resting energy expenditure in subjects with fibro-calculous pancreatic diabetes...

ORIGINAL ARTICLE

Resting energy expenditure in subjects withfibro-calculous pancreatic diabetesKishore Kumar BEHERA, 1 Mini JOSEPH,4 Sudeep Krishna SHETTY,1 Ashok CHACKO,2

Manoj Kumar SAHOO,2 Narayani V MAHENDRI,3 Veena NAIR,1 Shweta NADIG1 andNihal THOMAS1

Departments of 1Endocrinology, Diabetes & Metabolism, 2Gastrointestinal Sciences and 3Dietetics, Christian Medical College, Vellore and4Department of Home Science, Government College for Women, Trivandrum, India

Correspondence

Nihal Thomas, Department ofEndocrinology, Diabetes and MetabolismChristian Medical College, Vellore-632004,India.Tel: +91 416 2282528Fax: +91 416 4205844Email: [email protected]

Received 30 November 2012; revised 7May 2013; accepted 2 June 2013.

doi: 10.1111/1753-0407.12070

Abstract

Background: Fibro-calculous pancreatic diabetes is an indigenous disorderpresent in populations largely in tropical regions. Energy expenditure throughindirect calorimetry has not been studied in this disorder and may provideimportant clues as to the pathogenesis of diabetes in these patients.Methods: A total of 51 males in three groups comprising fibrocalculouspancreatic diabetes (FCPD) (group 1; n = 24), type 2 diabetes (group 2;n = 15) and healthy controls (group 3; n = 12) were studied. The body com-position was measured using Dual Energy X-ray Absorptiometry (DEXA)and the REE was estimated using indirect calorimetry. The predicted energyexpenditure (PEE) was calculated using three different equations.Results: Patients in both groups with diabetes had a higher mean waist-hipratio than the controls (P = 0.002). However patients with type 2 diabetes alonehad a significantly higher mean body mass index (P = 0.012), percentage of fat(P = 0.016) and total fat content (P = 0.031). There was no significant differ-ence in REE among the three groups. After adjustment of body mass index(BMI), the REE was significantly higher in patients with FCPD than in thosepatients with Type 2 diabetes. PEE correlated poorly with indirect calorimetry.Conclusions: Energy expenditure in patients with diabetes varies accordingto the composition and distribution of body fat and is lower in patients withFCPD. Standard predictive equations were not accurate for the assessment ofenergy expenditure in patients with FCPD. Further research is required torecommend specific nutritional therapy for this group of patients.

Keywords: energy expenditure, fibrocalculous pancreatic diabetes, indirectcalorimetry.

Introduction

Fibrocalculous pancreatic diabetes constitutes approxi-mately 1% of all subjects with diabetes and about 4% ofall patients with diabetes in the young.1 Malnutrition as

a consequence of chronic pancreatitis affects a greaternumber of patients as the duration of the disease and itsseverity progresses.2 The pathogenesis of malnutrition ismultifactorial. The most frequently observed factors area reduced energy intake due to pain, malabsorption of

Significant findings of the study: None of the long standing predictive equations accurately predict REE in thisgroup of patients. Nutritional management may not be uniform in the different subtypes of diabetes. Their higherenergy expenditure as seen from indirect calorimetry indicates that patients with pancreatic diabetes may requirea higher calorie intake to prevent malnutrition.What this study adds: Patients with fibrocalculous pancreatic diabetes have a higher energy expenditure thanpatients with type 2 diabetes associated with the presence of lower body fat content.

bs_bs_banner

Journal of Diabetes 6 (2014) 158–163

158 © 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

AS

IAT

RA

CK

mailto:[email protected]

nutrients due to exocrine pancreatic insufficiency andmetabolic derangements resulting from uncontrolleddiabetes. Owing to their under-nutrition and low leanbody mass, the restriction of calories is not an ideal wayto achieve glycemic control. A modified dietary regimenwith a more liberal calorie and protein intake when com-pared to patients with type 2 diabetes is required toovercome under-nutrition. However, the energy require-ments in these subjects have to be accurately assessedprior to modifying their calorie intake.

The energy expenditure of a person over an entire dayis divided into three different components:1. Basal metabolic rate (BMR), as defined as theminimum amount of energy required by the body whenin a state of physical and mental rest. BMR is measuredunder standardized conditions such as being at completerest in a pre-prandial state, and in a thermo neutralenvironment. If one of the conditions mentioned above isnot met (e.g. shorter time interval for fasting) the mea-surement is usually termed resting metabolic rate (RMR)or resting energy expenditure (REE). BMR accounts onan average for about three-quarters (60–75%) of totalenergy expenditure in most individuals.2. Diet induced thermogenesis (DIT) is the amount ofenergy used during the processes which involve diges-tion, absorption and transportation of nutrients andaccounts for about 10% of the total energy intake.3. Physical activity (PA) is the most variable componentthat indicates additional energy expenditure over andabove RMR and DIT. On average, PA accounts forabout 15–30% of the total daily energy expenditure.

There is a paucity of studies that have evaluated theenergy expenditure in subjects with chronic pancreatitis.Dickerson et al.3 assessed the REE in a small group ofhospitalized patients with chronic pancreatitis withoutdiabetes. Another study by Hébuterne et al.4 assessed theREE in patients with alcohol-related chronic pancreati-tis. The REE in ambulant, non-alcoholic subjects withpancreatitis and diabetes has not been studied hitherto.

Methods

Participants

Consecutive male patients attending the Endocrinologyor Gastroenterology out-patient department at our insti-tution and aged between 20 and 60 years fulfilling theeligibility criteria for chronic non-alcoholic pancreaticdiabetes (group 1) were recruited. Similarly eligible malesubjects who fulfilled the World Health Organization(WHO) diagnostic criteria for type 2 diabetes mellitus(group 2) and healthy controls (group 3) were alsorecruited.

Biochemical parameters

Fasting and post-meal venous plasma glucose and creati-nine were measured using reagents supplied by Roche(Manheim, Germany) on the Roche Modular P 800system. C-peptide was measured by chemiluminescenceImmunoassay; usingQ4 kits supplied by Siemens, on theImmulite 2000 system (%CV-7.5) (Siemens healthcareDiagnostic products, Llanberis, Gwynedd, UK). Glyco-sylated hemoglobin (HbA1c) was measured by ionexchange chromatography (HPLC) on BioRADVARIANT II Hemoglobin Testing System (Hercules,CA, USA) (%CV-3.1). Patients with pancreatic diabeteswere maintained on a high-fat diet (50 g/day) for 5 days(starting 2 days before stool collection) and 72-h stool fatwas estimated using a cut of value of less than 7 g/day asnormal.4

DEXA

The whole body composition was evaluated by using aDual Energy X-ray absorptiometry (DXA) scan with aHologic Delphi W (S/N 70471) DXA scanner.

Indirect calorimeter

Indirect calorimetry, the gold standard for measuringresting energy expenditure was used in our subjects.5 Theestimations were made any time during the day, providedmore than 2 h had elapsed after the last meal, so as tominimize measurement errors from diet-induced thermo-genesis. The oxygen consumption and carbon dioxideproduction were measured to calculate respiratory quo-tient [(RQ) = VCO2/VO2]. RQ within the normal physi-ological range confirms a consistent calorie intake by thestudy subjects. The REE was then calculated usingthe abbreviated Weir equation: 3.9(VO2) + 1.1 (VCO2) ×1.44 [(VO2− Oxygen intake (mL/min), VCO2-Carbondioxide output (mL/min)].

Predicted energy expenditure

The following equations were applied:1. Harris Benedict’s Equation

Male weight kg height cm6 76 age

: . . ( ) . ( ).

66 5 13 76 5 9+ × + × −×

This calorie formula takes height, weight, age, and sexinto consideration, which improves the accuracy of cal-culated energy expenditure when compared to calcula-tions based on total body weight alone.2. Cunningham’s Equation

370 21 6+ ×. ( )Fat free Mass FFM in kg

K.K. BEHERA et al. Energy expenditure in pancreatic diabetes

159© 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

AS

IAT

RA

CK

Since fat-free mass accounts for almost 80% of the varia-tions seen in RMR, this equation uses fat-free mass asthe predictor variable to estimate RMR.3. ICMR (Indian Council of Medical Research) Equa-tion (details in Table 1)

Prediction of accuracy

The accuracy of PEE against the gold standard wastested by calculating the percentage of the ratio betweenthe derived values (using equation) and the calorimeteryvalues. The percentage values were graded as “underpredictive” if <90%, “accurate” if between 90–110% and“over predictive” if >110%.

Statistical tests

Statistical analysis was performed using the SPSS11 soft-ware package. The correlation between two continuousvariables was established using Pearson’s correlation. Anindependent samples t-test was used to compare themeans of two continuous variables if they were normallydistributed and by Mann–Whitney U-(non-parametric)test if the distribution was skewed. “One-way anova”was used to establish a comparison within the threegroups and equal variances (between the two groups)were assessed by using Bonferroni Post-Hoc multiplecomparisons. Statistical significance was consideredwhen P-values were < 0.05.

Ethics

The study protocol was approved by the institutionalreview board. The study was approved and funded bythe Institutional Research Committee (IRB No. 6570dated 11/06/2008).

Results

A total of 51 male subjects were divided into three groups,namely fibrocalculous pancreatic diabetes (FCPD)(group 1; n = 24), type 2 diabetes mellitus (group 2; n = 15)and healthy controls (group 3; n = 12). Table 2 depicts themean age, height, weight, body mass index (BMI) andwaist-hip ratio. The subjects in all three groups differedsignificantly with regard to age. Patients with type 2diabetes were older than the others and the differences inthe mean age (Table 2) between the three groups wasstatistically significant (P = 0.001). A post-hoc analysisestablished that patients with type 2 diabetes had a sig-nificantly higher mean BMI of 21.34 ± 1.60 kg/m2;(P = 0.012) than the other two groups. The differences inwaist-hip ratio between these three groups were statisti-cally significant (P = 0.002) and in a post-hoc analysis thediabetic groups had a significantly higher mean waist-hipratio than those subjects in the control groups.

Table 3 illustrates the biochemical parameters of thestudy subjects whose HbA1c values were found to besignificantly higher in both the groups with diabetes(P = 0.001), as expected. There was no significant differ-ence in c-peptide values between group 1 and group 3.

Table 4 highlights the DEXA findings. The mean totalfat mass (in kg) and percentage of fat mass were signifi-cantly different among the groups. Post-hoc Bonferronianalysis showed that patients with type 2 diabetes had asignificantly higher fat percentage and total fat mass.The other components of body composition measuredby DXA bone mineral content (BMC) and lean bodymass were not significantly different among the groups.

The mean resting energy expenditure (REE) values asmeasured by indirect calorimetry (ICM) were not signifi-

Table 1 Indian Council of Medical Research equations for adultmales

Age (years) Prediction equation (Kcal/24 h)

18–30 14.5 × BW (kg) + 64530–60 10.9 × BW (kg) + 833>60 12.8 × BW (kg) + 463

BW, body weight in kg; Kcal, kilo calories.

Table 2 Demographic data of the study subjects represented with mean ± standard deviation (SD)

Parameters FCPD (n = 24) Type 2 DM (n = 15) Healthy (n = 12) P-value

Age (years) 36.04 ± 6.09 42.67 ± 7.63 20.83 ± 0.08 0.000*Height (cm) 166.02 ± 6.47 166.00 ± 4.98 169.70 ± 7.66 0.229Weight (kg) 53.32 ± 8.09 58.59 ± 5.06 55.24 ± 7.17 0.090BMI (kg/m2) 19.31 ± 2.43 21.34 ± 1.60 19.20 ± 2.17 0.012*Waist-Hip (W/H) ratio 0.91 ± 0.08 0.94 ± 0.59 0.84 ± 0.05 0.002*

*Indicates significant difference within the three groups using ANOVA.On Post Hoc Bonferroni analysis, the mean W/H ratio in fibrocalculous pancreatic diabetes (FCPD) subjects was lower than in type 2 diabetics(DM) and healthy controls. Body mass index (BMI) was significantly higher in type 2 diabetics.

Energy expenditure in pancreatic diabetes K.K. BEHERA et al.


AS

IAT

RA

CK

cantly different between the groups (P = 0.976) (Table 5).However, during analysis of covariance (Fig. 1), BMI wasfound to be the single major negative predictor of REEamong our study groups.

Discussion

This is the first ever study on energy expenditure con-ducted on subjects with FCPD. In the present study wehave compared two different groups with diabetes withhealthy controls. Anthropometric measures revealedthat all three groups of subjects in our study had anormal BMI. Nevertheless, the type 2 diabetes had a

significantly higher BMI; and DEXA revealed greatertotal body fat content and fat percentage when com-pared to the other two groups. Subjects with FCPD hada lower waist-hip ratio but did not differ significantlyfrom healthy adults in terms of height, weight, BMI andbody composition. We used the Asian cut off for BMI asmentioned by Gray et al.6

The PEE values derived using different equations didnot correlate with that of the gold standard. It was higherthan predicted by the equations indicating that none ofthese equations are accurate in measuring their restingenergy expenditure. Similar discrepancies were reportedwhen these equations were applied to different popula-

Table 3 Biochemical parameters of the study subjects

Biochemical Parameters (Normal range)

FCPD (n = 24) Type2 DM (n = 15) Healthy (n = 12)

P-valueGroup-1 Group-2 Group-3

Creatinine (44.2–123.76 μmol/L) 90.17 ± 20.33 87.52 ± 9.72 82.21 ± 5.30 0.339HbA1c (<6%) 9.45 ± 2.44 8.42 ± 1.80 5.58 ± 0.21 0.001*Fasting glucose (3.86–6.06 mmol/L) 6.01 (3.42–31.9) 9.37 (4.35–17.68) 4.68 (4.24–5.34) 0.006*Post-prandial glucose (4.4–7.71 mmol/L) 14.27 (4.74–45.29) 16.03 (8.82–22.15) 5.18 (2.80–6.34) 0.000*C-Peptide (0.36–1.65 nmol/L) 0.15 (0.60–1.19) – 0.18 (0.30–4.80) –

Stool fat (<18 g) 85.30 (8.20–93.50) – – –

*Indicates significant difference within the three groups using ANOVA.FCPD, fibrocalculous pancreatic disease; Type 2 DM, type 2 diabetes mellitus.

Table 4 Body composition by dual energy X-Rays absorptiometry (DEXA) represented with mean (standard deviation)

Parameters FCPD (n = 24) Type 2 DM (n = 15) Healthy (n = 12) P-value

BMC in kg 1.98 (0.36) 2.02 (0.21) 2.21 (0.27) 0.115Fat in kg 8.84 (4.26) 11.33 (2.71) 7.53 (4.69) 0.031*Lean body mass in kg 41.82 (4.44) 46.4 (3.77) 44.48 (4.03) 0.162Lean body mass + BMC in kg 43.81 (4.71) 45.66 (3.87) 46.69 (4.24) 0.152Total mass in kg 52.65 (7.05) 57.00 (4.84) 54.22 (7.04) 0.136Fat percentage 16.28 (6.42) 19.77 (4.28) 13.45(4.64) 0.016*

*Indicates significant difference within the three groups using ANOVA.On Post HOC Bonferroni analysis -Subjects with type 2 diabetes had significantly higher fat mass and fat percentage than healthy controls.BMC, bone mineral composition; FCPD, fibrocalculous pancreatic disease; Type 2 DM, type 2 diabetes mellitus.

Table 5 Resting energy expenditure (REE) as assessed by four methods

Resting energy expenditureby gold standard andequations

FCPD (n = 24) Type 2 DM (n = 15) Healthy (n = 12)

P-valueMean (SD) Mean (SD) Mean (SD)

REE by ICM 1628.37 (261.12) 1644.33 (304.98) 1645.91 (246.25) 0.976Harris& Benedict 1536.12 (139.29) 1563.71 (93.86) 1687.11 (124.16) 0.004*Cunningham 1316.31 (101.79) 1356.45 (83.61) 1378.59 (91.79) 0.152ICMR 1418.20 (117.43) 1494.00 (73.37) 1446.05 (104.07) 0.090

*Indicates significant difference among the three groups using ANOVA.The noted significance due to age difference between type 2 diabetes and healthy controls was nullified on post hoc analysis.FCPD, fibrocalculous pancreatic disease; Type 2 DM, Type 2 Diabetes Mellitus; REE, resting energy expenditure.



AS

IAT

RA

CK

tion groups. Studies done on hospitalized patients,7–9 andobese patients,10–13 have determined that these equationswere inadequate in predicting REE.

Contrary to expectations the unadjusted REE valuesas measured by indirect calorimetry did not vary signifi-cantly between our study groups. To understand thisphenomenon, the influence of confounding factors likeage and waist-hip ratio, which differed between the twotypes of diabetes was analyzed. Studies done previouslyhave shown that age-related changes contribute to barely2% of the variability observed in BMR.14 They evolvevery gradually from the second to the seventh decades oflife, in parallel with an increasing body fat and concomi-tantly declining lean body mass.14,15 The highest meanage in our study was only 42 years, which was seen inpatients with type 2 diabetes. Hence the differences in themean age even though significant would not have influ-enced the REE to any remarkable extent. The contribu-tion of fat mass to the variability of REE that has beenquoted in previous studies15 is relatively small and isaround 6%. Once again this supports the view14 that thecontribution of fat content in non-obese individuals (asdescribed in our study subjects) to the BMR is probablynegligible.

Fat-free mass, which is also referred to as lean bodymass has been considered to be the dominant factor14

influencing REE. It has been shown that fat-free masscontributes to around 65–85% of the variations in REE.15

This can be explained by the lack of a significant differ-ence in the lean body mass in our subjects.

During the transition from normal glucose tolerance(NGT) to impaired glucose tolerance (IGT) REEincreases by 4.2% and the progression to diabetes is

accompanied by a further 2.6% increase in REE.16 Thesefindings suggest that REE increases early during thedevelopment of diabetes, which indicates that thisincrease must be mediated by physiological factors otherthan changes in body composition.

Energy expenditure calculated by predictive equationsdid not correlate with indirect calorimetry. Therefore theusefulness of these equations in this category of subjectsis questionable.

The key limitation of our study is that there were alower number of controls and they were relativelyyounger in age. Some physiological factors such asendogenous glucose output (EGO), fasting insulin, andFFA concentration (but not glucose concentration) andinsulin-stimulated glucose disposal are known to influ-ence REE significantly.16 Our study has an obvious limi-tation, since we did not have data on these physiologicalfactors that may have given more clarity on their impacton REE. This is another obvious limitation of the study.

After adjustment for BMI, REE was lower in type 2diabetes patients compared to FCPD and healthy con-trols. Three of the available equations for measuringpredicted energy expenditure do not correlate with thegold standard.

The various physiological factors that influence REEin pancreatic diabetes need to be investigated further.This will help in deriving a more appropriate equation topredict REE in this specific category of patients.

Acknowledgement

The authors acknowledge and thank the ChristianMedical College & Hospital for the FLUID ResearchGrant (Grant no. 22X363), which provided funding forthis study.

Disclosure

The authors declare no conflict of interests.

References

1. Barman KK, Premalatha G, Mohan V. Tropical chronicpancreatitis. Postgrad Med J. 2003; 79: 606–15.

2. Pitchumoni CS. Special problems of tropical pancreatitis.Clin Gastroenterol. 1984; 13: 941–59.

3. Dickerson RN, Vehe KL, Mullen JL, Feurer ID. Restingenergy expenditure in patients with pancreatitis. CritCare Med. 1991; 19: 484–90.

4. Hébuterne X, Hastier P, Péroux JL, Zeboudj N, DelmontJP, Rampal P. Resting energy expenditure in patientswith alcoholic chronic pancreatitis. Dig Dis Sci. 1996; 41:533–9.

Figure 1 Body mass index (BMI) adjusted means of resting energyexpenditure (REE) among the three groups.

Energy expenditure in pancreatic diabetes K.K. BEHERA et al.


AS

IAT

RA

CK

5. McClave SA, Snider HL. Use of indirect calorimetry inclinical nutrition. Nutr Clin Pract. 1992; 7: 207–21.

6. Gray LJ, Yates T, Davies MJ et al. Defining Obesitycut-off points for migrant South Asians. PLoS ONE.2011; 6: e26464.

7. Reeves MM, Battistutta D, Capra S, Bauer J, DaviesPSW. Resting energy expenditure in patients with solidtumors undergoing anticancer therapy. Nutrition. 2006;22: 609–15.

8. Siirala W, Olkkola KT, Noponen T, Vuori A, Aantaa R.Predictive equations over-estimate the resting energyexpenditure in amyotrophic lateral sclerosis patients whoare dependent on invasive ventilation support. NutrMetab (Lond). 2010; 7: 70.

9. Stucky C-CH, Moncure M, Hise M, Gossage CM,Northrop D. How accurate are resting energy expendi-ture prediction equations in obese trauma andburn patients? JPEN J Parenter Enteral Nutr. 2008; 32:420–6.

10. O’Riordan CF, Metcalf BS, Perkins JM, Wilkin TJ. Reli-ability of energy expenditure prediction equations in theweight management clinic. J Hum Nutr Diet. 2010; 23:169–75.

11. Weijs PJM, Vansant GAAM. Validity of predictive equa-tions for resting energy expenditure in Belgian normalweight to morbid obese women. Clin Nutr. 2010; 29: 347–51.

12. Hofsteenge GH, Chinapaw MJM, Delemarre-van de Waal HA, Weijs PJM. Validation of predictiveequations for resting energy expenditure in obese adoles-cents. Am J Clin Nutr. 2010; 91: 1244–54.

13. Alves VGF, Da Rocha EEM, Gonzalez MC et al.Assessement of resting energy expenditure of obesepatients: Comparison of indirect calorimetry with formu-lae. Clin Nutr. 2009; 28: 299–304.

14. Johnstone AM, Murison SD, Duncan JS, Rance KA,Speakman JR. Factors influencing variation in basalmetabolic rate include fat-free mass, fat mass, age, andcirculating thyroxine but not sex, circulating leptin, ortriiodothyronine. Am J Clin Nutr. 2005; 82: 941–8.

15. Vaughan L, Zurlo F, Ravussin E. Aging and energyexpenditure. Am J Clin Nutr. 1991; 53: 821–5.

16. Weyer C, Bogardus C, Pratley RE. Metabolic factorscontributing to increased resting metabolic rate anddecreased insulin-induced thermogenesis during the devel-opment of type 2 diabetes. Diabetes. 1999; 48: 1607–14.



AS

IAT

RA

CK

resting energy expenditure in subjects with fibro-calculous pancreatic diabetes...

Documents