- Email: [email protected]
Resting energy expenditure in uremic, diabetic, and uremic diabetic subjects
Journal of Diabetes and Its Complications 18 (2004) 237 – 241
Resting energy expenditure in uremic, diabetic, and uremic diabetic subjects V. Rigalleau a,*, C. Lasseur b, S. Pe´cheur a, P. Chauveau b, C. Combe b, C. Perlemoine a, L. Baillet a, H. Gin a a
Nutrition-Diabe´tologie, Hopital Haut-Le´veˆque, Avenue de Magellan, 33600 Pessac, France b Ne´phrologie, Hopital Pellegrin, Place Ame´lie Raba-Le´on, 33000 Bordeaux, France Received 2 April 2003; received in revised form 26 May 2003; accepted 17 July 2003
Abstract We compared Harris and Benedict [H&B; Harris, J. A., & Benedict, F. G. (1919). A biometric study of basal metabolism in man. Washington, DC: Carnegie Institution of Washington. p. 279.] predicted resting energy expenditure (REE) to values measured by indirect calorimetry in normal, uremic, diabetic, and uremic diabetic subjects. Predicted REE were overestimated ( + 9.2%, P < .005) in uremic subjects, and underestimated ( 8.5%, P < .0001) in diabetic subjects. Uremic diabetic subjects were submitted to the opposite influences of diabetes and uremia on REE. Differences in body composition (lower fat-free mass in uremia and higher fat-free mass in diabetes) played a major role in these influences. In uremic diabetic subjects, predicted REE seemed well fitted to measured REE (biases < 2%), but they were less correlated, and limits of agreement between predicted and measured REE were large. Although their mean REE seems normal, prediction by the H&B equation leads to important individual errors in uremic diabetic subjects: direct measurement of energy expenditure by indirect calorimetry may be helpful to precise the adequate energy content of a diet for these subjects. D 2004 Elsevier Inc. All rights reserved. Keywords: Diabetes; Uremia; Resting energy expenditure
1. Introduction Dietary energy intakes about 35 kcal/kg/day are recommended in patients with End Stage Renal Diseases (National Kidney Foundation, 2000; Toigo, Aparicio, Attman, et al., 2000). These high levels may not be adequate in uremic diabetic subjects, who are often obese. Nondialyzed uremic diabetic subject do not present evidence of malnutrition despite higher resting energy expenditure (REE) than nondiabetic uremics (Avesani, Cuppari, Silva, et al., 2001). In diabetic hemodialyzed patients, the nutritional status is considerably variable: protein malnutrition, but also obesity, are both frequent (Cano, Roth, Aparicio, et al., 2002). The evaluation of energetic requirements comes critical to correctly replace lost calories when dietary proteins are
* Corresponding author. Tel.: +33-5-57-65-60-78; fax: +33-5-57-6560-79. E-mail address: [email protected] (V. Rigalleau). 1056-8727/04/$ – see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/S1056-8727(03)00077-1
restricted, which improves outcome in type 1 diabetic patients with diabetic nephropathy (Hansen, Tauber-Lassen, Jensenm, et al., 2002). REE can be measured by indirect calorimetry, or more simply evaluated by the Harris and Benedict (H&B) equation (Harris & Benedict, 1919): both methods have been proposed to determine the adequate energy content when a diet is prescribed to renal insufficient patients (Hirschberg & Kopple, 1990). However, both uremia and diabetes can affect energy metabolism: diabetic subjects have high REE (Nair, Halliday, & Garrow, 1984; Nair, Webster, & Garrow, 1986; Weyer, Bogardus, & Pratley, 1999), whereas normal (Bucciante, Senesi, Piva, et al., 1990; Monteon, Laidlaw, Shaib, et al., 1986) or low values (Olevitch, Bowers, & DeOreo, 1994; O’Sullivan, Lawson, Chan, et al., 2002) have been reported in uremia. These influences may both lead to biased evaluation of REE by the H&B equation, and the potentially resulting error for uremic diabetic subjects is not known; we therefore examinated the validity of the H&B equation in uremic, diabetic, and uremic diabetic
238
V. Rigalleau et al. / Journal of Diabetes and Its Complications 18 (2004) 237–241
Table 1 Main characteristics of normal, uremic, diabetic, diabetic with moderate renal failure and uremic diabetic subjects
n Gender (% male) Age (years) Diabetes type (% type 2) HbAlC (%) Body mass index (kg/m2) GFR (ml/min/1.73 m2)
Normal
Nondiabetic uremic
Nonuremic diabetic
Diabetic, moderate renal insufficiency
Uremic diabetic
16 69 40 F 12 – – 23.1 F 2.8 –
21 66 51 F 10 – – 25.1 F 4.6 13.6 F 5.2
18 72 58 F 11 80% 7.9 F 1.7 29.9 F 5.9 96.4 F 23.6
27 63 65 F 11 74 8.1 F 1.2 27.1 F 4.3 37.3 F 13.3
16 63 65 F 9 75 7.7 F 1.1 27.1 F 4.2 15.4 F 4.5
subjects, by comparing its results to values measured by indirect calorimetry.
2. Methods Results of H&B-derived REE were compared to direct measurements by indirect calorimetry in 98 subjects; their characteristics are summarized in Table 1: 18 nonuremic diabetic subjects (Glomerular Filtration Rate (GFR) > 60 ml/min/1.73 m2); 27 diabetic subjects with moderate renal insufficiency (20 V GFR V 60 ml/min/1.73 m2); 16 nondialyzed uremic diabetic subjects (GFR V 20 ml/ min/1.73 m2); 21 nondiabetic nondialyzed uremic subjects (GFR 13.6 F 5.2 ml/min/1.73 m2); and 16 normal subjects.
and foot. Respiratory exchanges were monitored at rest in the postabsorptive state, at 8:00 a.m. before breakfast after an overnight fast, during 45 min sessions in all subjects, with a Deltatrac metabolic monitor (Datex, Paris, France) calibrated with a reference gas before each session. Usual diet, physical activity, and medications of the subjects were not modified before or during the study. REE was derived from respiratory exchange measurements with conventional equations (Ferrannini, 1988). Results are expressed as mean F S.D. Predicted REE values (by H&B formula) were compared to measured (by indirect calorimetry) REE values by paired t tests and correlation coefficients were calculated in each group. Mean differences between predicted and measured REE values were expressed as %: 100 (predic(predicted measured REE)/measured REE, and compared between groups by unpaired t tests. A Bland & Altman procedure was also performed in each subgroup.
3. Results (Table 2, Figs. 1 and 2) GFR was assessed by a reference isotopic method (51CrEDTA clearance) in all (diabetic and nondiabetic) renal insufficient subjects. Body composition was predicted according to body mass index, age, and gender by Deurenberg’s equation (Deurenberg, Westrate, & Seidell, 1991), and was measured using bioimpedance analysis with a Thomasset and Boulier apparatus (L’Impulsion, He´rouville, France) (Boulier, Fricker, Thomasset, et al., 1990), with subcutaneous stainless needles placed on the opposite hand
As shown on Table 2, prediction of REE by the H&B formula was significantly biased in nondiabetic uremic subjects (overestimated REE) and nonuremic diabetic subjects (underestimated REE), whereas predicted values did not differ from measured values in uremic diabetic or normal subjects. The biases of REE prediction significantly differed between nondiabetic uremic and diabetic uremic subjects
Table 2 Predicted (H&B equation) and measured (indirect calorimetry) resting energy expenditure (kcal/24 h), predicted (Deurenberg equation) and measured (bioimpedance) % fat-free mass of normal, uremic, diabetic, diabetic with moderate renal failure and uremic diabetic subjects Normal Measured REE Predicted REE Bias by prediction of REE (%) Measured % fat-free mass Predicted % fat-free mass Bias by prediction of fat-free mass Measured REE/Fat-free Mass (kcal/ kg FFM/24 h)
Nondiabetic uremic
Nonuremic diabetic
Diabetic, moderate renal insufficiency
Uremic diabetic
1563 F 241 1595 F 239 + 2.1 F 4.5, ns 78.5 F 8.5 76.3 F 6.6 2.1 F 8.5% ns
1462 F 311 1586 F 322 + 9.2 F 11.7, P < .005 70.5 F 9.4 72.4 F 8.9 + 3.2 F 11.0% ns
1765 F 332 1607 F 284 8.5 F 7.5 P < .0001 72.1 F 7.0 63.8 F 9.4 10.5 F 11.7% P < .005
1472 F 229 1450 F 224 + 1.1 F 8.7 ns 65.9 F 8.5 64.4 F 8.1 1.4 F 7.2% ns
1543 F 250 1504 F 249 + 1.8 F 10.4 ns 66.9 F 6.9 64.4 F 9.5 3.8 F 7.5% ns
29.3 F 3.1
29.5 F 3.7
30.1 F 2.7
30.4 F 2.6
29.9 F 3.6
V. Rigalleau et al. / Journal of Diabetes and Its Complications 18 (2004) 237–241
239
Fig. 1. Measured (indirect calorimetry, Y axis) REE (kcal/day) plotted against predicted values (H&B equation, X axis) in nondiabetic uremic (crosses), nonuremic diabetic (squares) and uremic diabetic (crossed squares) subjects.
( P < .005), and between nonuremic diabetic and both groups of renal insufficient diabetic subjects (both P < .05). As shown in Table 2, prediction of body composition from weight, height, age, and gender by the Deurenberg’s equation led to similar biases: underestimation of percent fat-free mass in diabetic subjects ( P < .005), overestimation in nondiabetic uremic subjects (ns), and intermediary results in other groups. After normalization to measured fat-free mass, REE did not differ among uremic, diabetic, uremic diabetic, and normal subjects. Although mean predicted REE seemed well fitted to measured values in diabetic subjects with impaired renal function, they were less correlated as shown on Fig. 1. As a result, limits of agreement between calculated, and measured REE were therefore large in this population: 2SD was 260 kcal/day vs. 140 in normal subjects as shown on the Bland & Altman plots on Fig. 2.
4. Discussion In accordance with previous reports, we found that diabetic subjects have higher REE (Weyer et al., 1999), and nondialyzed uremic subjects lower REE (Bucciante et al., 1990; Olevitch et al., 1994), than predicted by the H&B equation. This widely used formula predicts REE from age, gender, weight, and height that are important predictors of fat-free mass. These parameters are indeed used to evaluate fatness by the Deurenberg equation, and we found similar biases when we compared directly measured fat-free mass by bioimpedance to their results
derived from Deurenberg equation in our subjects. Loss of lean body mass is a well-known consequence of kidney diseases (Mitch, 1999), and high lean body mass are expected in type 2 diabetic subjects due to obesity, but have also been reported in type 1 diabetes (Rosenfalck, Almdal, Gotfredsen, et al., 1997). These changes in body composition surely played a major role in the apparent modifications of REE as we found, as fat-free mass is the prime determinant of REE (Cunningham, 1991). After normalization to FFM, REE did not significantly differ between the subjects we studied, although the slight increase ( + 1 kcal/kgFFM/24 h, ns) in the diabetic groups is not far from the 2.5 – 5% higher values as compared to BMI-matched controls or in longitudinal studies (Weyer et al., 1999). In nondiabetic uremic subjects, normalization of REE by fat-free mass has lead to normal (Olevitch et al., 1994) or low (O’Sullivan et al., 2002) results when reported. Uremic diabetic subjects therefore have higher REE than their nondiabetic counterparts, as already reported (Avesani et al., 2001). As shown on Fig. 1, the diabetic uremic subjects we studied were submitted to opposite influences of uremia and diabetes on body composition, and hence, REE, and this occurred as soon as GFR was below 60 ml/min/1.73 m2. Their mean predicted REE was therefore well fitted to the measured value, but the validity of the H&B equation was reduced as shown by a lower correlation coefficient between predicted and measured REE. Limits of agreement between calculated and measured REE were therefore large in this population, so despite apparently normal mean REE, important errors are possible when the H&B equation is used in
240
V. Rigalleau et al. / Journal of Diabetes and Its Complications 18 (2004) 237–241
REE may increase by 15% in some subjects in that occurrence (Rigalleau, Combe, Blanchetier, et al., 1997), whereas lean body mass tends to decrease (Chauveau, Barthe, Rigalleau, et al., 1999).
References
Fig. 2. Bland & Altman plots of differences between predicted (H&B) vs. measured (indirect calorimetry) REE as a function of average REE by both methods in normal, uremic, diabetic, diabetic with moderate renal failure and uremic diabetic subjects.
uremic diabetic individuals. Measurement of REE by indirect calorimetry, or at least direct measurement of fat-free mass, can therefore be helpful for a precise dietary advice in these subjects. This may become crucial if a very-low-protein-supplemented diet is prescribed, as
Avesani, C. M., Cuppari, L., Silva, A. C., Sigulem, D. M., Cendoroglo, M., Sesso, R., & Draibe, S. A. (2001). Resting energy expenditure in predialysis diabetic patients. Nephrology, Dialysis, Transplantation, 16, 556 – 560. Boulier, A., Fricker, J., Thomasset, A. L., & Apfelbaum, M. (1990). Fatfree estimation by the two-electrode impedance method. American Journal of Clinical Nutrition, 52, 581 – 585. Bucciante, G., Senesi, G., Piva, M., Donato, D., Velluti, F., Milito, F., Spisani, C., & Bernardi, A. (1990). Resting metabolic rate by indirect calorimetry in uremic patients. Contributions to Nephrology, 81, 214 – 219. Cano, N., Roth, H., Aparicio, M., Azar, R., Canaud, B., Cheveau, P., Combe, C., Fouque, D., Laville, D., & Leverve, X. M. (2002). Malnutrition in hemodialysis diabetic patients: evaluation and prognostic influence. Kidney International, 62, 593 – 601. Chauveau, P., Barthe, N., Rigalleau, V., Ozenne, S., Castaing., F., Delclaux, C., de Precigout, V., Combe, C., & Aparicio, M. (1999). Outcome of nutritional status and body composition of uremic patients on very lowprotein diet. American Journal of Kidney Diseases, 34, 500 – 507. Cunningham, J. J. (1991). Body composition as a determinant of energy expenditure: a synthetic review and a proposed general prediction equation. American Journal of Clinical Nutrition, 54, 963 – 969. Deurenberg, P., Westrate, J. A., & Seidell, J. C. (1991). Body mass index as a measure of body fatness: age and gender specific prediction formulas. British Journal of Nutrition, 65, 105 – 114. Ferrannini, E. (1988). The theoretical basis of indirect calorimetry: a review. Metabolism, 37, 287 – 301. Hansen, H. P., Tauber-Lassen, E., Jensen, B. R., & Parving, H. H. (2002). Effect of dietary protein restriction on prognosis in patients with diabetic nephropathy. Kidney International, 62, 220 – 228. Harris, J. A., & Benedict, F. G. (1919). A biometric study of basal metabolism in man p. 279. Washington, DC: Carnegie Institution of Washington. Hirschberg, R. R., & Kopple, J. D. (1990). Energy requirements in patients with renal failure. Contributions to Nephrology, 81, 124 – 135. Mitch, W. E. (1999). Mechanisms causing loss of lean body mass in kidney disease. American Journal of Clinical Nutrition, 67, 359 – 366. Monteon, F. J., Laidlaw, S. A., Shaib, J. K., & Kopple, J. D. (1986). Energy expenditure in patients with chronic renal failure. Kidney International, 30, 741 – 747. Nair, K. S., Halliday, D., & Garrow, J. S. (1984). Increased energy expenditure in poorly controlled type 1 diabetic patients. Diabetologia, 27, 13 – 16. Nair, K. S., Webster, J., & Garrow, J. S. (1986). Effect of impaired glucose tolerance and type II diabetes on resting metabolic rate and thermic response to a glucose meal in obese women. Metabolism, 35, 640 – 644. National Kidney Foundation, J. S. (2000). Kidney disease outcomes quality initiative. Clinical practice guidelines for nutrition in chronic renal failure. I. Adult guidelines. B. Advanced chronic renal failure without dialysis. American Journal of Kidney Diseases, 35 (Suppl. 2), S56 – S65. Olevitch, L. R., Bowers, B. M., & DeOreo, P. B. (1994). Measurement of resting energy expenditure via indirect calorimetry among adult hemodialysis patients. Journal of Renal Nutrition, 4, 192 – 197. O’Sullivan, A. J., Lawson, J. A., Chan, M., & Kelly, J. J. (2002). Body composition and energy metabolism in chronic renal insufficiency. American Journal of Kidney Diseases, 39, 369 – 375.
V. Rigalleau et al. / Journal of Diabetes and Its Complications 18 (2004) 237–241 Rigalleau, V., Combe, C., Blanchetier, V., Aubertin, J., Aparicio, M., & Gin, H. (1997). Low protein diet in uremia: effects on glucose metabolism and energy production rate. Kidney International, 51, 1222 – 1227. Rosenfalck, A. M., Almdal, T., Gotfredsen, A., Viggers, L., & Hilsted, J. (1997). Increased lean body mass in patients with type 1 insulin dependent diabetes mellitus: the relationship with metabolic control. Endocrinology and Metabolism, 4, 241 – 246.
241
Toigo, G., Aparicio, M., Attman, P., Cano, N., Cianciaruso, B., Engel, B., Fouque, D., Heidland, A., Teplan, V., & Wanner, C. (2000). ESPEN consensus on nutritional treatment of patients with renal insufficiency (Part 1 of 2). Clinical Nutrition, 19, 197 – 207. Weyer, C., Bogardus, C., & Pratley, R. E. (1999). Metabolic factors contributing to increased resting metabolic rate and decreased insulin-induced thermogenesis during the development of type 2 diabetes. Diabetes, 48, 1607 – 1614.
Resting energy expenditure in uremic, diabetic, and uremic diabetic subjects V. Rigalleau a,*, C. Lasseur b, S. Pe´cheur a, P. Chauveau b, C. Combe b, C. Perlemoine a, L. Baillet a, H. Gin a a
Nutrition-Diabe´tologie, Hopital Haut-Le´veˆque, Avenue de Magellan, 33600 Pessac, France b Ne´phrologie, Hopital Pellegrin, Place Ame´lie Raba-Le´on, 33000 Bordeaux, France Received 2 April 2003; received in revised form 26 May 2003; accepted 17 July 2003
Abstract We compared Harris and Benedict [H&B; Harris, J. A., & Benedict, F. G. (1919). A biometric study of basal metabolism in man. Washington, DC: Carnegie Institution of Washington. p. 279.] predicted resting energy expenditure (REE) to values measured by indirect calorimetry in normal, uremic, diabetic, and uremic diabetic subjects. Predicted REE were overestimated ( + 9.2%, P < .005) in uremic subjects, and underestimated ( 8.5%, P < .0001) in diabetic subjects. Uremic diabetic subjects were submitted to the opposite influences of diabetes and uremia on REE. Differences in body composition (lower fat-free mass in uremia and higher fat-free mass in diabetes) played a major role in these influences. In uremic diabetic subjects, predicted REE seemed well fitted to measured REE (biases < 2%), but they were less correlated, and limits of agreement between predicted and measured REE were large. Although their mean REE seems normal, prediction by the H&B equation leads to important individual errors in uremic diabetic subjects: direct measurement of energy expenditure by indirect calorimetry may be helpful to precise the adequate energy content of a diet for these subjects. D 2004 Elsevier Inc. All rights reserved. Keywords: Diabetes; Uremia; Resting energy expenditure
1. Introduction Dietary energy intakes about 35 kcal/kg/day are recommended in patients with End Stage Renal Diseases (National Kidney Foundation, 2000; Toigo, Aparicio, Attman, et al., 2000). These high levels may not be adequate in uremic diabetic subjects, who are often obese. Nondialyzed uremic diabetic subject do not present evidence of malnutrition despite higher resting energy expenditure (REE) than nondiabetic uremics (Avesani, Cuppari, Silva, et al., 2001). In diabetic hemodialyzed patients, the nutritional status is considerably variable: protein malnutrition, but also obesity, are both frequent (Cano, Roth, Aparicio, et al., 2002). The evaluation of energetic requirements comes critical to correctly replace lost calories when dietary proteins are
* Corresponding author. Tel.: +33-5-57-65-60-78; fax: +33-5-57-6560-79. E-mail address: [email protected] (V. Rigalleau). 1056-8727/04/$ – see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/S1056-8727(03)00077-1
restricted, which improves outcome in type 1 diabetic patients with diabetic nephropathy (Hansen, Tauber-Lassen, Jensenm, et al., 2002). REE can be measured by indirect calorimetry, or more simply evaluated by the Harris and Benedict (H&B) equation (Harris & Benedict, 1919): both methods have been proposed to determine the adequate energy content when a diet is prescribed to renal insufficient patients (Hirschberg & Kopple, 1990). However, both uremia and diabetes can affect energy metabolism: diabetic subjects have high REE (Nair, Halliday, & Garrow, 1984; Nair, Webster, & Garrow, 1986; Weyer, Bogardus, & Pratley, 1999), whereas normal (Bucciante, Senesi, Piva, et al., 1990; Monteon, Laidlaw, Shaib, et al., 1986) or low values (Olevitch, Bowers, & DeOreo, 1994; O’Sullivan, Lawson, Chan, et al., 2002) have been reported in uremia. These influences may both lead to biased evaluation of REE by the H&B equation, and the potentially resulting error for uremic diabetic subjects is not known; we therefore examinated the validity of the H&B equation in uremic, diabetic, and uremic diabetic
238
V. Rigalleau et al. / Journal of Diabetes and Its Complications 18 (2004) 237–241
Table 1 Main characteristics of normal, uremic, diabetic, diabetic with moderate renal failure and uremic diabetic subjects
n Gender (% male) Age (years) Diabetes type (% type 2) HbAlC (%) Body mass index (kg/m2) GFR (ml/min/1.73 m2)
Normal
Nondiabetic uremic
Nonuremic diabetic
Diabetic, moderate renal insufficiency
Uremic diabetic
16 69 40 F 12 – – 23.1 F 2.8 –
21 66 51 F 10 – – 25.1 F 4.6 13.6 F 5.2
18 72 58 F 11 80% 7.9 F 1.7 29.9 F 5.9 96.4 F 23.6
27 63 65 F 11 74 8.1 F 1.2 27.1 F 4.3 37.3 F 13.3
16 63 65 F 9 75 7.7 F 1.1 27.1 F 4.2 15.4 F 4.5
subjects, by comparing its results to values measured by indirect calorimetry.
2. Methods Results of H&B-derived REE were compared to direct measurements by indirect calorimetry in 98 subjects; their characteristics are summarized in Table 1: 18 nonuremic diabetic subjects (Glomerular Filtration Rate (GFR) > 60 ml/min/1.73 m2); 27 diabetic subjects with moderate renal insufficiency (20 V GFR V 60 ml/min/1.73 m2); 16 nondialyzed uremic diabetic subjects (GFR V 20 ml/ min/1.73 m2); 21 nondiabetic nondialyzed uremic subjects (GFR 13.6 F 5.2 ml/min/1.73 m2); and 16 normal subjects.
and foot. Respiratory exchanges were monitored at rest in the postabsorptive state, at 8:00 a.m. before breakfast after an overnight fast, during 45 min sessions in all subjects, with a Deltatrac metabolic monitor (Datex, Paris, France) calibrated with a reference gas before each session. Usual diet, physical activity, and medications of the subjects were not modified before or during the study. REE was derived from respiratory exchange measurements with conventional equations (Ferrannini, 1988). Results are expressed as mean F S.D. Predicted REE values (by H&B formula) were compared to measured (by indirect calorimetry) REE values by paired t tests and correlation coefficients were calculated in each group. Mean differences between predicted and measured REE values were expressed as %: 100 (predic(predicted measured REE)/measured REE, and compared between groups by unpaired t tests. A Bland & Altman procedure was also performed in each subgroup.
3. Results (Table 2, Figs. 1 and 2) GFR was assessed by a reference isotopic method (51CrEDTA clearance) in all (diabetic and nondiabetic) renal insufficient subjects. Body composition was predicted according to body mass index, age, and gender by Deurenberg’s equation (Deurenberg, Westrate, & Seidell, 1991), and was measured using bioimpedance analysis with a Thomasset and Boulier apparatus (L’Impulsion, He´rouville, France) (Boulier, Fricker, Thomasset, et al., 1990), with subcutaneous stainless needles placed on the opposite hand
As shown on Table 2, prediction of REE by the H&B formula was significantly biased in nondiabetic uremic subjects (overestimated REE) and nonuremic diabetic subjects (underestimated REE), whereas predicted values did not differ from measured values in uremic diabetic or normal subjects. The biases of REE prediction significantly differed between nondiabetic uremic and diabetic uremic subjects
Table 2 Predicted (H&B equation) and measured (indirect calorimetry) resting energy expenditure (kcal/24 h), predicted (Deurenberg equation) and measured (bioimpedance) % fat-free mass of normal, uremic, diabetic, diabetic with moderate renal failure and uremic diabetic subjects Normal Measured REE Predicted REE Bias by prediction of REE (%) Measured % fat-free mass Predicted % fat-free mass Bias by prediction of fat-free mass Measured REE/Fat-free Mass (kcal/ kg FFM/24 h)
Nondiabetic uremic
Nonuremic diabetic
Diabetic, moderate renal insufficiency
Uremic diabetic
1563 F 241 1595 F 239 + 2.1 F 4.5, ns 78.5 F 8.5 76.3 F 6.6 2.1 F 8.5% ns
1462 F 311 1586 F 322 + 9.2 F 11.7, P < .005 70.5 F 9.4 72.4 F 8.9 + 3.2 F 11.0% ns
1765 F 332 1607 F 284 8.5 F 7.5 P < .0001 72.1 F 7.0 63.8 F 9.4 10.5 F 11.7% P < .005
1472 F 229 1450 F 224 + 1.1 F 8.7 ns 65.9 F 8.5 64.4 F 8.1 1.4 F 7.2% ns
1543 F 250 1504 F 249 + 1.8 F 10.4 ns 66.9 F 6.9 64.4 F 9.5 3.8 F 7.5% ns
29.3 F 3.1
29.5 F 3.7
30.1 F 2.7
30.4 F 2.6
29.9 F 3.6
V. Rigalleau et al. / Journal of Diabetes and Its Complications 18 (2004) 237–241
239
Fig. 1. Measured (indirect calorimetry, Y axis) REE (kcal/day) plotted against predicted values (H&B equation, X axis) in nondiabetic uremic (crosses), nonuremic diabetic (squares) and uremic diabetic (crossed squares) subjects.
( P < .005), and between nonuremic diabetic and both groups of renal insufficient diabetic subjects (both P < .05). As shown in Table 2, prediction of body composition from weight, height, age, and gender by the Deurenberg’s equation led to similar biases: underestimation of percent fat-free mass in diabetic subjects ( P < .005), overestimation in nondiabetic uremic subjects (ns), and intermediary results in other groups. After normalization to measured fat-free mass, REE did not differ among uremic, diabetic, uremic diabetic, and normal subjects. Although mean predicted REE seemed well fitted to measured values in diabetic subjects with impaired renal function, they were less correlated as shown on Fig. 1. As a result, limits of agreement between calculated, and measured REE were therefore large in this population: 2SD was 260 kcal/day vs. 140 in normal subjects as shown on the Bland & Altman plots on Fig. 2.
4. Discussion In accordance with previous reports, we found that diabetic subjects have higher REE (Weyer et al., 1999), and nondialyzed uremic subjects lower REE (Bucciante et al., 1990; Olevitch et al., 1994), than predicted by the H&B equation. This widely used formula predicts REE from age, gender, weight, and height that are important predictors of fat-free mass. These parameters are indeed used to evaluate fatness by the Deurenberg equation, and we found similar biases when we compared directly measured fat-free mass by bioimpedance to their results
derived from Deurenberg equation in our subjects. Loss of lean body mass is a well-known consequence of kidney diseases (Mitch, 1999), and high lean body mass are expected in type 2 diabetic subjects due to obesity, but have also been reported in type 1 diabetes (Rosenfalck, Almdal, Gotfredsen, et al., 1997). These changes in body composition surely played a major role in the apparent modifications of REE as we found, as fat-free mass is the prime determinant of REE (Cunningham, 1991). After normalization to FFM, REE did not significantly differ between the subjects we studied, although the slight increase ( + 1 kcal/kgFFM/24 h, ns) in the diabetic groups is not far from the 2.5 – 5% higher values as compared to BMI-matched controls or in longitudinal studies (Weyer et al., 1999). In nondiabetic uremic subjects, normalization of REE by fat-free mass has lead to normal (Olevitch et al., 1994) or low (O’Sullivan et al., 2002) results when reported. Uremic diabetic subjects therefore have higher REE than their nondiabetic counterparts, as already reported (Avesani et al., 2001). As shown on Fig. 1, the diabetic uremic subjects we studied were submitted to opposite influences of uremia and diabetes on body composition, and hence, REE, and this occurred as soon as GFR was below 60 ml/min/1.73 m2. Their mean predicted REE was therefore well fitted to the measured value, but the validity of the H&B equation was reduced as shown by a lower correlation coefficient between predicted and measured REE. Limits of agreement between calculated and measured REE were therefore large in this population, so despite apparently normal mean REE, important errors are possible when the H&B equation is used in
240
V. Rigalleau et al. / Journal of Diabetes and Its Complications 18 (2004) 237–241
REE may increase by 15% in some subjects in that occurrence (Rigalleau, Combe, Blanchetier, et al., 1997), whereas lean body mass tends to decrease (Chauveau, Barthe, Rigalleau, et al., 1999).
References
Fig. 2. Bland & Altman plots of differences between predicted (H&B) vs. measured (indirect calorimetry) REE as a function of average REE by both methods in normal, uremic, diabetic, diabetic with moderate renal failure and uremic diabetic subjects.
uremic diabetic individuals. Measurement of REE by indirect calorimetry, or at least direct measurement of fat-free mass, can therefore be helpful for a precise dietary advice in these subjects. This may become crucial if a very-low-protein-supplemented diet is prescribed, as
Avesani, C. M., Cuppari, L., Silva, A. C., Sigulem, D. M., Cendoroglo, M., Sesso, R., & Draibe, S. A. (2001). Resting energy expenditure in predialysis diabetic patients. Nephrology, Dialysis, Transplantation, 16, 556 – 560. Boulier, A., Fricker, J., Thomasset, A. L., & Apfelbaum, M. (1990). Fatfree estimation by the two-electrode impedance method. American Journal of Clinical Nutrition, 52, 581 – 585. Bucciante, G., Senesi, G., Piva, M., Donato, D., Velluti, F., Milito, F., Spisani, C., & Bernardi, A. (1990). Resting metabolic rate by indirect calorimetry in uremic patients. Contributions to Nephrology, 81, 214 – 219. Cano, N., Roth, H., Aparicio, M., Azar, R., Canaud, B., Cheveau, P., Combe, C., Fouque, D., Laville, D., & Leverve, X. M. (2002). Malnutrition in hemodialysis diabetic patients: evaluation and prognostic influence. Kidney International, 62, 593 – 601. Chauveau, P., Barthe, N., Rigalleau, V., Ozenne, S., Castaing., F., Delclaux, C., de Precigout, V., Combe, C., & Aparicio, M. (1999). Outcome of nutritional status and body composition of uremic patients on very lowprotein diet. American Journal of Kidney Diseases, 34, 500 – 507. Cunningham, J. J. (1991). Body composition as a determinant of energy expenditure: a synthetic review and a proposed general prediction equation. American Journal of Clinical Nutrition, 54, 963 – 969. Deurenberg, P., Westrate, J. A., & Seidell, J. C. (1991). Body mass index as a measure of body fatness: age and gender specific prediction formulas. British Journal of Nutrition, 65, 105 – 114. Ferrannini, E. (1988). The theoretical basis of indirect calorimetry: a review. Metabolism, 37, 287 – 301. Hansen, H. P., Tauber-Lassen, E., Jensen, B. R., & Parving, H. H. (2002). Effect of dietary protein restriction on prognosis in patients with diabetic nephropathy. Kidney International, 62, 220 – 228. Harris, J. A., & Benedict, F. G. (1919). A biometric study of basal metabolism in man p. 279. Washington, DC: Carnegie Institution of Washington. Hirschberg, R. R., & Kopple, J. D. (1990). Energy requirements in patients with renal failure. Contributions to Nephrology, 81, 124 – 135. Mitch, W. E. (1999). Mechanisms causing loss of lean body mass in kidney disease. American Journal of Clinical Nutrition, 67, 359 – 366. Monteon, F. J., Laidlaw, S. A., Shaib, J. K., & Kopple, J. D. (1986). Energy expenditure in patients with chronic renal failure. Kidney International, 30, 741 – 747. Nair, K. S., Halliday, D., & Garrow, J. S. (1984). Increased energy expenditure in poorly controlled type 1 diabetic patients. Diabetologia, 27, 13 – 16. Nair, K. S., Webster, J., & Garrow, J. S. (1986). Effect of impaired glucose tolerance and type II diabetes on resting metabolic rate and thermic response to a glucose meal in obese women. Metabolism, 35, 640 – 644. National Kidney Foundation, J. S. (2000). Kidney disease outcomes quality initiative. Clinical practice guidelines for nutrition in chronic renal failure. I. Adult guidelines. B. Advanced chronic renal failure without dialysis. American Journal of Kidney Diseases, 35 (Suppl. 2), S56 – S65. Olevitch, L. R., Bowers, B. M., & DeOreo, P. B. (1994). Measurement of resting energy expenditure via indirect calorimetry among adult hemodialysis patients. Journal of Renal Nutrition, 4, 192 – 197. O’Sullivan, A. J., Lawson, J. A., Chan, M., & Kelly, J. J. (2002). Body composition and energy metabolism in chronic renal insufficiency. American Journal of Kidney Diseases, 39, 369 – 375.
V. Rigalleau et al. / Journal of Diabetes and Its Complications 18 (2004) 237–241 Rigalleau, V., Combe, C., Blanchetier, V., Aubertin, J., Aparicio, M., & Gin, H. (1997). Low protein diet in uremia: effects on glucose metabolism and energy production rate. Kidney International, 51, 1222 – 1227. Rosenfalck, A. M., Almdal, T., Gotfredsen, A., Viggers, L., & Hilsted, J. (1997). Increased lean body mass in patients with type 1 insulin dependent diabetes mellitus: the relationship with metabolic control. Endocrinology and Metabolism, 4, 241 – 246.
241
Toigo, G., Aparicio, M., Attman, P., Cano, N., Cianciaruso, B., Engel, B., Fouque, D., Heidland, A., Teplan, V., & Wanner, C. (2000). ESPEN consensus on nutritional treatment of patients with renal insufficiency (Part 1 of 2). Clinical Nutrition, 19, 197 – 207. Weyer, C., Bogardus, C., & Pratley, R. E. (1999). Metabolic factors contributing to increased resting metabolic rate and decreased insulin-induced thermogenesis during the development of type 2 diabetes. Diabetes, 48, 1607 – 1614.