Acta Padiatr &and 80: 938-943, 1991
Effect of Metabolic Control on Serum Protein Concentrations in Diabetes STEPHEN F. KEMP and J. PAUL FRINDIK From the Departments of Pediatrics and Biochemistry, University of Arkansas for Medical Sciences and Arkansas ChildrenS Hospital, Little Rock, AR, USA
ABSTRACT. Kemp, S. F. and Frindik, J. P. (Departments of Pediatrics and Biochemistry, University of Arkansas for Medical Sciences and Arkansas Children’s Hospital, Little Rock, Arkansas, USA). Effect of metabolic control on serum protein concentrations in diabetes. Acta Paediatr Scand 80: 938, 1991. Serum albumin, transferrin, transthyretin (prealbumin), and retinol binding protein concentrations were determined in 74 children with insulin-dependent diabetes mellitus before and after a 10-day camp session during which blood glucose concentrations were controlled. Initial concentrations of albumin and transferrin in the subjects were not different from those in 21 children and adults without diabetes, and did not change during the study period. Transthyretin and retinol binding protein concentrations were lower in subjects with diabetes than in the control population, and increased from 182 ?49 mg/l and 42.5? 13.4 mg/l to 232 271 mg/l and 47.2 2 13.5 mg/l, respectively. We observed correlations between the changes in transferrin, transthyretin, and retinol binding protein. Although reductions in glycated albumin and transferrin indicated improvement in blood glucose control, there was no correlation between changes in the glycated markers and the concentrations of serum transport proteins. Thus, serum protein concentrations were influenced by the metabolic control of diabetes, but did not directly reflect blood glucose. Key words: insulindependent diabetes mellitus, transport proteins, albumin, transthyretin, prealbumin, transferrin, retinol binding protein.
A number of serum proteins have been shown to be sensitive to nutritional state. Gebre-Medhin et al. ( 1) have recently reported that albumin and transthyretin (formerly prealbumin) concentrations are lower in children with insulin dependent diabetes mellitus (IDDM) than in a healthy control population. Marked changes in serum protein appear at the time of diagnosis and improve during the first two years of therapy, although levels do not reach those seen in controls (2). Also, correlations were not observed between hemoglobin A,, fasting plasma glucose, or urinary glucose excretion in this investigation. These reports of changes in serum protein concentrations in children with diabetes prompted us to re-examine data we had collected from a previous study to determine whether a 10-day period of improved glucose control altered serum protein concentrations in children with IDDM. MATERIALS AND METHODS Study participants included 74 children and young adults with IDDM, all attending a summer camp for children with diabetes. Their ages ranged from 8 to 23 years (mean 13 years). The control population comprised 20 healthy nondiabetic children and adults, aged 5 to 35 years (mean 19 years). Subject participation was solicited by informed parental consent and when appropriate, patient assent. These investigations were approved by the University of South Alabama Committee for the Protection of Human Subjects. Study protocol. Each participant performed blood glucose measurements using b G Chemstrips (Boeringer-Mannheim, Indianapolis, IN) four times daily, at one or two day intervals. Each afternoon the records were reviewed by the medical staff, and the insulin dose and diet were ordered for the next day with the aim of keeping blood glucose concentrations between
Serum protein concentrations in diabetes 939
Acta Paediatr Scand 80
4.4 and 8.3 mmol/l(80 and 150 mg/dl). Venous blood specimens were collected in nonanticoagulant-containing glass tubes upon arrival at the camp and after ten days. After clotting at room temperature for 30 min, the blood was promptly centrifuged (5000 x g for 15 min), and the serum separated and immediately frozen at -20°C until analysis. At the time of sample collection, a second specimen for determination of hemoglobin A,, level was obtained. Hemolysate was prepared immediately after collection, the sample stored at 4"C, and hemoglobin A,= determined within 24 hours. Determination of glycated proteins. Hemoglobin A,, was determined by high performance liquid chromatography (Helena Laboratories, Beaumont, TX). For this assay the intraassay coefficient of variation is 1.73 Yo and the interassay coefficient of variation is 0.27 Yo. Glycated serum proteins were separated using boronate-agarose gel (Glyc-Affin, Isolab, Akron, OH) as previously described (3). Albumin concentrations were determined using bromcresyl green (4). Retinol binding protein, transferrin, and transthyretin concentrations were quantitatively estimated by the single radial immunodiffusion technique of Mancini et al. (3,with prepared plates obtained from Calbiochem-Behring Corporation (La Jolla, CA). All analytical samples were assayed in duplicate and the mean concentration was reported as the test result. Statistical methods. Data are presented as the mean and standard deviation (SD). Comparisons of data between subject groups were performed using the Wilk-Shapiro test for normality of distribution, the F-test for equality of variance, and the two-tailed pairwise or unpaired Student t-test as appropriate. Level of significance accepted for all statistical analyses was a=0.05.
RESULTS
Concentrations of serum proteins. Concentrations of albumin, transferrin, transthyretin, and retinol binding protein in both healthy children and in children with IDDM are summarized Table 1. Albumin and transferrin concentrations in children with IDDM at either the pre- or post-treatment evaluation were not significantly different from those in our healthy control population. Following the ten day study period the mean concentration of these proteins was slightly higher than the pretreatment values, although the differences were not statistically significant. Initial concentrations of transthyretin and retinol binding protein were significantly lower in the subjects with IDDM than in the controls. In addition, concentrations of both of these proteins increased significantly during the ten day study period. Correlations between the changes in each of these serum proteins in the subjects with IDDM are summarized in Table 2. A significant linear relationship was Table 1 . Concentrations of selected serum proteins in healthy children and children
with diabetes IDDM Control
Albumin (g/l) Transferrin (g/l) Transthyretin (Prealbumin) (mg/l) Retinol binding protein (mg/l)
N
Concentration
N
Pre-treatment concentration
Post-treatment concentration
21 20
45.023.0 3.1320.73
74 37
44.625.8 2.9520.60
45.8 k6.9 3.0820.79
16
302 2 I30
53
182249"
28
55.6?20.2
37
42.5? 13.4'
232?71b
" Significantly different from control, p=0.0002. Significantly different from pre-treatment concentration, p=O.OOOl
' Significantly different from control, p=0.005.
Significantly different from pre-treatment concentration, p=O.OOl
.
47.22 13.5d
940 S. F. Kemp and J. P. Frindik 25.0
. z
I
13.5 19.2
7.8
Acta Paediatr Scand 80
1
1
0 0
-
a m LT
c -
2.1 -
B
,
0 1
-3.5 -9.2
0
-15.0 -
I
I
I
I
I
Fig. I. Correlation of change in concentration of retinol binding protein with change in concentration of transferrin. Changes in retinol binding protein and
transferrin concentrations were compared in 36 children with IDDM after ten days of intensified control at a summer camp for children with diabetes. Diagonal line, best fit by least squares
observed between the changes in transferrin with transthyretin and retinol binding protein (Fig. l ) , and also between change in transthyretin and retinol binding protein. Indicators of blood glucose control in children with IDDM.The degree of blood glucose control was determined by measurement of hemoglobin Alc at the beginning of the study, and determination of glycated albumin and transferrin. As we have previously shown (3), hemoglobin AIc was elevated (8.1 t0.2%, normals 3 to 6%) in the subjects with IDDM. Glycated albumin in these subjects was 1 6 . 4 t 0.6 O/o (significantly different than in nondiabetic control subjects, 8.7+.0.3%) at the beginning of the study, and decreased to 14.6+-0.5O/oafter 10 days. Likewise, glycated transferrin was 1 1.4 5 0.6 O/o (significantly different from that in nondiabetic control subjects, 3.8+-0.3O/o) at the beginning of the study and decreased to 8.2 +- 0.3 Yo after 10 days. In these same serum specimens there was no correlation between any of the markers of control and the concentrations of albumin, transferrin, transthyretin, or retinol binding protein either before or after the ten day study period.
DISCUSSION Concentrations of serum transport proteins such as albumin, transferrin, transthyretin, and retinol binding protein vary according to nutritional intake and energy Table 2. Spearman’s rho matrix: Changes in serum protein concentrations
A Albumin A Transferrin A Transthyretin
A Transferrin
A Transthyretin
A RBP
rho
P
rho
P
rho
P
0.2 1
0.2 1
0.10 0.015 -
0.76
-
0.26 0.40
0.01
-
0.53 0.57
0.001 0.001
-
Acta Paediatr Scand 80
Serum protein concentrations in diabetes 94 I
balance. Protein synthesis declines in protein-energy malnutrition and other catabolic states, and increases with improvement in nutrition and/or energy utilization (6). Infections, vitamin and mineral deficiencies, liver disease, and hypothyroidism may all lead to decreased synthesis of these transport proteins, while conditions such as severe burns, nephrotic syndrome, and protein-losing enteropathies increase or cause their loss (6). Although IDDM has been associated with greater than normal variation in many analytes (7), diabetes per se has not been associated with profound alterations in the concentration of most serum proteins. Concentrations of several transport proteins, however, have been shown to be altered ( I ) and affected over a long course of treatment (2). We examined serum protein concentrations in a group of children with IDDM over a ten day period of optimized metabolic control. As we have previously demonstrated, short term improvement in glycemic control in these subjects was reflected by a reduction in the degree of glycation of both serum albumin and transferrin (3). Although we did not observe differences in serum albumin and transferrin concentrations between the control and diabetic population at the beginning of the study, serum transthyretin (formerly prealbumin) and retinol binding protein concentrations were significantly lower in the subjects with IDDM (Table 1). These findings differ somewhat from the report of Gebre-Medhin et al. ( I ) in that albumin and transthyretin were decreased in their patients, while transferrin and retinol binding protein concentrations were within the normal range. We did not observe a change in serum albumin concentration following the ten days of improved metabolic control. A small but statistically insignificant increase in the transferrin concentration was observed. Albumin, a routinely used clinical marker for nutritional status, is synthesized in the liver at a daily rate of 120 to 200 mg/kg body weight (8). Because of the large total body pool of albumin (3 to 5 g/kg body weight) and long half-life (14 to 20 days) (8), the serum concentration remains virtually unchanged in the face of short term nutritional variation. The serum concentration of transferrin, a transport protein for Fe", is also reduced in severe protein-energy malnutrition (6). The small change observed in our subjects with IDDM over the ten day study period may reflect transferrin's relatively long halflife of 8 to 9 days (8, 9). Both transthyretin and retinol binding protein increased significantly over ten days in our subjects with diabetes, although not to the concentrations in the control subjects (Table I ) . A similar improvement in concentrations of albumin and transthyretin was demonstrated by Kobbah et al. (2) over two years of therapy in the subjects with IDDM. We also observed significant correlations between increases in transferrin, transthyretin, and retinol binding protein. Since retinol binding protein (half-life 0.5 days) (10) is bound to transthyretin (half-life 2-3 days) (8) and both proteins have a relatively rapid clearance from the blood, a similar response to protein energy changes in both transport protein concentrations would not be unexpected. We did not observe a correlation between concentrations of serum proteins and hemoglobin A,c, glycated albumin, or glycated transferrin; findings consistent with the previous report that serum protein concentrations in IDDM did not correlate with hemoglobin Ale, fasting plasma glucose, or urinary glucose excretion ( I ) . Further, we did not observe a significant correlation between changes in any of the serum proteins and changes in either glycated albumin or transferrin. These data suggest that changes in serum protein concentrations are dependent, in part, upon the metabolic control of diabetes, but do not directly reflect concentrations of blood glucose.
942 S. F. Kemp and J. P. Frindik
Acta Paediatr Scand 80
There are several possible explanations for the increase in serum protein during the ten day study period. The camp diet is carefully supervised; dietary intake is regulated to a greater extent than is usual at home. Consequently, the ten day study period represents improved nutritional status for some children. Golden ( 1 1) has summarized a number of studies involving various forms of protein energy restriction and refeeding. There appears to be no relationship between retinol binding protein levels and protein intake or body protein content in malnuourished children. In obese adults protein restriction for ten days does not alter transthyretin levels provided that energy intake (carbohydrate) is maintained. Alternatively, the addition of energy restriction is associated with a rapid decrease in transthyretin levels. A slight decrease in retinol binding protein levels with severe protein restriction (20 g protein per day), and also with energy restriction on a high protein diet does occur ( 1 2). Thus, transthyretin and retinol binding protein concentrations respond rapidly to energy changes, but not dietary protein changes. A second possible explanation resides with the optimization of insulin therapy. The children in this study were all receiving once or twice daily injections of mixtures of intermediate acting and short acting insulins. Such a regimen can prevent life-threatening ketoacidosis and is adequate in most children to provide reasonable blood sugar control on a daily basis. It does, however, produce relatively high insulin concentrations in the peripheral circulation with low amounts of insulin available to the liver, the primary site of transport protein synthesis. GebreMedhin et al. ( 1 ) have speculated that the deficiencies of various transport proteins in diabetes might be the result of inadequate insulinization of the liver as well as the marked daily fluctuations of relative hyper- and hypo-insulinemia. Insulin doses were carefully adjusted during the ten day study period in our subjects with IDDM, and thus, total insulin concentrations may have been more appropriate during that time. These changes in insulin administration alone would not be expected to improve the insulinization of the liver, however. A third possible explanation is an increase in exercise associated with the wide variety of daily scheduled activites during the study period. In the camp environment almost all the children were more physically active than at home. That exercise potentiates the effect of insulin has been appreciated for many years (1 3). Perhaps the adjustment in insulin doses coupled with increased insulin sensitivity due to exercise resulted in improvement in insulinization of the liver sufficient to cause increases in concentrations of transport proteins. Just as a rapid decrease in transthyretin and retinol binding protein levels indicates loss of available energy, an increase in transthyretin and retinol binding protein concentrations in our children at camp indicates improved fuel utilization. Such improved fuel use is reflected not only by improved blood sugar control, but also by increases in serum transport protein concentrations. This increase reflects increased hepatic protein synthesis and probably reflects improved insulinization of the liver. In addition to increasing our understanding of the metabolic changes in diabetes, these observations suggest that serum concentrations of transport proteins, especially transthyretin and retinol binding protein, may be clinically useful as indicators of fuel utilization in patients with diabetes. ACKNOWLEDGEMENTS We thank Helena Laboratories for determination of the hemoglobin A,, values and Isolab, Inc. for providing some of the materials used in separation of glycated and nonglycated proteins.
Acta Paediatr Scand 80
Serum protein concentrations in diabetes 943
REFERENCES I . Gebre-Medhin M, Ewald U, Tuvemo T. Reduced serum proteins in diabetic children on a twice-daily insulin schedule. Acta Paediatr Scand 1985; 74: 961-65. 2. Kobbah AM, Hellsing K, Tuvemo T. Early changes of some serum proteins and metals in diabetic children. Acta Paediatr Scand 1988; 77: 734-40. 3. Kemp SF, Creech RH, Horn TR. Glycosylated albumin and transferrin: short term markers of blood glucose control. J Pediatr 1984; 105: 394-98. 4. Doumas BT, Biggs HG. Determination of serum albumin. In: Copper GA, ed. Standard methods of clinical chemistry, vol. 7. New York: Academic Press, 1972: 175. 5 . Mancini G , Carbonara AO, Heremans JF. Immunochemical quantitation of antigens by singel radial immunodifussion. Int J Immunochem 1965; 2: 235. 6. Merritt RJ, Blackburn GL. Nutritional assessment and metabolic response to illness of the hospitalized child. In: Suskind RM, ed. Textbook of pediatric nutrition. New York: Raven Press, 1981 : 285-307. 7. Holzel WGE. Intra-individual variation of some analytes in serum of patients with insulin-dependent diabetes mellitus. Clin Chem 1987; 33: 57-61. 8. Schultz HE, Heremans JF. Turnover of the plasma proteins. In: Schultz HE, Heremans JF, eds. Molecular biology of the human proteins, vol. 1. New York: Elsevier, 1966: 450. 9. Awai M, Brown EB. Studies of the metabolism of 1311-labeled human transferrin. J Lab Clin Med 1963; 61: 363-96. 10. lngenbleek Y, Van Den Schreik HG, De Nayer P, Devisscher M. Albumin, transferrin and the thyroxine-binding prealbuminhetinol-binding protein (TBPA-RBP) complex in assessment of malnutrition. Clin Chim Acta 1975; 63: 61-67. 1 1 . Golden MHN. Transport proteins as indices of protein status. Am J Clin Nutr 1982; 35: 1 159-65. 12. Shetty PS, Watrasiewicz KE, Jung RT, James WPT. Rapid-turnover transport proteins: an index of subclinical protein-energy malnutrition. Lancet 1979; 11: 230-32. 13. Lawrence RD. The effects of exercise on insulin action in diabetes. Br Med J 1926; I : 648-52.
Submitted Jan. 2, 1990. Accepted Nov. 12, 1990 (S. F. K.) Department of Pediatric Endocrinology and Diabetes Arkansas Children’s Hospital 800 Marshall St. Little Rock, AR 72202-3591 USA
Effect of Metabolic Control on Serum Protein Concentrations in Diabetes STEPHEN F. KEMP and J. PAUL FRINDIK From the Departments of Pediatrics and Biochemistry, University of Arkansas for Medical Sciences and Arkansas ChildrenS Hospital, Little Rock, AR, USA
ABSTRACT. Kemp, S. F. and Frindik, J. P. (Departments of Pediatrics and Biochemistry, University of Arkansas for Medical Sciences and Arkansas Children’s Hospital, Little Rock, Arkansas, USA). Effect of metabolic control on serum protein concentrations in diabetes. Acta Paediatr Scand 80: 938, 1991. Serum albumin, transferrin, transthyretin (prealbumin), and retinol binding protein concentrations were determined in 74 children with insulin-dependent diabetes mellitus before and after a 10-day camp session during which blood glucose concentrations were controlled. Initial concentrations of albumin and transferrin in the subjects were not different from those in 21 children and adults without diabetes, and did not change during the study period. Transthyretin and retinol binding protein concentrations were lower in subjects with diabetes than in the control population, and increased from 182 ?49 mg/l and 42.5? 13.4 mg/l to 232 271 mg/l and 47.2 2 13.5 mg/l, respectively. We observed correlations between the changes in transferrin, transthyretin, and retinol binding protein. Although reductions in glycated albumin and transferrin indicated improvement in blood glucose control, there was no correlation between changes in the glycated markers and the concentrations of serum transport proteins. Thus, serum protein concentrations were influenced by the metabolic control of diabetes, but did not directly reflect blood glucose. Key words: insulindependent diabetes mellitus, transport proteins, albumin, transthyretin, prealbumin, transferrin, retinol binding protein.
A number of serum proteins have been shown to be sensitive to nutritional state. Gebre-Medhin et al. ( 1) have recently reported that albumin and transthyretin (formerly prealbumin) concentrations are lower in children with insulin dependent diabetes mellitus (IDDM) than in a healthy control population. Marked changes in serum protein appear at the time of diagnosis and improve during the first two years of therapy, although levels do not reach those seen in controls (2). Also, correlations were not observed between hemoglobin A,, fasting plasma glucose, or urinary glucose excretion in this investigation. These reports of changes in serum protein concentrations in children with diabetes prompted us to re-examine data we had collected from a previous study to determine whether a 10-day period of improved glucose control altered serum protein concentrations in children with IDDM. MATERIALS AND METHODS Study participants included 74 children and young adults with IDDM, all attending a summer camp for children with diabetes. Their ages ranged from 8 to 23 years (mean 13 years). The control population comprised 20 healthy nondiabetic children and adults, aged 5 to 35 years (mean 19 years). Subject participation was solicited by informed parental consent and when appropriate, patient assent. These investigations were approved by the University of South Alabama Committee for the Protection of Human Subjects. Study protocol. Each participant performed blood glucose measurements using b G Chemstrips (Boeringer-Mannheim, Indianapolis, IN) four times daily, at one or two day intervals. Each afternoon the records were reviewed by the medical staff, and the insulin dose and diet were ordered for the next day with the aim of keeping blood glucose concentrations between
Serum protein concentrations in diabetes 939
Acta Paediatr Scand 80
4.4 and 8.3 mmol/l(80 and 150 mg/dl). Venous blood specimens were collected in nonanticoagulant-containing glass tubes upon arrival at the camp and after ten days. After clotting at room temperature for 30 min, the blood was promptly centrifuged (5000 x g for 15 min), and the serum separated and immediately frozen at -20°C until analysis. At the time of sample collection, a second specimen for determination of hemoglobin A,, level was obtained. Hemolysate was prepared immediately after collection, the sample stored at 4"C, and hemoglobin A,= determined within 24 hours. Determination of glycated proteins. Hemoglobin A,, was determined by high performance liquid chromatography (Helena Laboratories, Beaumont, TX). For this assay the intraassay coefficient of variation is 1.73 Yo and the interassay coefficient of variation is 0.27 Yo. Glycated serum proteins were separated using boronate-agarose gel (Glyc-Affin, Isolab, Akron, OH) as previously described (3). Albumin concentrations were determined using bromcresyl green (4). Retinol binding protein, transferrin, and transthyretin concentrations were quantitatively estimated by the single radial immunodiffusion technique of Mancini et al. (3,with prepared plates obtained from Calbiochem-Behring Corporation (La Jolla, CA). All analytical samples were assayed in duplicate and the mean concentration was reported as the test result. Statistical methods. Data are presented as the mean and standard deviation (SD). Comparisons of data between subject groups were performed using the Wilk-Shapiro test for normality of distribution, the F-test for equality of variance, and the two-tailed pairwise or unpaired Student t-test as appropriate. Level of significance accepted for all statistical analyses was a=0.05.
RESULTS
Concentrations of serum proteins. Concentrations of albumin, transferrin, transthyretin, and retinol binding protein in both healthy children and in children with IDDM are summarized Table 1. Albumin and transferrin concentrations in children with IDDM at either the pre- or post-treatment evaluation were not significantly different from those in our healthy control population. Following the ten day study period the mean concentration of these proteins was slightly higher than the pretreatment values, although the differences were not statistically significant. Initial concentrations of transthyretin and retinol binding protein were significantly lower in the subjects with IDDM than in the controls. In addition, concentrations of both of these proteins increased significantly during the ten day study period. Correlations between the changes in each of these serum proteins in the subjects with IDDM are summarized in Table 2. A significant linear relationship was Table 1 . Concentrations of selected serum proteins in healthy children and children
with diabetes IDDM Control
Albumin (g/l) Transferrin (g/l) Transthyretin (Prealbumin) (mg/l) Retinol binding protein (mg/l)
N
Concentration
N
Pre-treatment concentration
Post-treatment concentration
21 20
45.023.0 3.1320.73
74 37
44.625.8 2.9520.60
45.8 k6.9 3.0820.79
16
302 2 I30
53
182249"
28
55.6?20.2
37
42.5? 13.4'
232?71b
" Significantly different from control, p=0.0002. Significantly different from pre-treatment concentration, p=O.OOOl
' Significantly different from control, p=0.005.
Significantly different from pre-treatment concentration, p=O.OOl
.
47.22 13.5d
940 S. F. Kemp and J. P. Frindik 25.0
. z
I
13.5 19.2
7.8
Acta Paediatr Scand 80
1
1
0 0
-
a m LT
c -
2.1 -
B
,
0 1
-3.5 -9.2
0
-15.0 -
I
I
I
I
I
Fig. I. Correlation of change in concentration of retinol binding protein with change in concentration of transferrin. Changes in retinol binding protein and
transferrin concentrations were compared in 36 children with IDDM after ten days of intensified control at a summer camp for children with diabetes. Diagonal line, best fit by least squares
observed between the changes in transferrin with transthyretin and retinol binding protein (Fig. l ) , and also between change in transthyretin and retinol binding protein. Indicators of blood glucose control in children with IDDM.The degree of blood glucose control was determined by measurement of hemoglobin Alc at the beginning of the study, and determination of glycated albumin and transferrin. As we have previously shown (3), hemoglobin AIc was elevated (8.1 t0.2%, normals 3 to 6%) in the subjects with IDDM. Glycated albumin in these subjects was 1 6 . 4 t 0.6 O/o (significantly different than in nondiabetic control subjects, 8.7+.0.3%) at the beginning of the study, and decreased to 14.6+-0.5O/oafter 10 days. Likewise, glycated transferrin was 1 1.4 5 0.6 O/o (significantly different from that in nondiabetic control subjects, 3.8+-0.3O/o) at the beginning of the study and decreased to 8.2 +- 0.3 Yo after 10 days. In these same serum specimens there was no correlation between any of the markers of control and the concentrations of albumin, transferrin, transthyretin, or retinol binding protein either before or after the ten day study period.
DISCUSSION Concentrations of serum transport proteins such as albumin, transferrin, transthyretin, and retinol binding protein vary according to nutritional intake and energy Table 2. Spearman’s rho matrix: Changes in serum protein concentrations
A Albumin A Transferrin A Transthyretin
A Transferrin
A Transthyretin
A RBP
rho
P
rho
P
rho
P
0.2 1
0.2 1
0.10 0.015 -
0.76
-
0.26 0.40
0.01
-
0.53 0.57
0.001 0.001
-
Acta Paediatr Scand 80
Serum protein concentrations in diabetes 94 I
balance. Protein synthesis declines in protein-energy malnutrition and other catabolic states, and increases with improvement in nutrition and/or energy utilization (6). Infections, vitamin and mineral deficiencies, liver disease, and hypothyroidism may all lead to decreased synthesis of these transport proteins, while conditions such as severe burns, nephrotic syndrome, and protein-losing enteropathies increase or cause their loss (6). Although IDDM has been associated with greater than normal variation in many analytes (7), diabetes per se has not been associated with profound alterations in the concentration of most serum proteins. Concentrations of several transport proteins, however, have been shown to be altered ( I ) and affected over a long course of treatment (2). We examined serum protein concentrations in a group of children with IDDM over a ten day period of optimized metabolic control. As we have previously demonstrated, short term improvement in glycemic control in these subjects was reflected by a reduction in the degree of glycation of both serum albumin and transferrin (3). Although we did not observe differences in serum albumin and transferrin concentrations between the control and diabetic population at the beginning of the study, serum transthyretin (formerly prealbumin) and retinol binding protein concentrations were significantly lower in the subjects with IDDM (Table 1). These findings differ somewhat from the report of Gebre-Medhin et al. ( I ) in that albumin and transthyretin were decreased in their patients, while transferrin and retinol binding protein concentrations were within the normal range. We did not observe a change in serum albumin concentration following the ten days of improved metabolic control. A small but statistically insignificant increase in the transferrin concentration was observed. Albumin, a routinely used clinical marker for nutritional status, is synthesized in the liver at a daily rate of 120 to 200 mg/kg body weight (8). Because of the large total body pool of albumin (3 to 5 g/kg body weight) and long half-life (14 to 20 days) (8), the serum concentration remains virtually unchanged in the face of short term nutritional variation. The serum concentration of transferrin, a transport protein for Fe", is also reduced in severe protein-energy malnutrition (6). The small change observed in our subjects with IDDM over the ten day study period may reflect transferrin's relatively long halflife of 8 to 9 days (8, 9). Both transthyretin and retinol binding protein increased significantly over ten days in our subjects with diabetes, although not to the concentrations in the control subjects (Table I ) . A similar improvement in concentrations of albumin and transthyretin was demonstrated by Kobbah et al. (2) over two years of therapy in the subjects with IDDM. We also observed significant correlations between increases in transferrin, transthyretin, and retinol binding protein. Since retinol binding protein (half-life 0.5 days) (10) is bound to transthyretin (half-life 2-3 days) (8) and both proteins have a relatively rapid clearance from the blood, a similar response to protein energy changes in both transport protein concentrations would not be unexpected. We did not observe a correlation between concentrations of serum proteins and hemoglobin A,c, glycated albumin, or glycated transferrin; findings consistent with the previous report that serum protein concentrations in IDDM did not correlate with hemoglobin Ale, fasting plasma glucose, or urinary glucose excretion ( I ) . Further, we did not observe a significant correlation between changes in any of the serum proteins and changes in either glycated albumin or transferrin. These data suggest that changes in serum protein concentrations are dependent, in part, upon the metabolic control of diabetes, but do not directly reflect concentrations of blood glucose.
942 S. F. Kemp and J. P. Frindik
Acta Paediatr Scand 80
There are several possible explanations for the increase in serum protein during the ten day study period. The camp diet is carefully supervised; dietary intake is regulated to a greater extent than is usual at home. Consequently, the ten day study period represents improved nutritional status for some children. Golden ( 1 1) has summarized a number of studies involving various forms of protein energy restriction and refeeding. There appears to be no relationship between retinol binding protein levels and protein intake or body protein content in malnuourished children. In obese adults protein restriction for ten days does not alter transthyretin levels provided that energy intake (carbohydrate) is maintained. Alternatively, the addition of energy restriction is associated with a rapid decrease in transthyretin levels. A slight decrease in retinol binding protein levels with severe protein restriction (20 g protein per day), and also with energy restriction on a high protein diet does occur ( 1 2). Thus, transthyretin and retinol binding protein concentrations respond rapidly to energy changes, but not dietary protein changes. A second possible explanation resides with the optimization of insulin therapy. The children in this study were all receiving once or twice daily injections of mixtures of intermediate acting and short acting insulins. Such a regimen can prevent life-threatening ketoacidosis and is adequate in most children to provide reasonable blood sugar control on a daily basis. It does, however, produce relatively high insulin concentrations in the peripheral circulation with low amounts of insulin available to the liver, the primary site of transport protein synthesis. GebreMedhin et al. ( 1 ) have speculated that the deficiencies of various transport proteins in diabetes might be the result of inadequate insulinization of the liver as well as the marked daily fluctuations of relative hyper- and hypo-insulinemia. Insulin doses were carefully adjusted during the ten day study period in our subjects with IDDM, and thus, total insulin concentrations may have been more appropriate during that time. These changes in insulin administration alone would not be expected to improve the insulinization of the liver, however. A third possible explanation is an increase in exercise associated with the wide variety of daily scheduled activites during the study period. In the camp environment almost all the children were more physically active than at home. That exercise potentiates the effect of insulin has been appreciated for many years (1 3). Perhaps the adjustment in insulin doses coupled with increased insulin sensitivity due to exercise resulted in improvement in insulinization of the liver sufficient to cause increases in concentrations of transport proteins. Just as a rapid decrease in transthyretin and retinol binding protein levels indicates loss of available energy, an increase in transthyretin and retinol binding protein concentrations in our children at camp indicates improved fuel utilization. Such improved fuel use is reflected not only by improved blood sugar control, but also by increases in serum transport protein concentrations. This increase reflects increased hepatic protein synthesis and probably reflects improved insulinization of the liver. In addition to increasing our understanding of the metabolic changes in diabetes, these observations suggest that serum concentrations of transport proteins, especially transthyretin and retinol binding protein, may be clinically useful as indicators of fuel utilization in patients with diabetes. ACKNOWLEDGEMENTS We thank Helena Laboratories for determination of the hemoglobin A,, values and Isolab, Inc. for providing some of the materials used in separation of glycated and nonglycated proteins.
Acta Paediatr Scand 80
Serum protein concentrations in diabetes 943
REFERENCES I . Gebre-Medhin M, Ewald U, Tuvemo T. Reduced serum proteins in diabetic children on a twice-daily insulin schedule. Acta Paediatr Scand 1985; 74: 961-65. 2. Kobbah AM, Hellsing K, Tuvemo T. Early changes of some serum proteins and metals in diabetic children. Acta Paediatr Scand 1988; 77: 734-40. 3. Kemp SF, Creech RH, Horn TR. Glycosylated albumin and transferrin: short term markers of blood glucose control. J Pediatr 1984; 105: 394-98. 4. Doumas BT, Biggs HG. Determination of serum albumin. In: Copper GA, ed. Standard methods of clinical chemistry, vol. 7. New York: Academic Press, 1972: 175. 5 . Mancini G , Carbonara AO, Heremans JF. Immunochemical quantitation of antigens by singel radial immunodifussion. Int J Immunochem 1965; 2: 235. 6. Merritt RJ, Blackburn GL. Nutritional assessment and metabolic response to illness of the hospitalized child. In: Suskind RM, ed. Textbook of pediatric nutrition. New York: Raven Press, 1981 : 285-307. 7. Holzel WGE. Intra-individual variation of some analytes in serum of patients with insulin-dependent diabetes mellitus. Clin Chem 1987; 33: 57-61. 8. Schultz HE, Heremans JF. Turnover of the plasma proteins. In: Schultz HE, Heremans JF, eds. Molecular biology of the human proteins, vol. 1. New York: Elsevier, 1966: 450. 9. Awai M, Brown EB. Studies of the metabolism of 1311-labeled human transferrin. J Lab Clin Med 1963; 61: 363-96. 10. lngenbleek Y, Van Den Schreik HG, De Nayer P, Devisscher M. Albumin, transferrin and the thyroxine-binding prealbuminhetinol-binding protein (TBPA-RBP) complex in assessment of malnutrition. Clin Chim Acta 1975; 63: 61-67. 1 1 . Golden MHN. Transport proteins as indices of protein status. Am J Clin Nutr 1982; 35: 1 159-65. 12. Shetty PS, Watrasiewicz KE, Jung RT, James WPT. Rapid-turnover transport proteins: an index of subclinical protein-energy malnutrition. Lancet 1979; 11: 230-32. 13. Lawrence RD. The effects of exercise on insulin action in diabetes. Br Med J 1926; I : 648-52.
Submitted Jan. 2, 1990. Accepted Nov. 12, 1990 (S. F. K.) Department of Pediatric Endocrinology and Diabetes Arkansas Children’s Hospital 800 Marshall St. Little Rock, AR 72202-3591 USA
