PREDICTIVE VALUE OF LOW FERRITIN IN OLDER PERSONS WITH ANEMIA WITH AND WITHOUT INFLAMMATION: THE LEIDEN 85-PLUS STUDY

To the Editor: Iron deficiency is a common cause of anemia, being found in more than 15% of older persons with anemia.^1,2 Serum ferritin levels strongly correlate with body iron stores³ and are considered the best noninvasive test for the diagnosis of iron deficiency.^2,4 Ferritin therefore plays a central role in diagnostic and therapeutic algorithms for iron deficiency anemia in clinical practice,⁵ although ferritin is also an acute-phase protein and may be high with acute and chronic inflammatory conditions such as infections, rheumatoid arthritis, and cancer.² Because the prevalence of inflammatory conditions is very high in old age,⁶ it is not clear whether low ferritin can be used as a marker of low iron status in old age. Therefore, the association between low ferritin levels and anemia in old age in the presence and absence of inflammation was investigated.

METHODS

The present study is embedded in the Leiden 85-plus Study, a population-based prospective follow-up study of 512 85-year-old inhabitants of Leiden, the Netherlands.⁷ Participants who used iron supplements were excluded from the present study. At age 85, ferritin levels were determined using an immunological assay (E170, Roche, Almere, the Netherlands). Low serum ferritin was defined as ferritin less than 20 μg/L for men and less than 15 μg/L for women.⁸ C-reactive protein (CRP) levels were measured using a Hitachi 747 automated analyzer (Hitachi, Tokyo, Japan). High CRP was defined as CRP greater than 5 mg/L. Hemoglobin levels and mean corpuscular volume (MCV) were determined annually (aged 85–90) using an automated analysis system (Coulter Counter, Coulter Electronics, Hialeah, FL). Anemia was defined according to criteria of the World Health Organization (hemoglobin <130 g/L for men and <120 g/L for women).⁹

RESULTS

One hundred seventy-eight participants (34.8%) were male. The prevalence of anemia was 23.8% (n=122); 35 participants (6.8%) had low ferritin levels. Participants with low ferritin levels had more than twice the risk of having anemia than participants with normal ferritin levels (odds ratio (OR)=2.2, 95% CI=1.1–4.5). In participants with high CRP levels (n=171), low ferritin was associated with a risk of anemia that was 7 times as great (OR 7.0, 95% CI=1.4–34.9). No significant association was found between low ferritin and anemia in participants with normal CRP levels (n=341, OR 1.7, 95% CI=0.7–4.2).

The lowest hemoglobin levels were found in participants with low ferritin and high CRP levels (mean (SE) hemoglobin level 99 g/L (11 g/L) in men and 112 g/L (8 g/L) in women, Figure 1). Similar results were found for MCV (data not shown). Low ferritin was also associated with an additional decline in hemoglobin level (additional annual change in hemoglobin −2 g/L, 95% CI=−4.0 to −0.3 g/L) and MCV (additional annual change in MCV −0.78 fL, 95% CI=−1.3 to −0.22 fL) in the years thereafter. Again, these associations were most apparent in participants with high CRP levels. In this subgroup, low ferritin was associated with an additional annual decline in hemoglobin level of 5 g/L (additional annual change in hemoglobin level −5 g/L, 95% CI=−10 to −0.4 g/L) and an additional annual decline in MCV of 2.2 fL (additional annual change in MCV −2.2 fL, 95% CI=−3.6 to −0.71 fL).

Figure 1

Mean hemoglobin levels according to ferritin status and C-reactive protein (CRP) status for men and women at age 85. Low ferritin was defined as ferritin <20 μg/L for men and <15 μg/L for women. High CRP was defined as CRP>5 mg/L. ANOVA=analysis of variance.

DISCUSSION

The present study shows that low ferritin is associated with anemia in old age. This association is most prominent in older individuals with signs of an inflammatory host response in the plasma.

The diagnostic value of serum ferritin levels to detect iron deficiency in patients with anemia, infection, and inflammation has been questioned because of ferritin's “acute phase” properties. The findings presented here show the significance of measuring ferritin levels in older individuals, especially in those with infection or inflammation. In these patients, a low level of ferritin is a specific marker of iron status because of its “acute phase” properties; iron status must be poor when low ferritin levels are found in the presence of systemic inflammation.

How can the occurrence of low ferritin levels in the presence of inflammation be explained? Older patients with gastrointestinal tumors or chronic inflammatory diseases will have low iron stores because of gastrointestinal blood loss, malnutrition, or malabsorption of food-bound iron. In these individuals, iron stores may have become too low to facilitate a rise in ferritin in response to inflammation. Upregulation of hepcidin—the main regulator of iron homeostasis—is an alternative molecular pathway to explain these findings. Proinflammatory cytokines, particularly interleukin 6, induce the production and secretion of hepcidin by hepatocytes. Hepcidin binds to the membrane protein ferroprotein and induces its internalization and degradation in lysosomes, blocking the export of iron from cells.¹⁰ Although a preliminary analysis in the Invecchiare in Chianti Study could not demonstrate higher urinary hepcidin levels in older individuals with anemia of inflammation,¹¹ the hypothesis should be tested in other population-based studies, preferably using serum hepcidin assays, which have recently become available.¹²

ACKNOWLEDGMENTS

Conflicts of interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this paper.

The Leiden 85-plus Study was partly funded by an unrestricted grant from the Dutch Ministry of Health, Welfare and Sports. The Leiden 85-plus Study is a collaborative project of the Department of Gerontology and Geriatrics (RGJ Westendorp) and the Department of Public Health and Primary Care (J Gussekloo) of the Leiden University Medical Center, Leiden, the Netherlands.

Author Contributions: Professor Westendorp had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: J. Gussekloo and R.G.J. Westendorp. Acquisition of data: J. Gussekloo, A.J.M. de Craen, G.J. Blauw, and R.G.J. Westendorp. Analysis and interpretation of data: W.P.J. den Elzen, J. Gussekloo, J.M. Willems, A.J.M. de Craen, G.J. Blauw, W.J.J. Assendelft, and R.G.J. Westendorp. Drafting of the manuscript: W.P.J. den Elzen, J. Gussekloo, and R.G.J. Westendorp. Critical revision of the manuscript for important intellectual content: W.P.J. den Elzen, J. Gussekloo, J.M. Willems, A.J.M. de Craen, G.J. Blauw, W.J.J. Assendelft, and R.G.J. Westendorp. Statistical analysis: W.P.J. den Elzen., J. Gussekloo, A.J.M. de Craen, and R.G.J. Westendorp. Obtained funding: J. Gussekloo and R.G.J. Westendorp. Administrative, technical, or material support: W.P.J. den Elzen, J. Gussekloo, A.J.M. de Craen, and G.J. Blauw. Study supervision: J. Gussekloo, A.J.M. de Craen, W.J.J. Assendelft, and R.G.J. Westendorp.

Sponsor's Role: The sponsor had no role in the design, methods, subject recruitment, data collections, analysis, or preparation of the letter.