Whole-body protein kinetics in marasmus and kwashiorkor during acute infection13

Mark J Manary, Robin L Broadhead, and Kevin E Yarasheski

ABSTRACT Marasmus and kwashiorkor are clinically distinct manifestations of severe malnutrition. This study tested the hypothesis that rates of whole-body protein synthesis and breakdown are higher in marasmus than in kwashiorkor during acute infection. We measured whole-body protein kinetics using stable isotope tracers in eight children with marasmus and acute infection (pneumonia or malaria) to determine the rate of appearance of urea and leucine in plasma. Serum concentrations of total protein, albumin, and C-reactive protein were also measured. These findings were compared with those reported previously for 13 children with kwashiorkor (including marasmic kwashiorkor) and acute infection who were studied with the same methods. HIV infection was present in 10 of 21 children. Rates of protein breakdown and synthesis were higher in marasmus than in kwashiorkor (227 59 compared with 103 30 mol leucine kg 1 h 1 and 216 60 compared with 97 30 mol leucine kg 1 h 1, P < 0.001). The concentration of globulin (total protein minus albumin) was higher in marasmus than kwashiorkor (40 17 compared with 25 7 g/L, P 0.01), but C-reactive protein was not different (73 79 compared with 83 89 mg/L). HIV infection and body composition did not explain the differences between marasmus and kwashiorkor. The accelerated rate of protein turnover in children with marasmus and acute infection requires further investigation. Am J Clin Nutr 1998;67:12059. KEY WORDS Malnutrition, protein metabolism, infection, stable isotopes, mass spectrometry, marasmus, kwashiorkor, children, leucine, albumin, C-reactive protein INTRODUCTION Marasmus and kwashiorkor are the two primary clinical manifestations of diets deficient in energy and protein. Marasmus is characterized by wasting and decreased physical activity, but with stimulation, mental status and appetite are relatively normal (1, 2). Kwashiorkor, including marasmic kwashiorkor, is characterized by massive edema of the hands and feet, profound irritability, anorexia, a desquamative rash, hair discoloration, and a large fatty liver (3). The factors determining whether kwashiorkor or marasmus develops remain unknown. Marasmus has been described as an adaptation to inadequate energy and protein intake and kwashiorkor as a dysadaptation (4). Chronic infections, such as infection with HIV and tuberculosis, pathologi1 From the Department of Pediatrics, Washington University School of Medicine, St Louis; the Department of Paediatrics, College of Medicine, University of Malawi, Blantyre; and the Division of Metabolism, Washington University School of Medicine, St Louis. 2 Supported by the International Atomic Energy Agency (7894/RI/RB), Washington University Mass Spectrometry Resource (NIH RR00954 and NIH DK49393), and the Fulbright Scholarship Program of the US Information Agency. 3 Address reprint requests to MJ Manary, Department of Pediatrics, St Louis Childrens Hospital, One Childrens Place, St Louis, MO 63110-1077. E-mail: manary@kidsa1.wustl.edu. Received September 24, 1997. Accepted for publication December 15, 1997.

cally compromise nutritional status and contribute to the global burden of childhood malnutrition, particularly in Africa. Malnutrition is associated with a decrease in the rates of whole-body protein synthesis and breakdown (5). This is believed to be an adaptive mechanism to conserve energy and amino acids when they are in scarce supply (6). Previous reports of protein kinetics in childhood malnutrition did not distinguish between children with kwashiorkor and those with marasmus. During the physiologic stress of acute infection, the rate of whole-body protein turnover increases (7). This is an appropriate response to help the individual respond successfully to the infection. The increase in the rate of protein breakdown is presumed to provide amino acids for the synthesis of acute phase proteins (8). Recently, we reported that, early in recovery, children with kwashiorkor and pneumonia have lower rates of whole-body protein synthesis and breakdown than do uninfected children with kwashiorkor (9). This study was designed to test the hypothesis that during acute infection marasmic children have higher rates of protein turnover than do children with kwashiorkor. SUBJECTS AND METHODS Children were chosen from those < 7 y of age admitted to the nutrition ward of Queen Elizabeth Central Hospital in Blantyre, Malawi, with marasmus (or undernutrition) and pneumonia. Pneumonia was diagnosed by standard criteria of cough and a respiratory rate > 40/min (10). Respiratory signs and symptoms were present for < 24 h. Although all children studied met the diagnostic criteria for pneumonia, it has been shown that these

Downloaded from www.ajcn.org by guest on December 26, 2011

Am J Clin Nutr 1998;67:12059. Printed in USA. 1998 American Society for Clinical Nutrition

1205

1206

MANARY ET AL their weight from admission over 10 d before they started to gain weight; thus, the lowest weight may be a better estimate of lean body weight. In none of these children was body composition assessed by standard techniques such as bioelectrical impedance analysis or isotopic measurement of body water spaces because the equipment needed to perform these measurements is not available in Malawi; skinfold thickness measurements are unreliable in edematous malnutrition and thus also were not performed (13, 14). Protein kinetics and protein concentrations were compared by Students t test, with group data expressed as means SDs. A P value < 0.05 was considered to be signicant. RESULTS Eight children with marasmus or undernutrition were studied, six with pneumonia and two with falciparum malaria (Table 1). Rates of whole-body protein breakdown and synthesis were significantly higher in children with marasmus than in those with kwashiorkor (Table 2). Whereas rates of urea appearance were not significantly different between the two groups (P = 0.10), the power of this comparison was low. Globulin concentrations were higher in marasmus than kwashiorkor, but CRP concentrations were not (Table 1). In the 10 HIV-infected children (six with marasmus and four with kwashiorkor), rates of whole-body protein turnover were greater in children with marasmus (Table 3). DISCUSSION Rates of whole-body protein breakdown and protein synthesis are greater in children with marasmus than in those with kwashiorkor during acute infection. It has been reported that children with marasmus appropriately reduce their rates of whole-body protein turnover to conserve energy and amino acids (5). The present study suggests that rates of whole-body protein breakdown and synthesis can increase in children with marasmus under the physiologic stress of infection. In contrast, rates of protein turnover do

criteria do not readily distinguish malaria from pneumonia in Malawi (11). Children were thus further evaluated with a chest X-ray and a thick blood film and the results of these tests used to diagnose pneumonia or malaria. The X-rays were interpreted by a pediatric radiologist at Washington University School of Medicine without knowledge of the childs clinical history. Because microbiologic and serologic laboratory support were unavailable, no attempt was made to identify the etiologic agent of pneumonia. Each child was tested for HIV infection with a serum enzyme-linked immunosorbent assay. The study was conducted on Thursdays from 21 November 1994 to 12 January 1995. All eligible marasmic children admitted on Thursdays were included, none were arbitrarily excluded. The study was approved by the Health Sciences Research Committee of Malawi and the Human Studies Committee of Washington University in St Louis. Marasmic children were initially offered a diet providing 1.2 g protein kg 1 d 1 and 336 kJ kg 1 d 1, and their dietary intake was liberalized after 48 h if their clinical condition improved. Informed consent to participate in the study was obtained from the caretakers (usually mothers or grandmothers). Children were studied overnight during an 8-h period when they had no scheduled feedings. They were moved to a separate research ward and observed closely by a pediatrician and nurse throughout the duration of the study. The infusion protocol, analytic techniques, and method of calculation of protein breakdown, protein synthesis, and leucine oxidation rates were described previously (9). Briefly, all children were given a primed, constant, intravenous infusion of [1-13C]leucine and [15N2]urea in the postabsorptive state. The children ingested only sips of water during the infusion. The [13C]leucine enrichment in serum was used to calculate the leucine rate of appearance, which corresponds to the rate of whole-body protein breakdown in the postabsorptive state. Serum [15N2]urea enrichment was used to calculate the rate of urea appearance, which is proportional to the rate of leucine oxidation. This assumes that in the postabsorptive state a fixed fraction of nitrogen in urea is from leucine (12). C-reactive protein (CRP), albumin, and total serum protein concentrations were also measured in each marasmic child in the postabsorptive state. CRP was measured by rate nephelometry (Array 360; Beckman, High Wycombe, United Kingdom) and total serum protein and albumin by spectrophotometry (Vitros 250 Analyzer; Johnson and Johnson, Rochester, NY). The globulin fraction of protein was calculated as the difference between total protein and albumin. Thirteen children with kwashiorkor and pneumonia were studied previously with the same protocol (9). These children were used as a comparison group to evaluate differences in protein metabolism between these two forms of severe malnutrition. The same diet was initially offered to children with kwashiorkor, all children were studied a short time after admission (in most cases a few hours), and all children were acutely infected. Frozen serum samples from 11 of the children with kwashiorkor were analyzed for CRP, albumin, and total protein. To determine how body composition affected the calculations of protein kinetics, rates of protein synthesis, protein breakdown, and amino acid oxidation were expressed by using ve different denominators: per child, per kilogram using weight at the time of the study, per kilogram using the childs lowest recorded hospital weight, per body mass index (kg/m2), and per centimeter in length. Children treated for kwashiorkor in Malawi typically lost 8% of

Downloaded from www.ajcn.org by guest on December 26, 2011

TABLE 1 Comparison of children with acute infection and marasmus or kwashiorkor1 Marasmus (n = 5M, 3F) Age (mo) Number with each infection Pneumonia Malaria Number HIV positive Time in hospital when studied (d) Number with marasmic kwashiorkor Number who died in the hospital Weight-for-age (% of CGASR standard) Height-for-age (% of CGASR standard) Weight-for-height (% of CGASR standard) Body mass index (kg/m2) Albumin (g/L) Globulin (g/L) CRP (mg/L) 26 152 6 2 6 0.5 10 0 1 55 9 86 6 72 12 13.0 1.4 24 5 40 17 73 79 Kwashiorkor (n = 6M, 7F) 28 21 12 1 4 1.9 2.0 6 3 61 11 88 3 74 10 12.8 1.5 16 63 25 73 83 89

1 CGASR, Coles Growth Assessment Slide Rule (15); CRP, C-reactive protein. Data for children with kwashiorkor were reported previously (9). 2 x SD. 3 Signicantly different from marasmus, P 0.01.

PROTEIN KINETICS IN MARASMUS AND KWASHIORKOR

TABLE 2 Protein kinetics in children with acute infection and marasmus or kwashiorkor1 Marasmus (n = 5M, 3F) Kwashiorkor (n = 6M, 7F) Age (mo) Number with each infection Pneumonia Malaria Time in hospital when studied (d) Weight-for-age (% of CGASR standard) Height-for-age (% of CGASR standard) Weight-for-height (% of CGASR standard) Body mass index (kg/m2) Albumin (g/L) Globulin (g/L) CRP (mg/L) Protein breakdown ( mol leucine kg 1 h 1) ( mol leucine BMI 1 h 1) Protein synthesis ( mol leucine kg 1 h 1) ( mol leucine BMI 1 h 1) Urea appearance ( mol urea kg 1 h 1) ( mol urea BMI 1 h 1) Leucine oxidation ( mol leucine kg 1 h 1) ( mol leucine BMI 1 h 1)

1207

TABLE 3 Comparison of HIV-infected children with kwashiorkor and marasmus during an episode of acute systemic infection1 Marasmus (n = 3M, 3F) 28 162 5 1 0.2 0.4 55 8 86 6 75 8 13.3 1.4 23 5 44 16 85 87 218 63 124 46 210 63 119 45 74 44 42 26 7.9 4.7 4.5 2.8 Kwashiorkor (n = 1M, 3F) 28 12 4 0 0.5 1.0 58 10 86 4 73 8 12.6 0.9 16 3 30 10 114 120 104 473 59 423 100 443 56 393 45 39 30 34 4.8 4.2 3.2 3.6 Downloaded from www.ajcn.org by guest on December 26, 2011

Protein breakdown ( mol leucine kg 1 h 1) 227 59 103 302 ( mol leucine child 1 h 1) 1710 640 930 4602 1 1 ( mol leucine low kg h ) 227 59 112 332 ( mol leucine BMI 1 h 1) 126 42 64 272 1 1 ( mol leucine cm h ) 22.4 4.4 11.0 7.62 Protein synthesis ( mol leucine kg 1 h 1) 216 60 97 302 ( mol leucine child 1 h 1) 1620 613 878 4462 ( mol leucine low kg 1 h 1) 216 60 105 332 ( mol leucine BMI 1 h 1) 122 42 60 272 1 1 ( mol leucine cm h ) 21.6 4.5 10.4 7.82 Urea appearance ( mol urea kg 1 h 1) 102 85 56 36 ( mol urea child 1 h 1) 814 804 486 309 ( mol urea low kg 1 h 1) 102 85 61 36 ( mol urea BMI 1 h 1) 41 23 36 22 ( mol urea cm 1 h 1) 6.3 4.3 7.3 4.1 Leucine oxidation ( mol leucine kg 1 h 1) 10.9 9.1 6.0 3.8 ( mol leucine child 1 h 1) 87 86 52 33 ( mol leucine low kg 1 h 1) 10.9 9.1 6.5 3.9 ( mol leucine BMI 1 h 1) 4.4 2.5 3.9 2.4 ( mol leucine cm 1 h 1) 0.67 0.46 0.78 0.44 1 x SD. Low kg refers to the childs lowest recorded hospital weight. Data for children with kwashiorkor were reported previously (9). 2 Signicantly different from marasmus, P < 0.001.

1 CGASR, Coles Growth Assessment Slide Rule (15). CRP, C-reactive protein. Data for children with kwashiorkor were reported previously (9). 2 x SD. 3 Signicantly different from marasmus, P < 0.05.

not show the capacity to increase in acutely infected children with kwashiorkor (9). It is unknown, however, which proteins are synthesized and degraded at an accelerated rate during marasmus and acute infection. One might expect that the accelerated rates of protein turnover would be associated with a more vigorous acute phase response. We found higher concentrations of globulin in children with marasmus, but did not nd higher concentrations of CRP. This nding is consistent with previous reports (16). Only eight marasmic children were studied, so these findings must be considered preliminary. These eight children may not be representative of all children with marasmus. Additionally, the infectious agent responsible for the pneumonia in each of these children was unknown. The potential limitations of the analytic techniques used when applied to this population have been discussed previously (9). Urea appearance was used as an indirect measure of amino acid oxidation in this study. We believe this to be a valid measure because these children were fed a relatively low protein intake (and were studied during the postabsorptive state) and received broad-spectrum antibiotics, which likely decreased gastrointestinal bacterial flora and subsequent urea recycling. The high SD of urea appearance may be an artifact of one childs measurement, which was > 2 SDs above the mean. Malnutrition and acute infection affect tens of millions of African children, and whole-body protein kinetics are technically very difficult to measure in children from sub-Saharan Africa. This is the first report of protein kinetics comparing marasmus and kwashiorkor and it is unlikely that such data will be reported in large, well-controlled studies.

Despite the dramatic differences in appearance between marasmus and kwashiorkor, whole-body composition in the two conditions is similar (17). In both forms of severe malnutrition noncollagen protein is depleted by 50% (18). Water totals 79% of body composition in marasmus and 82% in kwashiorkor, the primary distinction being that in kwashiorkor water moves from the intracellular space to the extracellular space (19). Bodycomposition differences have confounded comparisons of protein kinetics between well-nourished and chronically undernourished adults (20). Comparisons of metabolic rates of malnourished and well-nourished children have been confounded by the denominator chosen (21). We compared protein kinetics in two forms of severe malnutrition, without a well-nourished control group. The differences in protein kinetics between marasmus and kwashiorkor remained significant whatever denominator was used (Table 2). The magnitude of the differences that we found was so large that there would have needed to be a 33% difference in lean body mass between the two groups to account for the differences in protein turnover. This is unlikely, so differences in body composition between children with kwashiorkor and those with marasmus probably did not confound our results. Infection with HIV is a confounding factor affecting the comparison of these two groups. The six marasmic HIV-infected children described here acquired HIV perinatally or through breast-feeding, and their marasmus may have been a manifestation of HIV infection. Accelerated rates of whole-body protein turnover with progressive immunosuppression was reported in HIV-infected adults without acute systemic infection. This sug-

1208

MANARY ET AL response in children with marasmus, allowing for a better outcome. The consequences of this more normal metabolic response in marasmus need further study.
We thank the Malaria Project of the Wellcome Trust in Blantyre, Malawi, for the use of its facilities; Jan Crowley for analytic assistance; Gary Shackleford for reading the chest X-rays; Sharon Bader for the CRP measurements; and Michael Whyte for the albumin and total protein measurements.

gests that HIV infection alone promotes whole-body protein catabolism (22, 23). In contrast, an earlier report found lower rates of protein turnover in HIV-infected adults without acute infection than in HIV-negative adults (24). There have been no reports of the effects of HIV infection on the rate of whole-body protein turnover in children, and it is unclear whether HIV infection alone alters whole-body protein kinetics. In the present study, the differences in whole-body protein kinetics between marasmus and kwashiorkor remained when only HIV-infected children were considered (Table 3). This suggests that although HIV infection may alter protein kinetics, the form of severe malnutrition does so as well. Millions of children in sub-Saharan Africa are currently infected with HIV and the effects of HIV infection on protein metabolism in malnourished children deserve further study. Tomkins et al (25) determined whole-body protein turnover with an oral [15N]glycine tracer in six children with undernutrition and acute infection (mean percentage weight-for-age: 64%). The rate of whole-body protein synthesis was 10.7 g kg 1 d 1, which was not significantly different from protein turnover in well-nourished children. In the same study, six children with moderate malnutrition (five with kwashiorkor) had significantly lower rates of protein synthesis of 6.3 g kg 1 d 1. With use of the same oral [15N]glycine method, three marasmic children without infection were studied before nutritional therapy. The rate of protein synthesis in these three children was 3.2 g kg 1 d 1 (5). Together, these findings are consistent with ours; wasted children have an increased rate of whole-body protein synthesis and degradation in response to acute infection, but children with kwashiorkor do not. Our children were more malnourished than those studied by Tomkins et al (mean percentage weight-for-age in the present study: 55%) but had a similar accelerated rate of protein synthesis of 8.5 g kg 1 d 1 (assuming that leucine composes 8.1% of body amino acids and the hourly rate that was measured persisted for 24 h). However, the value of this comparison is limited because Tomkins et al combined data from children with kwashiorkor and marasmus, studied children in the fed state, used a different amino acid tracer, and did not study severely malnourished children during acute infection. Wholebody protein kinetics have not been studied in normal children under 7 y of age using the same methods, but older children (712 y) have rates of protein breakdown of 135240 mol leucine kg 1 h 1, within the range we found for marasmic children with acute infection (26, 27). Among the children with marasmus, the mean amount of leucine released into the plasma from whole-body protein breakdown and used in new protein synthesis was 96%. This suggests that large amounts of nitrogen are not wasted, as was found in well-nourished adults during acute infection. Despite their HIV and acute infection these marasmic children continued to conserve amino acids in the postabsorptive state. This preliminary evidence suggests that during acute infection marasmic children have an accelerated rate of whole-body protein turnover, but children with kwashiorkor do not. Perhaps this is indicative of a more appropriate acute phase response, although we do not have direct evidence for this because we did not measure the fractional synthesis rate of any one protein. Children with marasmus and infection have a lower case-fatality rate than children with kwashiorkor and infection (28, 29). Perhaps the accelerated protein kinetics improves the immunologic

REFERENCES
1. McLaren DS. A fresh look at protein-calorie malnutrition. Lancet 1966;2:4858. 2. McLaren DS, Pellett P, Read W. A simple scoring system for classifying the severe forms of protein-calorie malnutrition of early childhood. Lancet 1967;1:5335. 3. Williams CD. Kwashiorkor. Lancet 1935;2:11512. 4. Rao KSJ. Evolution of kwashiorkor and marasmus. Lancet 1974;1:70911. 5. Golden MHN, Waterlow JC, Picou D. Protein turnover, synthesis and breakdown before and after recovery from protein-energy malnutrition. Clin Sci 1977;53:4737. 6. Waterlow JC. Metabolic adaptation to low intakes of energy and protein. Annu Rev Nutr 1986;6:495526. 7. Long CL, Jeevanandam M, Kim BM, Kinney JM. Whole body protein synthesis and catabolism in septic man. Am J Clin Nutr 1977; 30:13404. 8. Reeds PJ, Fjeld CR, Jahoor F. Do the differences between the amino acid compositions of acute-phase and muscle proteins have a bearing on nitrogen loss in traumatic stress? J Nutr 1994;124:90610. 9. Manary MJ, Brewster DR, Broadhead RL, Crowley JR, Fjeld CR, Yarasheski KE. Protein metabolism in children with edematous malnutrition and acute lower respiratory infection. Am J Clin Nutr 1997;65:100510. 10. Berman S, Simoes EAF, Lanata C. Respiratory rate and pneumonia in infancy. Arch Dis Child 1991;66:814. 11. Redd SC, Bloland PB, Kazembe PN, Patrick E, Tembenu R, Campbell CC. Usefulness of clinical case-definitions in guiding therapy for African children with malaria or pneumonia. Lancet 1992;340:11403. 12. Mitton SG, Calder AG, Garlick PJ. Protein turnover rates in sick, premature neonates during the first few days of life. Pediatr Res 1991;30:41822. 13. Ellis KJ, Nichols BL. Body composition. Adv Pediatr 1993; 40:15984. 14. Duggan MB, Milner RDG. Composition of weight gain by Kenyan children during recovery from measles. Hum Nutr Clin Nutr 1986;40:17383. 15. Cole TJ, Donnet ML, Stanfield WP. Weight-for-height indices to assess nutritional statusa new index on a slide-rule. Am J Clin Nutr 1981;34:193543. 16. Doherty JF, Golden MHN, Raynes JG, Griffin GE, McAdam KPWJ. Acute-phase protein response is impaired in severely malnourished children. Clin Sci 1993;84:16975. 17. Hansen JDL, Brinkman GL, Bowie MD. Body composition in protein-calorie malnutrition. S Afr J Med 1965;33:4915. 18. Picou D, Halliday D, Garrow JS. Total body protein, collagen and non-collagen protein in infantile protein malnutrition. Clin Sci 1966;30:34551. 19. Garrow JS, Smith R, Ward EE. Electrolyte metabolism in severe infantile malnutrition. Oxford, United Kingdom: Pergamon Press, 1968. 20. Soares MJ, Piers LS, Shetty PS, Jackson AA, Waterlow JC. Whole body protein turnover in chronically undernourished individuals. Clin Sci 1994;86:4416. 21. Brooke OG, Cocks T. Resting metabolic rate in malnourished babies

Downloaded from www.ajcn.org by guest on December 26, 2011

PROTEIN KINETICS IN MARASMUS AND KWASHIORKOR

in relation to total body potassium. Acta Paediatr Scand 1974;63:81725. 22. Lieberman S, Butterfield G, Harrison D, Hoffman A. Anabolic effects of recombinant insulin-like growth factor-I in cachectic patients with the acquired immunodeficiency syndrome. J Clin Endocrinol Metab 1994;78:40410. 23. Macallan D, McNurlan M, Milne E, Calder A, Garlick P, Griffin G. Whole-body protein turnover from leucine kinetics and the response to nutrition in human immunodeficiency virus infection. Am J Clin Nutr 1995;61:81826. 24. Stein T, Nutinsky C, Condoluci D, Schluter M, Leskiw M. Protein and energy substrate metabolism in AIDS patients. Metabolism 1990;39:87681.

1209

25. Tomkins AM, Garlick PJ, Schofield WN, Waterlow JC. The combined effects of infection and malnutrition on protein metabolism in children. Clin Sci 1983;65:31324. 26. Salman EK, Haymond MW, Bayne E, et al. Protein and energy metabolism in prepubertal children with sickle cell anemia. Pediatr Res 1996;40:3440. 27. Arslanian SA, Kalhan SC. Protein turnover during puberty in normal children. Am J Physiol 1996;270:E7984 28. Friedland IR. Bacteraemia in severely malnourished children. Ann Trop Paediatr 1992;12:43340. 29. McLaren DS, Shirajian E, Loshkajian H, Shadarevian S. Short-term prognosis in protein-calorie malnutrition. Am J Clin Nutr 1969; 22:86370.

Downloaded from www.ajcn.org by guest on December 26, 2011