Iron is an essential nutrient, but like any good thing, too much of it may do harm. According to an autopsy study published February 18 in Molecular Psychiatry, people who had had dementia and moderate to high burdens of plaques and tangles had more iron in their temporal cortices than those with less pathology. Does this iron do anything? It correlated with cognitive decline in the years prior to death, but whether it accelerated that decline is unclear. The researchers, led by Ashley Bush of the University of Melbourne and Martha Morris of Rush University Medical Center in Chicago, suggest that brain iron kills neurons via ferroptosis, a cell-death pathway driven by reactive forms of the element.

  • Only people with clinical and pathological hallmarks of Alzheimer’s disease had more cortical iron.
  • Among people with AD pathology, cortical iron correlated with cognitive decline.
  • This means iron rises late in AD pathogenesis.

Bush and colleagues had previously cast brain iron, in the form of ferritin, as a predictor of Alzheimer’s disease progression, and proposed that the element facilitates the genetic risk imposed by ApoE4 (Ayton et al., 2015). However, the current study found normal brain levels of iron in people who had high plaque and tangle burden but no dementia, suggesting the element ticks up late in the disease process. Because iron levels were only measured at autopsy, the study leaves unclear when in the course of disease iron levels started to rise.

The role of iron in AD has long been debated but never drawn much consensus in the field. Recent work has tried to link markers of brain iron and cognitive decline, and neuropathological studies have detected more iron in the brains of people with AD than controls, but the studies were too small to tease out the relationship between cortical iron, AD neuropathology, and clinical symptoms (Tao et al., 2014). 

First author Scott Ayton and colleagues used data from the Religious Orders Study and Memory and Aging Project cohorts (ROSMAP), in which participants were tracked for biomarker and cognitive changes during life, and their brains autopsied. The researchers quantified iron levels in gray matter from the inferior temporal cortices and cerebella of 209 people in the cohort. They found more iron in the cortex, but not cerebellum, in people who had a moderate to high burden of plaques and tangles, but only if they had been living with dementia. People who had been cognitively normal had normal iron levels, even if they had substantial AD pathology. Iron levels were also normal among 14 people who had been given a clinical dementia diagnosis but had little to no AD pathology. Ayton et al. also found that in people with high burdens of plaques and tangles, as assessed by National Institute on Aging/Reagan scale, cortical iron levels correlated with the rate of cognitive decline in their last 10 to 12 years of life. However, AD pathological burden itself was by far the greatest predictor of antemortem cognitive decline.

“While excess iron has been implicated in AD either as a cause or result for several decades, this study is one of the first reporting that the amount of increased iron correlates to the rate of cognitive decline,” commented Jeff Bulte of Johns Hopkins School of Medicine in Baltimore.

What it means is unclear. Iron might be an independent marker of synaptic loss or neuronal death, or it might somehow contribute to either or both, the authors suggest. Among people with substantial AD pathology, iron did not correlate with plaque burden, but did correlate with tangle burden, as judged by silver staining of sections from multiple cortical regions. However, a statistical mediation analysis indicated iron correlated with cognitive decline largely independently of tau.

In a joint comment to Alzforum, Neil Telling of Keele University and Joanna Collingwood of Warwick University, both in the U.K., wrote that the lack of a strong correlation between brain iron and neuropathological burden was expected. They believe the chemical form of iron, as opposed to its total amount, is important for toxicity. For example, they cited work from their group indicating that Aβ peptides interact with iron and trigger its reduction into a ferrous form, which can react with peptides to produce hydroxyl radicals (Collingwood et al., 2008; Everett et al., 2018). “It is likely to be the amount of reactive iron that is available to drive overproduction of radical species that is the crucial factor, rather than the total amount of iron at that site,” they wrote.

Bush noted that the iron associated with Aβ plaques represents but a small fraction of the total iron in the brain, most of which is inside cells. He believes this intracellular iron harms neurons, and accumulates because trafficking of amyloid precursor protein (APP), a purported iron transporter, falters in AD (Feb 2012 news; Lumsden et al., 2018). Iron is the key component of the ferroptosis cell-death pathway, in which the metal drives accumulation of harmful lipid hydroperoxides, which Bush said are abundant in the AD brain (Stockwell et al., 2017). 

Bush is involved in a multicenter, Phase 2 clinical trial testing the iron chelator drug deferiprone in people with mild AD (see This follows trials of clioquinol, which were canceled due to a toxic contaminant (Jan 2004 news; Apr 2005 news). The second-generation compound PTB2 failed to slow cognitive decline in people with AD (Apr 2014 news).—Jessica Shugart


  1. This is an interesting study that reinforces prior observations of altered regional brain iron levels in individuals with Alzheimer’s disease compared with age-matched controls (e.g. Tao et al., 2014). Significantly, this study presents new evidence that iron in a specific region of the cortex correlates with the rate of cognitive decline of the subjects. This postmortem evidence is important to inform and validate reported clinical measures of the relationship between cognitive decline and iron levels, determined indirectly via magnetic resonance Imaging (e.g. Du et al., 2018); it demonstrates how brain iron monitoring may be incorporated into disease progression monitoring of Alzheimer’s disease, to track cognitive change in tandem with physicochemical change in the tissues of the brain, and this is particularly important baseline information to evaluate the impact of treatments that affect iron metabolism.

    As Ayton and colleagues discuss in this study, measures of total iron concentration alone do not prove a role for iron in disease progression; in this respect their work evidences correlation rather than causation. Indeed, from the perspective of Alzheimer’s disease as a proteinopathy, it is interesting to note that the present study reports only a weak correlation between iron levels and the hallmark plaques and tangles.

    However, a critical factor that is frequently overlooked is that, with iron as an essential element in the brain, it is only when the tightly regulated trafficking and storage of iron is disrupted that iron might be expected to cause damage to cells and subsequent atrophy. Altered protein expression and/or function may impact bioavailability (e.g. Visanji et al., 2013). To understand the contribution of iron to neurodegeneration, it is important to evaluate the chemical form of the iron, identifying where it changes from its normal biological forms to more chemically reactive forms, creating potential to cause neuronal damage and subsequent atrophy.

    Many previous studies have indicated that within tissues, the total iron concentration alone is not the critical factor. Instead, it is that iron associated with pathological brain features (i.e. the abnormal protein deposits—neuritic plaques and neurofibrillary tangles) is disproportionately in more chemically reactive forms. Of particular importance, there is strong evidence from in vitro, animal model, and human postmortem studies, including work from our group (e.g. Collingwood et al., 2008; Everett et al., 2014; Telling et al., 2017; Everett et al., 2018), that the interaction of peptides such as amyloid-β with iron is responsible for local elevations of chemically reactive iron. Indeed, the mechanisms involving iron that might propel cognitive decline, which Ayton and colleagues consider in their study, are consistent with iron chemistry being perturbed in this way. Therefore, a key finding: that within the regions studied there was only a weak correlation of total iron levels with the pathological burden in the brain tissue, is not unexpected and is consistent with prior work (e.g. House et al., 2008): It is likely to be the amount of reactive iron that is available to drive overproduction of radical species that is the crucial factor, rather than the total amount of iron at that site.

    In summary, the findings of this study from Ayton and colleagues highlight the scope to use brain iron as an important biological marker of Alzheimer’s disease progression. The reported regional iron concentrations, and the weak correlation with pathological burden, reconfirm the value of differentiating the forms of iron present, to determine if there is evidence for disrupted iron bioavailability and chemistry to the detriment of the tissue.


    . Increased Iron Deposition on Brain Quantitative Susceptibility Mapping Correlates with Decreased Cognitive Function in Alzheimer's Disease. ACS Chem Neurosci. 2018 Jul 18;9(7):1849-1857. Epub 2018 May 15 PubMed.

    . Ferrous iron formation following the co-aggregation of ferric iron and the Alzheimer's disease peptide β-amyloid (1-42). J R Soc Interface. 2014 Jun 6;11(95):20140165. Print 2014 Jun 6 PubMed.

    . Nanoscale synchrotron X-ray speciation of iron and calcium compounds in amyloid plaque cores from Alzheimer's disease subjects. Nanoscale. 2018 Jul 5;10(25):11782-11796. PubMed.

    . 1.4T study of proton magnetic relaxation rates, iron concentrations, and plaque burden in Alzheimer's disease and control postmortem brain tissue. Magn Reson Med. 2008 Jul;60(1):41-52. PubMed.

    . Perturbed iron distribution in Alzheimer's disease serum, cerebrospinal fluid, and selected brain regions: a systematic review and meta-analysis. J Alzheimers Dis. 2014;42(2):679-90. PubMed.

    . Iron Biochemistry is Correlated with Amyloid Plaque Morphology in an Established Mouse Model of Alzheimer's Disease. Cell Chem Biol. 2017 Oct 19;24(10):1205-1215.e3. Epub 2017 Sep 7 PubMed.

    . Iron Deficiency in Parkinsonism: Region-Specific Iron Dysregulation in Parkinson's Disease and Multiple System Atrophy. J Parkinsons Dis. 2013 Oct 10; PubMed.

  2. Drs. Telling and Collingwood make excellent points. When measuring elemental composition of tissue in bulk, as we did, we do not achieve more discrimination than total metal levels. The values do not differentiate between iron species, microscopic location, and ligands, each of which is informative, as their previous work has shown. With postmortem tissue, as in our study, oxidation can induce artifacts of transition metal speciation, but nonetheless it is still possible to make meaningful comparisons for the chemical content of disease compared to healthy tissue with adequate controls. This is technically more laborious, but well worth investigating. We hope our current findings will stimulate more work in this area.

  3. The article by Ayton et al. is pivotal to our understanding of iron’s role in Alzheimer’s disease (AD) pathology and dementia. This exciting work connects another important recent research on microRNA (miR-346) to the field. Let us try to iron out the kinks between AD, cognitive decline, iron, and miR-346.

    It is beginning to dawn on us all that AD is not equivalent to the neuropathology associated with it. The separation of pathology from actual cognitive decline may need to be explained before AD can be reliably treated. If the net outcome in life quality and function is the same with or without pathology and depends on other factors, those "factors" may offer a more practical way to treat AD once it is diagnosed. After all, the field’s consensus is approaching the conclusion that it may be too late to treat AD based on pathology-associated molecules once it has advanced enough to be diagnosed. However, if the pathology sets up a necessary but not sufficient condition for actual AD to develop, traits or factors that then produce sufficiency could provide a way out of the “all diagnoses are too late” trap. Ayton et al. (in Ashley Bush’s lab) may have found, if not the escape hatch, at least an escape hatch. That is, deficiencies in Fe efflux could be a functional medical target at any stage of AD, not merely when the disorder is still undetectable.

    In terms of Fe homeostasis, as a brain ages, it may accumulate fragility in the form of AD-associated pathology. However, so long as Fe efflux mechanisms are still vigorous, AD could possibly be avoided. If Fe efflux becomes too weak, the Fe interacts with a fragile brain and produces AD via several mechanisms as discussed in this forum. Even adding additional efflux after the fact could be insufficient to reverse or even halt the damage.

    One important Fe efflux molecule is APP (Venkataramani et al., 2018; Rogers  et al., 2016). It may be counterintuitive to suggest that an APP deficiency that occurs after AD pathology aggregates have formed could contribute to dementia, but some of our work in human brain tissue specimens at different AD stages suggests that changes in APP levels vary by Braak stage in a non-linear fashion (Long et al., 2012). 

    We cannot say ourselves what the levels of APP would be in human brains with no cognitive decline but strong AD pathology versus those with full-fledged AD (pathology plus dementia). Nevertheless, an apparent disruption of Fe metabolism associated with dementia suggests a possible role for APP dysregulation to do more than potentially make more raw material for Aβ.

    Fe-based regulation of APP operates through at least two molecules, iron-response protein 1 (IRP1) (Venkataramani et al., 2018), and microRNA-346 (miR-346) (Long et al., 2019). These two operate antagonistically, with IRP1 suppressing and miR-346 enhancing translation, both at the same site in the APP mRNA 5'-UTR. It is also possible that the above interaction could stabilize the Fe efflux apparatus with other partners. Notably, levels of miR-346 are disrupted in AD brains (Long et al., 2019). We were, therefore, interested to see this unique and novel association between Fe and dementia in high-pathology brains. We hope that such a discovery is further used to investigate these brains’ concurrent levels of miR-346 and APP to test the hypothesis that disrupting the one disrupts the other in full AD (pathology plus dementia), and this disruption associates with Fe elevation and dementia in AD.


    . Manganese causes neurotoxic iron accumulation via translational repression of amyloid precursor protein and H-Ferritin. J Neurochem. 2018 Dec;147(6):831-848. Epub 2018 Nov 19 PubMed.

    . A role for amyloid precursor protein translation to restore iron homeostasis and ameliorate lead (Pb) neurotoxicity. J Neurochem. 2016 Aug;138(3):479-94. PubMed.

    . MicroRNA-153 Physiologically Inhibits Expression of Amyloid-β Precursor Protein in Cultured Human Fetal Brain Cells and Is Dysregulated in a Subset of Alzheimer Disease Patients. J Biol Chem. 2012 Sep 7;287(37):31298-310. PubMed.

    . Novel upregulation of amyloid-β precursor protein (APP) by microRNA-346 via targeting of APP mRNA 5'-untranslated region: Implications in Alzheimer's disease. Mol Psychiatry. 2019 Mar;24(3):345-363. Epub 2018 Nov 23 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Tau Joins APP in the Ironworks
  2. Pilot Study Suggests Clioquinol Benefits AD Patients
  3. Side-Reactions Sideswipe Clioquinol—Prevent Clinical Trial
  4. PBT2 Takes a Dive in Phase 2 Alzheimer’s Trial

Therapeutics Citations

  1. Clioquinol
  2. PBT2

Paper Citations

  1. . Ferritin levels in the cerebrospinal fluid predict Alzheimer's disease outcomes and are regulated by APOE. Nat Commun. 2015 May 19;6:6760. PubMed.
  2. . Perturbed iron distribution in Alzheimer's disease serum, cerebrospinal fluid, and selected brain regions: a systematic review and meta-analysis. J Alzheimers Dis. 2014;42(2):679-90. PubMed.
  3. . Three-dimensional tomographic imaging and characterization of iron compounds within Alzheimer's plaque core material. J Alzheimers Dis. 2008 Jun;14(2):235-45. PubMed.
  4. . Nanoscale synchrotron X-ray speciation of iron and calcium compounds in amyloid plaque cores from Alzheimer's disease subjects. Nanoscale. 2018 Jul 5;10(25):11782-11796. PubMed.
  5. . Dysregulation of Neuronal Iron Homeostasis as an Alternative Unifying Effect of Mutations Causing Familial Alzheimer's Disease. Front Neurosci. 2018;12:533. Epub 2018 Aug 13 PubMed.
  6. . Ferroptosis: A Regulated Cell Death Nexus Linking Metabolism, Redox Biology, and Disease. Cell. 2017 Oct 5;171(2):273-285. PubMed.

External Citations


Further Reading

Primary Papers

  1. . Brain iron is associated with accelerated cognitive decline in people with Alzheimer pathology. Mol Psychiatry. 2019 Feb 18; PubMed.