| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
Departments of
1
Clinical Chemistry and
2 Hematology, TUCH–Laboratories, Turku University Central Hospital, PO Box 52, FIN-20520 Turku, Finland;
3
Institute of Clinical Medicine, General Practice, University of Turku, FIN-20014 Turku, Finland;
4
Härkätie Health Center, PO Box 51, FIN-21421 Lieto, Finland;
5
Satakunta Central Hospital, Sairaalantie 3, FIN-28500 Pori, Finland
aauthor for correspondence: fax 358-2-3133920, e-mail Timo.Takala{at}tyks.fi
In recent years both the concept and diagnosis of iron deficiency (ID) have evolved considerably. The introduction of new laboratory tests, especially soluble transferrin receptor (sTfR), has enabled the identification of storage iron depletion, iron-deficient erythropoiesis, iron deficiency anemia (IDA), and functional ID as readily distinguishable clinical conditions (1)(2)(3)(4)(5). Efficient use of the new tests, however, requires that the respective reference populations be appropriately characterized. A problem of conventionally derived reference intervals (RIs) is that the reference groups may contain individuals with subclinical ID (6). Especially in populations with an increased risk of ID, this may widen RIs and impair the sensitivity of these laboratory tests (7).
Efforts have been aimed at modeling the development of subclinical ID and, subsequently, toward establishing truly health-related RIs for adults, children, and adolescents (2)(7)(8)(9)(10)(11). Studies performed in the elderly have used highly variable criteria for the selection of reference populations (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23). Therefore, despite the particular clinical interests regarding iron status in the elderly, no comprehensive studies covering the issue of subclinical ID are available.
This study was undertaken to investigate how subclinical ID influences the RIs of hematologic indices, sTfR, and other laboratory tests defining iron status in the elderly. To achieve this, we selected a general reference group of individuals 
65 years of age and produced RIs for hematologic indices, sTfR, and other laboratory tests. We also formed an iron-replete subgroup by excluding individuals with storage iron depletion. The differences between values from the general reference group and the iron-replete subgroup were used to investigate the extent of storage iron depletion and its impact on the RIs of the studied tests.
This study was a part of a larger Lieto study and was carried out in Lieto, a semi-industrialized rural community in southwestern Finland. A total of 1260 (82%) elderly Lieto residents born during or before 1933 were enrolled (533 men and 727 women). The participants were interviewed and clinically examined. Their diseases were recorded with the diagnosis codes listed in the 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10), and the medications were recorded. Venous blood samples were taken after an overnight fast from all participants. Written informed consent was obtained from all participants, and the Joint Commission on Ethics of the Turku University and Turku University Central Hospital approved the study protocol.
We selected retrospectively a general reference group according to the recommendations of the IFCC (24). In this group, we included those individuals who did not have diseases considered to interfere with hematologic indices, sTfR, and other laboratory tests (detailed exclusion criteria can be found in Table 1A
, which is available as a data supplement with the online version of this Technical Brief at http://www.clinchem.org/content/vol48/issue9/). In addition to several diagnoses, we used increased C-reactive protein (10 mg/L) and creatinine (125 µmol/L in women and 135 µmol/L in men) concentrations as separate exclusion criteria for the reference group. The only medication we used to exclude individuals from the general reference group was iron supplementation. The iron-replete subgroup was defined from the general reference group by including only individuals with serum ferritin values 22 µg/L (8). Separate RIs were calculated for the general reference group and the iron-replete subgroup.
|
Hemoglobin (Hb), hematocrit, erythrocyte count, mean corpuscular Hb, and mean corpuscular volume were measured on an automated analyzer (Advia 120; Bayer Diagnostics). Plasma sTfR assays were performed by use of an automated immunoturbidimetric method (2). Serum ferritin was measured with an automated time-resolved immunofluorometric assay (Autodelfia; Wallac). Serum transferrin was measured on a Behring Nephelometer (Behringwerke AG). Serum iron was measured by a ferrozine reaction (Roche Diagnostics GmbH). Transferrin saturation was calculated according to the formula: Transferrin saturation = [serum iron/(25 x serum transferrin)] x 100%. The sTfR/log ferritin ratio (TfR-F index) was calculated according to the formula: TfR-F index = sTfR/log serum ferritin.
We performed statistical analyses with the SAS®, Ver. 8.0, system for Windows (SAS Institute Inc.). Sex differences were analyzed with the Wilcoxon rank-sum test. The P values presented are two-sided. RIs were calculated by the nonparametric bootstrap estimation of reference limits and their confidence intervals with RefVal, Ver. 3.43, software (H.E. Solberg, Department of Clinical Chemistry, Rikshospitalet, Oslo, Norway) (25). The association of age with laboratory test results was evaluated by use of the Spearman linear correlation.
A total of 547 individuals (220 men and 327 women) were included in the general reference group. Of these, 465 (207 men and 258 women) were included in the iron-replete subgroup. The median ages were 71.6 years in the general reference group and 71.1 years in the iron-replete subgroup, and the age range was 65–100 years in both groups.
A statistically significant difference (P <0.01) between values for men and women was seen in all tests except sTfR (P = 0.32). For this reason, the RIs were estimated separately for both sexes for all of the analytes but sTfR. Listed in Table 1
are the medians, 95% reference limits, and the 90% confidence intervals of the reference limits for the laboratory tests studied in both reference groups. Several findings imply that we were able to efficiently exclude patients with subclinical ID by use of a ferritin concentration of 22 µg/L as a cutoff value. The first finding was that the upper limit of the RI for sTfR was clearly lower when calculated based on the iron-replete subgroup (0.95–2.41 mg/L) compared with the respective limit for the general reference group (0.96–3.03 mg/L). Similar but less prominent findings were seen in the upper reference limit for serum transferrin and in the lower limits for serum iron and transferrin saturation. The second finding was that the sex differences of these limits clearly narrowed down, which suggests that ID is more prevalent in women. The third finding supporting the exclusion of patients with subclinical ID from the reference groups was that the upper reference limit for sTfR was virtually unchanged when an alternative ferritin concentration of 100 µg/L was used as a cutoff value (n = 169; data not shown). This is in concordance with the suggestion of Guyatt et al. (26) that elderly patients with ferritin concentrations >100 µg/L do not have ID. The differences in the RIs for Hb and red cell indices between the iron-replete subgroup and the general reference group were minimal, as was expected because these tests should not be significantly influenced by subclinical ID.
In the general reference group, a slight statistically significant (P <0.05) positive correlation between age and sTfR was found, whereas statistically significant (P <0.05) negative correlations were observed between age and Hb, erythrocyte count, hematocrit, TfR-F index, serum ferritin (men), and serum iron (men). These trends seemed to be partly attributable to subclinical ID because correlations were diminished as the population was narrowed down to the iron-replete subgroup. Fig. 1
demonstrates how the median values for Hb and sTfR changed with age in the general reference group. Despite of these slight trends, we think that the use of single reference limits in populations 65 years or older is justified to assure the sensitivity of these tests to detect ID and anemia.
|
ID has been reported to be quite common in the elderly (27)(28). Advancing ID can be divided into three stages according to concentrations of serum ferritin, sTfR, and Hb (4)(8). In our study, 15% of the whole initial population (n = 1260) was deemed to be storage iron-deficient (stage I, ferritin <22 µg/L). This prevalence of storage ID may be underestimated because the whole initial population included patients with diseases that may increase ferritin concentrations irrespective of current iron status. As evaluated according to iron-replete reference values for the elderly (Table 1
), 7.0% presented with increased sTfR concentrations as a signal for iron-deficient erythropoiesis (stage II). Altogether, 7.1% were anemic, and 26% of these individuals (1.8% of the whole initial population) had IDA (stage III) according to increased sTfR and low Hb (<128 mg/L in men and <117 mg/L in women, i.e., lower reference limits of our hospital). These prevalences were lower than we expected. On the other hand, in a recent study, Fleming et al. (29) found ID only in 2.7% of the elderly participants and IDA in 1.2%, albeit with different criteria for ID. Furthermore, there is no evidence of a physiologic age-related decrease in erythropoietic potential (30).
Our findings support the idea that current RIs are often derived from population samples containing individuals with ID. These findings together with earlier reports imply that there are grounds as well as means for improving the sensitivity of laboratory tests to diagnose ID. This is of clinical importance in the elderly, in whom ID often heralds severe underlying diseases, such as ulcers or malignancies. In practice, this should motivate the production and use of iron-replete reference values of sTfR and other markers of iron status.
Acknowledgments
This study was funded by grants from Turku University, the Medical Research Foundation of the Turku University Central Hospital, and the Finnish Association of Hematology. One of the authors, P. Suominen, has been employed part-time as a medical consultant for Orion Diagnostica, manufacturer of the sTfR assay used in this study.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
