di
Anna Giulia Cattaneo, M.D.

Abstract
Anemia appears to be frequent in aged people, with higher prevalence of anemia of chronic diseases (ACD). However, wide discrepancy exists between different epidemiological studies, because of differences in sample selection and probably the low degree of accuracy in differential diagnosis. In recent years, attention has been focused on the search of simple, possibly non-expensive, powerful methods to reach accurate diagnosis. Among them, the possibility to routinely measure s-MMA in folate and vitamin B12 deficit, or sTfR to discriminate between ACD and iron deficiency anemia (IDA) in microcytic anemias, or the use of advanced, robuste statistic to evaluate simple and routinary laboratory values by means of multivariate analysis, appears to be most promising. Finding of new markers of pre-anemic states and the study of their evolutive processes should be of interest for prevention.

The prevalence of anemia in the elderly, even as a mild condition, not associated to evident hematological pathologies, is an interesting parameter for its possible link with avoidable conditions (like inadequate food, vitamins and microelements assumption and assimilation), and for its impact with the welfare of frail organisms. The clinical relevance of this disorder, as well as underliyng causes, are not completely understood, while anemia and mortality appears to be directly related in aging (1,2). No markers of alterations of erythrocyte physiology and production clearly related to aging have been stated, nor even is it clear if laboratory hematological values need to be adjusted to the age. Authors do not agree with the attitude to treat all the Hb and /or Hct variations in the oldest people, but the incomplete knowledge of basic mechanisms could make the prevention of true pathological forms ineffective.

Erithrocyte physiology in the elderly has not been studied in all its aspects. An unusual finding we have observed in old people (age: 73-101 years) free from anemic fingerprints, is reduced microviscosity (eta ) of hemoglobin-free erithrocyte ghosts, measured as fluorescence anisotropy of a rod-like probe (1,6 diphenyl - 1,3,5 - hexatriene, DPH). Steady-state temperature-dependent variations appears to be similar to that observed in younger people. The reported parameter (eta ) is strictly dependent, in experimental conditions, from lipids composition of the bilayer of cellular envelope. While not directly related to alterations of elastic properties of the red cell, abnormal is proved to alter protein aggregation in the bilayer and to be in dependance of the presence of lipoperoxidants or lipoperoxidation products in the bilayer, conditions that are able to influence rheological properties of red cells (3). In our patients, no evidences of altered routine and hematological parameters, included the osmotic resistance of red blood cells, has been found, but rheological parameters have not been studied. The clinical relevance of the alteration we have observed remains unexplained: it should be considered a marker of aging itself, or of an even undetectable form of anemia.

Epidemiological studies
Everyone professionally involved in the care of aged people could note the presence of anemic fingerprints in such patients as an unusually frequent alteration, but the study of the true clinical relevance of this phenomenon, its frequence and severity, requires the use of well planned sample selection and robust statistics. Results could be intuitively differents if subjects are choiced between hospital inpatients, nursing home residents, outpatients affering to a specialized clinic or to a general practitioner, or even randomly selected between residents in a defined region. Other characteristics of selected people, like sex prevalence, altitude of the abitual residence, usual alimentation and motion, and others, could be influent on results.
From epidemiological studies on large numbers of aged subjects, performed in the last decades, wide discrepancy has been observed, expecially when attention has been focused on the prominent types of anemia: the anemia of chronic diseases (ACD) appears to be the most frequent diagnosis in old subjects, while the prevalence of iron deficiency anemia (IDA) varies from 0 or 0.4% up to 22% (4,5). Recent studies confirm the low frequency of this type of anemia in old people, at least in western countries, in which the contrary - and namely a tendency to higher iron stores - seems to be the most prominent feature (6). Even if diagnostic criteria of anemia (7) are not fulfilled, lower levels of hemoglobin or altered hematological parameters are frequently observed in old subjects. According to several authors the large number of cases in which underlying causes of anemia are not found suggest a need for a more accurate diagnostic procedures instead of age-adjusted hematological values (1,8,9). There is disagreement about the necessity to treat or not the mildest cases without evident causality, but the prevalent trend is not to treat. However, in selected cases, a precocious treatment seems to ameliorate the prognosis and even reduce the mortality: it should be the case, for example, of old, anemic subjects with acute miocardial infarction (10), patients with early chronic renal failure (CRF)(11), and those with poor food, vitamin or oligoelements assumption (12,13). A summary of the most preminent epidemiological studies is reported in Table 1.

Ethiopathological diagnosis

In addition to the classical diagnostic approach to anemic patients, exhaustive in the majority of cases, a number of specific tests have been proposed to improve diagnosis, extend it to subclinical cases, or simplify the diagnostic iter. The most interesting among them are those requiring simplified, minimally expensive and invasive procedures, and those focused on specific aspects of the types of anemias more frequently seen later in life (Table 2).
The possibility to use well planned, powerful computerized analysis is of great interest in all diagnostic fields, included that here discussed. As an exemple, the multivariate analysis proposed by Shiga et al. (22) provided good diagnostic approximation requiring only routinely and automatically measured parameters: a rapid screen should be provided, limiting the use of more specialistic and expensive tests to true pathological findings of relevance. A strong link between Hb and Hct, as between MCV and MCHC has been demonstrated; diagnostic accuracy of 77% has been reached with only one hypothesis admitted, higher score (100%) with two most probable hypothesis allowed. However, by introducing a third variable (such age and sex) the predictivity with one hypothesis admitted is improved. This statistical test should be validate in untested conditions (like subjects treated for anemia or related diseases): aging is one of them.

By considering the most frequent types of anemia seen in old patients, it appears to be important, in the view of its therapeutic implications, to use a sensitive test to discriminate iron deficiency anemia (IDA) from anemia of chronic diseases (ACD), or to assess the influence of IDA on ACD, when they cohexist. A recently proposed version of soluble transferrin receptors (sTfR) determination for the routine laboratory, and its ratio to ferritin seems highly promising in this regard (23).

A special remark should be reserved to the anemia of chronic renal failure (CRF). The prevalence of CRF is so high in the elderly, that it is not incorrect to consider it as a general problem. When anemia is also present, it should be due to a variety of causes, often coexisting: occult blood losses through the urologic apparatus can cause IDA, chronic pyelonephritis is associated with ACD, hyperhemolysis, chronic DIC and hyperparathyroidism are additional causes of anemia. However, the so-called anemia of CRF is well characterized as a hypoproliferative syndrome in which marrow hypoplasia is secondary to low levels of erythropoietin (EPO), probably linked to parenchymal damage of the kidney. The therapeutical response to exogenous EPO differentiates this syndrome from that due to ACD, but it could be lacking if the marrow is refractory because of iron deficiency. For this reason, accurate diagnosis of IDA concomitant with CRF is important.

Vitamin B12 and folate deficiency with its consequences and causes is of special interest in older patients. Low serum levels of folate have been found in 5 or 29 % old patients with psychiatric disorders, being lower in subjects with dementia (30); erithrocytic folate was low in 13% of 200 patients of a geriatric unit (25). One reason for that could be poor nutrition, in addition to gastric and enteric hyatrogenic alterations. Prevalence of low B12 in large samples of differently selected aged subjects varied between 2,2 and 40.5 % (19,25,30,31). Only in few cases, however, values are found in true pathological range or with hematological signs of the defect (19,25,31). The necessity to introduce age-modified values of B12 is discussed, but not advised (32). On the contrary, the introduction of a test like MMA determination in urine (31) or in serum (19, 24,26,28,32) as a means for subclinical and early diagnosis appears to be more promising. The importance to detect as early as possible a deficiency of B12 in the elderly is motivated also by the implications ascribed to it in determining cognitive impairement and dementia, in the case of subclinical defect too. The use of a modified Schilling's test, namely that performed by the administration of B12 bound to nutritional proteins (32), or the desoxyuridine suppression test (24-27) should be confined to cases requiring more extensive study of underlying causes, or of therapeutical response.

Another class of diseases associated with anemia, more frequently seen in the elderly, are the myelodysplastic syndromes, very rare in the young, in which they are usually secondary to toxic agents. The French-American-British (FAB) classification, recently validated by WHO (33), requires bone marrow biopsy: it is an aid to describe the disease and guide therapeutical choices, but has only weak prognostic value (refractory anemia, RA, can be fatal for extreme pancytopenia, whithout reaching the leukemic state). Therapy, however, is often disappointing, expecially in the old patient. The possibility to genetically characterize the cells, and to point out patterns of "in vitro" growth linked to prognosis, improves our knowledge of the syndrome, while the MRD phenotyping adds some more precision to the therapeutical approach (29). The prognostic value of apoptotic vs hypoproliferative cells has beeen recently stated in idiopatic myelofibrosis (34)

In conclusion, accurate diagnosis of all causes of anemia even in the old and very old subject is desirable, and stigmatizing low Hb levels as "physiological" and unavoidable consequences of aging appears to be naïve. As a consequence there is no need for modified range values in older people in the majority of cases. Better examination of individual cases could in the future reduce the prevalence of anemia due to undiagnosed causes in this class of age, and improve the therapeutical approach. In contrast, the search for precocious markers of hematologic alterations. In the same manner, all new diagnostic procedures, requiring minimally expensive and invasive techniques, and useful for extensive and routine application, leading to improved diagnosis, should be quickly standardized and introduced into practice.

Addendum
An alphabetic list of abbreviations used in this work follows:
ACD: anemia of chronic diseases; B12: vitamin B12, or cobalamin; BFU-E: erythroid burst-forming unit; CFU-E: erythroid colony-forming unit; CMML: chronic myeloid monomyelocitic leukemia; CRF: chronic renal failure; DIC: disseminated intravascular coagulation; 2,3 DPG: 2,3 diphosphoglucose; DPH: 1,6 diphenyl-1,3,5-hexatriene; : microviscosity (poise); EPO: erythropoietin; FAB: French-American-British; G6PDH: glucose-6-pohsphate dehydrogenase; Fe: iron; Hb: hemoglobin concentration (g/dl); Hct: % hematocrit, IDA: iron deficiency anemia; IL-1 : interleukin 1- ; IRD: iron reserve depletion; MCH: mean corpuscular hemoglobin; MCHC: mean corpuscular hemoglobin concentration; MCV: mean corpuscular volume; MMA: methylmalonic acid; MRD: multiple resistant drug; PP: platelets; RA: refractory anemia; RAEB: refractory anemia with excess of blats; RAEB-T: refractory anemia with excess of blasts transformed; RARS: RA with ringed sideroblasts; RBC: red blood cells; s-: serum; sTfR: soluble Transferrin receptors; TIBC: total iron binding capacity; TNF : tumor necrosis factor ; u-: urinary; WBC: white blood cells; WHO (or OMS): World Health Organization (or Organisation Mondial de la Santè).

Table 1 Prevalence of anemia in old subjects

(a) WHO criteria for anemia: Hb 13 g/dl for ?, Hb 12 g/dl for ? (WHO - Nutritional anemias - Techn. Rep. Ser. 503, 1972.
See Addendum for abbreviations.

Table 2 Tests for anemia in addition to classical diagnostic procedures.

Abbreviations in Addendum.

References
1. Izaks GJ et al. (1999): JAMA 281: 1714-1717
2. Kikuchi M et al. (2001): J.Am.Geriatr.Soc. 49: 1226-1228
3. Shinitzky M, Inbar M (1976): Biochim.Biophys.Acta 433: 133-149
4. Lipschitz DA et al. (1981): Am.J.Hematol. 11: 47-54
5. Milman N, Schultz-Larsen K (1994): Aging Clin.Exp.Res. 6:97-103
6. Fleming DJ et al. (2001): Am.J.Clin.Nutr. 73: 638-646
7. WHO: Nutritional anemias. Techn.Rep.Series 503,1972
8. Nilsson-Ehle H et al.(1988): Acta Med.Scand.224:595-604
9. Joosten E et al. (1992): Gerontology 38:111-117
10. Wu WC et al. (2001): N.Engl.J.Med. 345: 1230-1236
11. Kazmi WH et al. (2001): Am.J.Kidney Dis. 38: 803-812
12. Klesges LM et al.(2001): Am.J. Public Health 91: 68-75
13. Mitrache C et al. (2001): Ann.Hematol. 80: 295-298
14. Myers AM (1968): Lancet 2: 261-263
15. Elwood PC et al. (1971): Br. J.Haematol. 21: 557-563
16. Mattila K et al.(1986): Scand.J.Clin.Invest.46:411-415
17. Milman N et al. (1990): Scand. J. Clin.Lab.Invest. 50: 77-83
18. Inelmen EM et al.(1994):Aging Clin.Exp.Res. 6:81-89
19. Lindenbaum J. et al. (1994): Am J. Clin.Nutr. 60: 2-11
20. Nilsson-Ehle H. et al. (2000): Eur.J.Haematol. 65: 297-305
21. Spykerelle Y et al. (2000): Gastroenterol. Clin. Biol. 24: 709-713
22. Shiga S. et al. (1997): Am.J.Hematol. 54: 108-117
23. Mast AE et al. (1998): Clin. Chem. 44: 45-51
24. Beck WS (1991): Ann.Rev.Med. 42: 311-322
25. Blundell EL et al. (1985): J.Clin.Pathol. 38: 1178-1184
26. Colon-Otero G et al. (1992): Med.Clin.North Am. 76:581-597
27. Karnaze DS, Carmel R (1987): Arch.Int.Med. 147: 429-431
28. Stabler SP et al. (1997): Am.J.Clin.Nutr. 66: 741-749
29. Besa EC (1992): Med.Clin.North Am. 76: 599-619
30. Nilsson K et al. (1994): Eur.J.Clin.Invest. 24: 600-606
31. Norman EJ, Morrison JA (1993): Am.J.Med. 94: 589-594
32. Carmel R (1997): Am.J.Clin.Nutr. 66: 750-759
33. Germing U. et al. (2000): Leuk.Res. 24: 983-992
34. Kvasnicka HM et al (1999): Ann. Hematol. 78: 65-72

Present address:
DBSF, Università dell'Insubria
Via J.H.Dunant,3 - 21100 Varese