An anemia of Alzheimer's disease


Lower hemoglobin is associated with cognitive impairment and Alzheimer's disease (AD). Since brain iron homeostasis is perturbed in AD, we investigated whether this is peripherally reflected in the hematological and related blood chemistry values from the Australian Imaging Biomarker and Lifestyle (AIBL) study (a community-based, cross-sectional cohort comprising 768 healthy controls (HC), 133 participants with mild cognitive impairment (MCI) and 211 participants with AD). We found that individuals with AD had significantly lower hemoglobin, mean cell hemoglobin concentrations, packed cell volume and higher erythrocyte sedimentation rates (adjusted for age, gender, APOE-ɛ4 and site). In AD, plasma iron, transferrin, transferrin saturation and red cell folate levels exhibited a significant distortion of their customary relationship to hemoglobin levels. There was a strong association between anemia and AD (adjusted odds ratio (OR)=2.43, confidence interval (CI) (1.31, 4.54)). Moreover, AD emerged as a strong risk factor for anemia on step-down regression, even when controlling for all other available explanations for anemia (adjusted OR=3.41, 95% CI (1.68, 6.92)). These data indicated that AD is complicated by anemia, which may itself contribute to cognitive decline.


Worldwide, 115.4 million cases of Alzheimer's disease (AD) are expected by 2050,1 an impetus to understand this incurable dementia. AD and unexplained anemia are the most common diagnoses in nursing homes, both with a prevalence of ≈45% in this setting.2,3 Several reports have associated lower hemoglobin and anemia on a cross-sectional basis with poorer cognition,4, 5, 6 and recently low hemoglobin has featured on a biomarker panel with high diagnostic accuracy for AD.7 Two prospective studies of elderly populations have reported that anemia and lower levels of hemoglobin were associated with an approximately twofold increased hazard for developing AD over ≈3 years.2,8 These studies speculated that low hemoglobin could be a systemic manifestation of AD.

Hemoglobin is the most abundant iron-protein in the body but is also synthesized in cortical neurons, where iron metabolism is severely perturbed in AD. Neocortical neurons in AD have decreased hemoglobin levels,9 and elevated iron levels that promote oxidative damage.10,11 Genetic evidence supporting a role for abnormal iron-associated metabolism in AD includes a risk polymorphism of hemoglobin12 and a synergistic risk effect for variants of iron-regulatory genes transferrin with HFE.13,14 Amyloid precursor protein has an important role in cellular iron export.11,15 An increase in IRP1 binding to the amyloid precursor protein iron responsive element in erythrocytes in AD compared to age-matched controls indicates that the central abnormality of hemoglobin and iron regulation in AD might be reflected in the blood.16

In the present study, we investigated the association of AD with anemia and the causes of anemia in AD. We analyzed iron-related blood biochemistry, hemoglobin and other blood chemistry values that may impact on anemia in the large Australian Imaging Biomarker and Lifestyle (AIBL) cohort of healthy controls (HC), mild cognitive impairment (MCI) and AD patients.

Materials and methods

Study design and participants

Recruitment and characteristics of this cohort were previously described.17 AIBL is a two-site (Melbourne, Perth, Australia), prospective study of aging, integrating data from neuroimaging, biomarkers, clinical and neuropsychological measurements and lifestyle information. Volunteers are in cohorts of: (a) AD by NINCDS-ADRDA criteria, (b) Mild cognitive impairment (MCI, associated with an increased risk for developing AD) and (c) cognitively healthy individuals (healthy controls, HC) (Supplementary Material, sections Participant Recruitment and Cohort Size for details).

AIBL was approved by the institutional ethics committees of Austin Health, St. Vincent’s Health, Hollywood Private Hospital, and Edith Cowan University. Written informed consent was obtained from all study participants.


Baseline blood samples were taken from overnight fasting participants, and fractionated or forwarded to clinical pathology laboratories for analysis as described1 (Supplementary Material, section Biochemistry for details). Anemia diagnosis is detailed in Supplementary Material. All participants were given the Food Frequency Questionnaire developed by the Cancer Council of Victoria (Supplementary Material, Iron and Folate Dietary Intake for details).

Statistical analysis

Analysis was performed with R version 2.15.2 (packages, Supplementary Material, sections Statistical Analysis). Pearson’s χ2 assessed differences in the distributions of the missing data, genders and APOE-ɛ4 incidence. Analysis of variance (ANOVA) assessed difference in age across the clinical classifications, followed by Tukey honest significant differences to test the pair-wise comparisons.

Clinical pathology data were checked for normality by inspection of histograms and quartile-quartile plots. For those that deviated from normality, Box–Cox18 analysis was performed. The transformed data were then checked for normality.

Analysis of covariance (ANCOVA) was used with age, gender and site as confounding variables (APOE-ɛ4 genotype was initially included, but was identified as a non-significant main effect and was not included in the final models) to test the difference between the clinical classifications and continuous data. False discovery rate19 (FDR) was used to correct for multiple testing. For the analysis of iron, dietary iron was also included as a confounding variable. The analysis of both serum and red cell folate were also adjusted for dietary folate. The results presented are the adjusted means and standard errors. The pair-wise analysis was corrected for multiple testing by controlling for FDR. Adjusted P-values are presented (Supplementary Material, section Statistical Analysis for details).

Fisher Exact tests were used to analyze the number of subjects above or below a given reference range between two groups. The P-values reported are FDR-adjusted for multiple testing.

Logistic regression was used to assess the odds ratio (OR) of having AD given the anemia state (true/false), and vice versa, that is, OR of being anemic based on the cognitive classification (HC/AD) (Supplementary Material, section Statistical Analysis for details). To compare the log odds estimates between two logistic regression models, model with all subjects included and one with the subject with anemia of inflammation removed, bootstrapping with 1000 replicates was performed.


Data from the initial baseline cohort17 were analyzed. The HC group was significantly younger than the MCI and AD groups (P<0.001), and the MCI group was significantly younger than the AD group (P<0.01). There were significantly more females in the combined cohort, particularly in the HC and AD groups. The AD group had significantly more APOE-ɛ4 allele carriers compared to the HC (P<0.001) and the MCI (P=0.033) groups, and the MCI group had significantly more ɛ4 carriers than the HC group (P<0.001) (Supplementary Table 1). There were small numbers of missing biochemical data, but not biased to any clinical group (Supplementary Figure 1).

We initially examined 23 clinicopathology test results deemed relevant to red cell synthesis and iron metabolism. None of the adjusted (gender, site, age) mean values lay outside of the clinical reference ranges, but ANCOVA revealed significant differences across the three clinical groups for hemoglobin, mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), packed cell volume (PCV), red cell folate (decreased in AD), erythrocyte sedimentation rate (ESR) and haptoglobin (increased in AD), with a trend of decreased plasma iron in AD (Table 1). While transferrin was not significantly different across the clinical groups (P=0.905), transferrin levels declined more steeply with age in AD (P=0.045) (Supplementary Figure 2).

Table 1 Hematological and associated clinical chemistry mean values in the clinical categories

Pair-wise comparison of these eight values of interest across the clinical groups (Figures 1a–h) revealed that iron, hemoglobin, MCH, MCHC, and PCV were significantly decreased in AD compared to HC, and ESR was significantly elevated in AD compared to HC. MCI values lay between the HC and AD values for iron, hemoglobin, MCH and MCHC. Only for MCHC were the differences significant between HC and MCI as well as AD and MCI (HC>MCI>AD). ESR and PCV were significantly different only between MCI and AD (MCI≈HC). In the ANCOVA for plasma iron, dietary iron intake was not a significant confounder and was removed from the final model. While haptoglobin showed a significant difference across the clinical groups, the individual pair-wise comparisons showed only a trend towards increase in AD (Figure 1h).

Figure 1

Changes in hematological and associated clinicopathology adjusted values in Alzheimer’s disease. Shown are pair-wise comparisons of clinicopathology values that are significantly different (on analysis of covariance (ANCOVA), Table 1) or showing a trend (iron) across the cohorts. (a) iron, (b) hemoglobin, (c) mean cell hematocrit, (d) mean cell hemoglobin concentration, (e) erythrocyte sediment rate, (f) packed cell volume, (g) red cell folate, previously reported,28 (h) haptoglobin, (i) the ratio of red cell to serum folate (rfol/sfol). Shown are the adjusted means (±s.e.) and P-values from ANCOVA adjusted for site, age and gender. HC, healthy controls; MCI, mild cognitive impairment, AD, Alzheimer's disease.

PowerPoint slide

Since there was a significant decrease in hemoglobin levels and red cell hemoglobin content in the AD cohort, we analyzed the association between anemia (WHO criteria20: males hemoglobin<130 g l−1, females hemoglobin<120 g l−1) and AD by logistic regression, finding a strong unadjusted association (OR=5.94, 95%CI (3.13, 9.94), P<0.001). Adjusted for known major AD risk factors, age, gender and APOE-ɛ4 status, the OR for being diagnosed as AD if anemic in this cohort was 2.43 (95%CI (1.30, 4.54), P=0.005) (Table 2a), consistent with previous reports.4,8,21. Conversely, higher MCHC was protective for AD (OR=0.62, 95%CI(0.48–0.80), P<0.001, adjusted for age, gender and APOE-ɛ4 status).

Table 2 Logistic regression analysis of risk factors

We then interrogated the reciprocal OR, for being anemic if diagnosed as AD, and controlled for confounders for the risk of anemia. As expected, increasing red cell distribution width, urea, ESR and erythropoietin, and decreasing WCC, iron and ceruloplasmin all showed an association with anemia within the combined HC/AD cohort (Table 2b). Surprisingly, increasing ferritin was associated with anemia, as was increasing red cell folate (but only in AD cases). AD emerged as a strong risk factor for anemia, even when controlling for all other available explanations for anemia (adjusted OR=3.41, 95%CI (1.68, 6.92), P<0.001) (Table 2b). Neither iron nor folate dietary intake were confounders. Similarly, although zinc and copper deficiency are uncommon causes of anemia, plasma zinc and copper were not significant terms in the OR model (not shown). Non-steroidal anti-inflammatory drugs (NSAIDs) may contribute to anemia by promoting GI bleeding, although we found no evidence of blood loss to account for the increased risk of anemia in the AD cohort. Nevertheless, χ2 tests indicated no bias in the number of subjects on NSAIDs compared to those not on NSAIDs between the anemic and non-anemic groups within AD and HC cohorts (P=0.552, P=0.200, respectively, Supplementary Table S2). Acetylcholinesterase inhibitors can also potentially cause GI bleeding (by Mallory–Weiss syndrome). χ2 testing showed no bias in the usage of acetylcholinesterase inhibitors between anemic AD and non-anemic AD subjects (P=0.204, Supplementary Table S3).

Taken together, these data indicate that AD is associated with a proclivity to anemia that could not be explained by known risk factors. To explore possible explanations for the proclivity towards anemia identified in the AD cohort, we analyzed the proportional prevalence of abnormal clinical pathology results (outside the reference range). These were indeed more prevalent in the AD group compared to the HC and MCI groups (13 out of 29 tests, expanded to include white cell sub-population counts, Supplementary Table 4), with no abnormality being more prevalent in the HC cohort. Abnormal results in MCI patients lay in prevalence between HC and AD abnormalities.

The older age of the AD cohort could account for the increased prevalence of abnormal tests, so we created age-matched MCI and HC groups (Supplementary Table 1B) and repeated this analysis. The prevalence of abnormalities for the AD cohort remained significantly greater than HC for low levels of hemoglobin (18.54 vs 7.47%, AD:HC), PCV (20.49 vs 12.03%), transferrin saturation (9.27 vs 3.78%), serum iron (5.85 vs 1.26%), and serum folate (8.33 vs 1.67%), with elevations of ESR (46.70 vs 33.33%), haptoglobin (42.51 vs 30.25%), ferritin (17.56 vs 7.98%) and TSH (9.80 vs 3.77%) (Table 3). Differences in dietary intake could not account for the increased prevalence of serum folate and iron abnormalities observed.

Table 3 Frequency of abnormal clinicopathology results in age-matched clinical cohorts

Confining this analysis to age-matched anemia cases only, no clinical pathology abnormality was significantly more prevalent between HC and AD cohorts (Supplementary Table 5). ANCOVA of the hematology and biochemical data, controlling for age and gender, revealed small differences between anemic AD and HC subjects (Supplementary Table 6) for MCHC, lymphocyte count, bilirubin, and transferrin saturation (males only), but did not reveal a clearly different signature between the AD and HC anemic subjects.

As reticulocyte counts were not available, we used plasma erythropoietin levels as an indication of drive for red cell production. As expected, elevated erythropoietin levels were observed in the anemic cohort (P=0.003, Supplementary Figure 3), and associated with anemia risk (OR=1.45 CI (1.10, 1.92), Table 2b). ANCOVA, controlling for age, gender, site and clinical classification, revealed no significant difference in erythropoietin across the neurological categories (P=0.103, Table 1).

When anemia cases were categorized on the basis of clinical pathology abnormalities, anemia of chronic inflammation (AI) was the only anemia category significantly more prevalent in the AD group (28.9% of types of anemia) compared with the age-matched HC group (16.7%, P<0.001; Supplementary Figure 4 and Supplementary Table 7).

AI is characterized by an elevation of interleukin-6 (IL-6).22 However, IL-6 was not elevated in the AD cohort (Table 1). ANCOVA (adjusted for age, gender and site) revealed an elevation of IL-6 in subjects with anemia (P=0.005), but no interaction between IL-6, anemia and neurological category (P=0.826). To determine the extent that the increased risk for anemia in the AD group is accounted for by AI, we compared two logistic models, model 1 (m1) using all subjects and model 2 (m2) where the AI subjects were removed. There was no change in the AD OR for anemia (OR difference (m1−m2) −0.181 (95% CI (−1.79, 1.67), P=0.807)) when the AI subjects were removed. The risk for anemia in the AD cohort using m2 was still strong (adjusted OR=3.59, 95% CI (1.68, 6.95), P=0.002). Therefore, AI does not explain the increased prevalence of anemia in AD.

These findings suggest that AD patients may have a defect in hemoglobin synthesis of unknown etiology. To characterize this, we examined the relationships between levels of hemoglobin and its substrates: plasma iron, transferrin, transferrin saturation, haptoglobin and red cell folate. Using multiple regression analysis (controlling for age, gender and clinical classification), the relationship between hemoglobin (as a product) and each factor (as a precursor) showed an inverted-U-shaped quadratic association (P<0.001) for each clinical cohort. The associations of hemoglobin as a product of either plasma iron, transferrin level, transferrin saturation or red cell folate were all significantly altered in the AD group (P<0.001, Figure 2) in that hemoglobin levels dropped precipitously as the substrate levels approached the margins of the normal range. The haptoglobin–hemoglobin relationship was not affected by clinical classification (P=0.5472). For the relationship between hemoglobin and red cell folate, there was a steeper decline in hemoglobin levels in AD patients who had red cell folate levels above 1000 nmol l−1 (Figure 2d), recapitulating the surprising observation that elevated red cell folate increased the risk for anemia in AD patients only (Table 2b). These relationships were not significantly perturbed in the MCI cohort.

Figure 2

The relationship of blood hemoglobin to precursor levels is perturbed in AD. Lines of best fit from linear regression models for hemoglobin as a product of (a) plasma iron, (b) plasma transferrin (quadratic relationship), (c) transferrin saturation, (d) red cell folate. Each shows a quadratic relationship. For all plots, age is fixed at the population mean (72.2 years). The gray shaded regions are the 95% confidence intervals. Solid lines—healthy controls, dotted lines—mild cognitive impairment, dashed lines—Alzheimer's disease. The red vertical lines indicate the min and max of the reference ranges.

PowerPoint slide

The puzzling association of higher levels of red cell folate with anemia (Table 2b, Figure 2d), together with significantly lower red cell folate levels (−13%, Table 1, Figure 1g) in the AD cohort, could reflect anomalous folate metabolism in AD. Serum folate levels were not significantly changed in AD (Table 1). So we analyzed the age-matched cohort for participants who had the mismatched combination of abnormally low serum folate with normal (or elevated) red cell folate. This combination of results was indeed uncommon in the HC group (0.5%) but was significantly more frequent in the MCI (3.8%, P=0.005) and AD (6.6%, P<0.001) groups (Supplementary Table 8). Indeed, in AD it was more common than abnormally low red cell folate (Supplementary Table 8). In addition, the mean ratio of red cell to serum folate, reflecting the exchange of folate between intracellular and extracellular blood compartments, was significantly elevated in anemic AD subjects (Figure 1i), with no change in anemic HC and MCI subjects, Since nutritional intake of folate did not explain the increased prevalence of anemia in AD, these data implicate altered exchange of folate between serum and red cells as potentially contributory to the increased risk for anemia in AD.


Anemia is an established risk factor for cognitive loss, causally attributed to low oxygenation of obligatorily aerobic cortical tissue.2,4,5,21 In the AIBL cohort, people with anemia were 2.40-fold more likely to have AD (adjusting for age and APOE-ɛ4), and anemia emerged as a stronger potential risk factor for AD than aging 5 years (Table 2a). These findings are consistent with a prospective study that reported anemia as a risk factor for AD, accelerating cognitive decline to a degree similar to aging 12 years.8 Therefore, identifying and treating anemia in older age could be imperative. However, our findings indicate that management of anemia in AD may not be straightforward. AI was more common in the AD cohort, but neither AI nor other known causes could explain the increased risk for anemia or lower hemoglobin levels in AD.

Idiopathic anemia becomes prevalent in advanced age, and indeed, in the AIBL cohort, reversible causes of anemia were rare and idiopathic anemia was the most common cause (Supplementary Figure 4 and Supplementary Table 7). Is it possible that an impairment of hemoglobin production is a systemic manifestation of AD, and that as subclinical AD pathology also becomes prevalent above the age of 60, could idiopathic anemia itself be etiologically related to AD? Abnormalities of red cells have been reported in studies of smaller AD cohorts.23, 24, 25, 26 We found that on average, AD red cells are smaller, have less concentrated hemoglobin, sediment faster and have lower folate levels than HC red cells (Figure 1). For these results, MCI red cells have values between AD and HC.

As with idiopathic anemia, we could not ascertain a clear etiological explanation for the lowering of hemoglobin in AD, but we detected changes in systemic iron, folate metabolism, thyroid, and inflammation that could adversely impact hemoglobin production. Importantly, the relationships between hemoglobin and its iron-related precursors (serum iron, transferrin and transferrin saturation) were perturbed in AD (Figures 2a and c). Unlike HC and MCI, hemoglobin levels in AD declined significantly unless serum iron, transferrin and transferrin saturation were in the middle of the normal range. Therefore, iron-related hemoglobin production may be brittle in AD. The significantly greater prevalence of abnormally high serum ferritin in AD (17.65%) compared to age-matched HC subjects (7.98%, Table 3) together with significantly lower plasma iron levels (Figure 1a) and increased prevalences of abnormally low serum iron and transferrin saturation (Table 3) in AD, argue that iron accumulates in tissue stores and is not adequately mobilized in AD. This is consistent with iron accumulation in the neocortex in AD.11

Folate deficiency is a risk factor for both dementia and anemia, and while we observed folate changes in AD, they could not be accounted for by nutritional deficiency or medication. Serum folate levels were not significantly changed In the AD group compared to the age-matched HC group, but subnormal levels were significantly more prevalent (8.33% vs 1.67%, Table 3). This might be consistent with a nutritional problem, but red cell folate is a better reflection of long-term deficiency27 and while red cell folate levels were decreased in the AD group (Figure 1g),28 subnormal levels were uncommon and not more prevalent in AD than in HC (Table 3). Alternatively, a systemic defect in folate metabolism that could contribute to the increased prevalence of anemia in AD was evidenced by the brittle relationship between hemoglobin levels and red cell folate in AD (Figure 2d), as well as the puzzling observation that increasing red cell folate strongly increases the risk for anemia in AD but not HC subjects (Table 2b). Abnormal folate metabolism has been implicated in AD pathogenesis. A genome-wide association study identified a significant association between polymorphisms of methylenetetrahydrofolate dehydrogenase (NADP+ dependent) 1-like protein and the risk for late-onset AD.29 A proteomic study found that methylenetetrahydrofolate dehydrogenase 1 levels are 32-fold elevated in erythrocytes in AD compared to controls.26 Folate is susceptible to oxidation, notably by elevated cytosolic iron.30

Hypothyroidism might also contribute to the increased incidence of anemia in AD since abnormally elevated TSH was more common (9.80%) than age-matched HC (3.77%, Table 3). However, adjusted mean TSH levels were not lower in the AD cohort, indicating that the contribution of hypothyroidism to anemia in AD is limited.

Inflammation may have a role in lowering hemoglobin in AD, since we observed elevated ESR, suppressed serum iron, elevated ferritin and suppressed transferrin in the AD cohort (Tables 1 and 3), and AI was the only diagnostic category for anemia that was significantly more prevalent (+12.2%) in the AD cohort compared to HC (Supplementary Figure 4 and Supplementary Table 7). However, logistic regression analysis excluding AI cases still determined that AD was a risk factor for anemia, and inflammation does not explain the differential tendency to lower hemoglobin levels in AD that we have observed. However, hepcidin and CRP levels in AD may be valuable in appraising the tendency to anemia.

Therefore, the etiology of the 3.41-fold increase in anemia in AD involves perturbances of multiple systems for hemoglobin synthesis. Hemolysis could be excluded since haptoglobin was elevated in AD (Table 1), as others have also reported.31 While the mechanism of this idiopathic hemoglobin defect in AD remains to be determined, we note that Aβ binds hemoglobin,32 and hemoglobin accumulates in the hallmark cortical amyloid plaque and congophilic angiopathy of AD.33 Aβ is enriched in red cell membranes in AD compared to healthy controls,34,35 and oxidizes red cell hemoglobin,36 potentially contributing to anemia in AD.

With the caveat that this is a cross-sectional study and the subjects were recruited via advertising, and so not a true random sample of the Australian population, our findings identify AD as a novel risk factor for anemia. As low hemoglobin itself impairs cognition and is a risk factor for AD, the hemoglobin lesion of AD may not be an epiphenomenon. The interrelationship between AD and hemoglobin warrants closer investigation as a potential target for intervention.


  1. 1

    Prince M, Jackson J . World Alzheimer report 2009. Alzheimer's disease International, London, UK;

  2. 2

    Atti AR, Palmer K, Volpato S, Zuliani G, Winblad B, Fratiglioni L . Anaemia increases the risk of dementia in cognitively intact elderly. Neurobiol Aging 2006; 27: 278–284.

    CAS  Article  Google Scholar 

  3. 3

    Van Rensbergen G, Nawrot T . Medical conditions of nursing home admissions. BMC Geriatr 2010; 10: 46.

    Article  Google Scholar 

  4. 4

    Shah RC, Wilson RS, Tang Y, Dong X, Murray A, Bennett DA . Relation of hemoglobin to level of cognitive function in older persons. Neuroepidemiology 2009; 32: 40–46.

    Article  Google Scholar 

  5. 5

    Deal JA, Carlson MC, Xue Q-L, Fried LP . Chaves PHM. Anemia and 9-year domain-specific cognitive decline in community-dwelling older women: The Women's Health and Aging Study II. J Am Geriatr Soc 2009; 57: 1604–1611.

    Article  Google Scholar 

  6. 6

    Denny SD, Kuchibhatla MN, Cohen HJ . Impact of anemia on mortality, cognition, and function in community-dwelling elderly. Am J Med 2006; 119: 327–334.

    Article  Google Scholar 

  7. 7

    Doecke JD, Laws SM, Faux NG, Wilson W, Burnham SC, Lam CP et al. Blood-based protein biomarkers for diagnosis of Alzheimer disease. Arch Neurol 2012; 69: 1318–1325.

    Article  Google Scholar 

  8. 8

    Shah RC, Buchman AS, Wilson RS, Leurgans SE, Bennett DA . Hemoglobin level in older persons and incident Alzheimer disease: prospective cohort analysis. Neurology 2011; 77: 219–226.

    CAS  Article  Google Scholar 

  9. 9

    Ferrer I, Gomez A, Carmona M, Huesa G, Porta S, Riera-Codina M et al. Neuronal hemoglobin is reduced in Alzheimer's disease, Argyrophilic Grain disease, Parkinson's disease, and dementia with Lewy bodies. J Alzheimers Dis 2011; 23: 537–550.

    CAS  Article  Google Scholar 

  10. 10

    Smith MA, Harris PL, Sayre LM, Perry G . Iron accumulation in Alzheimer disease is a source of redox-generated free radicals. Proc Natl Acad Sci USA 1997; 94: 9866–9868.

    CAS  Article  Google Scholar 

  11. 11

    Duce JA, Tsatsanis A, Cater MA, James SA, Robb E, Wikhe K et al. Iron-export ferroxidase activity of beta- amyloid precursor protein is inhibited by zinc in Alzheimer's disease. Cell 2010; 142: 857–867.

    CAS  Article  Google Scholar 

  12. 12

    Perry RT, Gearhart DA, Wiener HW, Harrell LE, Barton JC, Kutlar A et al. Hemoglobin binding to A beta and HBG2 SNP association suggest a role in Alzheimer's disease. Neurobiol Aging 2008; 29: 185–193.

    CAS  Article  Google Scholar 

  13. 13

    Kauwe JSK, Bertelsen S, Mayo K, Cruchaga C, Abraham R, Hollingworth P et al. Suggestive synergy between genetic variants in TF and HFE as risk factors for Alzheimer's disease. Am J Med Genet B Neuropsychiatr Genet 2010; 153B: 955–959.

    CAS  Article  Google Scholar 

  14. 14

    Lehmann DJ, Schuur M, Warden DR, Hammond N, Belbin O, Kölsch H et al. Transferrin and HFE genes interact in Alzheimer's disease risk: the Epistasis Project. Neurobiol Aging 2012; 33: 202.e1–202.e13.

    CAS  Article  Google Scholar 

  15. 15

    Lei P, Ayton S, Finkelstein DI, Spoerri L, Ciccotosto GD, Wright DK et al. Tau deficiency induces parkinsonism with dementia by impairing APP-mediated iron export. Nat Med 2012; 18: 291–295.

    CAS  Article  Google Scholar 

  16. 16

    Cho HH, Cahill CM, Vanderburg CR, Scherzer CR, Wang B, Huang X et al. Selective translational control of the Alzheimer amyloid precursor protein transcript by iron regulatory protein-1. J Biol Chem 2010; 285: 31217–31232.

    CAS  Article  Google Scholar 

  17. 17

    Ellis K, Bush A, Darby D, De Fazio D, Foster J, Hudson P et al. The Australian Imaging, Biomarkers and Lifestyle (AIBL) study of aging: methodology and baseline characteristics of 1112 individuals recruited for a longitudinal study of Alzheimer's disease. Int Psychogeriatr 2009; 21: 672–687.

    Article  Google Scholar 

  18. 18

    Box G, Cox D . An analysis of transformations. J R Stat Soc Ser B 1964; 26: 211–243.

    Google Scholar 

  19. 19

    Benjamini Y, Hochberg Y . Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B 1995; 57: 289–300.

    Google Scholar 

  20. 20

    World Health Organization. Iron Deficiency Anaemia: Assessment, Prevention, and Control: A Guide For Programme Managers. UNICEF United Nations University: WHO, Geneva, Switzerland, 2001.

  21. 21

    Beard CM, Kokmen E, O'Brien PC, Anía BJ, Melton LJ . Risk of Alzheimer's disease among elderly patients with anemia: population-based investigations in Olmsted County, Minnesota. Ann Epidemiol 1997; 7: 219–224.

    CAS  Article  Google Scholar 

  22. 22

    Maccio A, Madeddu C, Massa D, Mudu MC, Lusso MR, Gramignano G et al. Hemoglobin levels correlate with interleukin-6 levels in patients with advanced untreated epithelial ovarian cancer: role of inflammation in cancer-related anemia. Blood 2005; 106: 362–367.

    CAS  Article  Google Scholar 

  23. 23

    Ajmani RS, Metter EJ, Jaykumar R, Ingram DK, Spangler EL, Abugo OO et al. Hemodynamic changes during aging associated with cerebral blood flow and impaired cognitive function. Neurobiol Aging 2000; 21: 257–269.

    CAS  Article  Google Scholar 

  24. 24

    De Franceschi L, Olivieri O, Corrocher R . Erythrocyte aging in neurodegenerative disorders. Cell Mol Biol 2004; 50: 179–185.

    CAS  PubMed  Google Scholar 

  25. 25

    Mohanty JG, Eckley DM, Williamson JD, Launer LJ, Rifkind JM . Do red blood cell-beta-amyloid interactions alter oxygen delivery in Alzheimer's disease? Adv Exp Med Biol 2008; 614: 29–35.

    CAS  Article  Google Scholar 

  26. 26

    Mohanty JG, Shukla HD, Williamson JD, Launer LJ, Saxena S, Rifkind JM . Alterations in the red blood cell membrane proteome in Alzheimer's subjects reflect disease-related changes and provide insight into altered cell morphology. Proteome Sci 2010; 8: 11.

    Article  Google Scholar 

  27. 27

    Amos R, Dawson D, Fish D, Lemming R, Linnell J . Guidelines on the investigation and diagnosis of cobalamin and folate deficiencies. A publication of the British Committee for Standards in Haematology. BCSH General Haematology Test Force. Clin Lab Haematol 1994; 16: 101–115.

    Article  Google Scholar 

  28. 28

    Faux NG, Ellis KA, Porter L, Fowler CJ, Laws SM, Martins RN et al. Homocysteine, vitamin B12, and folic acid levels in Alzheimer's disease, mild cognitive impairment, and healthy elderly: baseline characteristics in subjects of the Australian Imaging Biomarker Lifestyle study. J Alzheimers Dis 2011; 27: 909–922.

    CAS  Article  Google Scholar 

  29. 29

    Naj AC, Beecham GW, Martin ER, Gallins PJ, Powell EH, Konidari I et al. Dementia revealed: novel chromosome 6 locus for late-onset Alzheimer disease provides genetic evidence for folate-pathway abnormalities. PLoS Genet 2010; 6: 9e1001130.

    Article  Google Scholar 

  30. 30

    Suh JR, Oppenheim EW, Girgis S, Stover PJ . Purification and properties of a folate-catabolizing enzyme. J Biol Chem 2000; 275: 35646–35655.

    CAS  Article  Google Scholar 

  31. 31

    Zhang R, Barker L, Pinchev D, Marshall J, Rasamoelisolo M, Smith C et al. Mining biomarkers in human sera using proteomic tools. Proteomics 2004; 4: 244–256.

    CAS  Article  Google Scholar 

  32. 32

    Dyrks T, Dyrks E, Hartmann T, Masters C, Beyreuther K . Amyloidogenicity of beta A4 and beta A4-bearing amyloid protein precursor fragments by metal-catalyzed oxidation. J Biol Chem 1992; 267: 18210–18217.

    CAS  Google Scholar 

  33. 33

    Chuang JY, Lee CW, Shih YH, Yang T, Yu L, Kuo YM . Interactions between amyloid-beta and hemoglobin: implications for amyloid plaque formation in Alzheimer's disease. PLoS One 2012; 7: e33120.

    CAS  Article  Google Scholar 

  34. 34

    Rogers J, Li R, Mastroeni D, Grover A, Leonard B, Ahern G et al. Peripheral clearance of amyloid beta peptide by complement C3-dependent adherence to erythrocytes. Neurobiol Aging 2006; 27: 1733–1739.

    CAS  Article  Google Scholar 

  35. 35

    Villemagne VL, Perez KA, Pike KE, Kok WM, Rowe CC, White AR et al. Blood-borne amyloid-beta dimer correlates with clinical markers of Alzheimer's disease. J Neurosci 2010; 30: 6315–6322.

    CAS  Article  Google Scholar 

  36. 36

    Jayakumar R, Kusiak JW, Chrest FJ, Demehin AA, Murali J, Wersto RP et al. Red cell perturbations by amyloid beta-protein. Biochim Biophys Acta 2003; 1622: 20–28.

    CAS  Article  Google Scholar 

Download references


Core funding for the study was provided by CSIRO, which was supplemented by in-kind contributions from the study partners (see The research was also supported by the Science Industry and Endowment Fund (see, the Cooperative Research Center for Mental Health, and the National Health and Medical Research Council (NHMRC) via the Dementia Collaborative Research Centres program (DCRC2). The Florey Institute acknowledges the funding support from the Victorian Government’s Operational Infrastructure Support program. NGF is supported by a NHMRC training fellowship. AIB is supported by NHMRC Australia Fellowship. Pfizer International has contributed financial support to assist with analysis of blood samples and to further the AIBL research program. The McCusker Foundation has contributed financial and in-kind support to AIBL. Alzheimer’s Australia (Victoria and Western Australia) assisted with promotion of the study and the screening of telephone calls from volunteers. The AIBL team thanks the following clinicians who referred patients with AD and/or MCI to the study: Associate Professor Brian Chambers, Professor Edmond Chiu, Dr Roger Clarnette, Associate Professor David Darby, Dr Mary Davison, Dr John Drago, Dr Peter Drysdale, Dr Jacqui Gilbert, Dr Kwang Lim, Professor Nicola Lautenschlager, Dr Dina LoGiudice, Dr Peter McCardle, Dr Steve McFarlane, Dr Alastair Mander, Dr John Merory, Professor Daniel O’Connor, Dr Ron Scholes, Dr Mathew Samuel, Dr Darshan Trivedi, Dr Peter Panegyres and Associate Professor Michael Woodward.

Author information




Corresponding author

Correspondence to A I Bush.

Ethics declarations

Competing interests

AIB and CLM are shareholders in Prana Biotechnology Limited. AIB is a shareholder in Cogstate Limited and Mesoblast Limited.

Additional information

Supplementary Information accompanies the paper on the Molecular Psychiatry website

Supplementary information

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Faux, N., Rembach, A., Wiley, J. et al. An anemia of Alzheimer's disease. Mol Psychiatry 19, 1227–1234 (2014).

Download citation


  • alzheimer's disease
  • anemia
  • folate
  • hematology
  • hemoglobin
  • iron

Further reading