Serum transferrin and ferritin in pubertal boys: relations to body growth, pubertal stage, erythropoiesis, and iron deficiency13

Transcription

1 Serum transferrin and ferritin in pubertal boys: relations to body growth, pubertal stage, erythropoiesis, and iron deficiency13 Raija Anttila and Martti A Siimes ABSTRACT We studied serum transferrin and ferritin concentrations in relation to individual body growth, stage of puberty, blood hemoglobin, and red blood cell iron (RBCI) in 6 prepubertal or early pubertal boys at 3-mo intervals for 18 mo. One-third of the boys had increased serum transferrin concentrations and almost all had decreased ferritin concentrations during the followup. No change in mean transfemn was observed but the individual 18-mo increments in transferrin correlated positively with the increments in hemoglobin (r =.55, P <.1) and in estimated RBCI (r =.31, P.2). Serum transferrin remained stable at different genital stages, but ferritin was lower in the pubertal than in the prepubertal boys. Transferrin concentrations at 18 mo correlated positively with the preceding weight velocities. The rise in transferrin did not lead to an increase in iron-deficiency anemia. In contrast, transferrmn rose in boys whose hemoglobin increased. In pubertal boys with relatively ample iron status, serum transferrin may be an indicator of increased availability of iron for erythropoiesis. The declining ferritin concentration indicates that part of the extra iron is mobilized through redistribution from stores to red blood cell mass and is generally associated with greatly increasing absorption. Thus, the pubertal changes in transferrmn and ferritin are not necessarily indications of iron deficiency. Am J Clin Nutr l996;63; KEY WORDS Transferrin, ferritin, erythropoiesis, iron deficiency, puberty, boys INTRODUCTION Transferrins are glycoproteins that bind iron reversibly. They transport iron between sites of absorption, storage, and utilization. Bone marrow has the most active need for iron for erythropoiesis, which is the most important determinant of the rate of plasma iron turnover. The concentration of transferrin in serum increases in iron-deficiency anemia and during pregnancy. Transferrin decreases with iron overload or lack of protein. Transferrin acquires iron by three systems: intestinal uptake, recycling of iron released in the turnover of ironcontaining enzymes and proteins, especially hemoglobin, and mobilization of iron stores (1). The concentration of ferritin in serum gives a quantitative measure of the amount of stored iron in normal subjects and it is low when there is an iron deficiency and high when there is an iron overload (2, 3). The amount of iron in body stores is the main factor controlling the absorption of iron from the gastrointestinal tract. There is also some evidence that the rate of erythropoiesis participates in the regulation of iron absorption (4). Male puberty is an interesting time in many respects. Rapid growth of the body and muscles at the end of puberty requires a great deal of iron, and the achievement of adult testosterone concentrations is associated with an increase in erythropoiesis and in the hemoglobin concentration (5). Thus, iron status is likely to change and iron deficiency may develop and it is difficult to decide what changes are physiologic reflections of growth and accelerated hematopoiesis and what changes reflect the development of genuine iron deficiency. These changes may follow one another within a few years and increase the individual variability in iron needs. We studied serum transferrin and ferritin concentrations in relation to individual body growth, stage of puberty, and blood hemoglobin and red blood cell iron (RBCI) concentrations in 6 healthy prepubertal or early pubertal boys at 3-mo intervals for 18 mo. In this setting, serum transferrin may be associated with a physiologic increase in erythropoiesis and serum ferritin may reflect redistribution of iron from stores to red blood cell mass, although both are related to the degree of iron deficiency. SUBJECTS AND METHODS Subjects An opportunity to participate in the study was offered to 84 boys of the same secondary school level in three schools near the Children s Hospital, Helsinki. The study was approved by the parents, the school authorities, those responsible for school health care, and the Ethics Committee of the Children s Hospital. Written information about the study was sent to the families. Of the 84 boys, 62 agreed to participate. Two boys decided not to continue, which left 6 boys for the study. I From the Children s Hospital, Helsinki, University Central Hospital, Helsinki. 2 Supported by grants from the Foundation for Pediatric Research, Helsinki. 3 Reprints not available. Address correspondence to R Anttila, Chilthen s Hospital, University of Helsinki, FIN-29 Helsinki, Finland. Received January 12, Accepted for publication September 21, Downloaded from by guest on July 14, 211 Am J Clin Nutr 1996;63: Printed in USA American Society for Clinical Nutrition 179

2 18 ANT11LA AND SIIMES Medical histories and physical examinations indicated that all the boys were healthy except for three, who used seasonal medication for asthma. Epilepsy was diagnosed in one boy during the study and he started taking oxicarbamazepine. These three boys were included in the study. At the beginning of the study the mean age of the boys was ±.5 y. Of the 6 boys, 35 had mean testicular volumes <2. ml and were considered prepubertal. The other 25 boys had testicular volumes ranging from 2. to 6.8 ml. During the 1 8-mo study all boys were studied seven times at 3-mo intervals. Three boys were considered to be iron-deficient during the first 6 mo of the study and started taking iron medication, one after the first visit and two at 6 mo. The first boy had an initial hemoglobin value of 124 g/l, a mean corpuscular volume (MCV) of 86 if., a serum iron concentration of 9.3 mol/l, a serum transferrin concentration of 3. g/l, transferrin saturation of 1 2%, and a serum ferritin concentration of S.tg/L. At 6 mo in one boy, the hemoglobin concentration had changed from 129 to 12 g/l, the MCV from 86 to 85 fl, the serum iron concentration from to 1.5 amol/l, the serum transferrin concentration from 3. 1 to 4.5 gil, transfemn saturation from 15% to 9%, and the serum ferritin concentration from 21 to 23 p.g/l. In another boy the hemoglobin value had changed from 139 to 133 g/l, the MCV from 81 to 8 fl, the serum iron concentration from 33.7 to 1.8.amol/L, the serum transferrin concentration from 3.5 to 3.9 g/l, and transferrin saturation from 37% to 1 1%; the serum ferritin concentration remained stable at 18.ag/L. All three of these boys were included in the analyses. Subsequently, as the study advanced, we realized how difficult it was to assess the need for iron medication on the basis of these indexes. Thus, after 6 mo into the study, no subjects started taking iron medication on the basis of these clinical indicators. Methods At each examination, body weight was measured on the same platform scale, with the boys wearing small shorts only. Height was measured on a Harpenden scale (British Indicators, Ltd, London) with a precision of.1 cm. Genital and pubic-hair stages were assessed according to the method of Tanner (6). The length and width of the testicles were also measured with a ruler to the nearest millimeter while the subjects were in a lying position. Testicular volume was calculated by using the following formula:.52 X longitudinal axis (in cm) X transverse axis (in cm2) (7). Mean testicular volumes were used for analyses. We estimated changes in muscle mass by determining the greatest diameter of the right quadriceps muscle using an ultrasound method (8). Venous blood samples were drawn between 83 and 14. Hemoglobin concentrations were measured with a Coulter Counter T 89 (Coulter Electronics, Hialeah, FL). Serum transferrin concentrations were measured by an immunoturbidimetric method with a Hitachi analyzer (Tokyo), with antihumantransferrin from Orion Diagnostica (Espoo, Finland), and with an IMPRO calibrator (Labquality, Helsinki) calibrated with US National Reference Preparation for Human Specific Serum Proteins. The cv at a concentration of g/l (n = 36) was 5.4% and at 2.75 g/l (n = 216) was 4.6%. Serum ferritin concentrations were determined by a two-site chemiluminometric immunoassay with an automatic analyzer (ACS: l8;ciba Corning, Halstead, United Kingdom). The day-to-day variation of ferritin concentrations calculated from control samples was 5.9%. The circulating red blood cell mass was calculated as the hematocrit value X blood volume (64 mllkg) x weight. RBCI was calculated as red blood cell mass X 1.1 (9). Statistics Student s unpaired and paired t tests, analysis of variance (ANOVA), and simple-regression analysis were used for statistical analyses. Values are expressed as means ± SEMs. Transfemn and ferritin were also analyzed by multiple-regression analysis with height, weight, testicular volume, height velocity, and weight velocity as confounding factors. Because of skewness, we used the logarithmic transformation of the serum ferritin values before the analyses. RESULTS The mean values for the boys body height and weight, testicular volume, and quadriceps muscle thickness increased (Table 1). As the pubertal development of the boys progressed during the 18-mo follow-up, the number of prepubertai boys Downloaded from by guest on July 14, 211 TABLE 1 Growth and iron indexes in 6 healthy boys at 6-mo intervals, by age Omo 6mo l2mo l8mo ll.7y l2.2y l2.6y l3.2y Body height (cm) ± ± ± ± 1.12 Body weight (kg) 4.1 ± ± ± ± 1.12 Testicular volume (ml) 2.4 ± ± ± ±.62 Quadriceps muscle (cm) 2.9 ± ± ± ±.72 Hemoglobin (g/l) 13 ± ± ± ±.92 MCV (fl) 85 ±.4 85 ±.4 84 ±.4 84 ±.42 Serum iron (mol/l) 16.2 ± ± ± ± 73 Serum transferrmn (g/l) 2.9 ± ± ± ± #{216}#{216}54 Transferrin saturation (%) 22 ±.9 18 ±.9 21 ±.9 23 ± 1.12 Serum femtin (Lg/L) 35 ± ± ± ± 1.42 Estimated RBCI (mg) 111 ± ± ± ± 362 i ± SEM. MCV, mean corpuscular volume; RBCI, red blood cell iron. 2-4 Consecutive analyses by ANOVA: 2 p <.1, - p <.1, p < o.os.

3 TRANSFERRIN AND FERRITIN IN PUBERTAL BOYS 181 decreased from 35 to 4 (Table 2). We observed a small but significant elevation in the mean hemoglobin value from 13 to 134 g/l. When values were grouped according to pubertal stage, the mean hemoglobin value of the most advanced 12 boys rose from 13 1 to 1 38 g/l. The mean serum transferrin, iron, transferrin iron saturation, and MCV values in the whole study group remained unchanged. The values, including serum transferrin, at different genital stages also remained stable. The mean ferritin concentration of all the boys decreased from 35 to 22 j.ag/l during the study. The mean ferritin concentration was higher in the prepubertal boys than in the pubertal boys. Five boys had iron deficiency at 18 mo when only the cutoff point for ferritin (1 j.agfl) was used. When two cutoff points were used to determine iron deficiency-a serum transferrin saturation value < 16% and a serum ferritin concentration <1 j.ag/l-zero boys were iron-deficient. When a serum transferrin saturation value < 16% and a serum ferritin concentration < 12,i.g/L were used as the cutoff points, only 2 (3.3%) of the entire group of 6 boys were iron-deficient. Although we observed no change in mean serum transferrin during the study, the individual 18-mo increments in transferrin varied from -.4 to +.4 gil. All boys with an increase in transferrin also had increased hemoglobin concentrations. In those boys whose hemoglobin decreased, serum transferrin was unchanged or decreased (Figure 1; Table 3). Furthermore, we found a positive correlation between the increments in transferrin and hemoglobin (r =.55, P <.1) and between the increments in transferrmn and estimated RBCI (r =.31, P.2) (Figure 1). The relation between increments in transferrin and hemoglobin was independent of the changes in body height, weight, testicular volume, transferrin saturation, and MCV (Table 3). Serum transferrin correlated positively with body weight at 18 mo (r =.3, P <.5) and with preceding weight velocity (r.44, P <.1; Table 4). Transferrin was also analyzed by multiple-regression analysis with body height, body weight, and testicular volume at 18 mo, and with the preceding height and weight velocities as confounding factors. Only weight velocity remained significant (P =.2). In contrast with transferrin, the mean serum ferritin concentration decreased significantly during the study (Table 1). Virtually every boy with an increase in RBCI also demonstrated a concomitant decrease in serum ferritin, but the individual 18-mo increments in ferritin did not correlate with the increments in hemoglobin or RBCI (Figure 1). Only a weak association was found between the increments in serum transferrin : 2 #{149} t ,/Z? 9 o GD r.55 P <.1 ocaca cr1 a p p CaO ao Hemoglobin, gil r -.8 p -.5 (98 g o o % ooa Iooo Oc A RBCI, r -.31 P -.2 r -.5 P -.7 FIGURE 1. The 18-mo increments in hemoglobin and in red blood cell iron (RBCI) in relation to the increments in transferrmn and ferritin in healthy boys. and ferritin (r =.26, P =.47). Serum ferritin was also analyzed by multiple-regression analysis with body height, body weight, and testicular volume at 18 mo, and preceding height and weight velocities as confounding factors. With these adjustments only body weight remained significant (P =.2). DISCUSSION One-third of the boys had increased serum transferrin concentrations and almost all of the boys had decreased serum ferritin concentrations during the 18-mo follow-up, but very few of the boys developed iron deficiency as defined by a low ferritin concentration and low transferrmn saturation. The mdividual rises in serum transferrin and declines in ferritin were only moderately associated with each other. Finally, the rise in serum transferrmn did not lead to increasing evidence of irondeficiency anemia. On the contrary, the boys whose serum transferrin concentrations increased also had increased hemoglobin concentrations. We conclude that during puberty there Downloaded from by guest on July 14, 211 TABLE 2 Mean hemoglobin, transferrin, and ferritin concentrations in 6 boys with genital stages 1-4 (G1-G4) at the start of the study and after 18 mo Genital stage Gl G2 G3 G4 Hemoglobin (g/l) mo 13. ± 1. [35] 131. ± 1.5 [25] mo 128. ±.9 [3] 132. ± 2.2 [11] 134. ± 1.2 [35] 137. ± 2.3 [11] Transferrin (g/l) mo 2.8 ±.1 [35] 2.9 ±.1 [25] mo 2.4 ±.2 [3] 2.9 ±.1 [11] 2.8 ±.1 [35] 2.9 ±.12 [11] Ferritin (Lg/L) mo 35. ± 3.1 [35] 36. ± 3.2 [25] mo 38. ± 7. [3] 22. ± 2.62 [11] 2. ± 1.8 [35] 21. ± 2.8 [11] S ± SEM; n in brackets. 2 Significantly different from Gl, P <.5.

4 182 ANTFILA AND SIIMES TABLE 3 Increments in body height, weight, testicular volume, mean corpuscular volume (MCV), transferrin saturation, and ferritin in relation to increments in hemoglobin and transferrmn L Hemoglobin < z Hemog lobin L Transferrmn Transfemn > Transferrin Transferrin > (n 18) (n = ) (n 22) (n 2) Body height (cm) 9. ±.6-1. ± ±.6 Body weight (kg) 7.7 ± ± ±.7 Testicular volume (ml) 4.7 ± ± ±.6 L MCV (if.) -.7 ± ± ±.4 L Transferrmn saturation (%).7 ± ± 2. ± 1.9 Ferritin (Lg/L) -1.6 ± ± ± 352 ± SEM. 2 Significantly different from ferritin values for the other two groups, P <.5. may be a physiologic increase in transferrin that is likely related to stimulated erythropoiesis and is not necessarily an indication of iron deficiency. The boys also demonstrated large decrements in storage iron, as evidenced by their serum ferritin concentrations, which were dependent on individual pubertal development. The lack of correlation between hemoglobin and ferritin increments was not surprising because during growth penods the expansion of blood volume and ned blood cell mass can be the major drain of stored iron even when the hemoglobin concentration is not elevated. There are two processes going on during puberty: an increase in red blood cell mass related to blood volume expansion and an increase in hemoglobin concentration. Transferrmn seems to be more sensitive to the latter process. During pregnancy, serum transferrmn increases and serum fernitin decreases. A rise in transferrmn is seen even when iron supplementation is effective. An increase in plasma volume coincides with a rise in red blood cell mass, and a considerable amount of iron is needed for the growing fetus. A rise in maternal serum transferrmn is not only a sign of iron deficiency, but is also speculated to be an indication of increasing erythropoiesis (1-12). By analogy, in our study of pubertal boys with relatively ample iron status, the mdividual rise in serum transferrin may also reflect increasing iron turnover for needs of erythropoiesis. This possibility is supported by findings that transferrmn may be an independent and effective growth-promoting agent in vitro (1). Whether transferrmn is also a growth factor in vivo is controversial. Misaki et al (13) in their study of 13- to 15-y-old boys noticed that the serum transferrmn concentration was correlated with height velocity, alkaline phosphatase activity, and insulin-like growth factor 1 concentration. They concluded that in normal boys serum transferrmn may be a marker of skeletal growth. In our study, serum transferrin increased in the individuals whose hemoglobin concentrations had increased. The correlation of transferrmn with weight velocity also fit with the view that transferrmn is linked with increasing erythropoiesis. Yet, RBCI increased in essentially every individual in the study, even though only one-half of the boys had increases in transferrin concentrations. During male puberty, erythropoiesis is unusually active and may increase an average of 2.5-fold above the previous level (14). A part of the increasing erythropoiesis is needed for a rise in hemoglobin concentration, but growth in body height, and especially in weight and muscles, causes a marked increase in blood volume and consequently places heavy demands on iron availability. In individuals with active physical training the rise in red blood cell volume may result in an additional requirement for iron. To examine the influence of erythropoiesis on iron absorption, radioiron-absorption tests were performed in normal subjects before and after a course of recombinant erythropoietin (15). The authors noticed a striking enhancement of iron absorption after regular erythropoietin administration and concluded it to be related to the combined effect of diminished iron stores and augmented erythropoiesis. The decreasing serum fernitin concentrations found in our study indicate that part of the extra iron required is mobilized through redistribution of Downloaded from by guest on July 14, 211 TABLE 4 Correlation coefficients (r) between transfemn and ferritin, and the various criteria of growth Serum transferrin Log serum ferritin 6mo l2mo l8mo 6mo l2mo l8mo Body height Height velocity Body weight Weight velocity.5l.4 O Testicular volume P <.5. 2P <.1. 3P <.1.

5 TRANSFERRIN AND FERRITIN IN PUBERTAL BOYS 183 iron from stores to red blood cell mass. However, it is probable that the bulk is obtained by greatly increasing the intestinal absorption of iron because decreasing stones lead to improvement of iron absorption to supply most of the growing needs of erythropoiesis. REFERENCES I. Dc long, van Dijk IP, van Eijk HG. The biology of transferrmn. Clin Chim Acta 199;19:l lacobs A, Miller F, Worwood M, Beamish MR. Wardrop CA. Ferritin in the serum of normal subjects and patients with iron deficiency and iron overload. Br Med I 1972;4: Worwood M. Femtin. Blood Rev 199;4: Cook JD, Dassenko 5, Skikne BS. Serum transferrmn receptor as an index of iron absorption. Br I Haematol 199:75: Thomsen K, Riis B, Krabbe 5, Christiansen C. Testosterone regulates the haemogbobin concentration in male puberty. Acta Paediatr Scand I 986;75: Tanner IM. Growth at adolescence. 2nd ed. Oxford, United Kingdom: Blackwell Scientific Publications, Hansen P, With TK. Clinical measurements of the testes in boys and men. Acta Med Scand 1952:266(suppl): U 8. Koskebo EK, Kivisaari LM, Saarmnen UM, Siimes MA. Quantitation of muscles and fat by ultrasonography: a useful method in the assessment of malnutrition in children. Acta Paediatr Scand 1991 ;8: Green R, Charlton R, Seftel NH, et al. Body iron excretion in man. Am I Med 1968;45: Kaneshige E. Serum femtin as an assessment of iron stores and other hematologic parameters during pregnancy. Obstet Gynecol ;57: Romsbo I, Haram K, Sagen N, Augensen K. Iron requirement in normal pregnancy as assessed by serum ferritin, serum transfemn saturation and erythrocyte protoporphyrin determinations. Br I Obstet Gynecol 1983;9: Blunden RW, Casey GI, Giorgio P. Ho IQK, Petrucco OM, Kimber RI. The effect of normal and high dose iron supplementation on serum ferritin levels during pregnancy. I Obstet Gynecol 198 l;2: Misaki M, Shima T, Yano Y, Okazaki T. Serum transferrmn as a marker of bone growth in boys: correlation with serum alkaline phosphatase activity, plasma insulin-like growth factor 1 and rate of growth in height. Horm Metab Res 1991 ;23: Dallman PR. Changing iron needs from birth through adolescence. In: Fomon SI, Zlotkin 5, eds. Nutritional anemias. New York: Raven Press, 1992: Skikne BS, Cook ID. Effect of enhanced erythropoiesis on iron absorption. I Lab Clin Med 1992;12:746-5l. Downloaded from by guest on July 14, 211