You are viewing a javascript disabled version of the site. Please enable Javascript for this site to function properly.
Go to headerGo to navigationGo to searchGo to contentsGo to footer
In content section. Select this link to jump to navigation

No Genetic Overlap Between Circulating Iron Levels and Alzheimer’s Disease

Abstract

Iron deposition in the brain is a prominent feature of Alzheimer’s disease (AD). Recently, peripheral iron measures have also been shown to be associated with AD status. However, it is not known whether these associations are causal: do elevated or depleted iron levels throughout life have an effect on AD risk? We evaluate the effects of peripheral iron on AD risk using a genetic profile score approach by testing whether variants affecting iron, transferrin, or ferritin levels selected from GWAS meta-analysis of approximately 24,000 individuals are also associated with AD risk in an independent case-control cohort (n∼10,000). Conversely, we test whether AD risk variants from a GWAS meta-analysis of approximately 54,000 account for any variance in iron measures (n∼9,000). We do not identify a genetic relationship, suggesting that peripheral iron is not causal in the initiation of AD pathology.

INTRODUCTION

Iron is the most abundant metal in the brain, where it is vital for neurotransmitter synthesis, myelination of neurons, and energy generation by mitochondria [1]. However excess iron contributes to the generation of reactive oxygen species, and consequent tissue damage [2]. Dysfunctional brain iron homeostasis is believed to play an important role in Alzheimer’s disease (AD) [3]. Iron accumulation is seen in the AD postmortem brain [4] and iron content correlates with disease duration and Mini-Mental State Examination (MMSE) score [5, 6]. Individuals with mild cognitive impairment (MCI) with high risk of AD, showed higher cortical iron in vivo using MRI (measured using quantitative susceptibility mapping techniques), which spatially co-localized with Aβ plaques and correlated with higher plaque load [7]. In addition, transferrin (an iron transport protein) and ferritin (an intracellular iron storage protein) are both elevated in AD brain tissue in neurodegenerative regions [8]. Ferritin levels in cerebrospinal fluid (CSF) negatively correlated with cognitive performance and predicted conversion from MCI to AD [9]. Ferritin levels were also associated with CSF apolipoprotein E levels and were elevated by the AD risk allele, APOE ɛ4, suggesting that ferritin may reflect the mechanism by which APOE ɛ4 is a risk factor for AD.

Iron trafficking across the blood-brain barrier is tightly regulated and early studies suggested that the brain is protected from systemic fluctuations in iron, with a lack of correlation between liver and brain iron concentrations postmortem [10, 11]. Animal studies went on to challenge this view, showing that excess dietary iron increased brain iron levels in specific brain regions [12]. Quantitative MRI studies measuring the proton transverse relaxation rate (R2) now allow iron concentrations to be assessed in the brain in vivo. One such study in cognitively normal elderly men found that iron levels in basal ganglia structures were correlated with serum iron measures [13]. In an investigation in the large Australian Imaging Biomarker and Lifestyle (AIBL) cohort of healthy controls, MCI and AD patients had disturbed brain iron metabolism reflected in the periphery by a decrease in plasma iron and hemoglobin [14], which was due to a deficiency of iron-loading onto transferrin [15]. Several mechanisms have been suggested to cause dysregulation of iron transport across the blood-brain barrier in AD including the involvement of amyloid-β protein precursor fragments and chronic inflammation [11]. A deficit in brain iron trafficking, which is essential for heme formation, neurotransmitter synthesis, and myelination of axons, could contribute to the pathophysiology of AD. But results are inconsistent, with two meta-analyses having differing conclusions on whether differences in circulating iron levels can be detected between AD cases and controls, and reporting heterogeneity between studies [16, 17].

It is clear that iron dysregulation has a role in AD, and that to a limited extent plasma iron might reflect changes in brain iron levels, but there has been little investigation of whether peripheral iron levels over the long-term affect risk of AD. Apart from the lack of suitable and adequately powered prospective studies, a limitation of observational studies is the inability to distinguish between causal associations and those due to confounding and reverse causation. A systematic review found that, in a limited number of trials, testing whether depletion or supplementation of iron changed a person’s risk of AD provided no conclusive evidence, and that additional studies arenecessary [18].

Drug development and randomized clinical trials are expensive and take many years to reach fruition, especially for a slowly progressive disease where treatment needs to start early in the disease course. An alternative approach, which overcomes the problem of reverse causation, is Mendelian Randomization (MR). Here the genetic variants affecting the putative causal variable are used as instrumental variables to test for an effect on disease risk. A demonstration that genetic polymorphisms known to modify the phenotype level also modify disease risk provides indirect evidence of a causal association between phenotype and disease. MR analysis has the following assumptions: firstly, the genetic variant used is only associated with the risk factor of interest; secondly, it is independent of all confounding variables; and, finally, there is no causal pathway leading from the genetic variant to the disease except through the risk factor of interest. For highly polygenic traits, a large number of genetic polymorphisms can be combined to explain a larger proportion of the variance of the trait. The large numbers of variants included means that some are likely to violate the assumptions for a MR analysis. But a lack of association between appropriate SNPs and the outcome, given a dataset large enough to give reasonable power suggests that there is no causal relationship. A shared genetic basis indicates either, pleiotropy where a variant affects multiple traits independently, or a causal relationship between the two correlated traits; with the requirement that any potential confounders must be taken into account. If a shared genetic basis is found, then a quantitative MR approach would then be required to compare direct and mediated paths between variants affecting the postulated causal variables and the outcome. This method has been widely used, both confirming and refuting suggested causal relationships based on epidemiological findings [19]. For example, this approach has had significant success in clarifying relationships between lipid levels and ischemic heart disease [20]. In addition, a recent study compared 42 traits or diseases with available large genome-wide association studies (GWAS) where, among other findings, the authors found evidence that an increased body mass index causally increases triglyceride levels [21].

MR was recently used to test for an effect of serum iron on Parkinson’s disease (PD) risk, using three genetic variants influencing iron levels (HFE rs1800562, HFE rs1799945, and TMPRSS6 rs855791) [22]. The combined MR estimate showed a statistically significant protective effect of increased serum iron in PD, suggesting that over the course of a life time, alteration in tissue iron homeostasis reflected by a decrease in serum iron levels is on the causal pathway in the pathogenesis of PD. Twelve iron associated SNPs identified though GWAS were also used to investigate the role of iron in atherosclerosis, and identified a potential causal role in women [23].

Table 1

Alzheimer’s disease case-control cohort data sets. The AD cohorts which contributed data to the assessment of the effect of iron genetic profile scores to risk of AD. The APOE ɛ4 frequency is shown for the individuals where APOE genotype data was available, with the sample size in brackets. AD, Alzheimer’s disease; CN, controls

CohortsN AD casesN ControlsMean Age (range, SD)% FemaleAPOE ɛ4 Frequency
Genetic and Environmental Risk for Alzheimer’s disease (GERAD1) [43]2,36194279.064.6AD = 0.33 (n = 2,183)
(60–108, 7.7)CN = 0.13 (n = 906)
Innovative Medicines in Europe (AddNeuroMed) [44]22328077.559.8AD = 0.33 (n = 217)
(60–98, 6.9)CN = 0.15 (n = 143)
Kings Health Partners- Dementia Case Register (KPH-DCR) [45]648579.559.7AD = 0.38 (n = 52)
(61–93, 6.8)CN = 0.14 (n = 65)
Alzheimer’s Disease Neuroimaging Initiative (ADNI) [46]16520576.344.9AD = 0.42 (n = 165)
(60–91, 6.0)CN = 0.14 (n = 204)
Wellcome Trust Case Control Consortium 1958 British Birth Cohort (WTCCC2) [47]04,9265449.7CN = 0.16 (n = 4,862)
(all 54)

Single genetic variants that influence serum iron levels have not been shown to have a large effect on AD risk. The transferrin genetic variant C2 has been investigated and shown to have a small but significant association (OR = 1.11, 95% CI 1.05 to 1.17, in a meta-analysis of 19 studies [24]). Several studies previously reported an increased frequency of the HFE H63D (rs1799945) mutation in AD patients [25], but these findings have not been replicated in the largest AD GWAS meta-analysis [26]. There is evidence of interaction effects, which would not be apparent in GWAS meta-analyses, involving H63D and APOE ɛ4 alleles where the combination appears to affect age of onset and, to a lesser extent, risk [27–29].

Since several genes are well characterized for their impact on peripheral iron variation, we sought to determine their combined causal effect on AD risk. We test the effect of a large number of genetic variants affecting the iron-related measures of serum iron concentration, transferrin (the major iron transporter), ferritin (which reflects iron storage in bone marrow), and transferrin saturation (ratio between serum iron and total iron binding capacity) on AD risk, in combination using a genetic profile score (GPS) approach. Variants are selected from an iron GWAS meta-analysis discovery cohort [30] (n = 23,986) and tested in large independent target AD case-control datasets (n = 9,251). In addition, we test for the converse scenario, whether those at a high genetic risk for AD have higher peripheral iron levels throughout life, using SNPs identified by the AD GWAS meta-analysis discovery cohort [26] (from the International Genomics of Alzheimer’s Project, IGAP n = 54,162) in an independent population-based target sample with available iron measures (n = 8,893). Previously an AD polygenic score analysis has shown that disease prediction accuracy is greatest including SNPs with p value <0.5. Including the full polygenic score significantly improved prediction over use of APOE alone where including both APOE and PRS gave AUC = 78.2% [31]. Examples of the AD PRS based on the IGAP discovery analysis demonstrating genetic overlap with other traits include neuroimaging measures of subcortical brain volumes, plasma C-reactive protein, and lipids [32, 33]. Finally, to confirm our findings using an alternative method, we used SNP effect concordance analysis (SECA) with only the discovery datasets, to examine whether SNPs found to be associated with the serum iron measures are enriched within associated SNPs with AD risk, and vice versa.

MATERIAL AND METHODS

Subjects

The AD case-control cohort comprises the datasets shown in Table 1. All individuals were of European descent and all AD case-control cohort individuals were age ≥60 years. Controls were screened for dementia using either MMSE or ADAS-cog and were determined to be free from characteristic AD plaques at neuropathological examination or had a Braak score ≤2.5. Individuals with AD met criteria for either probable (NINCDS-ADRDA, DSM-IV) or definite (CERAD) AD. Individuals classed as MCI were excluded. The WTCCC2 1958 BC samples are population samples aged 54 years at collection and are included as unscreened controls in this analysis.

The population-based sample set comprises (a) adult twins, their spouses, and first degree relatives who volunteered for studies on risk factors or biomarkers for physical or psychiatric conditions (n = 8,380); (b) people with self-reported endometriosis and unaffected relatives (n = 830) [34, 35]. The mean age is 47 years (ranged 15–92 years) with 62% female. Biochemical markers of iron status were measured using standard clinical methods on Roche/Hitachi 917 or Modular P analyzers [30]. Serum iron was measured by colorimetry with Ferrozine reagent, serum transferrin by immunoturbidimetry, and ferritin by latex particle immunoturbidimetry. Transferrin saturation was calculated from the iron and transferrin results. The values for ferritin were log transformed to produce a normal distribution.

Genetic profile scores

GPS for serum iron, transferrin, transferrin saturation, and ferritin (log) were calculated in target AD case-control cohorts, using stage 1 summary data from the discovery sample of a GWAS meta-analysis combining 11 population-based studies of biochemical markers of iron status, with a sample size of 23,986 [30] using the method previously described ([36] and Supplementary Methods). In brief, linkage disequilibrium-based clumping was used to select SNPs in the discovery data, providing the most significantly associated SNP available in the target data set. The total score is calculated by the number of risk alleles weighted by the standardized per-allele effects for p value thresholds of 1×10–6, 1×10–4, 1×10–3, 0.01, 0.05, 0.1, 0.5, and 1 (all SNPs)(Supplementary Table 1).

The AD GPS was generated in the target population-based cohort using summary data from the AD GWAS meta-analysis from the IGAP discovery sample consisting of 17,008 AD cases and 37,154 controls [26]. GPS were calculated as described above, with the number of risk alleles weighted by the effect on AD risk (log odds ratio). All APOE associated signal was removed and APOE genotype assessed separately.

APOE genotype

In the AD cohorts, a subset of samples have available APOE genotypes (Table 1) inferred from rs429358 and rs7412 SNPs genotyped using TaqMan SNP genotyping assays. In the Australian dataset, APOE genotype was estimated from imputed rs429358 and rs7412 SNP genotypes (Supplementary Methods).

GPS association analysis

In the AD cohort data sets, we tested for an association between iron, transferrin, transferrin saturation, and ferritin GPS at each p value threshold with AD case-control status using logistic regression (performed in STATA v11) controlling for age, sex, and four ancestry principal components. Results for each dataset were combined in a meta-analysis allowing a test for between study heterogeneity (STATA METAN specifying a random effects model). Finally, all datasets were combined in a mega analysis also controlling for study. In addition, we separately assessed the effect of the three iron level influencing variants that have previously been shown to associate with PD risk [22]. We tested for an association with the following SNPs: HFE rs1800562, HFE rs1799945, and TMPRSS6 rs855791 using logistic regression under an additive model and then combined the three variants in a GPS. To investigate any potential interaction effect of APOE ɛ4 genotype, we also repeated these analyses controlling for APOE ɛ4 carrier status and also in APOE ɛ4 positive and APOE ɛ4 negative groups.

In the population-based dataset, we tested for an association of AD GPS and number of APOE ɛ4 alleles with peripheral iron measures (iron, transferrin, transferrin saturation, and ferritin) using Genome-wide Efficient Mixed Model Association algorithm (GEMMA) software [37]. The sample contains related individuals including monozygotic and dizygotic twin pairs, and other first degree relatives. We used linear mixed model regression using the likelihood ratio test, including sex, age, and four ancestry principal components as covariates and controlling for family structure using a genetic relatedness matrix estimated from genome-wide genotypes.

SNP effect concordance analysis

We carried out SECA analysis of large scale GWAS meta-analysis summary statistics to examine the genetic overlap between AD and each iron measure using the default approach [38]. SECA allows a larger sample size to be examined without the need for individual level genotype data. The GWAS meta-analysis results for AD (meta-analysis n = 74,046) [26] and iron measures (iron, transferrin, transferrin saturation, and ferritin; meta-analysis n = 23,986) [30] were used to test for an excess of SNPs associated in the AD and iron phenotype data sets, and whether the SNP effect directions are concordant. SNP effects across the two GWAS summary results were aligned (AD and iron) to the same effect allele and independent SNPs were extracted via LD clumping identifying a subset of independent SNPs with the most significant p-values in the AD dataset. Restricting to SNPs associated with p1≤0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0 in the AD dataset, exact binomial statistical tests determine whether there is an excess of SNPs associated in both datasets for the subset of SNPs associated with p2≤0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0 in the iron dataset. Fisher’s exact test is then used to determine whether there is an excess of SNPs where the effect directions are concordant across the datasets for each p value subset.

Due to the larger sample size the AD GWAS summary statistics were initially used as dataset 1, and each of the iron measures as dataset 2, providing the greatest possible power. Because the analysis is restricted to those SNPs which are most highly associated in dataset 1, we also repeated the analysis with the iron GWAS summary statistics as dataset 1 (in case of a scenario where SNPs strongly affecting iron phenotypes had an effect on AD risk, but SNPs strongly affecting AD risk did not affect iron phenotypes).

RESULTS

GPS analysis

The discovery GWAS meta-analysis datasets used in the study contain large sample sizes (in total 54,162 for AD and 23,986 for serum iron status) and show both AD and serum iron measures to have a strong polygenic components [27, 31]. For serum iron measures using replication cohorts, the lead SNPs at the 11 significant loci explained 3.4, 7.2, 6.7, and 0.9% of the phenotypic variance for iron, transferrin, saturation, and (log-transformed) ferritin, respectively [30]. There is large deviation from the expected distribution of association test statistics compared to observed values, with association signals observed far below the level of genome-wide significance (Fig. 1). Therefore, using SNPs below genome-wide significance will increase power to detect an association.

Fig.1

Q-Q plots of the association p-values from the discovery GWAS meta-analyses. Including the GWAS meta-analysis of biochemical markers of iron status [30] and the International Genomics of Alzheimer’s Project [26]. SNPs in the APOE region (within 500 kb either side of APOE locus) are excluded from the AD plot. The red line is the line of equivalence, observed = expected.

Q-Q plots of the association p-values from the discovery GWAS meta-analyses. Including the GWAS meta-analysis of biochemical markers of iron status [30] and the International Genomics of Alzheimer’s Project [26]. SNPs in the APOE region (within 500 kb either side of APOE locus) are excluded from the AD plot. The red line is the line of equivalence, observed = expected.

Association analysis conducted in each AD disease case-control data set identified no effect of any serum iron status GPS (serum iron, transferrin, ferritin, and transferrin saturation) on AD risk, and the meta-analysis identified no significant between study heterogeneity (Supplementary Figure 1). When combined in a mega analysis no effect of any serum iron status GPS (serum iron, transferrin, ferritin, and transferrin saturation) on AD risk was identified with a sample size of 6,381 controls and 2,870 AD cases (Table 3). Controlling for APOE genotype did not significantly affect the results, and no significant association was identified in separate APOE ɛ4 carrier and non-carrier groups (data not shown). Previously three iron level influencing genetic variants (HFE rs1800562, HFE rs1799945, and TMPRSS6 rs855791) have been shown to be associated with PD risk [22]. There was no association of these SNPs with AD status in our dataset and no interaction identified with APOE ɛ4 status (Supplementary Table 2). In addition, the GPS constructed from these three SNPs did not have an effect on AD risk (Supplementary Table 2).

There was no association of AD GPS or APOE ɛ4 with any peripheral iron measure (Table 4).

Table 2

Serum iron measures in the Australian data set

Serum measureNMeanRangeSD
Iron (μmol/L)8,75119.540.10–50.506.74
Transferrin Saturation (%)8,80028.710.12–95.310.80
Transferrin (g/L)8,8912.821.40–5.190.44
Ferritin (log10) (μg/L)8,8922.000.00–3.260.50
Table 3

The association of serum iron measure genetic profile score (GPS) at different p value thresholds with AD risk. The association analysis was carried out using logistic regression controlling for sex, age, four ancestry principal components, and study. β,standardized Beta

GPSAssociation with AD risk (n = 9,251)
  βSEp
Ironp≤10.040.030.278
p≤0.50.030.030.365
p≤0.10.010.030.868
p≤0.050.020.030.638
p≤0.01–0.010.030.695
p≤0.001–0.010.030.839
p≤0.00010.020.030.624
p≤0.0000010.020.330.632
Transferrinp≤10.030.030.291
  Saturationp≤0.50.030.030.330
p≤0.10.030.030.381
p≤0.050.020.030.584
p≤0.010.020.030.510
p≤0.0010.020.030.590
p≤0.00010.020.030.628
p≤0.0000010.030.030.408
Transferrinp≤10.000.030.933
p≤0.50.000.030.950
p≤0.10.020.030.589
p≤0.050.010.030.797
p≤0.01–0.020.030.517
p≤0.001–0.030.030.299
p≤0.0001–0.030.030.404
p≤0.000001–0.020.030.467
Ferritinp≤10.020.030.577
p≤0.50.030.040.465
p≤0.10.030.040.465
p≤0.050.050.040.196
p≤0.010.030.030.347
p≤0.0010.030.030.355
p≤0.00010.030.030.377
p≤0.0000010.040.030.170
Table 4

The association of AD GPS at different p value thresholds (excluding APOE) and number of APOE ɛ4 alleles with iron phenotypes. The association analysis was carried out using linear mixed models implemented in GEMMA (genome-wide efficient mixed-model association) [37] using the likelihood ratio test. Family relationships were controlled for using a genetic relatedness matrix estimated from genotypes. Sex, age, and four ancestry principal components were also included as covariates. β, standardized Beta

Serum IronAD GPSNβSEp
Measure
Ironp≤18,7510.020.010.153
p≤0.58,7510.020.010.148
p≤0.18,7510.010.010.349
p≤0.058,7510.010.010.594
p≤0.018,7510.000.010.747
p≤0.0018,7510.010.010.405
p≤0.00018,7510.010.010.615
p≤0.0000018,7510.020.010.119
APOE ɛ48,4940.000.010.843
Transferrinp≤18,800371.45224.200.097
  Saturationp≤0.58,800201.12136.430.140
p≤0.18,80046.4054.110.391
p≤0.058,80013.3738.990.732
p≤0.018,8002.8218.460.878
p≤0.0018,8000.766.580.908
p≤0.00018,8000.252.150.908
p≤0.0000018,8003.191.270.012
APOE ɛ48,5310.020.020.477
Transferrinp≤18,891–218.75225.190.331
p≤0.58,891–78.29137.030.568
p≤0.18,8919.8654.360.856
p≤0.058,89123.1239.160.555
p≤0.018,8915.8718.520.751
p≤0.0018,89116.296.580.013
p≤0.00018,8914.972.150.021
p≤0.0000018,891–1.771.280.166
APOE ɛ48,619–0.020.020.466
Ferritinp≤18,892156.22192.510.417
p≤0.58,89281.98117.140.484
p≤0.18,89235.6146.420.442
p≤0.058,8927.4933.470.822
p≤0.018,89211.0515.850.485
p≤0.0018,8922.535.640.654
p≤0.00018,892–0.641.840.728
p≤0.0000018,8920.851.090.435
APOE ɛ48,6210.010.020.486

SNP effect concordance analysis

In agreement with the GPS analysis, we did not identify any significant pleiotropy between datasets or concordant effects using SECA. We tested for an excess of SNPs associated with AD also associating with iron phenotypes. Using a binomial test, we compared the AD dataset with each of the iron phenotype datasets in turn examining 144 SNP subsets (testing twelve p value threshold combinations). No SNP sets were found to have nominally significant pleiotropy (Fig. 2). Using Fisher’s test, we also tested for an excess of SNPs where the effect directions (BETA) are concordant between SNP subsets in each dataset. Again, we identified no significant concordance (Supplementary Figure 2). Additionally, no significant pleiotropy or concordant effects were seen when switching the primary dataset, i.e., testing for an excess of SNPs associated with each iron phenotype also associating with AD.

Fig.2

Genetic overlap between dataset 1 (AD) and dataset 2 (Serum iron). In the SECA analysis, exact binomial statistical tests are performed to determine whether there is an excess of SNPs associated in both datasets for 144 SNP subsets from 12×12 p-value threshold combinations. A binomial test ‘heatmap’ plot is generated to graphically summarize the proportion of SNP subsets with an excess [observed(obs)≥expected (exp)] or deficit (obs<exp) number of associated SNPs, and empirical p-values (adjusted for testing all 144 subsets) are calculated via permutation.

Genetic overlap between dataset 1 (AD) and dataset 2 (Serum iron). In the SECA analysis, exact binomial statistical tests are performed to determine whether there is an excess of SNPs associated in both datasets for 144 SNP subsets from 12×12 p-value threshold combinations. A binomial test ‘heatmap’ plot is generated to graphically summarize the proportion of SNP subsets with an excess [observed(obs)≥expected (exp)] or deficit (obs<exp) number of associated SNPs, and empirical p-values (adjusted for testing all 144 subsets) are calculated via permutation.

DISCUSSION

It is becoming increasingly clear from investigations of iron homeostasis and recent advances in iron imaging methods that iron dysregulation is an important feature of AD, and therefore lowering of iron content in the brain is a potential therapeutic target [39]. But there is limited understanding of the importance of peripheral iron levels in AD risk, and whether prolonged increased or decreased iron levels may be a risk factor for AD. We investigated whether there is a shared genetic basis between AD and peripheral iron levels using a PRS approach. We identified no effect of genetic variants affecting peripheral iron biomarkers (including iron, transferrin, transferrin saturation, and ferritin) on AD risk. Nor did we find increased serum iron levels in those who are at increased genetic risk of developing AD, including both APOE ɛ4 carriers and those with a higher load of other common risk variants. In addition, in an investigation of the genetic overlap between AD and each iron measure, we do not find any significant overlap of genetic loci from the results of large-scale GWAS meta-analysis studies.

Taken together, our results suggest that the causes of variation in brain iron that might contribute to AD are distinct from those causing variations in circulating iron (serum iron) or in iron stores in bone marrow or other organs (serum ferritin). Iron retention is complex in different organs, and our current data on peripheral iron measurement cannot exclude causation by other genes that affect iron levels in the brain that are not reflected by serum values. In addition, the peripheral iron measurements used are standard clinical pathology measures. Non-standard and possibly more direct measures (such as transferrin saturation using size exclusion chromatography-inductively coupled plasma-mass spectrometry) have been shown to be more sensitive to differences in the blood between AD patients and controls [15].

It is also possible that, even if iron is not a primary cause of increase in AD risk, it accumulates after the initiation of cell damage by other mechanisms, and exacerbates it. Evidence for this comes from recent work showing that once Aβ forms aggregates they induce iron accumulation [40]. Iron-related therapies could still be relevant for patients who are in the early stages of AD.

Iron accumulation in tissues is a feature of many diseases, and may prove to be causal for some. Our current results for AD are in contrast to previous evidence of a causal association of increased peripheral iron measures with a decreased risk of PD [22]. The authors hypothesized that low peripheral iron may decrease neuronal iron storage though a reduction in ferritin, resulting in free iron accumulation in the brain. To investigate whether a similar effect exists for AD, we tested a larger number of iron-affecting variants against the most recent GWAS meta-analysis on AD risk. These explain a larger proportion of the variance and therefore we would expect them to have more power to detect anyeffect.

However, our analysis has limitations that need to be considered. Firstly, the multi-SNP GPS includes a large number of genetic variants of unknown effect or multiple effects; therefore we cannot rule out that as well as affecting iron levels, some also affect AD risk though other pathways and could potentially do so in opposite directions. To attempt to address this issue, we also tested for an effect of three genetic variants (in HFE and TMPRSS6) known to have a direct role in peripheral iron levels and previously shown to have an effect on PD risk [22], where we also did not find an effect. In addition, we cannot rule out the possibility that other genomic variations, such as epigenetic dysregulation, affect iron levels which are then causal for AD.

Secondly, as in other complex diseases and phenotypes, discovered genetic variants only represent a small proportion of the variance in both iron levels and AD risk. This study utilizes summary data from the two largest GWAS meta-analysis discovery cohorts for both AD and biochemical markers of iron status (total sample sizes of 54,162 and 23,986, respectively [26, 30]) to compute comprehensive GPS. In addition, the GPS were applied to large samples with individual level genotype and phenotype data (For AD cases-control: 2,813 AD cases, and 6,438 controls (of which 4,926 are unscreened for AD, aged 54), and ≥8,751 for iron measures). Even so, we cannot rule out a small effect that is not detectable with this sample size.

Thirdly, effects on iron in relevant brain areas may differ from effects on circulating iron or iron in other organs. Previous studies identified an association between iron accumulation in the basal ganglia of elderly men and peripheral iron measures [13]. However, only 9% of the variance of CSF ferritin can be explained by plasma ferritin [9], highlighting the separation between these compartments. It is also possible that there are genetic loci more relevant to iron-homeostasis in elderly people, as the sample used to construct the iron phenotypes GPS have a mean age of 47.

Our results suggest that there is not a causal connection between lifetime peripheral iron measures and increased risk of AD. We did not replicate the previous finding of an effect of HFE SNPs on risk of AD and an epistatic interaction for risk with APOE ɛ4 genotype, but we cannot yet rule out an association of HFE SNPs with AD age of onset or phenotypic interactions [25, 27, 28].

It has been suggested that public recommendations for AD risk reduction should caution the use of iron supplementation for those whom it is not required [18, 41, 42]. Dietary patterns such as a Mediterranean diet and reduced red meat consumption that associate with lower AD risk do tend to have a low iron intake, but also have other unrelated health benefits for example high intake of vegetables and low saturated fat. Consistent with our genetic findings, there is no clear evidence that dietary intervention affecting iron intake alters the risk of AD [18]. More work is needed to assess the effect of iron on the progression (as opposed to the initiation) and age of onset of AD.

In conclusion, although iron deposition is an important feature of AD brain tissues, these results suggest that there is not a significant causal relationship between lifetime peripheral iron levels and AD.

ACKNOWLEDGMENTS

Genetic and Environmental Risk for Alzheimer’s Disease Consortium (GERAD1) Collaborators: Denise Harold1, Rebecca Sims1, Amy Gerrish1, Jade Chapman1, Valentina Escott-Price1, Nandini Badarinarayan1, Richard Abraham1, Paul Hollingworth1, Marian Hamshere1, Jaspreet Singh Pahwa1, Kimberley Dowzell1, Amy Williams1, Nicola Jones1, Charlene Thomas1, Alexandra Stretton1, Angharad Morgan1, Kate Williams1, Sarah Taylor1, Simon Lovestone2, John Powell3, Petroula Proitsi3, Michelle K Lupton3, Carol Brayne4, David C. Rubinsztein5, Michael Gill6, Brian Lawlor6, Aoibhinn Lynch6, Kevin Morgan7, Kristelle Brown7, Peter Passmore8, David Craig8, Bernadette McGuinness8, Janet A Johnston8, Stephen Todd8, Clive Holmes9, David Mann10, A. David Smith11, Seth Love12, Patrick G. Kehoe12, John Hardy13, Rita Guerreiro14,15, Andrew Singleton14, Simon Mead16, Nick Fox17, Martin Rossor17, John Collinge16, Wolfgang Maier18, Frank Jessen18, Reiner Heun18, Britta Schürmann18,19, Alfredo Ramirez18, Tim Becker20, Christine Herold20, André Lacour20, Dmitriy Drichel20, Hendrik van den Bussche21, Isabella Heuser22, Johannes Kornhuber23, Jens Wiltfang24, Martin Dichgans25,26, Lutz Frölich27, Harald Hampel28, Michael Hüll29, Dan Rujescu30, Alison Goate31, John S.K. Kauwe32, Carlos Cruchaga33, Petra Nowotny33, John C. Morris33, Kevin Mayo33, Gill Livingston34, Nicholas J. Bass34, Hugh Gurling34, Andrew McQuillin34, Rhian Gwilliam35, Panagiotis Deloukas35, Markus M. Nöthen20, Peter Holmans1, Michael ODonovan1, Michael J.Owen1, Julie Williams1.

1Medical Research Council (MRC) Centre for Neuropsychiatric Genetics and Genomics, Neurosciences and Mental Health Research Institute, Department of Psychological Medicine and Neurology, School of Medicine, Cardiff University, Cardiff, UK; 2Department of Psychiatry, Medical Sciences Division, University of Oxford, Oxford, UK; 3Kings College London, Institute of Psychiatry, Department of Neuroscience, De Crespigny Park, Denmark Hill, London, UK; 4Institute of Public Health, University of Cambridge, Cambridge, UK; 5Cambridge Institute for Medical Research, University of Cambridge, Cambridge, UK; 6Mercers Institute for Research on Aging, St. James Hospital and Trinity College, Dublin, Ireland; 7Institute of Genetics, Queens Medical Centre, University of Nottingham, UK; 8Ageing Group, Centre for Public Health, School of Medicine, Dentistry and Biomedical Sciences, Queens University Belfast, UK; 9Division of Clinical Neurosciences, School of Medicine, University of Southampton, Southampton, UK; 10Clinical Neuroscience Research Group, Greater Manchester Neurosciences Centre, University of Manchester, Salford, UK; 11Oxford Project to Investigate Memory and Ageing (OPTIMA), University of Oxford, Department of Pharmacology, Mansfield Road, Oxford, UK; 12University of Bristol Institute of Clinical Neurosciences, School of Clinical Sciences, Frenchay Hospital, Bristol, UK; 13Department of Molecular Neuroscience and Reta Lilla Weston Laboratories, Institute of Neurology, UCL, London, UK; 14Laboratory of Neurogenetics, National Institute on Aging, National Institutes of Health, Bethesda, Maryland, USA; 15Department of Molecular Neuroscience, Institute of Neurology, University College London, Queen Square, London, UK; 16MRC Prion Unit, Department of Neurodegenerative Disease, UCL Institute of Neurology, London, UK; 17Dementia Research Centre, Department of Neurodegenerative Diseases, University College London, Institute of Neurology, London, UK; 18Department of Psychiatry, University of Bonn, Bonn, Germany; 19Institute for Molecular Psychiatry, University of Bonn, Bonn, Germany; 20Department of Genomics, Life & Brain Center, University of Bonn, Bonn, Germany; 21Institute of Primary Medical Care, University Medical Center Hamburg-Eppendorf, Germany; 22Department of Psychiatry, Charité Berlin, Germany; 23Department of Psychiatry, Friedrich-Alexander-University Erlangen-Nürnberg, Germany; 24Department of Psychiatry and Psychotherapy, University Medical Center (UMG), Georg-August-University, Göttingen, Germany; 25Institute for Stroke and Dementia Research, Klinikum der Universität München, Munich, Germany; 26Department of Neurology, Klinikum der Universität München, Munich, Germany; 27Central Institute of Mental Health, Medical Faculty Mannheim, University of Heidelberg, Germany; 28Institute for Memory and Alzheimer’s Disease & INSERM, Sorbonne Universities, Pierre and Marie Curie University, Paris, France and Institute for Brain and Spinal Cord Disorders (ICM), Department of Neurology, Hospital of Pitié-Salpétrière, Paris, France; 29Centre for Geriatric Medicine and Section of Gerontopsychiatry and Neuropsychology, Medical School, University of Freiburg, Germany; 30Department of Psychiatry, University of Halle, Halle, Germany; 31Neuroscience Department, Icahn School of Medicine at Mount Sinai, New York, USA; 32Department of Biology, Brigham Young University, Provo, UT, USA; 33Departments of Psychiatry, Neurology and Genetics, Washington University School of Medicine, St Louis, MO, USA; 34Department of Mental Health Sciences, University College London, UK; 35The Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK.

For the GERAD1 Consortium: Cardiff University was supported by the Wellcome Trust, Medical Research Council (MRC), Alzheimer’s Research UK (ARUK), and the Welsh Assembly Government. ARUK supported sample collections at the Kings College London; the South West Dementia Bank; and Universities of Cambridge, Nottingham, Manchester, and Belfast. The Belfast group acknowledges support from the Alzheimer’s Society, Ulster Garden Villages, N. Ireland R&D Office, and the Royal College of Physicians/Dunhill Medical Trust. The MRC and Mercer’s Institute for Research on Ageing supported the Trinity College group. The South West Dementia Brain Bank acknowledges support from Bristol Research into Alzheimer’s and Care of the Elderly. The Charles Wolfson Charitable Trust supported the OPTIMA group. Washington University was funded by NIH grants, Barnes Jewish Foundation, and the Charles and Joanne Knight Alzheimer’s Research Initiative. Patient recruitment for the MRC Prion Unit/UCL Department of Neurodegenerative Disease collection was supported by the UCLH/UCL Biomedical Centre. LASER-AD was funded by Lundbeck SA. The Bonn group was supported by the German Federal Ministry of Education and Research (BMBF), Competence Network Dementia and Competence Network Degenerative Dementia, and by the Alfried Krupp von Bohlen und Halbach-Stiftung. The GERAD1 Consortium also used samples ascertained by the NIMH AD Genetics Initiative.

AddNeuroMed is part of InnoMed (Innovative Medicines in Europe), an integrated project funded by the European Union of the Sixth Framework program priority ((FP6-2004-LIFESCIHEALTH-5); the Alzheimer’s Research Trust UK; the John and Lucille van Geest Foundation; and the NIHR Biomedical Research Centre for Mental Health and Biomedical Research Unit for Dementia at the South London, Maudsley NHS Foundation Trust and Kings College London, and a joint infrastructure grant from Guy’s and St Thomas’ Charity and the Maudsley Charity; Academy of Finland, Kuopio University Hospital (HS) and funding from UEF-BRAIN (HS).

The Kings Health Partners Dementia Case Register is funded by the National Institute for Health Research (NIHR) Mental Health Biomedical Research Centre and Dementia Unit at South London and Maudsley NHS Foundation Trust and (Institute of Psychiatry, Psychology and Neuroscience) King’s College London.

Data collection and sharing for ADNI was funded by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2-0012). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and through generous contributions from the following: Alzheimer’s Association; Alzheimer’s Drug Discovery Foundation; Araclon Biotech; BioClinica, Inc.; Biogen Idec Inc.; Bristol-Myers Squibb Company; Eisai Inc.; Elan Pharmaceuticals, Inc.; Eli Lilly and Company; EuroImmun; F. Hoffmann-La Roche Ltd and its affiliated company Genentech, Inc.; Fujirebio; GE Healthcare;; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development, LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Medpace, Inc.; Merck & Co., Inc.; Meso Scale Diagnostics, LLC.; NeuroRx Research; Neurotrack Technologies; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Synarc Inc.; and Takeda Pharmaceutical Company. The Canadian Institutes of Health Research is providing funds to support ADNI clinical sites in Canada. Private sector contributions are facilitated by the Foundation for the National Institutes of Health (http://www.fnih.org). The grantee organization is the Northern California Institute for Research and Education, and the study is coordinated by the Alzheimer’s Disease Cooperative Study at the University of California, San Diego. ADNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California.

This study makes use of data generated by the Wellcome Trust Case-Control Consortium. A full list of the investigators who contributed to the generation of the data is available from http://www.wtccc.org.uk. Funding for the project was provided by the Wellcome Trust under award 076113 and 085475” and cite the relevant primary WTCCC publication (details of which can be found on the WTCCC website).

For the Australian study, we acknowledge funding from the Australian National Health and Medical Research Council (NHMRC grants 241944, 389875, 389891, 389892, 389938, 442915, 442981, 496739 and 552485), US National Institutes of Health (NIH grants AA07535, AA10248, AA014041, AA011998, AA013320, AA013321, AA017688, DA012854), and the Australian Research Council (ARC grant DP0770096). MKL is supported by a Perpetual Foundation Wilson Fellowship for early career researchers.

BB was supported by the Australian National Health and Medical Research Council (APP1084417 and APP1079583).

We thank the International Genomics of Alzheimer’s Project (IGAP) for providing AD meta-analysis summary results data for these analyses. The investigators within IGAP contributed to the design and implementation of IGAP and/or provided data but did not participate in analysis or writing of this report. IGAP was made possible by the generous participation of the control subjects, the patients, and their families. The i–Select chips were funded by the French National Foundation on Alzheimer’s disease and related disorders. EADI was supported by the LABEX (laboratory of excellence program investment for the future) DISTALZ grant, Inserm, Institut Pasteur de Lille, Université de Lille 2 and the Lille University Hospital. GERAD was supported by the Medical Research Council (Grant n° 503480), Alzheimer’s Research UK (Grant n° 503176), the Wellcome Trust (Grant n° 082604/2/07/Z) and German Federal Ministry of Education and Research (BMBF): Competence Network Dementia (CND) grant n° 01GI0102, 01GI0711, 01GI0420. CHARGE was partly supported by the NIH/NIA grant R01 AG033193 and the NIA AG081220 and AGES contract N01–AG–12100, the NHLBI grant R01 HL105756, the Icelandic Heart Association, and the Erasmus Medical Centre and Erasmus University. ADGC was supported by the NIH/NIA grants: U01 AG032984, U24 AG021886, U01 AG016976, and the Alzheimer’s Association grant ADGC–10–196728.

We acknowledge the Genetics of Iron Status Consortium as the source of the iron meta-analysis SNP association data. The following individuals are part of the Iron Status Consortium and therefore contributed to the design and implementation of the iron meta-analysis study but did not participate in analysis or writing of this paper (unless named as authors).

Genetics of Iron Status Consortium: Beben Benyamin1,2, Tonu Esko3,4, Janina S. Ried5, Aparna Radhakrishnan6, Sita H. Vermeulen7,8, Michela Traglia9,10, Martin Gögele11, Denise Anderson12, Linda Broer13,14, Clara Podmore15, Jianán Luan15, Zoltan Kutalik16,17, Serena Sanna18, Peter van der Meer19, Toshiko Tanaka20, Fudi Wang21, Harm-Jan Westra22, Lude Franke22, Evelin Mihailov3,23, Lili Milani3, Jonas Hälldin3, Juliane Winkelmann24,25,26,27, Thomas Meitinger26,28, Joachim Thiery29,30, Annette Peters31,32, Melanie Waldenberger31,32, Augusto Rendon6,33,34, Jennifer Jolley6,33, Jennifer Sambrook6,33, Lambertus A. Kiemeney7,35, Fred C. Sweep36, Cinzia F. Sala37, Christine Schwienbacher11, Irene Pichler11, Jennie Hui38,39, Ayse Demirkan13,40, Aaron Isaacs13,41, Najaf Amin13, Maristella Steri18, Geŕard Waeber42, Niek Verweij19, Joseph E. Powell1,43, Dale R. Nyholt2, Andrew C. Heath44, Pamela A.F. Madden44, Peter M. Visscher1,43, Margaret J. Wright2, Grant W. Montgomery2, Nicholas G. Martin2, Dena Hernandez45, Stefania Bandinelli46, Pim van der Harst19,47,48, Manuela Uda18, Peter Vollenweider42, Robert A. Scott15, Claudia Langenberg15, Nicholas J. Wareham15, InterAct Consortium, Cornelia van Duijn13,41,49, John Beilby38,39, Peter P. Pramstaller11,50, Andrew A. Hicks11, Willem H. Ouwehand6,33, Konrad Oexle28, Christian Gieger5,31,32, Andres Metspalu3, Clara Camaschella9,51, Daniela Toniolo9, Dorine W. Swinkels36, John B. Whitfield2.

1The University of Queensland, Queensland Brain Institute, Brisbane, Queensland, Australia; 2QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia; 3Estonian Genome Center, University of Tartu, Tartu, Estonia; 4The Broad Institute of Massachusetts Institute of Technology and Harvard University, Cambridge, Massachusetts, USA; 5Institute of Genetic Epidemiology, Helmholtz Zentrum München, Neuherberg, Germany; 6Department of Haematology, University of Cambridge, Cambridge, UK; 7Department for Health Evidence, Radboud University Medical Centre, Nijmegen, The Netherlands; 8Department of Genetics, Radboud University Medical Centre, Nijmegen, The Netherlands; 9Division of Genetics and Cell Biology, San Raffaele Scientific Institute, Milano, Italy; 10Institute for Maternal and Child Health-IRCCS “Burlo Garofolo”- Trieste, Trieste, Italy; 11Center for Biomedicine, European Academy of Bolzano/Bozen (EURAC), Bolzano, Italy; 12Telethon Institute for Child Health Research, Centre for Child Health Research, The University of Western Australia, West Perth, Western Australia, Australia; 13Department of Epidemiology, Subdivision Genetic Epidemiology, Erasmus Medical Center, Rotterdam, The Netherlands; 14Department of Internal Medicine, Erasmus Medical Center, Rotterdam, The Netherlands; 15MRC Epidemiology Unit, University of Cambridge School of Clinical Medicine, Institute of Metabolic Science, Cambridge Biomedical Campus, Cambridge, UK; 16Institute of Social and Preventive Medicine (IUMSP), Centre Hospitalier Universitaire Vaudois (CHUV), Lausanne, Switzerland; 17Swiss Institute of Bioinformatics, Lausanne, Switzerland; 18Istituto di Ricerca Genetica e Biomedica (IRGB), Consiglio Nazionale delle Ricerche (CNR), Monserrato, Cagliari, Italy; 19University of Groningen, University Medical Center Groningen, Department of Cardiology, Groningen, The Netherlands; 20Translational Gerontology Branch, National Institute on Aging, Baltimore, Maryland, USA; 21School of Public Health, Zhejiang University, Hangzhou, P. R. China; 22Department of Genetics, University Medical Center Groningen, Groningen, The Netherlands; 23Institute of Molecular and Cell Biology, University of Tartu, Tartu, Estonia; 24Department of Neurology and Neurosciences and Center for Sleep Sciences and Medicine, Stanford University, Palo Alto, California, USA; 25Neurologische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universitat München, Munich, Germany; 26Institut für Humangenetik, Helmholtz Zentrum München, Neuherberg, Germany; 27Munich Cluster for Systems Neurology (SyNergy), Munich, Germany; 28Institute of Human Genetics, Klinikum Rechts der Isar, Technische Universitat München, München, Germany; 29Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics, Universitatsklinikum Leipzig, Leipzig, Germany; 30LIFE Research Center for Civilization Diseases, Medical Faculty, University Leipzig, Leipzig, Germany; 31Institute of Epidemiology II, Helmholtz Zentrum München - German Research Center for Environmental Health, Neuherberg, Germany; 32Research Unit of Molecular Epidemiology, Helmholtz Zentrum München –German Research Center for Environmental Health, Neuherberg, Germany; 33NHS Blood and Transplant, Cambridge Biomedical Campus, Cambridge, UK; 34Medical Research Council Biostatistics Unit, Cambridge Biomedical Campus, Cambridge, UK; 35Department of Urology, Radboud University Medical Centre, Nijmegen, The Netherlands; 36Laboratory of Genetic, Endocrine and Metabolic Diseases (LGEM), Department of Laboratory Medicine, Radboud University Medical Centre, Nijmegen, The Netherlands; 37Division of Genetics and Cell Biology, San Raffaele Scientific Institute, Milano, Italy; 38PathWest Laboratory Medicine of WA, Nedlands, Western Australia, Australia; 39Schools of Pathology and Laboratory Medicine, and Population Health, The University of Western Australia, Nedlands, Western Australia, Australia; 40Department of Human Genetics, Leiden University Medical Center, Leiden, The Netherlands; 41Center of Medical Systems Biology, Leiden, The Netherlands; 42Department of Medicine, Internal Medicine, CHUV, University of Lausanne, Lausanne, Switzerland; 43University of Queensland Diamantina Institute, University of Queensland, Princess Alexandra Hospital, Brisbane, Queensland, Australia; 44Department of Psychiatry, Washington University, St. Louis, Missouri, USA; 45Laboratory of Neurogenetics, National Institute on Aging, Bethesda, Maryland, USA; 46Geriatric Unit, Azienda Sanitaria Firenze (ASF), Firenze, Italy; 47Durrer Center for Cardiogenetic Research, ICIN-Netherlands Heart Institute, Utrecht, The Netherlands; 48University of Groningen, University Medical Center Groningen, Department of Genetics, Groningen, The Netherlands; 49Member of Netherlands Consortium for Healthy Aging sponsored by Netherlands Genomics Initiative, Leiden, The Netherlands; 50Department of Neurology, General Central Hospital, Bolzano, Italy; 51Vita-Salute University, Milan, Italy.

Authors’ disclosures available online (http://j-alz.com/manuscript-disclosures/17-0027r3).

Appendices

The supplementary material is available in the electronic version of this article: http://dx.doi.org/10.3233/JAD-170027.

REFERENCES

[1] 

Gerlach M , Ben-Shachar D , Riederer P , Youdim MB ((1994) ) Altered brain metabolism of iron as a cause of neurodegenerative diseases? J Neurochem 63: , 793–807.

[2] 

Padurariu M , Ciobica A , Lefter R , Serban IL , Stefanescu C , Chirita R ((2013) ) The oxidative stress hypothesis in Alzheimer’s disease. Psychiatr Danub 25: , 401–409.

[3] 

Hare D , Ayton S , Bush A , Lei P ((2013) ) A delicate balance: Iron metabolism and diseases of the brain. Front Aging Neurosci 5: , 34.

[4] 

Antharam V , Collingwood JF , Bullivant JP , Davidson MR , Chandra S , Mikhaylova A , Finnegan ME , Batich C , Forder JR , Dobson J ((2012) ) High field magnetic resonance microscopy of the human hippocampus in Alzheimer’s disease: Quantitative imaging and correlation with iron. Neuroimage 59: , 1249–1260.

[5] 

Zhu WZ , Zhong WD , Wang W , Zhan CJ , Wang CY , Qi JP , Wang JZ , Lei T ((2009) ) Quantitative MR phase-corrected imaging to investigate increased brain iron deposition of patients with Alzheimer disease. Radiology 253: , 497–504.

[6] 

Ding B , Chen KM , Ling HW , Sun F , Li X , Wan T , Chai WM , Zhang H , Zhan Y , Guan YJ ((2009) ) Correlation of iron in the hippocampus with MMSE in patients with Alzheimer’s disease. J Magn Reson Imaging 29: , 793–798.

[7] 

van Bergen JM , Li X , Hua J , Schreiner SJ , Steininger SC , Quevenco FC , Wyss M , Gietl AF , Treyer V , Leh SE , Buck F , Nitsch RM , Pruessmann KP , van Zijl PC , Hock C , Unschuld PG ((2016) ) Colocalization of cerebral iron with Amyloid beta in Mild Cognitive Impairment. Sci Rep 6: , 35514.

[8] 

Crichton RR , Dexter DT , Ward RJ ((2011) ) Brain iron metabolism and its perturbation in neurological diseases. J Neural Transm (Vienna) 118: , 301–314.

[9] 

Ayton S , Faux NG , Bush A , Alzheimer’s Dis Neuroimaging Initiative ((2015) ) Ferritin levels in the cerebrospinal fluid predict Alzheimer’s disease outcomes and are regulated by APOE. Nat Commun 6: , 6760.

[10] 

Koeppen AH ((2003) ) A brief history of brain iron research. J Neurol Sci 207: , 95–97.

[11] 

McCarthy RC , Kosman DJ ((2015) ) Mechanisms and regulation of iron trafficking across the capillary endothelial cells of the blood-brain barrier. Front Mol Neurosci 8: , 31.

[12] 

Pinero DJ , Li NQ , Connor JR , Beard JL ((2000) ) Variations in dietary iron alter brain iron metabolism in developing rats. J Nutr 130: , 254–263.

[13] 

House MJ , St Pierre TG , Milward EA , Bruce DG , Olynyk JK ((2010) ) Relationship between brain R(2) and liver and serum iron concentrations in elderly men. Magn Reson Med 63: , 275–281.

[14] 

Faux NG , Rembach A , Wiley J , Ellis KA , Ames D , Fowler CJ , Martins RN , Pertile KK , Rumble RL , Trounson B , Masters CL , Bush AI ((2014) ) An anemia of Alzheimer’s disease. Mol Psychiatry 19: , 1227–1234.

[15] 

Hare DJ , Doecke JD , Faux NG , Rembach A , Volitakis I , Fowler CJ , Grimm R , Doble PA , Cherny RA , Masters CL , Bush AI , Roberts BR ((2015) ) Decreased plasma iron in Alzheimer’s disease is due to transferrin desaturation. ACS Chem Neurosci 6: , 398–402.

[16] 

Tao Y , Wang Y , Rogers JT , Wang F ((2014) ) Perturbed iron distribution in Alzheimer’s disease serum, cerebrospinal fluid, and selected brain regions: A systematic review and meta-analysis. J Alzheimers Dis 42: , 679–690.

[17] 

Wang ZX , Tan L , Wang HF , Ma J , Liu J , Tan MS , Sun JH , Zhu XC , Jiang T , Yu JT ((2015) ) Serum iron, zinc, and copper levels in patients with Alzheimer’s disease: A replication study and meta-analyses. J Alzheimers Dis 47: , 565–581.

[18] 

Loef M , Walach H ((2012) ) Copper and iron in Alzheimer’s disease: A systematic review and its dietary implications. Br J Nutr 107: , 7–19.

[19] 

Dudbridge F ((2016) ) Polygenic epidemiology. Genet Epidemiol 40: , 268–272.

[20] 

Burgess S , Freitag DF , Khan H , Gorman DN , Thompson SG ((2014) ) Using multivariable Mendelian randomization to disentangle the causal effects of lipid fractions. PLoS One 9: , e108891.

[21] 

Pickrell JK , Berisa T , Liu JZ , Segurel L , Tung JY , Hinds DA ((2016) ) Detection and interpretation of shared genetic influences on 42 human traits. Nat Genet 48: , 709–717.

[22] 

Pichler I , Del Greco MF , Gogele M , Lill CM , Bertram L , Do CB , Eriksson N , Foroud T , Myers RH , Nalls M , Keller MF , Benyamin B , Whitfield JB , Pramstaller PP , Hicks AA , Thompson JR , Minelli C ((2013) ) Serum iron levels and the risk of Parkinson disease: A mendelian randomization study. PLoS Med 10: , e1001462.

[23] 

Galesloot TE , Janss LL , Burgess S , Kiemeney LA , den Heijer M , de Graaf J , Holewijn S , Benyamin B , Whitfield JB , Swinkels DW , Vermeulen SH ((2015) ) Iron and hepcidin as risk factors in atherosclerosis: What do the genes say? BMC Genet 16: , 79.

[24] 

Wang Y , Xu S , Liu Z , Lai C , Xie Z , Zhao C , Wei Y , Bi JZ ((2013) ) Meta-analysis on the association between the TF gene rs1049296 and AD. Can J Neurol Sci 40: , 691–697.

[25] 

Ali-Rahmani F , Schengrund CL , Connor JR ((2014) ) HFE gene variants, iron, and lipids: A novel connection in Alzheimer’s disease. Front Pharmacol 5: , 165.

[26] 

Lambert JC , Ibrahim-Verbaas CA , Harold D , Naj AC , Sims R , Bellenguez C , DeStafano AL , Bis JC , Beecham GW , Grenier-Boley B , Russo G , Thorton-Wells TA , Jones N , Smith AV , Chouraki V , Thomas C , Ikram MA , Zelenika D , Vardarajan BN , Kamatani Y , Lin CF , Gerrish A , Schmidt H , Kunkle B , Dunstan ML , Ruiz A , Bihoreau MT , Choi SH , Reitz C , Pasquier F , Cruchaga C , Craig D , Amin N , Berr C , Lopez OL , De Jager PL , Deramecourt V , Johnston JA , Evans D , Lovestone S , Letenneur L , Morón FJ , Rubinsztein DC , Eiriksdottir G , Sleegers K , Goate AM , Fiévet N , Huentelman MW , Gill M , Brown K , Kamboh MI , Keller L , Barberger-Gateau P , McGuiness B , Larson EB , Green R , Myers AJ , Dufouil C , Todd S , Wallon D , Love S , Rogaeva E , Gallacher J , St George-Hyslop P , Clarimon J , Lleo A , Bayer A , Tsuang DW , Yu L , Tsolaki M , Bossú P , Spalletta G , Proitsi P , Collinge J , Sorbi S , Sanchez-Garcia F , Fox NC , Hardy J , Deniz Naranjo MC , Bosco P , Clarke R , Brayne C , Galimberti D , Mancuso M , Matthews F ; European Alzheimer’s DiseaseInitiative (EADI); Genetic and Environmental Risk in Alzheimer’sDisease; Alzheimer’s Disease Genetic Consortium; Cohorts for Heartand Aging Research in Genomic Epidemiology, Moebus S , Mecocci P , Del Zompo M , Maier W , Hampel H , Pilotto A , Bullido M , Panza F , Caffarra P , Nacmias B , Gilbert JR , Mayhaus M , Lannefelt L , Hakonarson H , Pichler S , Carrasquillo MM , Ingelsson M , Beekly D , Alvarez V , Zou F , Valladares O , Younkin SG , Coto E , Hamilton-Nelson KL , Gu W , Razquin C , Pastor P , Mateo I , Owen MJ , Faber KM , Jonsson PV , Combarros O , O’Donovan MC , Cantwell LB , Soininen H , Blacker D , Mead S , Mosley TH Jr , Bennett DA , Harris TB , Fratiglioni L , Holmes C , de Bruijn RF , Passmore P , Montine TJ , Bettens K , Rotter JI , Brice A , Morgan K , Foroud TM , Kukull WA , Hannequin D , Powell JF , Nalls MA , Ritchie K , Lunetta KL , Kauwe JS , Boerwinkle E , Riemenschneider M , Boada M , Hiltuenen M , Martin ER , Schmidt R , Rujescu D , Wang LS , Dartigues JF , Mayeux R , Tzourio C , Hofman A , Nöthen MM , Graff C , Psaty BM , Jones L , Haines JL , Holmans PA , Lathrop M , Pericak-Vance MA , Launer LJ , Farrer LA , vanDuijn CM , Van Broeckhoven C , Moskvina V , Seshadri S , Williams J , Schellenberg GD , Amouyel P ((2013) ) Meta-analysis of 74,046individuals identifies 11 new susceptibility loci for Alzheimer’sdisease. Nat Genet 45: , 1452–1458.

[27] 

Percy M , Moalem S , Garcia A , Somerville MJ , Hicks M , Andrews D , Azad A , Schwarz P , Beheshti Zavareh R , Birkan R , Choo C , Chow V , Dhaliwal S , Duda V , Kupferschmidt AL , Lam K , Lightman D , Machalek K , Mar W , Nguyen F , Rytwinski PJ , Svara E , Tran M , Wheeler K , Yeung L , Zanibbi K , Zener R , Ziraldo M , Freedman M ((2008) ) Involvement of ApoE E4 and H63D in sporadic Alzheimer’s disease in a folate-supplemented Ontario population. J Alzheimers Dis 14: , 69–84.

[28] 

Percy M , Somerville MJ , Hicks M , Garcia A , Colelli T , Wright E , Kitaygorodsky J , Jiang A , Ho V , Parpia A , Wong MK ((2014) ) Risk factors for development of dementia in a unique six-year cohort study. I. An exploratory, pilot study of involvement of the E4 allele of apolipoprotein E, mutations of the hemochromatosis-HFE gene, type 2 diabetes, and stroke. J Alzheimers Dis 38: , 907–922.

[29] 

Alizadeh BZ , Njajou OT , Millan MR , Hofman A , Breteler MM , van Duijn CM ((2009) ) HFE variants, APOE and Alzheimer’s disease: Findings from the population-based Rotterdam study. Neurobiol Aging 30: , 330–332.

[30] 

Benyamin B , Esko T , Ried JS , Radhakrishnan A , Vermeulen SH , Traglia M , Gogele M , Anderson D , Broer L , Podmore C , Luan J , Kutalik Z , Sanna S , van der Meer P , Tanaka T , Wang F , Westra HJ , Franke L , Mihailov E , Milani L , Haldin J , Winkelmann J , Meitinger T , Thiery J , Peters A , Waldenberger M , Rendon A , Jolley J , Sambrook J , Kiemeney LA , Sweep FC , Sala CF , Schwienbacher C , Pichler I , Hui J , Demirkan A , Isaacs A , Amin N , Steri M , Waeber G , Verweij N , Powell JE , Nyholt DR , Heath AC , Madden PA , Visscher PM , Wright MJ , Montgomery GW , Martin NG , Hernandez D , Bandinelli S , van der Harst P , Uda M , Vollenweider P , Scott RA , Langenberg C , Wareham NJ , InterAct C , van Duijn C , Beilby J , Pramstaller PP , Hicks AA , Ouwehand WH , Oexle K , Gieger C , Metspalu A , Camaschella C , Toniolo D , Swinkels DW , Whitfield JB ((2014) ) Novel loci affecting iron homeostasis and their effects in individuals at risk for hemochromatosis. Nat Commun 5: , 4926.

[31] 

Escott-Price V , Sims R , Bannister C , Harold D , Vronskaya M , Majounie E , Badarinarayan N , Morgan K , Passmore P , Holmes C , Powell J , Brayne C , Gill M , Mead S , Goate A , Cruchaga C , Lambert JC , van Duijn C , Maier W , Ramirez A , Holmans P , Jones L , Hardy J , Seshadri S , Schellenberg GD , Amouyel P , Williams J ((2015) ) Common polygenic variation enhances risk prediction for Alzheimer’s disease. Brain 138: , 3673–3684.

[32] 

Lupton MK , Strike L , Hansell NK , Wen W , Mather KA , Armstrong NJ , Thalamuthu A , McMahon KL , de Zubicaray GI , Assareh AA , Simmons A , Proitsi P , Powell JF , Montgomery GW , Hibar DP , Westman E , Tsolaki M , Kloszewska I , Soininen H , Mecocci P , Velas B , Lovestone S , Brodaty H , Ames D , Trollor JN , Martin NG , Thompson PM , Sachdev PS , Wright MJ ((2016) ) The effect of increased genetic risk for Alzheimer’s disease on hippocampal and amygdala volume. Neurobiol Aging 40: , 68–77.

[33] 

Desikan RS , Schork AJ , Wang Y , Thompson WK , Dehghan A , Ridker PM , Chasman DI , McEvoy LK , Holland D , Chen CH , Karow DS , Brewer JB , Hess CP , Williams J , Sims R , O’Donovan MC , Choi SH , Bis JC , Ikram MA , Gudnason V , DeStefano AL , van der Lee SJ , Psaty BM , van Duijn CM , Launer L , Seshadri S , Pericak-Vance MA , Mayeux R , Haines JL , Farrer LA , Hardy J , Ulstein ID , Aarsland D , Fladby T , White LR , Sando SB , Rongve A , Witoelar A , Djurovic S , Hyman BT , Snaedal J , Steinberg S , Stefansson H , Stefansson K , Schellenberg GD , Andreassen OA , Dale AM ((2015) ) Polygenic overlap between c-reactive protein, plasma lipids, and Alzheimer disease. Circulation 131: , 2061–2069.

[34] 

Benyamin B , Ferreira MA , Willemsen G , Gordon S , Middelberg RP , McEvoy BP , Hottenga JJ , Henders AK , Campbell MJ , Wallace L , Frazer IH , Heath AC , de Geus EJ , Nyholt DR , Visscher PM , Penninx BW , Boomsma DI , Martin NG , Montgomery GW , Whitfield JB ((2009) ) Common variants in TMPRSS6 are associated with iron status and erythrocyte volume. Nat Genet 41: , 1173–1175.

[35] 

Painter JN , Anderson CA , Nyholt DR , Macgregor S , Lin J , Lee SH , Lambert A , Zhao ZZ , Roseman F , Guo Q , Gordon SD , Wallace L , Henders AK , Visscher PM , Kraft P , Martin NG , Morris AP , Treloar SA , Kennedy SH , Missmer SA , Montgomery GW , Zondervan KT ((2011) ) Genome-wide association study identifies a locus at 7p15.2 associated with endometriosis. Nat Genet 43: , 51–54.

[36] 

Purcell SM , Wray NR , Stone JL , Visscher PM , O’Donovan MC , Sullivan PF , Sklar P ((2009) ) Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature 460: , 748–752.

[37] 

Zhou X , Stephens M ((2012) ) Genome-wide efficient mixed-model analysis for association studies. Nat Genet 44: , 821–824.

[38] 

Nyholt DR ((2014) ) SECA: SNP effect concordance analysis using genome-wide association summary results. Bioinformatics 30: , 2086–2088.

[39] 

Belaidi AA , Bush AI ((2016) ) Iron neurochemistry in Alzheimer’s disease and Parkinson’s disease: Targets for therapeutics. J Neurochem 139: (Suppl 1), 179–197.

[40] 

Everett J , Cespedes E , Shelford LR , Exley C , Collingwood JF , Dobson J , van der Laan G , Jenkins CA , Arenholz E , Telling ND ((2014) ) Ferrous iron formation followingthe co- aggregation of ferric iron and theAlzheimer’s disease peptide beta-amyloid (1-42). J R Soc Interfac 11: , 20140165.

[41] 

Barnard ND , Bush AI , Ceccarelli A , Cooper J , de Jager CA , Erickson KI , Fraser G , Kesler S , Levin SM , Lucey B , Morris MC , Squitti R ((2014) ) Dietary and lifestyle guidelines for the prevention of Alzheimer’s disease. Neurobiol Aging 35: (Suppl 2), S74–S78.

[42] 

Hare DJ , Arora M , Jenkins NL , Finkelstein DI , Doble PA , Bush AI ((2015) ) Is early-life iron exposure critical in neurodegeneration? Nat Rev Neurol 11: , 536–544.

[43] 

Harold D , Abraham R , Hollingworth P , Sims R , Gerrish A , Hamshere ML , Pahwa JS , Moskvina V , Dowzell K , Williams A , Jones N , Thomas C , Stretton A , Morgan AR , Lovestone S , Powell J , Proitsi P , Lupton MK , Brayne C , Rubinsztein DC , Gill M , Lawlor B , Lynch A , Morgan K , Brown KS , Passmore PA , Craig D , McGuinness B , Todd S , Holmes C , Mann D , Smith AD , Love S , Kehoe PG , Hardy J , Mead S , Fox N , Rossor M , Collinge J , Maier W , Jessen F , Schurmann B , van den Bussche H , Heuser I , Kornhuber J , Wiltfang J , Dichgans M , Frolich L , Hampel H , Hull M , Rujescu D , Goate AM , Kauwe JS , Cruchaga C , Nowotny P , Morris JC , Mayo K , Sleegers K , Bettens K , Engelborghs S , De Deyn PP , Van Broeckhoven C , Livingston G , Bass NJ , Gurling H , McQuillin A , Gwilliam R , Deloukas P , Al-Chalabi A , Shaw CE , Tsolaki M , Singleton AB , Guerreiro R , Muhleisen TW , Nothen MM , Moebus S , Jockel KH , Klopp N , Wichmann HE , Carrasquillo MM , Pankratz VS , Younkin SG , Holmans PA , O’Donovan M , Owen MJ , Williams J ((2009) ) Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer’s disease. Nat Genet 41: , 1088–1093.

[44] 

Lovestone S , Francis P , Kloszewska I , Mecocci P , Simmons A , Soininen H , Spenger C , Tsolaki M , Vellas B , Wahlund LO , Ward M , AddNeuroMed Consortium ((2009) ) AddNeuroMed–the European collaboration for the discovery of novel biomarkers for Alzheimer’s disease. Ann N Y Acad Sci 1180: , 36–46.

[45] 

Hye A , Riddoch-Contreras J , Baird AL , Ashton NJ , Bazenet C , Leung R , Westman E , Simmons A , Dobson R , Sattlecker M , Lupton M , Lunnon K , Keohane A , Ward M , Pike I , Zucht HD , Pepin D , Zheng W , Tunnicliffe A , Richardson J , Gauthier S , Soininen H , Kłoszewska I , Mecocci P , Tsolaki M , Vellas B , Lovestone S ((2014) ) Plasma proteins predict conversion to dementia from prodromal disease. Alzheimers Dement 10: , 799–807.

[46] 

Mueller SG , Weiner MW , Thal LJ , Petersen RC , Jack CR , Jagust W , Trojanowski JQ , Toga AW , Beckett L ((2005) ) Ways toward an early diagnosis in Alzheimer’s disease: The Alzheimer’s Disease Neuroimaging Initiative (ADNI). Alzheimers Dement 1: , 55–66.

[47] 

Wellcome Trust Case Control Consortium ((2007) ) Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 447: , 661–678.