Reference Values for ... - Clinical Chemistry

0 downloads 0 Views 853KB Size Report
Departments of' Pediatrics,2 Bio Statistics, and3 Clinical Chem- istry, Academic Hospital and ... the time of blood sampling they were free of acute infections, as judged by clinical ..... Nelson textbook of pediatrics. 13th ed. Philadelphia: WB ...
CLIN. CHEM. 34/12, 2444-2447 (1988)

Reference Values for FructosamineConcentrationsin Children’sSera: Influenceof Protein Concentration,Age, and Sex J. D Scheppe MP. Derde,2P. Goubert,3and F. Gorus Fructosamine and protein (total and fractionated) were measured in the serum of 170 normal children, ages two weeks to 15 years. The mean fructosamine concentration was 2.12 mmol/L, 5% lower than the mean value observed for adults. We observed no sex-related difference in fructosamine values, but saw a pronounced age dependency of reference values. For children younger than three years, the mean concentration of fructosamine was 15% lower than in adults, but glycated protein concentrations increased with age, reaching essentially adult values by age six years. Expressing fructosamine concentrations per gram of serum total protein or of albumin weakened the influence of age, but did not eliminate it completely. We report reference intervals for fructosamine concentrations in children’s sera.

Additional Keyphrases: pediatric chemistry related effects

sex- and age-

diabetes

Excessive non-enzymatic glycation of body proteins, a consequence of protracted hyperglycemia, is one of the major biochemical processes believed to be implicated in the chronic complications of diabetes mellitus (1-4). Measurements of the concentrations of glycated hemoglobin and proteins in serum, generated by the same process, can provide us with objective indices of control of hyperglycemia within the preceding three weeks (for serum proteins) to two months (for hemoglobin). Thus such measurements are clinically useful in avoiding long-term complications of diabetes (1-4). To assess fluctuations in concentrations of the abovementioned analytes, numerous assay methods have been proposed, based on principles as diverse as photometry, affinity chromatography, electrophoresis, isoelectric focusing, and various inimunoassays. Unfortunately, all so farwith the possible exception of HPLC-have been hampered by a lack of reproducibility, detracting from their potential to precisely reflect the prevailing mean concentrations of glucose in the patient’s serum (3, 5). A recent colorimetric method, which measures the ketoamine linkages within glycated serum proteins by their ability to reduce nitroblue tetrazolium at alkaline pH, has emerged as an attractive analytical procedure, combining the following advantages: absence of sample pretreatment, excellent reproducibility, ease of automation, small sample requirement, and lack of influence by variant hemoglobulins or altered erythrocyte kinetics (4, 6). This method therefore seems especially indicated for the monitoring of newborns and pregnant women (4). No detailed reference data for children have yet Departments of’ Pediatrics,2 Bio Statistics, and3 Clinical Chemistry, Academic Hospital and Medical School of the “Vrje Universiteit Brussel,” Laarbeeklaan 101, B-lOgO Brussels, Belgium. 4Address correspondence to this author, at the Academic Hospital of the “Vr*je Universiteit Brussel.” Presented in part at the 5th Meeting of the “Soci#{233}t#{233} Francaise de Recherche en P#{233}diatrie,” Toulouse, France, September 1987. Received July 12, 1988; accepted September 8, 1988. 2444

CLINICAL CHEMISTRY, Vol. 34, No. 12,1988

been reported (7), although these might be expected to be age-dependent in view of the positive correlation between the concentrations of serum protein and fructosainine observed in adults and the reported age-dependent changes in protein concentrations and homeostasis in children (8,9). We therefore undertook the present study to establish reference ranges for concentrations of fructosamine in children’s sera and to investigate possible dependency of these ranges on children’s serum protein content, age, and sex.

Materials and Methods Subjects. We studied 170 well-nourished children (114 boys and 56 girls) without personal or familial history of diabetes mellitus. The subjects’ ages ranged between two weeks and 15 years (mean value: six years and eight months). The children were admitted for minor surgical interventions such as circumcision or fracture reposition. At the time of blood sampling they were free of acute infections, as judged by clinical and biological indicators (normal temperature, leukocyte count, and erythrocyte sedimentation rate) and were taking no medication. Further biological selection criteria included having normal concentrations of the following analytes: plasma glucose (fasting or 2 h postprandial), serum proteins (total and fractionated), and serum prealbuinin (transthyretin). Serum samples from 23 healthy blood donors were used to establish the reference interval for adults. Collection of sera. Blood was sampled between 0900 and 1200 hours on the day of admission by venipuncture of the antecubital vein. Samples were collected into commercial plastic tubes (Sarstedt, Haasrode, Belgium), allowed to clot at room temperature, and centrifuged within 1 h at 2000 x g for 15 mm. The sera were stored at -20#{176}C until analyzed (within a week of collection). No icteric or lipemic sera were included. Fructosaniine assay. Fructosaniine in serum was measured with a commercial kit (kit 07-1121-7; Roche Diagnostics, Basel, Switzerland), adapted to a Cobas-Bio (Roche) centrifugal analyzer, according to the instructions outlined in the package insert. The calibrator of the reagent kit was glycated bovine serum albumin, which was assigned a nominal fructosaniine value in terms of desoxymorpholinofructose equivalents by comparison with a primary desoxymorpholinofructose standard. By this methodology, fi-uctosamine concentrations in healthy blood donors (ages 18-60 y) ranged between 1.61 and 268 mmol/L (mean 2.23, SD 0.27, n = 23). The day-to-day CV was 2.1% in this concentration range, even when different batches of reagents were used. For 27 diabetic patients selected without conscious bias, concentrations correlated moderately well with glycated hemoglobin data obtained by isoelectric focusing (10) (r = 0.85, by linear regression), in agreement with previous findings (11, 12). That the correlation is not better is understandable, because the different analytes being measured reflect similar phenomena but over a different time scale (1-4).

Other analytical techniques. Glucose was determined with a specific glucose oxidase method (Boehringer, Mannheim, F.R.G.), adapted to an RA-1000 “random-access” analyzer (Technicon, Dublin, Ireland). The biuret method for total protein (Merck, Darmstadt, F.R.G.) was also used with the HA-bOO. For electrophoresis of serum proteins we used an Olympus Hyte 200 apparatus (Olympus, Hamburg, F.R.GJ with cellulose acetate as carrier and Ponceau Red as the protein stain. This fully automatic instrument directly measures the concentrations of albumin, and of the alpha1-, alpha2-, beta-, and gamma-globulins by densitometric scanrnng of the protein proffles. Prealbumin concentrations were estimated with an ICS rate nephelometer and ‘PAB” reagent kits (both from Beckman Instruments, Brea, CA). The mean concentration of total protein and albumin in the blood donor group (n = 23) amounted to 69.2 g/L (SD 5.0) and 44.3 g/L (SD 4.7), respectively. Statistical analysis. To assess the impact of the subjects’ age, sex, or concentrations of serum proteins on their concentrations of fructosamine in serum, we used analysis of variance or analysis of covariance. The Kolmogorov-Smirnov test was used to assess the gaussian distribution of fructosamine values in the different age groups; Fructosamine, protein, and albumin concentrations measured for the children’s group and the adult blood donors were compared by means of the Mann-Whitney U test. Significance of the differences between the means of various age groups was determined by the Student-Newman-Keuls procedure (13). The relation between concentrations of different analytes in serum and the patients’ ages was evaluated by computing Pearson correlation coefficients. All statistical tests were carried out two-sided at the 0.05 level of significance. Reference limits for fructosamine concentrations in each age group were established according to the IFCC recommendations by computing the 0.025 and 0.975 fractiles and their 90% confidence limits with a parametric method (14). Results As shown in Table 1, the children’s mean fructosamine concentration was 2.12 (SD 0.22) mmol/L, which is significantly lower than the mean value for adults: 2.23 (SD 0.27) mmol/L (P = 0.02; Mann-Whitney U test). By contrast, the mean total protein and albumin concentrations for the juvenile group as a whole (Table 1) were not significantly different from adult values. We saw no significant sexrelated differences in mean fructosamine concentration or in mean age, protein concentration, protein proffle, or fructosa-

Table 1. ConcentratIon of Fructosamlne and Some Other Serum Analytes In 170 ChIldren Age, y

6.7

SD 4.2

Fructosamine, mmol/L Total protein, g/L Prealbumin, g/L Albumin,g/L

2.12

0.22

1.50-2.70

70.7 0.18

4.7

50.0-81.0 0.06-0.32

48.2 1.7

34.0-56.0 1.0-3.0

9.0

3.1 0.5 1.2 0.7 2.4

3.0 4.4 0.25 12.0

0.3 0.4 0.07 3.0

2.4-3.7

Mean

Alpha1-globulin,g/L Alpha2-globulin,g/L Beta-globulin,g/L Gamma-globulin,

6.2 5.9

g/l

0.04

Range 0.0-15.8

3.0-10.0 3.0-8.0

3.0-16.0

Ratios(mmol/g): Fructosamine/total

protein (xl 0_2)

Fructosamine/albumin (xl 0-2) Fructosamine/gamma-globulin

Fructosamine/prealbumin

3.5-5.6 0.13-0.63

6.0-28.0

mine/protein (fraction) ratios in the children’s group. Analysis of covariance excluded the possibility that differences between results for boys and girls in mean concentrations of fructosamine might be masked by differences in age distribution within each sex group. Therefore, in the subsequent data analysis we will not further distinguish between boys and girls. We used Pearson correlation coefficients to probe for possible interdependency of the various analytes assessed in the children studied. As shown in Table 2, there was a significant positive correlation (r values between 0.34 and 055) between fructosamine concentrations on the one hand and age, concentrations of total protein, prealbumin, albumiii, and gamma-globulin on the other. In agreement with previous reports (9), a similar relationship existed between these protein fractions and subjects’ age (r values between 0.42 and 0.52, Table 2). Expressing fructosamine concentrations per unit of total protein, albumin, or ganuna-globulin concentration did reduce-but did not abolish-their agedependency (Table 2), whereas normalization of the fructosamine concentration toward prealbumin concentrations excluded the influence of age (Table 2) at the expense of a grossly broadened reference interval (see Table 1). To define better the time-course of the increase in fructoasinine during childhood, we established a scatter-graph representing assayed serum concentrations vs age at sampling (Figure la), so we could visually select appropriate age classes that could be further compared. By linear regression analysis, the equation of the regression line was computed to be fructosamine (mmoIJL) = 1.89 + 0.03 age (years) (r = 0.66; P