ORIGINAL RESEARCH

Dietary Intake of Patients With ChronicKidney Disease Entering the LORD Trial:Adjusting for UnderreportingRobert G. Fassett, MBBS,*† Iain K. Robertson, MPH,‡ Dominic P. Geraghty, PhD,‡Madeleine J. Ball, MD,‡ and Jeff S. Coombes, PhD*

Objective: The study objective was to determine the dietary intake of patients with chronic kidney disease beforeand after filtering for suspected underreporters and to investigate the impact of underreporting on the interpretationof diet data.

Design: This was a cross-sectional study.Setting: The study included outpatients from hospitals and clinics in Northern Tasmania, Australia.Patients: Data from 113 patients enrolled in the Lipid Lowering and Onset of Renal Disease trial were used in this

study. Patients with serum creatinine greater than 120 mmol/L were included, and those taking lipid-loweringmedication were excluded.

Methods: Patients completed a 4-day self-report diet diary, and FoodWorks software was used to determinetheir daily intake of energy, macronutrients, and specific micronutrients. Diet diaries were assessed for likelyunderreporting using the Goldberg cutoff approach with a ratio of energy intake to estimated resting energyexpenditure of 1.27. Nutrient intakes were compared with current National Kidney Foundation’s Kidney DiseaseOutcomes Quality Initiative guidelines, World Health Organization recommendations, recommended daily allow-ances, and daily values adjusted for energy intake.

Results: Demographics of the patients were as follows: male/female, 71/42; age (mean ! standard deviation),60 ! 15 years; body mass index, 28.6 ! 6.0 kg/m2, and serum creatinine, 223.4 ! 110.0 mmol/L. According to thecriteria, 80 patients (70.8%) were underreporting their energy intake. Underreporters were more likely to be femaleand younger, and have a higher body mass index and elevated serum creatinine. In all patients, daily energy intake(89.6 ! 32.4 kJ/kg) was lower than recommended (125-145 kJ/kg); however, this was not the case for validreporters (128.3 ! 23.7 kJ/kg). Protein intake was higher (0.9 ! 0.3 g/kg) than recommended (0.75 g/kg) in allpatients and even higher (1.2 ! 0.3 g/kg) in valid reporters. Mean calcium, zinc, and dietary fiber intakes were allbelow recommendations in all patients, but these differences were not apparent in valid reporters.

Conclusion: Interpreting self-report diet diary data from patients with chronic kidney disease without attemptingto exclude underreporters will lead to erroneous conclusions, especially in respect to energy, protein, dietary fiber,calcium, and zinc intakes.© 2007 by the National Kidney Foundation, Inc.

*School of Human Movement Studies, University of Queensland, Brisbane, Australia.†Renal Research Tasmania, Launceston General Hospital, Launceston, Tasmania, Australia.‡School of Human Life Sciences, University of Tasmania, Launceston, Tasmania, Australia.This project was supported by a grant from the Clifford Craig Medical Research Trust, C. Prosser Green Trust. Pfizer Pharmaceuticals

provided atorvastatin and placebo for the LORD trial.Address reprint requests to Jeff S. Coombes, PhD, School of Human Movement Studies, Rm 535 Connell Building, University of

Queensland, St. Lucia, Queensland, Australia 4072. E-mail: jcoombes@hms.uq.edu.au.© 2007 by the National Kidney Foundation, Inc.1051-2276/07/1704-0002$32.00/0doi:10.1053/j.jrn.2007.04.004

Journal of Renal Nutrition, Vol 17, No 4 ( July), 2007: pp 235–242 235

NUTRITION IS KNOWN to play a majorrole in the preservation of kidney func-

tion, the reduction of cardiac comorbidities, andthe overall well-being of patients with chronickidney disease (CKD), with deterioration in nu-tritional status known to precede end-stage kid-ney disease.1 Although it is recommended thatregular monitoring of the patient’s nutritionalstatus should be a routine component of predi-alysis care,1 measuring dietary intake is difficult. Aself-report multiday diet diary is acknowledged asone of the best approaches to quantify energy andnutrient intake. However, it is also well recog-nized that there are considerable sources of errorthat may effect interpretation of the data.2

The validity of the data from self-report dietdiaries can be assessed by determining the ratio ofenergy intake (EI) with resting energy expendi-ture (REE), and a joint report organized by theWorld Health Organization (WHO) determinedthat a cutoff limit of 1.27 is likely to identifyindividuals who underreport on diet diaries.3 Thefrequency of underreporting has been assessed innumerous studies in many subject groups, includ-ing children,4 adolescents,5 athletes,6 obese sub-jects,7 and the general population.8 Estimates varyfrom as low as 2%9 to 85%,10 with most studiesreporting the incidence to be approximately20%.11 In patients with CKD, a recent studyusing the 1.27 EI:REE cutoff reported a muchgreater prevalence of underreporting (72.5%),12

and this was similar to a previous study of patientson hemodialysis (61%).13

In the general population, underreporting hasled to erroneous conclusions regarding nutrientintake.14,15 The aims of this study were to con-firm the extent of underreporting in patients withCKD and to assess the impact of underreportingon dietary conclusions.

MethodsStudy Population

Data from 113 patients with CKD enrolled inthe Lipid lowering and Onset of Renal Disease(LORD) trial were included in this analysis. Thetrial is a randomized placebo-controlled studyinvestigating the effects of 10 mg of atorvastatinon the progression of kidney disease in patientswith CKD.16 The inclusion criterion for the trialis a serum creatinine greater than 120 !mol/L.

Exclusion criteria were patients already takinglipid-lowering medication or women with child-birth potential. Patient characteristics are pro-vided in Table 1. All patients were white, and theprimary causes of kidney disease, according to theclassification system used by the Australian andNew Zealand Dialysis Transplant Registry, werehypertension (n " 37), glomerulonephritis (n "31), reflux nephropathy (n " 10), polycystickidney disease (n " 9), diabetic nephropathy (n" 8), analgesic nephropathy (n " 1), and mis-cellaneous (n " 17). The study was approved bythe Ethics Committee of the Launceston GeneralHospital in accordance with the guidelines of theDeclaration of Helsinki. Patient recruitment oc-curred from January 2002 to January 2005.

After informed consent was obtained, patientswere provided with a 4-day diet diary as part ofbaseline measures for the LORD trial. Patientsthen had a 3-month run-in phase before theystarted taking either the hydroxymethylglutaryl-coenzyme A reductase inhibitor, atorvastatin (10mg), or placebo. At the end of the run-in phase,patients provided fasting blood and 24-hour urinesamples. Less than 5 days after these samples wereobtained, patients attended a routine clinic visitwhere height and weight were measured and thecompleted diet diaries were returned. Patientswere then provided with the tablets to begin theintervention.

All clinical chemistry measures were com-pleted in public and private pathology laborato-ries accredited with the National Association ofTesting Authorities (Australia). Full blood countswere performed on a Coulter MaxM (Coulter,Luton, UK). Total cholesterol, high-density li-poprotein cholesterol, triglycerides, urinarychemistry, electrolytes, and liver enzymes weremeasured on an Olympus AU-600 automatedanalyzer (Olympus Mishima, Shizuoka, Japan).Low-density lipoprotein cholesterol was calcu-lated using the Freidwald formula.17 Estimatedglomerular filtration rate was determined usingthe Cockcroft and Gault equation,18 and creati-nine clearance was calculated from the 24-hoururine collection.

Diet DiariesThe 4-day diet diary contained photographs of

common foods (e.g., steak, pasta, rice, potatoes) onplates with their corresponding mass. For each food,

FASSETT ET AL236

three different portions were displayed (large, me-dium, and small). This diary had been previouslyvalidated against 7-day weighed records.19 Whenpatients were provided with the diaries, they weregiven detailed instructions by one of two trainednurses on what was expected, and specific exampleswere given. They were asked to complete the diaryentry at the time of food consumption, with theinstruction that the 4 days should be as close to theirnext pathology visit as possible. This was toward theend of the run-in phase, approximately 3 monthsafter they were given the diary. They were asked toinclude 3 weekdays and 1 weekend day. The diarieswere analyzed using version 4.0 of FoodWorks(Xyris, Brisbane, Australia). Dietary supplementswere also entered into the analysis with the specificsupplement details obtained at either the baseline orend of run-in visit. Energy and protein intakes werebased on actual body weight. It should be noted thatthis database is not complete for all nutrients in allfoods, especially for vitamin and minerals.

Assessing Valid ReportersThe ratio between EI and estimated REE was

used to assess the accuracy of each patient’s dietdiary. A cutoff value of 1.27 was selected on thebasis of a joint report from the Food and Agri-culture Organization, WHO, and the UnitedNations University.3 Estimated energy expendi-

ture was determined using the Schofield equa-tions20 with an adjustment for physical activitylevels.21 Physical activity was assessed using theActive Australia Questionnaire.22

Data AnalysisData are displayed as the mean ! standard

deviation (or 95% confidence intervals) and com-pared with current National Kidney Foundation’sKidney Disease Outcomes Quality Initiative (K/DOQI) guidelines,1 WHO recommendations,23

recommended daily allowances (RDA), and dailyvalues (DV)23 adjusted for EI. For the determi-nation of normal, overweight, and obese, therecommendations from the WHO were used,with a healthy body mass index (BMI) at 18.5 to24.9, a BMI greater than 25 indicating over-weight, and a BMI greater than 30 indicatingobesity.24 T tests were used to compare validreporters and underreporters. Significance wasassumed when P less than .05, and all statisticalanalyses were performed using Version 8.2 ofSTATA (Statacorp, College Station, TX).

ResultsThe mean ! standard deviation EI/resting

metabolic rate [RMR] of all patients was 1.08 !0.34. After using the Goldberg cutoff approach

Table 1. Patient Characteristics (Mean ! Standard Deviation)

VariableAll

N " 113Valid Reporters

n " 33Underreporters

n " 80

Sex (female) n " 42, 37% n " 8, 24% n " 34, 43%†Age (y) 60 ! 15 65 ! 10 59 ! 16*Body mass (kg) 81.8 ! 17.3 74.5 ! 12.3 84.8 ! 18.2†Body mass index (kg/m2) 28.6 ! 6.0 25.4 ! 3.4 30.0 ! 6.4†Systolic blood pressure (mm Hg) 143.4 ! 20.2 145.5 ! 24.3 142.5 ! 18.3Diastolic blood pressure (mm Hg) 81.0 ! 12.9 82.1 ! 15.2 80.6 ! 11.9Hemoglobin (g/L) 130.6 ! 17.0 134.8 ! 18.7 128.9 ! 16.1Hematocrit (L/L) 0.39 ! 0.05 0.41 ! 0.1 0.39 ! 0.05Estimated glomerular filtration rate (mL/min/1.73 m2) 40.3 ! 19.4 38.9 ! 14.9 40.8 ! 21.0Urinary protein excretion (g/d) 1.28 ! 1.94 0.74 ! 0.87 1.50 ! 2.20Creatinine clearance (mL/sec) 0.68 ! 0.32 0.77 ! 0.27 0.64 ! 0.30*Serum creatinine (!mol/L) 223.4 ! 110.0 187.9 ! 60.4 238.0 ! 122.0*Total cholesterol (mmol/L) 5.6 ! 1.4 5.8 ! 1.4 5.6 ! 1.4HDL cholesterol (mmol/L) 1.2 ! 0.4 1.3 ! 0.4 1.2 ! 0.3*LDL cholesterol (mmol/L) 3.5 ! 1.1 3.6 ! 1.1 3.4 ! 1.1Total cholesterol/HDL cholesterol ratio 4.9 ! 1.6 4.6 ! 1.7 5.0 ! 1.5Triglycerides (mmol/L) 2.2 ! 1.4 1.9 ! 0.9 2.4 ! 1.6Serum albumin (g/L) 40.2 ! 4.5 41.5 ! 3.8 39.6 ! 4.7

HDL, high-density lipoprotein; LDL, low-density lipoprotein.*Significantly different from valid reporters (P # .05).†Significantly different from valid reporters (P # .01).

DIETARY INTAKE OF PATIENTS WITH CKD 237

and an EI/RMR of 1.27, it was calculated that 80patients (70.8%) were underreporting their usualEIs. Indeed, 53 of all the patients reported theirEI to be below their estimated RMR. The esti-mated EI/RMR of the underreporters was 0.90! 0.20. This effectively meant that they reportedconsuming 30% less than their energy needs.

To allow for the possibility that overweightand obese patients may have been intentionallyeating less energy to decrease body weight, addi-tional analyses were carried out using an RMRderived from healthy body weight calculated us-ing a BMI of 25. With this approach, only 9 ofthe 80 underreporters would have been removedfrom this list, further supporting the notion thatthe diet diaries were not being accurately com-pleted.

Characteristics of all patients, valid and under-reporters, are shown in Table 1. There weresignificantly (P # .05) more female (43% vs. 24%)and younger (59 vs. 65 years) underreporters. Inaddition, underreporters had a greater body mass(85 vs. 75 kg), BMI (30 vs. 25 kg/m2), and serumcreatinine (238 vs. 187 !mol/L). There was atrend (P " .058) for higher urinary protein ex-cretion in underreporters (1.5 vs. 0.7 g/day). Inregard to BMI, 12% of the valid reporters wereobese, 30% were overweight, 55% were of nor-mal weight, and 3% were underweight, and in

those underreporting, 39% were obese, 35% wereoverweight, and 26% were of normal weight.

Figure 1 summarizes the findings from thedietary analyses of all patients (data also in Table2). From these data it would be concluded thatthese patients with CKD had energy, total car-bohydrate (%), complex carbohydrate, dietary fi-ber, calcium, and zinc intakes significantly lower(P # .001) than recommendations. In addition,the percentage of energy from fat and totalprotein intake were significantly higher than rec-ommended by the WHO and K/DOQI, respec-tively. However, after excluding the underre-porters, the only nutrient that was still below arecommended level was grams of complex car-bohydrate. Dietary protein and percent fat werestill above recommended levels.

DiscussionThe major findings of this study are that a large

percentage (70.8%) of the 113 patients with CKDunderreported their EI using a 4-day self-reportdiet diary and that conclusions based on theseanalyses are likely to be inaccurate.

Our findings are in agreement with Avesani etal.,12 who used the same cutoff of 1.27 EI/REEand found that 72.5% of patients with CKDunderreported their EI. The incidence of under-

Figure 1. Nutrient intakes as a percentage of recommendations (mean 95% confidence interval). *Complexcarbohydrates, fat, saturated fat, and cholesterol are DVs adjusted for EI using the K/DOQI recommendation(125 kJ/d). Protein, fat, saturated fat, and cholesterol intakes are all recommended to be below the indicated100%, thus the different shading. Box plots show 10th, 25th, 50th, 75th, and 90th percentiles. CHO,carbohydrate; DV, daily value; EI, energy intake; K/DOQI, National Kidney Foundation Kidney DiseaseOutcomes Quality Initiative.

FASSETT ET AL238

reporting seems to be much higher in patientswith CKD than in the general population. Areview by Hill and Davies11 provides a compre-hensive list of studies investigating the validity ofself-reported EI, with the majority of studiesfinding that the incidence of underreporting isapproximately 20%.

The percentage of overweight or obese pa-tients underreporting was 74% compared with42% of the valid reporters. The associationbetween underreporting and increased body

mass/BMI is consistent with the literature.12,25

It may be that the patients were restricting theirEI to lose weight. However, when an idealweight (based on a BMI of 25) was used, andthe EI/RMR ratio was recalculated, only 9 ofthe 80 underreporters would have been re-moved from the list of underreporters. Further-more, the EI/RMR ratio of the 80 underre-porters was only 0.90. This indicates that theunderreporters are either considerably reducingtheir EI or inaccurately reporting their food

Table 2. Dietary Intakes of Patients With Chronic Kidney Disease (Mean ! SD)

Daily Intakes Recommendation

All Patients Valid Reporters

Daily Intake(Mean ! SD)

No. of Patients (of 113)Below/exceeding*

Recommendation (%)Daily Intake

(Mean ! SD)

No. of Patients (of 33)Below/exceeding*

Recommendation (%)

Energy (kJ/kg) 125–145 kJ/kg† (K/DOQI) 89.6 ! 32.4 95 (84%) 128.3 ! 23.7 16 (48%)Protein (%)* 10%–15% total energy

intake (WHO)18.2 ! 3.5 86 (76%) 16.3 ! 3.3 17 (57%)

Protein (g/kg)* 0.75 g/kg† (K/DOQI) 0.9 ! 0.3 81 (72%) 1.2 ! 0.3 33 (100%)Carbohydrate (%) 50%–75% (WHO) 46.4 ! 7.0 76 (67%) 48.2 ! 6.4 19 (56%)Complex

carbohydrate (g)Adjusted DV‡367 g for All334 for VR

111.3 ! 44.9 113 (100%) 153.3 ! 40.4 33 (100%)

Fat (%)* 15%–30% (WHO) 31.6 ! 6.4 69 (61%) 32.5 ! 5.3 24 (73%)Fat (g)* Adjusted DV§

79 g for All72 g for VR

58.9 ! 21.7 26 (23%) 80.9 ! 19.1 24 (73%)

Saturated fat (g)* Adjusted DV!24 g for All22 g for VR

24.2 ! 10.2 50 (44%) 33.1 ! 10.0 28 (85%)

Cholesterol (mg)* Adjusted DV¶367 mg for All334 mg for VR

209.4 ! 96.4 11 (9%) 254.5 ! 105.5 6 (18%)

Vitamin C (mg) 30–40 (RDA) 118.1 ! 88.8 0 (0%) 155.1 ! 90.0 0 (0%)Vitamin A eq. (!g) 750 (RDA) 910.7 ! 623.4 52 (46%) 1269.1 ! 872.4 5 (15%)Total folate (!g) 200 (RDA) 280.9 ! 108.4 25 (22%) 366.8 ! 119.8 0 (0%)Thiamin (mg) 0.7–1.1 (RDA) 1.6 ! 0.7 14 (12%) 2.1 ! 0.7 0 (0%)Riboflavin (mg) 1.0–1.7 (RDA) 1.8 ! 0.9 27 (24%) 2.3 ! 0.9 0 (0%)Niacin (mg) 11–19 (RDA) 32.5 ! 10.1 1 (0%) 23.0 ! 6.8 0 (0%)Sodium (mg) Range 920–2300 (RDA) 2326.7 ! 916.2 $49 (43%)

#4 (4%)3112.5 ! 951.4 $26 (79%)

#0 (0%)Calcium (mg) 800–1000 (RDA) 703.9 ! 319.4 $16 (14%)

#79 (70%)940.9 ! 335.4 $12 (36%)

#14 (42%)Magnesium (mg) 270–320 (RDA) 263.7 ! 95.4 $27 (24%)

#69 (61%)349.4 ! 104.7 $18 (55%)

#7 (21%)Potassium (mg) Range 1950–5460 (RDA) 2725.4 ! 1005.0 $2 (2%)

#23 (20%)3508.9 ! 1186.5 $2 (6%)

#0 (0%)Iron (mg) 7–12 (RDA) 10.7 ! 4.1 $32 (28%)

#18 (16%)14.1 ! 4.8 $21 (64%)

#0 (0%)Zinc (mg) 12 (RDA) 9.2 ! 2.6 94 (83%) 10.9 ! 2.7 23 (68%)Phosphorus (mg) 1000 (RDA) 1179.3 ! 403.9 38 (34%) 1545.9 ! 386.2 1 (3%)Dietary fiber (g) Range 27–40 (WHO) 19.8 ! 8.9 $5 (4%)

#94 (83%)28.0 ! 10.2 $5 (15%)

#17 (52%)

SD, standard deviation; K/DOQI, National Kidney Foundation Kidney Disease Outcomes Quality Initiative; RDA, recommended daily allowance; WHO, WorldHealth Organization; DV, daily value; VR, valid reporter.

*Protein, fat, saturated fat, and cholesterol intakes are all recommended to be at or below recommendations.†Recommended intakes for energy and protein intake are based on actual body weight.‡DV for complex carbohydrate intake " 300 g/d for a 8368 kJ/d intake. This was adjusted for each individual using the recommended K/DOQI minimum energy

intake (125 kJ/d).§DV for fat intake is less than 65 g/d for a 8368 kJ/d intake. This was adjusted for each individual using the recommended K/DOQI minimum energy intake

(125 kJ/d).!DV for saturated fat intake is less than 20 g/d for a 8368 kJ/d intake. This was adjusted for each individual using the recommended K/DOQI minimum energy

intake (125 kJ/d).¶DV for cholesterol intake is less than 300 mg/d for a 8368 kJ/d intake. This was adjusted for each individual using the recommended K/DOQI minimum energy

intake (125 kJ/d).

DIETARY INTAKE OF PATIENTS WITH CKD 239

intake. On the basis of evidence from numer-ous other studies and longitudinal data frompatients with CKD showing that reporting lowEIs is not associated with long-term weightloss,12 it is postulated that the patients areconsuming more energy than they are report-ing. Potential reasons for underreporting arenumerous and a recent review article by Mau-rer et al.25 discusses the complex psychosocialand behavioral characteristics that may explainunderreporting.

Other characteristics of the underreporters inthis study were as follows: female, younger age,higher serum creatinine (lower creatinine clear-ance), and a trend (P " .058) for elevated urinaryprotein excretion. Attitudes to food provide themost likely explanation for diet-recording behav-ior26 and the finding that female patients withCKD are more likely to underreport is consistentwith the literature from the general popula-tion.11,25,26 It is believed to be due to increasedbody and diet consciousness and an associateddesire to convey a positive image when complet-ing diet diaries and thereby misrepresenting theirintake.

The finding that younger patients were morelikely to be underreporting (mean age differencebetween valid and underreporters was 6 years) isunusual compared with previous studies.11 A pos-sible explanation is that younger patients might nowbe much busier than older individuals because ofdiffering work commitments and therefore mayhave been more likely to forget or not devoteenough time to accurately completing the diary.

Underreporters had elevated serum creatinineand lower creatinine clearance and a trend (P ".06) for elevated urinary protein excretion. Thissuggests that patients with less kidney function aremore likely to underreport, which may also berelated to a desire to relay a more positive imageof their food consumption. To our knowledge,there have been no other reports on the relation-ships between the extent of a disease and thedegree of underreporting.

The other major finding of this study was thatconclusions based on the analysis of all patientswould be false if there is no attempt to determinethe accuracy of the reporting. Other studies havealso found that dietary conclusions can be signif-icantly altered by the inclusion or exclusion ofunderreporters.14,15 The main discrepancy in ourstudy was EI. After the first analysis with all 113

patients, it would have been concluded that thisgroup of patients are energy deficient to the pointthat they are only obtaining 70% of their energyneeds (using the lowest value from the recom-mended range). Because, on average, the groupwas overweight, a decrease in EI may be benefi-cial. However, it is likely that many of thesepatients are not accurately reporting their foodintake. Therefore, an initial conclusion of a po-tentially beneficial low EI is incorrect.

There is much debate regarding the recom-mended protein intake for patients with CKD.27

If low-protein diets are beneficial to patients withCKD, then the findings of this study indicate thata greater percentage of patients are consumingmore than that recommended by K/DOQI.1

Indeed, when all patients were analyzed, 72% areabove the protein recommendation of 0.75 g/kgbody mass. However, when underreporters areexcluded, all 33 patients (100%) are above thislevel. Qualitative analysis of the diet diariesshowed that a large number of individuals con-sumed red meat on a regular basis, suggesting thata good strategy to decrease dietary protein intakewould be to advise a limitation of red meatconsumption. However, additional care must betaken to avoid malnutrition with low proteinintakes (#0.6 g/kg body mass).

Mean total carbohydrate (%), complex carbo-hydrate, dietary fiber, calcium, and zinc intakewere lower than recommended in the entirestudy group. However, when the underreporterswere excluded, the only nutrient that was stillbelow a recommended level was grams of com-plex carbohydrate. Because the percentage totalcarbohydrate intakes were only slightly below theWHO lower-limit recommendation of 50%, thisindicated a high intake of simple sugars. All di-etary guidelines recommend low intakes of sim-ple sugars, and dietary advice, such as readingfood labels and avoiding foods with sugar, honey,sucrose, dextrose, or corn syrup listed as the firstor second ingredient, would be beneficial to thesepatients.

A number of factors may limit the externalvalidity of our findings. The patients with CKDincluded in this analysis are from a trial investi-gating the effects of atorvastatin on kidney func-tion. These individuals were not taking lipid-lowering medication at the time this study wascarried out, which may not be representative ofall patients with CKD. The use of a 4-day diet

FASSETT ET AL240

diary to collect food information is recognized ashaving a number of limitations, such as difficultiesin estimating food quantities, compliance withregularly recording foods, and the use of unfamil-iar terms/foods that makes entry into a softwareprogram difficult.28 A number of strategies wereused to avoid these potential problems, includingproviding a diary with portion-size pictures thatwas previously validated against 7-day weighedrecords in a non-CKD population19 and askingthem to enter foods at the time of consumptionand providing them with detailed instructions bya trained individual.

Our estimation of energy expenditure was ob-tained using the Schofield formulae, and althoughthere are studies to report the validity of theseequations in healthy populations, they have notbeen validated in patients with CKD. However,our incidence of underreporting was strikinglysimilar to that found by Avesani et al.12 (71% vs.72.5%), who determined energy expenditure us-ing indirect calorimetry. Finally, the calculationof normalized protein nitrogen appearance wouldhave been a useful addition to investigating theprotein status of the patients. Unfortunately, inpatients with CKD this requires the additionalmeasure of urinary urea,29 which is not a routinemeasure in these patients.

ConclusionsThese data show that the population sample

with CKD studied have a high incidence (71%)of underreporting their dietary intake. Withoutaccounting for underreporting, dietary analysiswould lead to erroneous conclusions, especiallyin respect to energy, protein, dietary fiber, cal-cium, and zinc intakes. It is, however, of notethat there are some real dietary issues, and adviceon intake of protein and fat is important.

AcknowledgmentsThe authors thank Lisa Anderson, Marianne Smith, and

Krystal Chugg for their assistance in collecting data.

References1. Combe C, McCullough KP, Asano Y, et al: Kidney

Disease Outcomes Quality Initiative (K/DOQI) and the DialysisOutcomes and Practice Patterns Study (DOPPS): nutritionguidelines, indicators, and practices. Am J Kidney Dis 44:39-46,2004.

2. Black AE, Prentice AM, Goldberg GR, et al: Measure-ments of total energy expenditure provide insights into thevalidity of dietary measurements of energy intake. J Am DietAssoc 93:572-579, 1993.

3. Energy and Protein Requirements Technical Report SeriesNo. 724, Report of a joint Food and Agriculture Organization/World Health Organization/United Nations University meet-ing. World Health Organization, Geneva, 1985.

4. Davies PS, Coward WA, Gregory J, et al: Total energyexpenditure and energy intake in the pre-school child: a com-parison. Br J Nutr 72:13-20, 1994.

5. Bratteby LE, Sandhagen B, Fan H, et al: Total energyexpenditure and physical activity as assessed by the doubly labeledwater method in Swedish adolescents in whom energy intake wasunderestimated by 7-d diet records. Am J Clin Nutr 67:905-911,1998.

6. Edwards JE, Lindeman AK, Mikesky AE, et al: Energybalance in highly trained female endurance runners. Med SciSports Exerc 25:1398-1404, 1993.

7. Goris AH, Westerterp-Plantenga MS, Westerterp KR:Undereating and underrecording of habitual food intake in obesemen: selective underreporting of fat intake. Am J Clin Nutr71:130-134, 2000.

8. Livingstone MB, Prentice AM, Strain JJ, et al: Accuracy ofweighed dietary records in studies of diet and health. BMJ300:708-712, 1990.

9. Livingstone MB, Prentice AM, Coward WA, et al: Vali-dation of estimates of energy intake by weighed dietary recordand diet history in children and adolescents. Am J Clin Nutr56:29-35, 1992.

10. Novotny JA, Rumpler WV, Riddick H, et al: Personalitycharacteristics as predictors of underreporting of energy intake on24-hour dietary recall interviews. J Am Diet Assoc 103:1146-1151, 2003.

11. Hill RJ, Davies PS: The validity of self-reported energyintake as determined using the doubly labelled water technique.Br J Nutr 85:415-430, 2001.

12. Avesani CM, Kamimura MA, Draibe SA, et al: Is energyintake underestimated in nondialyzed chronic kidney diseasepatients? J Ren Nutr 15:159-165, 2005.

13. Kloppenburg WD, de Jong PE, Huisman RM: Thecontradiction of stable body mass despite low reported dietaryenergy intake in chronic haemodialysis patients. Nephrol DialTransplant 17:1628-1633, 2002.

14. Stallone DD, Brunner EJ, Bingham SA, et al: Dietaryassessment in Whitehall II: the influence of reporting bias onapparent socioeconomic variation in nutrient intakes. Eur J ClinNutr 51:815-825, 1997.

15. Macdiarmid JI, Vail A, Cade JE, et al: The sugar-fatrelationship revisited: differences in consumption between menand women of varying BMI. Int J Obes Relat Metab Disord22:1053-1061, 1998.

16. Fassett RG, Ball MJ, Robertson IK, et al: Baseline serumlipids and renal function in chronic kidney disease patientsentering the LORD trial. Int J Clin Pharmacol Ther 44:580-588,2006.

17. Friedewald WT, Levy RI, Fredrickson DS: Estimation ofthe concentration of low-density lipoprotein cholesterol inplasma, without use of the preparative ultracentrifuge. ClinChem 18:499-502, 1972.

DIETARY INTAKE OF PATIENTS WITH CKD 241

18. Cockcroft DW, Gault MH: Prediction of creatinine clear-ance from serum creatinine. Nephron 16:31-41, 1976.

19. Eddington J, Thorogood M, Geekie M, et al: Assessmentof nutritional intake using dietary records with estimatedweights. J Hum Nutr Diet 2:407-414, 1989.

20. Schofield WN: Predicting basal metabolic rate, new stan-dards and review of previous work. Hum Nutr Clin Nutr39(Suppl 1):5-41, 1985.

21. Black AE: Critical evaluation of energy intake using theGoldberg cut-off for energy intake: basal metabolic rate. Apractical guide to its calculation, use and limitations. Int J ObesRelat Metab Disord 24:1119-1130, 2000.

22. Brown WJ, Trost SG, Bauman A, et al: Test-retest reli-ability of four physical activity measures used in populationsurveys. J Sci Med Sport 7:205-215, 2004.

23. Sizer FS, Whitney EN: Nutrition: concepts and contro-versies. Thompson, Brooks Cole, Belmont, CA, 2005.

24. What are overweight and obesity? World Health Orga-nization Fact Sheet No. 311: Available at: http://www.who.int/mediacentre/factsheets/fs311/en/. Accessed January 29, 2007.

25. Maurer J, Taren DL, Teixeira PJ, et al: The psychosocialand behavioral characteristics related to energy misreporting.Nutr Rev 64:53-66, 2006.

26. Price GM, Paul AA, Cole TJ, et al: Characteristics of thelow-energy reporters in a longitudinal national dietary survey.Br J Nutr 77:833-851, 1997.

27. Johnson DW: Dietary protein restriction as a treatmentfor slowing chronic kidney disease progression: the case against.Nephrology (Carlton) 11:58-62, 2006.

28. Blumenkrantz MJ, Kopple JD, Gutman RA, et al: Meth-ods for assessing nutritional status of patients with renal failure.Am J Clin Nutr 33:1567-1585, 1980.

29. Abbas K: Using normalized protein nitrogen appearance(nPNA) in assessing nutrition. Nephrol Nurs J 33:677-678, 2006.

FASSETT ET AL242