Erythropoietin, GDF15, IL6, hepcidin and testosterone levels in a large cohort of elderly individuals with anaemia of known and unknown cause

Haemoglobin levels decline with increasing age in bothmen and women (1), with the prevalence of anaemiaincreasing sharply from approximately 5% at age 65 tomore than 20% of individuals at age 85 yrs (2). Whilemuch of this decline can be explained as a consequenceof comorbid conditions (nutritional deficiencies, inflam-matory disease, blood loss, renal disease or cancer), whenindividuals with identifiable causes of anaemia areexcluded from analysis, there is still a marked decline inhaemoglobin levels in the elderly – most notably in men(2, 3). Approximately one-third of the cases of anaemiain the elderly cannot be ascribed to known causes – andtherefore constitute a group with anaemia of unknownaetiology or AUE (4–6). There is considerable interest inlearning more about the possible causes of anaemia inthis group, in part because epidemiologic studies haveshown that anaemia (regardless of the cause) is associ-ated with increased morbidity and mortality in a varietyof contexts (7, 8). Among frail elderly, both frequencyof anaemia (approximately 50%) and the fraction ofunexplained anaemia (>40%) are even higher than innon-frail elderly (9).Proposed explanations for the increase in AUE seenwith advancing age include the following: (i) a stem cellor progenitor cell defect altering response to erythropoi-etic stimulation (10–12); (ii) insufficient erythropoieticdrive in response to anaemia (4, 5, 13, 14); (iii) anacquired abnormality in circulating red cells (15); and(iv) cytokine dysregulation affecting production orresponse to erythropoietin (EPO) or iron availability (6,AbstractEpidemiologic studies have documented an increasing frequency of anaemia in individuals 65 yrs andolder. Elderly individuals with anaemia have been categorised into the following: those with chronicdisease, those with iron, B12 or folate deficiency and those with anaemia of unknown aetiology (AUE).There is considerable interest and debate as to whether AUE has an inflammatory component, is causedby cytokine dysregulation affecting production or response to erythropoietin (EPO) or iron availability orrepresents a novel pathologic process. Here, we compare a large cohort of AUE cases with a matched,non-anaemic control group and with individuals who have anaemia of defined cause. IL-6, hepcidin,GDF15, EPO and testosterone levels were compared. IL6 and hepcidin levels did not differ significantlybetween AUE and control groups, indicating that inflammation or iron restriction is not central feature ofanaemia in this group. GDF15 levels were significantly elevated when comparing AUE with controls andwere markedly elevated in patients with renal disease. Testosterone levels were lower in men from theAUE group compared with non-anaemic controls. EPO levels in the AUE group were increased relative tocontrols but were inappropriately low for the degree of anaemia. Our data indicate that an impaired EPOresponse, in the absence of evidence for iron restriction or inflammation, is characteristic of AUE.