curve were 0.841 (95% CI: 0.751–0.932) for serum prealbumin, and 0.737 (95%

CI: 0.634–0.840) for serum albumin.

Conclusion. –

Our results showed a high prevalence of PEW among Haemodialysis patients. Also, our ﬁndings suggest that SGA, serum albumin and prealbumin may be relative appropriate and practical markers for assessing nutritional status in HD patients.

C

2017 Socie´te´ francophone de ne´phrologie, dialyse et transplantation. Published by Elsevier Masson SAS. All rights reserved.

* Correspondingauthor.

E-mailaddress:sailerachid@yahoo.fr(R.Saı¨le).

Available online at

ScienceDirect

www.sciencedirect.com

http://dx.doi.org/10.1016/j.nephro.2017.02.013

1769-7255/^C 2017Socie´te´ francophonedene´phrologie,dialyseettransplantation.PublishedbyElsevierMassonSAS.Allrightsreserved.

(2)

LDL-C LDL-cholesterol

NKF/KDOQI The National Kidney Foundation and Kidney Disease/

Dialysis Outcomes and Quality Initiative PEW protein-energy wasting

PEM protein-energy malnutrition SGA subjective global assessment TC total cholesterol

TG triglycerides

2. Introduction

Protein-energy wasting (PEW) represents one of the most serious complications of chronic kidney disease (CKD) [1] and its consequences, particularly for patients on maintenance dialysis, are devastating in terms of quality of life, morbidity, and mortality [2].

Several designations have been previously used for the syndrome of wasting such as uremic malnutrition, protein-energy malnutrition (PEM), or malnutrition-inflammation complex [3]. Hence, the confusing terminology used to describe the interrelated mechanisms causing wasting, malnutrition and inflammation in patients with CKD led the International Society of Renal Nutrition and Metabolism (ISRNM) to convene an expert panel to reexamine the nomenclature used for the diagnosis of the wasting syndrome, distinct from malnutrition and inflammation [4]. PEW is therefore the new terminology proposed to describe a state of decreased body stores of protein and energy fuels (i.e., body protein and fat mass) [5].

In haemodialysis (HD) patients, various studies using different criteria have been used to establish the presence of PEW and reported that the prevalence varies between 15 and 76%, according to the type of dialysis modality, nutritional assessment tools, and origin of the patient population [6]. Hence, several factors are associated with this high prevalence of malnutrition in HD patients, including recurrent illness, inadequate food intake, hormonal and gastrointestinal disorders, dietary restrictions, drugs that alter nutrient absorption, insufﬁcient dialysis, and constant presence of associated diseases. Furthermore, uremia, acidosis, and HD procedure per se are hypercatabolic and associated with the presence of an inﬂammatory state and socioeconomic and cultural aspects [7].

Thus, because of that high prevalence of PEW in HD patients, nutritional assessment should be performed to identify the risks and/or causes of deterioration of the nutritional status, and to establish a nutritional diagnosis [8]. However, there is not a method that can be considered as a gold standard to validate the scoring systems. The quest for a gold standard has resulted in many clinical scoring lists, tools, and parameters to diagnose malnutri- tion. The National Kidney Foundation/and Kidney Disease Outco- mes and Quality Initiative (NKF/KDOQI) recommends combining measures tools and endorses the use of subjective global assessment (SGA) as an effective, noninvasive, fast, easy, and inexpensive nutrition assessment tool [9]. The SGA is based on clinical history and physical examination, which are combined subjectively to form a global rating of well-nourished, moderately malnourished, or severely malnourished. Other clinical nutrition- related scoring lists that have been proposed to assess PEW include the Geriatric Nutritional Risk Index (GNRI) and the composite score on protein-energy nutritional status (cPENS) [10]. Furthermore, a number of more or less individual parameters have been associated with PEW, such as serum albumin, body mass index (BMI) [11] and the normalized protein nitrogen appearance (nPNA) rate [12].

However, which of those methods should be used to detect more precisely a patient with PEW is yet to be determined. The difﬁculty in establishing the best method to assess PEW lies in the fact that all such parameters have limitations when used in isolation [13]. In this sense, the ISRNM expert panel has proposed a set of criteria to identify PEW in CKD patients [5], and suggests that PEW can be diagnosed if at least three of the four categories are present [5]:

serum chemistry (low serum levels of albumin, prealbumin, or total cholesterol);

body mass (unintentional weight loss overtime, decreased BMI, or total body fat percentage);

muscle mass (decreased muscle mass over time, mid-arm muscle circumference, or creatinine appearance);

dietary intake (unintentional decreased protein or energy intake).

In Morocco, there is a lack of data regarding the nutritional status of HD patients because it is not a routine practice in hospital treatment centers. However, the prevalence of PEW and the appropriateness of diagnostic criteria have not yet been described using ISRNM criteria in our country. Therefore, the ﬁrst aim of our study was to assess the prevalence of PEW in patients on chronic HD by using of different nutritional assessment parameters, and at verifying which can identify the greatest number of HD patients with PEW. The second aim sought to investigate predictors of nutritional status in a haemodialysis center in Morocco.

3. Patients and methods

This is a cross-sectional analysis of the patients with end-stage renal disease (ESRD) undergoing maintenance HD in the Depart- ment of nephrology-transplantation and haemodialysis of the university hospital centre of Casablanca, Morocco. The study was conducted after an approval from the institutional review hospital board and a written informed consent was obtained from each patient prior to enrollment in the study.

Our study was conducted on 126 ESRD patients (60 men and 66 women; ranging from 18 to 65 years), treated with HD 3 times a week, for at least 4 hours per session. The inclusion criteria were patient’s age of 18 years and over, and HD treatment for at least the previous 6 months. Exclusion criteria were: dialysis vintage less than 6 months, major surgery within two weeks, an implantable cardioverter deﬁbrillator or pacemaker, acute intercurrent disease, and language barriers and physical or mental disability making participation unfeasible.

Patient data (age, gender, dialysis vintage, primary kidney disease and co-morbidity) were taken from patient records.

3.1. Diagnosis of PEW

PEW was identiﬁed in patients based on diagnostic criteria provided by the International Society of Renal Nutrition and Metabolism expert panel [5]. In our study, patients were assessed for serum albumin < 38 g/dL, BMI < 23 kg/m

²

, muscle wasting 10%

over 6 months and dietary energy intake < 25 kcal/kg ideal body weight. In addition, we assessed patients for potential markers of PEW as proposed by ISRNM in terms of body mass and composition measures: total body fat percentage, laboratory markers: serum prealbumin, serum cholesterol and CRP.

We also, proposed to include the SGA as an additional marker

for PEW. A more detailed description of this semi-quantitative

scoring system (SGA) based on history and physical examination

was later published by Detsky et al. [14]. The history focused on

(3)

7 variables, namely weight change in preceding 6 months, change in dietary intake, presence of gastro intestinal symptoms, change in functional capacity, subcutaneous loss of fat, muscle wasting and edema. A seven point scoring system was applied to the above 7 variables. Each SGA component is scored 1 to 7, with the lowest values indicative of severity. On the basis of subjective consider- ation of all scores from each component, an overall score is assigned to each patient: 1 to 2, severe PEW; 3 to 5, moderate to mild PEW; and 6 to 7, well-nourished.

3.2. Dietary nutrient intake assessment

For dietary assessment, each patient was interviewed for the consumption of food and beverages using 24-hour recall method and food frequency questionnaire. The three-day dietary recall included a dialysis day, a weekend day, and a non-dialysis day, were collected by interviews, during which the subjects were provided with a color pictures of common food and their servings in order to help them in estimating the real amounts of consumed food. The daily intake for each studied nutrient was calculated as the average of the three-day food records. The portion sizes of foods consumed by each patient were converted into percent carbohydrates (CHO), fats, proteins and energy.

Dietary energy and nutrient intakes were compared with the NKF/KDOQI recommendations to assess their adequacy [15]. The values of energy and protein intake were expressed in kg of current weight per day.

3.3. Blood sampling and biochemical measurements

Venous blood samples (about 5 mL) were obtained from patients after overnight fasting for a minimum of 12 hours and before the beginning of haemodialysis. The samples were centrifuged at 4000 rpm for 10 minutes to separate serum from blood cells. Serum lipids: Total cholesterol (TC), HDL-cholesterol (HDL-C), triglycerides (TG) were determined using commercially available test kits ‘‘SGM italia’’. LDL-cholesterol (LDL-C) was calculated using Friedwald’s formula [16]. Serum albumin was determined by the bromocresol green method. Both serum prealbumin and C-reactive protein (CRP) were measured by nephelometry using the MININEPH

^TM

of the Binding Site Group, Birmingham, UK.

3.4. Bioelectrical impedance analysis (BIA)

Anthropometric variables included dry weight (weight after dialysis session), body mass index (BMI; weight in kilograms/

height in square meters [kg/m

²

]), fat mass percentage and muscle mass percentage were measured using the Omron BF11 Body Composition Analyzer

^TM

. For the BIA measurements, the subject stood in an upright position with bare feet on the analyzer foot pads. The impedance between the two feet was measured while an alternating current (50 kHz and less than 500 m A) passed through the lower body.

3.5. Statistical analysis

Continuous variables are expressed as mean standard devia- tion. Categorical variables are expressed as percentages and compared using the Chi

²

test. Continuous variables were compared using the Student’s t-test for independent samples after verifying the normality of distribution or by analysis of variance (ANOVA) when comparing more groups. The relationships between varia- bles were assessed by Pearson’s correlation analysis and linear regression models. Receiver operating characteristic curve (ROC) analysis was used to quantify the assessing value of independent

parameters for nutritional status using SGA as the reference criteria. Differences with a P < 0.05 were considered statistically signiﬁcant.

4. Results

4.1. Patients characteristics and prevalence of PEW

The patients’ mean age was 44.82 14.01 years, and female patients predominated (52.38%). The clinical characteristics and baseline analysis of the study population and prevalence of PEW are shown in Table 1. The most frequent aetiology of chronic kidney disease (CKD) was glomerulonephritis hypertensive (25.40%), follo- wed by glomerulosclerosis (14.30%) and diabetes mellitus (3.96%).

In this study, based on SGA, 25.40% (n = 32) of HD patients were well-nourished (SGA-A), 60.32% (n = 76) were moderately mal- nourished (SGA-B) and 14.30% (n = 18) were severely malnour- ished (SGA-C). However, when using the criteria proposed by the ISRNM to estimate PEW only 36.50% (n = 46) of the patients were diagnosed with PEW (met three of the four criteria indicating malnutrition) (Table 2). In fact, the method that identiﬁed the greatest number of patients with PEW was dietary intake (n = 102;

80.95%), followed by SGA (n = 94; 74.62%), serum prealbumin (n = 90; 71.43%), BMI (n = 81; 64.28%), serum albumin (n = 76;

60.32%), and, at last, muscle mass (n = 38; 30.16%).

Because the SGA seems to be the method capable of detecting a high number of patients with PEW, the comparison analysis of

Table1

Demographicandclinicalcharacteristicsofthestudypatients.

Characteristics Allpatients(n=126)

n (%)

Gender

Women 66 52.38

Men 60 47.62

Age(years)

18–40 34 26.98

41–60 60 47.62

>60 32 25.40

Durationofhemodialysis(years)

10 49 38.88

>10 77 61.11

Primaryrenaldisease

Glomerulonephritishypertensive 32 25.40

Glomerulosclerosis 18 14.30

Diabetesmellitus 5 3.96

Othercauses 46 36.51

Indeﬁnite 25 19.81

Table2

Prevalenceofprotein-energywastingaccordingtoSGAandISRNMcriteria.

Characteristics Allpatients(n=126)

n (%)

PEWbyISRNMcriteria

Serumalbumin<3.8g/dL 76 60.32

Serumprealbumin<30mg/dL 94 74.60

Serumcholesterol<100mg/dL 9 7.14

BMI<23kg/m² 81 64.28

Musclewasting(10%over6months) 38 30.16

Totalbodyfat<10% 10 7.93

Dietaryproteinintake<0.8g/kg/day 98 77.77 Dietaryenergyintake<25kcal/day 102 80.95 Subjectiveglobalassessment

SGA-A 32 25.40

SGA-B 76 60.32

SGA-C 18 14.30

SGA:subjectiveglobalassessment;ISRNM:InternationalSocietyofRenalNutrition andMetabolism;BMI:bodymassindex.

(4)

dietary intake, biochemical parameters and body composition data of patients were assessed according to SGA grades, as presented in Table 3.

The mean age of patients and duration of HD were higher but not signiﬁcant in SGA grade C compared with patients in SGA grades A and B.

Regarding dietary intake, the mean dietary energy intake was 23.00 9.10 kcal/kg of current weight/day, which is below the value recommended for this population [27]. Likewise, the mean protein intake (0.80 0.30 g/kg of current weight/day) was also below the value recommended for HD patients. Although daily protein and fat consumption of the SGA-A group was higher than SGA-B and SGA-C groups (P = 0.002 and P = 0.03, respectively).

There was no signiﬁcant difference between groups for anthropometric measurements, body fat and muscle mass.

Patients of group C had signiﬁcantly lower BMI (P = 0.001) and body weight (P = 0.05) when compared to group B and A.

Moreover, along with the malnutrition severity, serum pre- albumin and albumin levels were decreased (P = 0.0001). Total protein revealed significant decreases when SGA grades A and B were compared with grade C. While, CRP level was significantly higher in malnourished patients (P = 0.01). There were no significant differences in phosphorus, calcium and lipids profile between well-nourished and malnourished patients (Table 3).

4.2. Clinical and biological factors associated with SGA

Pearson correlation showed a negative association between SGA and serum prealbumin (r = 0.54; P = 0.0001), serum albumin (r = 0.50; P = 0.001), energy intake (r = 0.34; P = 0.002), protein intake (r = 0.41; P = 0.0001), fat intake (r = 0.45; P = 0.01), carbohydrate (r = 0.30; P = 0.03), and a positive correlation between CRP (r = 0.65; P = 0.0001), indicating that more the degree of malnutrition increases, the levels of parameters falls,

except for CRP which is directly correlated with the degree of malnutrition. There was a negative, but not significant, correlation between BMI, TC, total protein and SGA score (Table 4). The SGA score had no significant correlation with age nor sex, suggesting that both males and females had an equal tendency toward malnutrition. However, age was inversely and significantly correlated with protein intake (r = 0.25; P < 0.001) and serum albumin (r = 0.30; P < 0.001). Moreover, serum albumin levels may fall modestly with a sustained decrease in dietary protein intake and may rise with increased protein or energy intake.

The above study had demonstrated that both serum albumin and prealbumin were relatively good markers for assessing nutritional status. The ROC curve analysis was used to analyze whether the two markers could be better predictors for nutritional

Table3

Foodintake,anthropometricmeasurementsandbiochemicalparametersofpatientsaccordingtoSGAscore.

Variables Overall

(meanSD)

SGA-A (meanSD)

SGA-B (meanSD)

SGA-C (meanSD)

P-value

Age(years) 44.8214.01 43.0413.60 43.9612.98 48.7816.68 NS

DurationofHD(months) 144.9574.82 131.1175.30 148.3273.22 150.0080.46 NS

Foodintake

Energy(kcal/day) 1306.63496.21 1614.33363.40 1260.61444.02 1163.72562.02 0.001**

Energy(kcal/kg) 23.009.10 28.568.80 22.157.26 20.3811.86 0.001**

Protein(g/day) 45.9520.28 58.8818.78 45.8317.76 33.9020.95 0.002**

Protein(g/kg) 0.800.37 1.080.47 0.760.28 0.580.33 0.001**

Fat(g/day) 46.6343.01 73.2156.96 45.8739.74 25.1119.95 0.03*

Carbohydrates(g/day) 173.7468.76 208.357.77 171.7462.34 148.6083.34 NS

Biochemicalparameters

Albumin(g/L) 37.535.22 41.054.06 38.034.75 33.023.80 0.0001***

Prealbumin(mg/L) 269.9496.87 352.8788.27 262.4695.62 217.1052.70 0.0001***

Calcium(mg/dL) 8.711.17 8.601.07 8.771.26 8.681.02 NS

Phosphorus(mg/dL) 4.241.74 3.821.57 4.641.94 4.241.74 NS

Totalprotein(g/L) 72.587.33 74.026.60 71.807.01 69.278.57 0.01*

CRP(mg/L) 32.7221.36 14.3812.78 29.6515.33 57.5620.06 0.001**

TC(g/L) 1.520.37 1.530.31 1.520.40 1.510.38 NS

LDL-c(g/L) 0.760.36 0.820.44 0.760.35 0.700.31 NS

HDL-c(g/L) 0.510.30 0.500.27 0.480.30 0.560.31 NS

TG(g/L) 1.370.63 1.200.52 1.470.68 1.240.52 NS

Haemoglobin(g/dL) 9.651.57 9.611.50 9.731.76 9.470.96 NS

Anthropometry

Height(cm) 161.387.50 162.545.46 161.3410.92 159.366.03 NS

Weight(kg) 59.0610.50 62.3414.48 57.1012.64 54.7410.37 0.05*

BMI(kg/m²) 21.813.98 24.536.08 21.403.25 19.202.62 0.001**

Musclemass(%) 33.107.97 32.688.33 32.338.36 31.559.03 NS

Bodyfat(%) 24.7812.02 24.8012.70 24.0411.71 26.8012.50 NS

Waistcircumference(cm) 85.2512.31 87.5012.32 85.3311.35 84.8112.47 NS

Hipcircumference(cm) 94.269.96 95.2510.96 94.039.88 93.239.71 NS

AllvaluesareexpressedasmeanSD.P-valuebasedonANOVAtestforparametriccontinuousvariableisusedforcomparisonamongthethreegroups:well-nourishedpatients, mildtomoderatemalnutritionandseveremalnutrition(*P<0.05).NS:non-signiﬁcant;SD:standarddeviation;HD:haemodialysis;CRP:C-reactiveprotein;TC:totalcholesterol;

TG:triglyceride;HDL-C:high-densitylipoprotein-cholesterol;LDL-C:low-densitylipoprotein-cholesterol;BMI:bodymassindex.*P<.05,**P<.001,***P<.0001.

Table4

CorrelationanalysisbetweenvariablesofPEWandsubjectiveglobalassessmentfor studyparticipants.

Parameters Pearsonindex(r) Statisticalprobability(P)

Age 0.10 NS

Gender 0.21 NS

Serumalbumin(g/L) 0.50 0.001

Serumprealbumin(mg/L) 0.54 0.001

Totalprotein 0.10 NS

TC(g/L) 0.02 NS

BMI(kg/m²) 0.10 NS

CRP(mg/L) 0.65 0.04

Musclemass(%) 0.15 NS

Totalbodyfat(%) 0.08 NS

Energy(kcal/kg/day) 0.34 0.002

Protein(g/kg/day) 0.41 0.0001

Fat(g/day) 0.45 0.01

Carbohydrates(g/day) 0.30 0.03

PEW:protein-energywasting;NS:non-signiﬁcant;CRP:C-reactiveprotein;TC:

totalcholesterol;BMI:bodymassindex;r:Pearson’scorrelationcoefﬁcients.

(5)

status using SGA as the reference standard (Fig. 1). The area under the curve (AUC) was 0.841 (95% CI: 0.751–0.932) for serum prealbumin (Fig. 1A), and 0.737 (95% CI: 0.634–0.840) for serum albumin (Fig. 1B). High level of serum albumin and prealbumin were shown to be the appropriate predictors for better nutritional status (P < 0.05). Serum albumin with a threshold value of 37.5 g/L provided 75.2% sensitivity and 26.3% speciﬁcity for the prediction of malnutrition. Serum prealbumin with a threshold value of 270.5 mg/L provided 87.5% sensitivity and 33.3% speciﬁcity for the prediction of malnutrition.

5. Discussion

This study analyzes for the ﬁrst time, the prevalence of PEW among HD patients by use of different nutritional assessment methods, including the criteria established by the ISRNM expert

panel and SGA questionnaire. The results of our study, showed a high prevalence of PEW ranged from 30% to 81% when assessed with the different methods in isolation. It is worth noting that, even in a small sample, all patients assessed were diagnosed as malnourished by use of at least one of the methods, emphasizing the high nutritional risk of the population studied.

The results showed that, our population of study is young; the mean age of patients was 44.82 14.01 years. This result was consistent with the ﬁndings of a number of studies in our country, but differed from others [17]. Though the causes of renal diseases are similar all over the world, their incidences differ signiﬁcantly among countries. The incidence and prevalence of ESRD attributed to diabetic nephropathy have increased over the years as reported in the regional and national registries. According to the investigation into the prevalence and the risk factors of the chronic kidney disease in Morocco (MaReMar), the leading causes of the kidney disease are the diabetes (32.8%), arterial hypertension (28.2%) and the urinary lithiasis (9.2%).

However, our study, which performed in only one center of hemodialysis in Morocco, showed that the hypertensive glomeru- lonephritis is the leading cause of nephropathies (25.40%), whereas only 3.96% of the patients have a diabetic nephropathy, but this result must be considered with caution because the causes of ESRD were unknown in 19.81% of our patients.

Concerning the nutritional status, according to the SGA, 25.40%

of our patients were well-nourished (SGA-A), 60.32% were moderately malnourished (SGA-B) and 14.28% were severely malnourished (SGA-C). This prevalence was similar to other studies who also reported a very high prevalence of malnutrition using the SGA, ranging from 31% to 75% [8,18]. However, the rate of severely malnourished patients in our study (14%) was very high to the percentage category ranking worldwide (6% to 9%) [19]. In contrast, a similar percentage was reported in a Sweden study in which 13% of participants were severely malnourished [20] and to the results of Habibe Sahin et al., who found severe malnutrition rate as 15.3% [21].

Because of the difﬁculty in diagnosing a patient with PEW, the ISRNM expert panel has proposed diagnostic criteria of PEW with four categories. When the criteria proposed by the ISRNM panel were applied, only 46 patients (36.50%) of the sample assessed were considered malnourished. Approximately, a similar result was reported in a Spanish study, which showed that 40% of their HD population met PEW criteria [22]. This prevalence was lower than that has been seen in European CONTRAST study (74%) [23]

and higher than that has been found in a Japanese study (15%) [6]. The relatively higher rate of malnutrition in our sample compared to previous may be a consequence of the variations between the countries that conducted those studies, comorbidi- ties, sample heterogeneity, diversity in dietary patterns, socioeco- nomic status, and medical care at hospitals from one country to another, or even within the same country [19,22].

However, there are several clinical, nutritional and biochemical parameters that may be indicative of PEW in patients on HD.

Concerning biochemical parameters, serum albumin is most frequently used to assess protein malnutrition, based on the concept that the level of serum albumin reflects the visceral protein status [24]. Hypoalbuminemia in HD patients can be difficult to interpret and may reflect protein malnutrition in clinically stable patients. In our study, serum albumin level was low (mean 37.53 5.22 g/L) indicating PEW. Also, like many other studies a statistically significant lower level of serum albumin was determined when comparing well-nourished to malnourished HD patients [21,25]. Hence, the United States Renal Data System (USRDS) database results have shown that every 1 g/dL fall in the serum albumin level is associated with 39% increase in the risk of cardiovascular death [26].

Fig.1.Receiveroperating characteristic(ROC)curve fordifferentbiochemical parametersinassessingnutritionalstatususingsubjectiveglobalassessment(SGA) asthereferencestandard(*P<0.05);AUC:theareaunderthecurve.Theidealtest wouldhave100%sensitivityand100%speciﬁcityandreachtheupperleftcornerof thegraph;atestwithnodiagnosisvaluewouldliealongthediagonalbetweenthe lowerleftcornerandtheupperrightcorner.A.ROCcurveforserumprealbumin.

B.ROCcurveforserumalbumin.

(6)

On the other hand, previous studies showed that serum prealbumin is like a mirror representing the true protein metabolic status under inflammatory conditions in HD patients, and is a more sensitive indicator for the nutrition status due to its shorter half- life which is about two to three days which is much shorter than that of serum albumin [27]. In our study, ROC analysis indicated the diagnostic efficiency of serum prealbumin and albumin in malnutrition. In fact, serum prealbumin showed relatively higher sensitivity (87.5% vs. 75.2%) and had a good compatibility with SGA compared to serum albumin. Moreover, comparing to serum albumin, it seemed that serum prealbumin could be better predictor. In addition, studies have shown that albumin levels decrease significantly, as age advances when compared to prealbumin [28], which is in agreement with our finding. However, both serum prealbuminand serum albumin had a low specificity to assess malnutrition (33.3%. vs. 26.3%). This may attribute to the fact that serum albumin can be influenced by other morbidity factors including infection, over-hydration, inflammation and chronic disease [29], or it can be due to the threshold value considered in our study. But, despite all, serum albumin should also be considered and the combination of both serum albumin and prealbumin may be as convenient as SGA for evaluation PEW of HD patients.

Another important cause of PEW in HD patients is chronic inflammation. It may induce any degree of severity of PEW, and may be associated with the most severe forms of PEW encountered in chronic HD patients [30]. Previous studies have reported that higher levels of proinflammatory cytokines contribute to increased lipolysis [31], muscle protein breakdown, and nitrogen loss, leading to sarcopenia and increased mortality in these patients [32]. In our study, C-reactive protein was significantly higher in malnourished patients and in well-nourished patients. In addition, chronic inflammation may be associated with inadequate dietary protein and energy intake because it can cause anorexia and may be can explain the high prevalence of PEW in our patients.

Evidence shows that the assessment of quality and quantity of food intake is another important step in the management and treatment of HD patients [33]. In a large cohort of adult patients with moderate to advanced CKD, Kopple [34] shows that dietary protein intake and nutritional status correlated directly with glomerular ﬁltration rate (GFR). Thus, nutritional guidelines suggest daily energy intake higher that 30–35 kcal/kg ideal body weight and daily protein intake higher than 1.1–1.2 g/kg ideal b.w.

[15]. Also, recent studies have shown that HD patients’ survival only becomes impaired when protein intake is below 0.9 g/kg/d and that survival is highest for a protein intake between 1.0 and 1.4 g/kg/d [29]. Our study showed that dietary intakes of energy and protein in a large percentage of HD patients (80.95% and 77.77%, respectively) were lower than recommended intakes and there was a signiﬁcant association between the prevalence of PEW and dietary intakes of energy and protein. Available literature shows that the most signiﬁcant cause for these inadequate intakes in HD patients are anorexia, reduced food consumption to prevent hyperkalemia or hyperphosphatemia, dental problems, further dietary restrictions because of underlying illnesses such as diabetes and physical or economic inability to purchase food [7,35].

Lastly, and no less importantly, bioelectrical impedance analysis is a rapid, reliable and noninvasive method that is used for determination of nutritional status of HD patients [36]. While the above technique, used in our study, assumes only two body compartments, contrary to dual-energy X-ray absorptiometry (DEXA), which can estimate three body compartments consisting of fat mass, lean body mass, and bone mass [37]. An updated search of the current literature found that DEXA is used as the reference standard for whole body composition analysis in research studies.

In addition, the more elaborate methods are costly and time- consuming, factors, which confine their use to a few research centers. Therefore, it is difficult to draw general conclusions by extrapolation of our findings, especially those of fat mass and lean body mass.

Many studies stress the protective effect of a higher BMI in dialysis patients; in particular, a BMI of 23 kg/m

²

or higher seems to reduce the risk of morbidity and mortality and is associated with improved survival [38]. Conversely, a BMI lower than 20 kg/m

²

, suggestive of reduced fat and lean body mass, is considered an index of malnutrition: this was detected only in a small percentage of our patients (15%). Among our patients, 64.28% had low BMI (< 23 kg/m

²

), i.e., at high risk of morbidity and mortality. Also, we have found a negative but not signiﬁcant correlation between BIA measurements and SGA, these results are not in agreement with previous studies [39].

Some weaknesses and limitations of our study should be noted.

This study was conducted at a single dialysis center with a small sample size. Furthermore, the quest to the preferred test to assess PEW should be further investigated in longitudinal and/or interventional studies to verify our results.

6. Conclusion

The prevalence of PEW is high (74.62% by SGA and 36.50% by ISRNM-based criteria) in Moroccan HD patients. Among the nutritional parameters used, we suggest that SGA, serum albumin and prealbumin may be relative appropriate and practical markers for assessing nutritional status in HD patients of our country.

Moreover, our ﬁnding highlights the need to implement and evaluate prevention strategies and management of PEW to increase the quality of life and decrease the morbidity and mortality of the HD patients.

Disclosure of interest

The authors declare that they have no competing interest.

Acknowledgements

This work was supported by the National Centre for Scientiﬁc and Technique research (CNRST).

References

[1]Gracia-IguacelC,Gonza´lez-ParraE,Barril-CuadradoG,SańchezR,EgidoJ, Ortiz-ArduańA,etal.Definingprotein-energywastingsyndromeinchronic kidney disease: prevalence and clinical implications. Nefrologia 2014;

34(suppl.4):507–19.

[2]FerozeU,NooriN,KovesdyCP,MolnarMZ,MartinDJ,Reina-PattonA,etal.

Quality-of-life andmortality in hemodialysis patients:roles of raceand nutritionalstatus.ClinJAmSocNephrol2011;6:1100–11.

[3]Kalantar-ZadehK,IkizlerTA,BlockG,AvramMM,KoppleJD.Malnutrition- inﬂammationcomplexsyndromeindialysispatients:causesandconsequen- ces.AmJKidneyDis2003;42(Suppl.5):864–81.

[4]RiellaMC.Nutritionalevaluationofpatientsreceivingdialysisfortheman- agementofprotein-energywasting:whatisoldandwhatisnew?JRenNutr 2013;23(Suppl.3):195–8.

[5]FouqueD,Kalantar-ZadehK,KoppleJ,CanoN,ChauveauP,CuppariL,etal.A proposednomenclatureanddiagnosticcriteriaforprotein-energywastingin acuteandchronickidneydisease.KidneyInt2008;73(Suppl.4):391–8.

[6]Sonoko Y,Yumiko S,MayuT,SayakaM,Yu SaitoMD,KazuakiM,etal.

Prevalenceofprotein-energywasting (PEW)andevaluation ofdiagnostic criteriaandetiologyinJapanesemaintenance hemodialysispatients.Asia PacJClinNutr2016;25(Suppl.2):292–9.

[7]Kalantar-ZadehK,BlockG,McAllisterCJ,HumphreysMH,KoppleJD.Appetite andinﬂammation,nutrition,anemia,andclinicaloutcomeinhemodialysis patients.AmJClinNutr2004;80:299–307.

[8]VeginePM,FernandesAC,TorresMR,SilvaMI,AvesaniCM.Assessmentof methodstoidentifyprotein-energywastinginpatientsonhemodialysis.JBras Nefrol2011;33(Suppl.1):39–44.

(7)

[9]KoppleJD.Clinicalpracticeguidelinesfornutritioninchronicrenalfailure.K/

DOQI,NationalKidneyFoundation.AmJKidneyDis2001;37(Suppl.2):

66–70.

[10]BeberashviliI,AzarA,SinuaniI,KadoshiH,ShapiroG,FeldmanL,etal.

Comparisonanalysisofnutritionalscoresforserialmonitoringofnutritional statusinhemodialysispatients.ClinJAmSocNephrol2013;8:443–51.

[11]Kalantar-ZadehK,StrejaE,MolnarMZ,LukowskyLR,KrishnanM,KovesdyCP, etal.Mortalitypredictionbysurrogatesofbodycomposition:anexamination oftheobesityparadoxinhemodialysispatientsusingcompositerankingscore analysis.AmJEpidemiol2012;175:793–803.

[12]RavelVA,MolnarMZ,StrejaE,KimJC,VictoroffA,JingJ,etal.Lowprotein nitrogenappearanceasasurrogateoflowdietaryproteinintakeisassociated withhigherall-causemortalityinmaintenancehemodialysispatients.JNutr 2013;143:1084–92.

[13]Kalantar-ZadehK.Protein-energywasting.Nutritioninkidneydisease.Nutri- tionandhealth,2nded.,NewYork:SpringerScienceBusinessmedia;2014.p.

219–30.

[14]DetskyAS,BakerJP,RourkeK,JohnstonN,WhitwellJ,MendelsonRA,etal.

Predictingnutritionassociatedcomplicationsforpatientsundergoinggastro- intestinalsurgery.JParenterEnteralNutr1987;11:8–13.

[15]JuillardL,Guebre-Egziabher F,FouqueD.Whatisthebeneﬁtofthenew EuropeanRecommendationsfordialysisinnutrition?NephrolTher2010;6:

S2–6.

[16]FriedewaledWT,LevyR,FedicksonD.Estimationoftheconcentrationoflow- densitylipoprotein-cholesterol inplasmawithoutuseofthe preparative ultracentrifuge.ClinChem1972;18:499–502.

[17]MpioI,CleaudC,ArkoucheW,LavilleM.Resultsoftherapeuticsstrategyof protein-energywastinginchronichemodialysis:aprospectivestudyduring 12months.NephrolTher2015;11:97–103.

[18]ElliotHA,RobbL.Computer-basedundernutritionscreeningtoolforhemodi- alysispatients.NephrolDialTransplant2009;1:1–6.

[19]TayyemRF,MrayyanMT,HeathDD,BawadiHA.Assessmentofnutritionalstatus amongESRDpatientsinJordanianhospitals.JRenNutr2008;18(Suppl.3):

281–7.

[20]QureshiAR,AlvestrandA,DanielssonA,Divino-FilhoJC,GutierrezA,Lindholm B,etal.Factorsinﬂuencingmalnutritioninhemodialysispatients:across- sectionalstudy.KidneyInt1998;53:773–82.

[21]SahinH,Y´nanc¸N,Katrancy´ D,O¨ zlemAslanN.Isthereacorrelationbetween subjectiveglobalassessmentandfoodintake,anthropometricmeasurement and biochemical parameters in nutritional assessment of haemodialysis patients?PakJMedSci2009;25(suppl.2):201–6.

[22]Gracia-IguacelC,Gonza´lez-ParraE,Pe´rez-Go´mezMV,Mahı´lloI,EgidoJ,Ortiz A,etal.Prevalenceofprotein-energywastingsyndromeanditsassociation withmortalityinhaemodialysispatientsinacentreinSpain.Nefrologia 2013;33:495–505.

[23]MazairacAH,DeWitGA,GrootemanMP,PenneEL,vanderWeerdNC,vanden DorpelMA, et al.Acomposite score ofprotein-energy nutritional status

predictsmortalityinhaemodialysispatientsnobetterthanitsindividual components.NephrolDialTransplant2011;26:1962–7.

[24]KaysenGA,DonBR.Factorsthataffectalbuminconcentrationindialysis patientsandtheirrelationshiptovasculardisease.KidneyInt2003;84:94–7.

[25]FaintuchJ,MoraisAA,SilvaMA,VidigalEJ,CostaRA,LyrioDC,etal.Nutritional proﬁle and inﬂammatory status of haemodialysis patients. Ren Fail 2006;28:295–301.

[26]JadejaY,KherV.Protein-energywastinginchronickidneydisease:anupdate withfocusonnutritionalinterventionstoimproveoutcomes.IndianJEndo- crinolMetab2012;16(Suppl.2):246–51.

[27]SreedharaR,AvramMM,BlancoM,BatishR,AvramMM,MittmanN.Pre- albuminis thebestnutritionalpredictorofsurvivalinhemodialysis and peritonealdialysis.AmJKidneyDis1996;28(Suppl.6):937–42.

[28]PicciniS,FairburnA,GillE,BudgeonCA,O’SullivanT.Predictorsofmalnutri- tioninAustralianhaemodialysispatientsandcomparisonofdietaryprotein intakestonationalguidelines.RSAJ2014;10(suppl.3):133–40.

[29]CanoN.Hemodialysis,inﬂammationandmalnutrition.Neurologia2001;21:5.

[30]DukkipatiR,KoppleJD.Causesandpreventionofprotein-energywastingin chronickidneyfailure.SeminNephrol2009;29(Suppl.1):39–49.

[31]LahrachH,EssiarabF,TiminouniM,HatimB,ElKhayatS,Er-RachdiL,etal.

AssociationofapolipoproteinEgenepolymorphismwithend-stagerenal diseaseandhyperlipidemiainpatientsonlong-termhemodialysis.RenFail 2014;36(suppl.10):1504–9.

[32]LahrachH,GhalimN,LebraziH,RamdaniB,Saı¨leR.Inﬂammation,cardiovascular risk and mortality among long-term haemodialysis patients. EJSR 2012;81:168–78.

[33]BrossR,NooriN,KovesdyCP,MuraliSB,BennerD,BlockG,etal.Dietary assessmentofindividualswithchronickidneydisease.SeminDial2010;23:

359–64.

[34]KoppleJD.McCollumAwardLecture,1996:protein-energymalnutritionin maintenancedialysispatients.AmJClinNutr1997;65:1544–57.

[35]FouqueD,VennegoorM,terWeeP,WannerC,BasciA,CanaudB,etal.EBPG guidelinesonnutrition.NephrolDialTransplant2007;22(Suppl.2):S45–87.

[36]Erdog˘anE,TutalE,UyarME,BalZ,DemirciBG,SayınB,etal.Reliabilityof bioelectricalimpedanceanalysisintheevaluationofthenutritionalstatusof hemodialysis patients–acomparisonwith MiniNutritionalAssessment.

TransplantProc2013;45(Suppl.10):3485–8.

[37]WabelP,ChamneyP,MoisslU,JirkaT.Importanceofwholebodybioimpe- dance spectroscopy for the management of ﬂuid balance. Blood Purif 2009;27:75–80.

[38]LeaveySF,McCulloughK,HeckingE,GoodkinD,PortFK,YoungEW.Body massindexandmortalityin‘‘healthier’’ascomparedwith‘‘sicker’’hemodialysispatients:resultsfromtheDialysisOutcomesandPracticePatternsStudy (DOPPS).NephrolDialTransplant2001;16:2386–94.

[39]ChenJ,PengH,YuanZ,ZhangK,XiaoL,HuangJ,etal.Combinationwith anthropometricmeasurementsandMQSGAtoassessnutritionalstatusin Chinesehemodialysispopulation.IntJMedSci2013;10(Suppl.8):974–80.