Low frequency of elevated serum transferrin saturation in elderly subjects

Clinica Chimica Acta 298 (2000) 181–186 www.elsevier.com / locate / clinchim

Short communication

Low frequency of elevated serum transferrin saturation in elderly subjects a, b b Carlo Franzini *, Aurelia Berlusconi , Chiara Favarelli , Simona Brambilla a a

Istituto di Scienze Biomediche Ospedale L. Sacco, Universita` degli Studi di Milano, Ospedale L. Sacco, Via G.B. Grassi 74, 20157 Milano, Italy b Laboratorio Analisi, Pio Albergo Trivulzio, Milano, Italy Received 14 October 1999; received in revised form 3 March 2000; accepted 12 March 2000

Abstract Serum transferrin saturation (TS) values were calculated on the basis of serum iron and transferrin (protein) measurements in a total of 2425 serum samples from six groups of subjects: individuals applying for selection as blood donors (M and F, median age 34 and 32 years); patients referring to the hospital laboratory for routine testing (M and F, median age 45 and 48 years); and elderly subjects living in a specialized institute (M and F, median age 76 and 82 years). In the first four groups the frequency of TS values , 15% and . 62% respectively, was substantially as expected, considering the average health conditions and sex. These results indirectly support the reliability of the measurement procedure. In the elderly group, however, the frequency of TS values . 62% was zero. Mean TS values in the elderly group (males and females) were significantly lower (P , 0.0001) than in the blood donors group and in the hospital patients one. This observation suggests a shortened survival in the presence of (unrecognized) iron overload, pointing out at the usefulness of iron overload screening using simple biochemical tests.  2000 Elsevier Science B.V. All rights reserved. Keywords: Transferrin saturation; Iron overload; Ageing

*Corresponding author. Tel.: 1 39-02-3579-9806; fax: 1 39-02-3820-1626. E-mail address: [email protected] (C. Franzini) 0009-8981 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0009-8981( 00 )00273-4

182

C. Franzini et al. / Clinica Chimica Acta 298 (2000) 181 – 186

1. Introduction Serum transferrin saturation (TS), is considered a reliable test for the screening of iron overload, whether due to hereditary haemochromatosis or to other causes [1], and it is usually calculated from iron and total iron-binding capacity (TIBC) measurements. In the last years, the advantages of the immunochemical measurement of serum transferrin (TRF, as a protein) compared with the measurement of TIBC have been discussed [2]: the availability of international reference preparation for proteins in human serum [3] supports the reliability of the former analytical approach. TIBC and TRF measurements have been eventually confirmed to be substantially equivalent [4], thus supporting the feasibility of calculating TS from serum iron and TRF measurements. We applied such an analytical approach to the measurement of TS in serum samples from elderly subjects, from candidate blood donors and from hospital patients.

2. Materials and methods In the study, 2425 serum samples were included, divided into three groups. The ‘‘Donors’’ group included 575 serum samples from individuals (313 males and 262 females) referring to our hospital’s blood bank for selection as blood donors. All of them underwent medical examination supplemented with routine biochemistry and haematology profiles, and were found to be in apparently good conditions. The ‘‘Hospital’’ group included 1219 serum samples from in- and out-patients (533 males and 686 females) referring to the hospital biochemistry laboratory for routine testing, regardless of their health / illness condition. The ‘‘Hospice’’ group included 631 serum samples from old subjects (216 males and 415 females) living in the Pio Albergo Trivulzio, a home for old people in Milano hosting old but substantially self-sufficient people, albeit suffering from some kind of chronic disease, as expected in such an advanced age. Data about the age of the subjects in each group are given in Table 1. Measurements (of serum iron and TRF) were performed daily on fresh samples at the laboratories of the two institutions: the collection of samples and the generation of analytical data spanned a time-interval of about 2 months. Serum iron and TRF concentrations were measured in the Hospital laboratory (Donors and Hospital groups) and in the Hospice laboratory (Hospice group). Iron was measured in both laboratories by automated end-point two-wavelengths photometry, with ferrozine as a chromogenic indicator. In both cases measurements were done with a Hitachi 917 Analyser (Roche Diagnostics, Milan, Italy) with reagents and calibrators supplied by the same company. TRF was measured by automated immunoturbidimetry, with either a Hitachi 917 (Hospital) or a Hitachi 911 (Hospice) analyser (Roche Diagnostics, Milan, Italy), with reagents

C. Franzini et al. / Clinica Chimica Acta 298 (2000) 181 – 186

183

Table 1 Age, frequency of TS values below and above selected cut-off values (15, 45 and 62%), and median TS values in the different groups, males and females Quantity

Age, interval (years)a Age, median (years) TS , 15% (%) TS . 45% (%) TS . 62% (%) TS, median (%) a

Males

Females

Donors (n 5 313)

Hospital (n 5 533)

Hospice (n 5 216)

Donors (n 5 262)

Hospital (n 5 686)

Hospice (n 5 415)

20–46 34 1.9 19.2 3.5 33.8

26–62 45 13.3 19.1 6.9 29.8

60–96 76 17.1 2.8 0 21.4

21–42 32 11.8 8.4 2.3 26.4

24–64 48 17.3 13.6 3.2 27.0

60–104 82 22.8 3.6 0 20.3

Overall interval.

and calibrators by the same company. Calibration was reported to be traceable to the international calibrator RPPHS / CRM 470 [3]. The two laboratories participated in the Regione Lombardia mandatory External Quality Assessment Scheme for Clinical Chemistry. The average bias (from the scheme’s consensus mean value) for iron measurement was (mean6S.D. from measurements on 6 different EQA samples) 98.763.8% (Hospital) and 98.963.8% (Hospice), the means’ difference being not quite significant (P 5 0.0857); corresponding figures for TRF were 96.061.8 and 100.665.2%, the means’ difference being not quite significant (P 5 0.0926). Monthly analytical between-series imprecision values in iron measurement (CV%, from daily internal quality control) were in the interval 1.00–2.14% (Hospital) and 1.05–1.92 (Hospice); corresponding figures for TRF were 1.81– 3.70% and 1.23–3.55%, respectively. The original mass concentration values (in mg / dL and mg / dL for iron and transferrin respectively) were converted into substance concentration values (in mmol / L) on the basis of 55.85 as the relative atomic mass of iron and 79570 as the relative molecular mass of transferrin [5]. TS values (in % units) were calculated as TS 5 100 / 2 3 [Fe] / [TRF], where [Fe] and [TRF] are the serum substance concentration values of iron and transferrin (in mmol / l) and where a stoichiometric ratio of two atoms of iron per molecule of transferrin is assumed. The results were statistically assessed by standard parametric and nonparametric tests. The significance of the differences among groups’ means was tested by ANOVA, that of the difference among EQA values by the Student’s t-test for paired data. Calculations were performed with an Excel spreadsheet program (Microsoft Corporation) and with the Instat statistical package (GraphPad Software, San Diego, CA, USA).

184

C. Franzini et al. / Clinica Chimica Acta 298 (2000) 181 – 186

Table 2 Mean values and standard deviations for iron and transferrin concentrations and for transferrin (TRF) saturation in the three groups, and statistical significance of the differences Quantity

S-Iron, mean6S.D. mmol/l S-Transferrin, mean6S.D. mmol/l TRF Saturation, mean6S.D.%

Males a

Females a

Donors n 5 313

Hospital n 5 533

Hospice n 5 216

Donors n 5 262

Hospital n 5 686

Hospice n 5 415

20.966.5( 1 )

15.468.1( 1 )

11.265.1

18.767.0( 1 )

14.966.9( 1 )

10.464.6

31.265.5( 1 )

25.167.5( 2 )

24.766.5

35.567.8( 1 )

30.067.8( 1 )

25.266.9

35.0613.6( 1 )

32.4617.2( 1 )

23.3610.0

28.4613.4( 1 )

28.5615.1( 1 )

21.569.4

a

Statistical significance of the differences with the relevant ‘‘Hospice’’ group (by ANOVA): ( 1 ) P , 0.0001 (extremely significant); ( 2 ) P 5 0.4928 (not significant).

3. Results The 15% and the 62% TS values are generally considered as the threshold limits for iron deficiency and iron overload respectively [6]; for homozygous hereditary haemochromatosis the 45% TS value is also suggested. The frequencies of values exceeding such limits in the three groups (males and females) were therefore calculated and are shown in Table 1, together with the age-data of the subjects in each group. Mean values and standard deviations for iron and TRFconcentration, and for transferrin saturation as well, are shown in Table 2.

4. Discussion As expected, the frequency of low ( , 15%) TS values is higher in females, as compared with males, within each of the three groups. Among both males and females, the smaller rates of low TS values observed in the Donors group are consistent with the substantially good conditions of health of the candidate blood donors. The rates of high TS values ( . 62%) in the Donors and Hospital groups, males and females, are also expected: they are consistent with the number of conditions associated with iron overload [1], and with the high frequency of unrecognized haemochromatosis [7]. Moreover, such rates are higher in the Hospital group (males and females, including patients suffering from a variety of diseases) if compared with the candidate blood donors; within each group they are lower in females. As a whole, these results indirectly support the reliability of calculating TS from iron and TRF measurements. However, the null frequency of TS values . 62% in the elderly group (males and females) was an unexpected finding: in a previous study heterozygous

C. Franzini et al. / Clinica Chimica Acta 298 (2000) 181 – 186

185

haemochromatosis, with TS in the interval 9–81%, was found to occur in a 61–90 years age group [8]. A small number of cases exceeding the 45% TS threshold value was observed in the elderly (Hospice) group, but however its frequency was much lower than in the other groups. Median TS values in the Donors group substantially agree with the mean values reported for normal subjects [8]. In comparison with the Donors group values, the median TS values in the Hospital group (males and females) show small differences (the higher frequencies of high and low values compensate each other), but in the Hospice group such median TS values are definitively lower. Mean TS values in Table 2 are somewhat different from median TS values in Table 1, due to non-normal distributions; however, the differences of mean values in the Hospice group versus the Donors and the Hospital groups are similar to the differences shown by the medians. The differences of the mean values in the Hospice group versus the Donors and the Hospital groups are statistically extremely significant in every case (P , 0.0001). Markedly lower values of mean iron concentration are mainly responsible for the lower mean values of TS in the Hospice group (Table 2). Such low iron values do not match proportionally low values of TRF, although the mean TRF values in the Hospice group are also significantly (P , 0.0001) lower than the mean TRF values in the other two groups, with the exception of the males– Hospice / males–Hospital match. Different explanations can be given of our findings. However, assuming that elevated TS means iron overload, it is tempting to speculate that patients with iron overload do not survive as much as people with normal iron status. Iron overload could represent an additional cumulative risk factor, contributing to decreased life expectancy and earlier onset of disability [9]. It would be of high practical value if this hypothesis proved correct. As a matter of fact costeffective simple means to screen for possible iron overload are at present available [8], e.g. by combining serum TS and ferritin measurements, possibly supplemented with the measurement of serum soluble transferrin receptor concentration, as recently reported [10]. Cost-effective simple means to correct iron overload are also available [8].

References [1] Witte DL, Crosby WH, Edwards CQ, Fairbanks VF, Mitros FA. Hereditary hemochromatosis. Clin Chim Acta 1996;245:139–200. [2] Beilby J, Olynyk J, Ching S, Prins A, Swanson N, Reed W, Harley H, Garcia-Webb P. Transferrin index: an alternative method for calculating the iron saturation of transferrin. Clin Chem 1992;38:2078–81.

186

C. Franzini et al. / Clinica Chimica Acta 298 (2000) 181 – 186

[3] Whicher JT, Ritchie RF, Myron Johnson A, Baudner S, Bienvenu J, Bilirup-Jensen S, Carlstrom A, Dati F, Milford Ward A, Svendsen PJ. New international reference preparation for proteins in human serum (RPPHS). Clin Chem 1994;40:934–8. [4] Gambino R, Desvarieux E, Orth M, Matan H, Ackattupathil T, Lijoi E, Wimmer C, Bower J, Gunter E. The relation between chemically measured total iron-binding capacity concentrations and immunologically measured transferrin concentrations in human serum. Clin Chem 1997;43:2408–12. [5] MacGillivray RTA, Mendez A, Shewale JG, Sinha SK, Lineback-Zins J, Brew K. The primary structure of human serum transferrin. J Biol Chem 1983;258:3543–53. [6] Baer DM, Simons JL, Staples RL, Rumore GJ, Morton CJ. Hemochromatosis screening in asymptomatic ambulatory men 30 years of age and older. Am J Med 1995;98:464–8. [7] Barton JC, Bertoli LF. Hemochromatosis: The genetic disorder of the twenty-first century. Nat Med 1996;2:394–5. [8] Bulaj ZJ, Griffen LM, Jorde LB, Edwards CQ, Kushner JP. Clinical and biochemical abnormalities in people heterozygous for hemochromatosis. New Eng J Med 1996;335:1799– 805. [9] Vita AJ, Terry RB, Hubert HB, Fries JF. Aging, health risks, and cumulative disability. New Eng J Med 1998;338:1035–41. [10] Looker AC, Loyevsky M, Gordeuk VR. Increased serum transferrin saturation is associated with lower serum transferrin receptor concentration. Clin Chem 1999;45:2191–9.