Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Genetic Loci Associated with Alzheimer’s Disease and Cerebrospinal Fluid Biomarkers in a Finnish Case-Control Cohort

  • Lyzel S. Elias-Sonnenschein ,

    Contributed equally to this work with: Lyzel S. Elias-Sonnenschein, Seppo Helisalmi

    Affiliation Department of Psychiatry and Neuropsychology, School for Mental Health and Neuroscience, Alzheimer Center Limburg, Maastricht University, Maastricht, The Netherlands

  • Seppo Helisalmi ,

    Contributed equally to this work with: Lyzel S. Elias-Sonnenschein, Seppo Helisalmi

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Teemu Natunen,

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Anette Hall,

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Teemu Paajanen,

    Current address: Cognition and Work Team, Finnish Institute of Occupational Health, Helsinki, Finland

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Sanna-Kaisa Herukka,

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Marjo Laitinen,

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Anne M. Remes,

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Anne M. Koivisto,

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Kari M. Mattila,

    Affiliation Department of Clinical Chemistry, Fimlab Laboratories and School of Medicine, University of Tampere, Tampere, Finland

  • Terho Lehtimäki,

    Affiliation Department of Clinical Chemistry, Fimlab Laboratories and School of Medicine, University of Tampere, Tampere, Finland

  • Frans R. J. Verhey,

    Affiliation Department of Psychiatry and Neuropsychology, School for Mental Health and Neuroscience, Alzheimer Center Limburg, Maastricht University, Maastricht, The Netherlands

  • Pieter Jelle Visser,

    Affiliations Department of Psychiatry and Neuropsychology, School for Mental Health and Neuroscience, Alzheimer Center Limburg, Maastricht University, Maastricht, The Netherlands, Department of Neurology, Alzheimer Center, VU University Medical Center Amsterdam, Amsterdam, The Netherlands

  • Hilkka Soininen,

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland

  • Mikko Hiltunen

    Affiliation Institute of Clinical Medicine-Neurology, University of Eastern Finland and Department of Neurology, Kuopio University Hospital, Kuopio, Finland



To understand the relation between risk genes for Alzheimer’s disease (AD) and their influence on biomarkers for AD, we examined the association of AD in the Finnish cohort with single nucleotide polymorphisms (SNPs) from top AlzGene loci, genome-wide association studies (GWAS), and candidate gene studies; and tested the correlation between these SNPs and AD markers Aβ1–42, total tau (t-tau), and phosphorylated tau (p-tau) in cerebrospinal fluid (CSF).


We tested 25 SNPs for genetic association with clinical AD in our cohort comprised of 890 AD patients and 701-age matched healthy controls using logistic regression. For the correlational study with biomarkers, we tested 36 SNPs in a subset of 222 AD patients with available CSF using mixed models. Statistical analyses were adjusted for age, gender and APOE status. False discovery rate for multiple testing was applied. All participants were from academic hospital and research institutions in Finland.


APOE-ε4, CLU rs11136000, and MS4A4A rs2304933 correlated with significantly decreased Aβ1–42 (corrected p<0.05). At an uncorrected p<0.05, PPP3R1 rs1868402 and MAPT rs2435211 were related with increased t-tau; while SORL1 rs73595277 and MAPT rs16940758, with increased p-tau. Only TOMM40 rs2075650 showed association with clinical AD after adjusting for APOE-ε4 (p = 0.007), but not after multiple test correction (p>0.05).


We provide evidence that APOE-ε4, CLU and MS4A4A, which have been identified in GWAS to be associated with AD, also significantly reduced CSF Aβ1–42 in AD. None of the other AlzGene and GWAS loci showed significant effects on CSF tau. The effects of other SNPs on CSF biomarkers and clinical AD diagnosis did not reach statistical significance. Our findings suggest that APOE-ε4, CLU and MS4A4A influence both AD risk and CSF Aβ1–42.


Alzheimer’s disease (AD) is a neurodegenerative disease with a complex etiology. Neuritic plaques mainly composed of aggregated β-amyloid (Aβ) and neurofibrillary tangles (NFTs) resulting from hyperphosporylated tau protein (p-tau) are pathological hallmarks of AD [1]. Aβ1–42 tends to aggregate more compared to other Aβ isoforms [2]. Total tau (t-tau) concentrations in cerebrospinal fluid (CSF) have been suggested to indicate the extent of neuronal damage, while p-tau levels reflected the phosphorylated state of tau [3]. P-tau and t-tau levels signal axonal degeneration [4], [5]. In AD, the concentration of CSF Aβ1–42 is decreased, which is supposed to reflect sequestration of Aβ1–42 in amyloid plaques in the brain [6], while t-tau and p-tau levels are increased [1].

The majority of AD cases have been reported to have a strong genetic component [7]. The apolipoprotein E (APOE) ε4 allele is the strongest known genetic risk factor for AD. Other high-risk genetic variants have been identified in genome-wide association studies (GWAS) (for review, see [8]). In addition, previous studies showed that a number of candidate genes correlated with Aβ or tau. However, the relation between these AD risk genes and AD biomarkers remains ambiguous. With the exception of APOE and translocase of outer mitochondrial membrane 40 homolog (TOMM40), single nucleotide polymorphisms (SNPs) identified in case-control GWAS with clinical AD as outcome have not been replicated in GWAS with biomarkers as outcome [9].

In the study, we first investigated whether previously reported genetic risk factors for AD were associated with AD risk in a Finnish case-control cohort. Second, we tested in the AD group the effects of these variants on the AD markers Aβ1–42 and tau in CSF. We selected SNPs from AlzGene and GWAS, and from candidate genes that previously showed a relation with CSF Aβ1–42 and tau.

Materials and Methods

Study Population

The Finnish-AD is a multicenter cohort comprised of 890 AD patients and 701 age-matched healthy controls from Kuopio, Oulu, and Tampere in Finland. All patients were diagnosed with probable AD according to the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) criteria [10]. AD patients with an early onset did not show conclusive evidence of autosomal dominant transmission or mutations in the amyloid precursor protein (APP), presenilin 1 (PSEN1), or presenilin 2 (PSEN2) genes. Control subjects had no symptoms of cognitive impairment based on clinical interview and neuropsychological examination. Of the 890 AD patients, 222 from Kuopio provided CSF for our study.

The Ethics Committee of the North-Savo Hospital District, Kuopio Finland, approved the study. The physician and/or the study nurse gave written information of the study and explained the study protocol to the patient and caretaker if available. All participants provided written informed consent. A next of kin, caretakers or guardians consented on the behalf of participants whose capacity to consent was compromised. In these cases also the patient's own opinion was asked and considered, and the patient was recruited in the study only when he or she also agreed with this. The ethics committee approved this informed consent procedure.

Gene Selection

We selected genes based on their reported association with AD or effect on CSF Aβ and tau. We included the 10 ‘top results’ loci from AlzGene, which is an online database providing meta-analyses of published genetic association studies (for AlzGene top results criteria, see [11]), and selected the most promising SNPs from GWAS or other candidate gene studies. APOE, CR1, BIN1, CD2AP, CLU, MS4A4E, MS4A6A, PICALM, ABCA7, and CD33 were from AlzGene (for gene names, see Table S1). Analyses with APOE alone were performed for reference purposes, as APOE is an established susceptibility gene for sporadic AD [11][13]. Functions of the AlzGene variants in relation to AD have been previously described [9], [14], [15]. MS4A4A, EXOC3L2 and MTHFD1L have been shown in GWAS to be associated with AD [16][18] but have not been included in the AlzGene top list to date.

CYP19A and TOMM40 have been shown in GWAS with biomarkers as outcome to be related to Aβ, whereas EPC2 and RELN were associated with tau [19][21]. CYP19A has also been reported to increase AD risk [22].

The candidate genes we selected that have been studied in relation to CSF biomarkers were ACE, IDE, MAPT, SORL1, CYP46A1, BDNF, TF, PPP3R1, and another TOMM40 polymorphism. The effects of ACE, MAPT, SORL, and TOMM40 on Aβ and tau have been reported in a review (see [9]). PPP3R1 has been correlated with increased p-tau [4]. BDNF was linked with decreased total Aβ while TF was related to decreased Aβ1–42/Aβ1–40 ratio [23]. CYP46A1 has been associated with AD [24] and correlated with Aβ1–42 [25]. IDE has been reported to decrease Aβ1–42 in AD [26]. Moreover, it has been associated with AD risk [27] and neuropathological Aβ deposition [28].

Thirty-six SNPs in 25 genes were included in the analyses with CSF biomarkers (Table S1). For the genetic association analyses, we tested 25 genetic variants, excluding those in ABCA7, BIN1, BDNF, CD2AP, CD33, CYP46, CLU, CR1, EPHA1, EXOC3L2, MS4A4E and MS4A6A because these genes were previously studied in the Finnish population and found to be associated with AD either in the Finnish cohort alone or in multicenter GWAS [16], [24], [29][31].


DNA was extracted from peripheral blood with EDTA and amplified using polymerase chain reaction technique. DNA samples were randomly placed on 384-well plates. Genotyping using Sequenom iPlex platform (Sequenom, Hamburg, Germany) was performed at University of Eastern Finland (UEF) in Kuopio. Patients and controls were dichotomized as APOE-ε4 carriers or noncarriers.

Quality control procedures included using duplicates and negative controls, filtering on individual sample and SNP call rate, and testing whether the SNPs were in Hardy-Weinberg equilibrium (HWE). Samples with an average call rate of 90% were included. SNPs for the association analysis were in Hardy-Weinberg equilibrium (p>0.001).

CSF Analysis

CSF was obtained through lumbar puncture performed at UEF and the Kuopio University Hospital. Aβ1–42, t-tau, and p-tau levels were measured using commercially available INNOTEST enzyme-linked immunosorbent assays (ELISAs) (Innotest ß-amyloid(1–42), Innotest hTau Ag, and Innotest Phospho-tau(181P), Innogenetics, Ghent, Belgium). All measurements were performed at UEF.

The relation between SNPs and biomarkers was assessed only for the AD group because the number of controls with available CSF (n = 30) was too small to allow for meaningful analyses.

Statistical Analysis

SPSS version 19 (Chicago, IL, USA) was used for the statistical analyses.

We performed power calculations for the genetic association analyses [32]. Simulation analysis yielded more than 80% power to achieve ∼1.3–1.5 risk effect at p = 0.05, indicating that our case-control sample size was sufficient to find moderate genetic association with AD. The same risk effect for the false discovery rate (FDR) corrected p = 0.005 yielded more than 50% power. For quantitative trait association [33], we computed ∼80% power for an effect size of 60–70 pg/ml for the dominant model at p = 0.05 and 50% power at FDR corrected p = 0.005. Similarly for t-tau, the effect size was 120–150 pg/ml and for p-tau 15–20 pg/ml for a power of 80% at p = 0.05 and 50% when FDR corrected (p = 0.005).

Genetic association with AD was examined using Pearson chi-square test for multiple group comparisons and binary logistic regression for pairwise comparisons in univariate and multivariate analysis adjusted for age, sex, and APOE-ε4 status. The relation between SNPs and biomarkers was assessed through mixed models for multiple group comparisons and pairwise comparisons, with correction for confounders. Normality was tested and assumed for biomarkers.

For the binary association, we first performed an overall test that provided information on genotype differences between cases and controls, and multiple comparison tests. For the quantitative association, we compared differences in mean biomarker values between genotypes in the overall test. Based on genotype frequencies, we opted to use a dominant model. Minor genotype frequencies for both genetic association and CSF biomarker analyses were 0–15% for 40–50% of the SNPs studied. This reduces statistical power of a recessive model in most SNPs. All analyses were first corrected for age and sex, and then repeated with corrections for age, sex, and APOE-ε4 status.

We used FDR correction for multiple testing, following the method of Benjamini and Hochberg [34]. Corrections were based on the number of SNPs tested and were performed separately for binary and quantitative associations. Observed p-values were ranked from smallest to largest. Adjusted p-values were successively computed in a step-up manner, starting from the second largest p-value, as follows: observed p-value (total number of SNPs tested/rank). Statistical significance was set at FDR adjusted p<0.05.


Characteristics of the Cohort

Table 1 shows the demographic characteristics of the participants. AD patients with available CSF did not differ from patients without CSF in terms of onset age (p = 0.07) and Mini-Mental State Examination [35] score (p = 0.77).

Genetic Association with AD

APOE conferred a significant AD risk of 6.30 times higher among ε4 allele carriers compared to noncarriers (p<0.0001, Table 2). Age and gender did not affect the results (OR = 6.25, 95% CI 5.22–8.15, p<0.0001).

Table 2. SNPs associated with Alzheimer’s disease risk.

TOMM40 rs157580, rs2075650, and rs8106922 indicated association with AD risk in the univariate analyses (FDR p<0.05). A protective effect was observed for G allele carriers of rs157580 and rs8106922, and a risk effect for G allele carriers of rs2075650. Only rs2075650 remained significant in the multivariate analysis (unadjusted p<0.007) but did not pass FDR correction (p>0.05).

Results for MAPT rs16940758 suggested no association with AD (FDR adjusted p>0.05), although AD and control groups differed in multiple comparison test. PPP3R1 rs1868402 and ACE rs4293 showed association with AD in multivariate analysis (unadjusted p = 0.03 for rs1868402 and p = 0.01 for rs4293) but did not remain significant after FDR adjustment. Overall and univariate analyses with rs1868402 and rs4293 were not significant.

Effects of SNPS on CSF Aβ1–42

APOE ε4 allele carriers had significantly reduced CSF Aβ1–42 (FDR adjusted p<0.05, Table 3). Apart from APOE, only CLU significantly affected Aβ1–42 among the AlzGene top loci. Carriers of rs11136000 major allele (C, risk allele in AlzGene meta-analysis) showed significantly decreased Aβ1–42 (FDR adjusted p<0.05).

Table 3. Effects of SNPs on CSF Aβ1–42 in Alzheimer’s disease.

Of the SNPs identified in GWAS but not in the top AlzGene loci, only MS4A4A correlated with Aβ1–42. Minor allele carriers (A, risk allele in GWAS) of rs2304933 showed significantly decreased Aβ1–42 levels (FDR adjusted p<0.05) compared to major allele (C) carriers. The correlation was strengthened when corrected for APOE-ε4 status.

Decreased Aβ1–42 levels among EXOC3L2 rs597668 minor allele carriers (C, risk allele in GWAS) were observed (unadjusted p = 0.02) but the correlation did not pass FDR filter (p>0.05).

Risk allele carriers of TOMM40 SNPs identified through GWAS with biomarkers as outcome and through other candidate gene studies had decreased Aβ1–42 concentrations, but this was not independent of APOE-ε4 status (FDR adjusted p>0.05).

The rest of the SNPs did not exhibit conclusive effects on Aβ1–42 concentrations.

Effects of SNPs on CSF Tau

None of the SNPs from the top AlzGene loci significantly correlated with CSF t-tau and p-tau. The strongest effect was for PICALM rs642949 and t-tau (p = 0.06, adjusted for age, sex, and APOE-ε4 status), with increased t-tau levels among minor allele (C) carriers (Table 4). APOE ε4 allele carriers also showed a nonsignificant increase in t-tau concentrations (p = 0.08).

Table 4. Effects of SNPs on CSF t-tau and p-tau in Alzheimer’s disease.

None of the polymorphisms in other GWAS with clinical AD as outcome were linked with t-tau and p-tau.

From GWAS SNPs with biomarker as outcome, none of the results attained statistical significance. EPC2 rs1374441 and t-tau showed the strongest effect. T-tau levels increased among minor allele (C, risk allele in GWAS) carriers (p = 0.06, corrected for age, sex, and APOE-ε4 status).

Of the SNPs selected from other candidate genes, marginal correlations at FDR unadjusted p<0.05 were obtained for polymorphisms in PPP3R1, SORL1, and MAPT (Table 4). T-tau levels increased among carriers of PPP3R1 rs1868402 major allele (T, FDR adjusted p>0.05). The effect was strengthened when corrected for APOE-ε4 status.

SORL1 rs73595277 minor allele carriers (G) had increased p-tau (FDR adjusted p>0.05). Effects of SORL1 slightly decreased with APOE-ε4 correction.

In MAPT, minor allele carriers of rs2435211 (T) had increased t-tau levels (unadjusted p = 0.03) and minor allele carriers of rs16940758 had increased p-tau levels (unadjusted p = 0.03). Correcting for APOE-ε4 status enhanced the effects. These results, however, did not remain significant after FDR correction (adjusted p>0.05 for both SNPs).


We performed a case-control genetic association analysis of 25 AD risk variants and tested the effects of 36 risk variants on CSF Aβ1–42, t-tau, and p-tau in the AD group.

The allele frequencies of genes from AlzGene meta-analysis of Caucasian ancestry were comparable with allele frequencies of most SNPs in our study. Only APOE and TOMM40 showed genetic association with AD. However, TOMM40 is in linkage disequilibrium with APOE and did not exhibit an effect independent of APOE, in accordance with recent evidence [36].

Of the AlzGene top loci, APOE and CLU correlated significantly with decreased Aβ1–42. Our result for APOE confirms previous findings [37]. Other studies on CLU have not found the same effect, which may be attributed to their smaller sample size [38], [39]. Consistent with GWAS findings, the minor allele of CLU exerted a protective effect on AD risk [29], and on CSF Aβ1–42 in the Finnish cohort. CLU did not significantly affect tau. CLU binds soluble Aβ and plays a role in Aβ clearance and aggregation [14], which could partly explain why it primarily affected Aβ.

PICALM affects Aβ concentration in the brain through endocytic processes. Rs3851179 has been reported to correlate with CSF Aβ1–42 in another study, the major allele (G) being the risk allele [39]. In our cohort, we found no correlation between rs3851179 and CSF markers. Instead, rs642949 minor allele carriers (C) had decreased Aβ1–42 and increased t-tau levels, although these were not statistically significant. The C allele of rs642949 has been reported to exert a risk effect in a case-control study [40].

APOE and the other top AlzGene loci did not correlate with t-tau and p-tau in our cohort.

Of the GWAS SNPs not in AlzGene top loci, MS4A4A showed significant correlation with Aβ1–42 but not with tau. This is a novel finding. MS4A4A belongs to the MS4A cluster [18]. Not much is known yet about MS4A4A rs2304933 but its effect on Aβ1–42 is consistent with our genetic association analysis results suggesting a risk effect of the minor (A) allele. The mechanisms by which MS4A4A affect CSF Aβ1–42 levels need further investigation.

The decrease in Aβ1–42 levels among C allele carriers of EXOC3L2 rs597668 is interesting. Although not statistically significant (FDR adjusted p>0.05), it conferred with a previous study identifying the C allele as a risk allele in the Finnish population [30]. The C allele has also been reported in another study to promote AD progression [41].

Among GWAS SNPs with biomarker as outcome, TOMM40 SNPs correlated with CSF Aβ1–42 but not with tau, which was consistent with previous findings [19], [20]. None of the TOMM40 SNPs remained significant after APOE correction.

The SNPs from candidate genes were not related to CSF Aβ1–42. For tau, the small effects we observed for variants in PPP3R1, SORL1, and MAPT (unadjusted p<0.05) were suggestive of a trend. PPP3R1 is a protein phosphatase and is the calcium binding regulatory subunit of calcineurin [42]. Calcineurin is involved in modulating tau phosphorylation [43]. A previous study found PPP3R1 to correlate with p-tau [4]. In our cohort, we found an effect on t-tau but the effect on p-tau was weaker. This difference in results could partly be attributed to variability in population.

SORL1 binds to ApoE and plays a role in Aβ1–42 production [37]. Numerous SNPs in SORL1 have been studied in relation to AD [44] and CSF biomarkers [37] but results so far have been inconclusive [9]. One study found that rs3824968 in SORL1 significantly reduced Aβ1–42 in AD [45] whereas another study reported no correlation [46]. SORL1 appears to exert small effects likely to be detected only in mega-analysis of pooled samples [47] or in haplotypes [48]. In general, single loci in SORL1 did not correlate with CSF Aβ1–42 and tau [9]. Our result relating SORL with p-tau may be due to chance.

MAPT codes for tau proteins. Aggregated hyperphosphorylated tau proteins are a component of NFTs. Consistent with the findings of another study, we found MAPT to correlate with CSF t-tau [49].

We noted a number of changes in the strength of correlation when we corrected for APOE. TOMM40 was no longer associated with AD and CSF Aβ1–42. This confirms previous finding that TOMM40 is in strong linkage disequilibrium with APOE. The effect of EXOC3L2 on CSF Aβ1–42 became nonsignificant. This suggests that the effect was not independent of APOE. Due to our small sample size, our result could also be a false positive. For the other SNPs, we found no or minor changes on either AD risk or correlation with CSF markers after APOE correction. This means that the effects of these SNPs were independent of APOE, or that our sample size was too small to detect interaction effects.

We found few SNP-related differences in CSF Aβ because Aβ may already be strongly decreased in AD patients who were already demented. The effect of genetic risk factors on amyloid metabolism may be more evident in predementia stages. Of interest is the weak correlation of MAPT and PPP3R1 with tau, which confirms the role of these SNPs in tau metabolism. MAPT and PPP3R1 also showed weak associations with clinical AD, which suggest that they contribute to dementia risk. In the analyses with biomarkers, none of the AlzGene or other GWAS SNPs were related to tau, suggesting that these SNPs have no clear effect on tau metabolism.

Our study had the unique design that we tested SNPs both for genetic association in a clinical case-control design and for correlation with CSF biomarkers. Another strength was the large selection of high-risk SNPs identified by GWAS or other candidate gene studies covering different possible pathophysiological pathways. The small sample size remained a limitation. We had sufficient power for finding moderate associations, but a larger sample is needed for detecting the very small effect sizes of the other SNPs studied, if these effects are present.

In conclusion, we provide evidence that APOE, CLU, and MS4A4A, which have been identified in GWAS to be associated with AD, also significantly affected CSF Aβ1–42. To our knowledge, ours is the first study to report on the correlation between MS4A4A and CSF Aβ1–42. None of the AD risk genes studied showed significant effects on CSF tau. The nonsignificant trends in PPP3R1 and MAPT in relation to tau may be due to our small sample size rather than genuine lack of risk effects. Collaboration on a larger scale is necessary to ascertain the effects of the aforementioned SNPs and identify reliable genetic risk variants for AD markers in CSF.

Supporting Information

Table S1.

Genetic variants included in the study. We listed here the 36 SNPs that were all tested for correlation with CSF Aβ1–42 and tau. We indicated which SNPs were excluded from the genetic association analysis because they have been previously genotyped and reported in genetic association studies including the Finnish cohort.


Author Contributions

Critical revision/comments of the paper: SH PJV TN AH TP SKH ML AMR AMK KMM TL FRJV HS MH. Conceived and designed the experiments: LSES PJV SH MH. Performed the experiments: SH SKH ML LSES. Analyzed the data: SH AH PJV LSES. Contributed reagents/materials/analysis tools: SH MH AMR TL KMM. Wrote the paper: LSES.


  1. 1. Blennow K, de Leon MJ, Zetterberg H (2006) Alzheimer’s disease. Lancet 368: 387–403.
  2. 2. Bu G (2009) Apolipoprotein E and its receptors in Alzheimer’s disease: pathways, pathogenesis and therapy. Nat Rev Neurosci 10: 333–344.
  3. 3. Bekris LM, Millard S, Lutz F, Li G, Galasko DR, et al. (2012) Tau phosphorylation pathway genes and cerebrospinal fluid tau levels in Alzheimer’s disease. Am J Med Genet B Neuropsychiatr Genet 159B: 874–883.
  4. 4. Cruchaga C, Kauwe JS, Mayo K, Spiegel N, Bertelsen S, et al.. (2010) SNPs associated with cerebrospinal fluid phospho-tau levels influence rate of decline in Alzheimer’s disease. PLoS Genet 6.
  5. 5. Spires-Jones TL, Stoothoff WH, de Calignon A, Jones PB, Hyman BT (2009) Tau pathophysiology in neurodegeneration: a tangled issue. Trends Neurosci 32: 150–159.
  6. 6. Fagan AM, Mintun MA, Mach RH, Lee SY, Dence CS, et al. (2006) Inverse relation between in vivo amyloid imaging load and cerebrospinal fluid Abeta42 in humans. Ann Neurol 59: 512–519.
  7. 7. Lambert JC, Amouyel P (2011) Genetics of Alzheimer’s disease: new evidences for an old hypothesis? Curr Opin Genet Dev 21: 295–301.
  8. 8. Bertram L, Lill CM, Tanzi RE (2010) The genetics of Alzheimer disease: back to the future. Neuron 68: 270–281.
  9. 9. Elias-Sonnenschein LS, Bertram L, Visser PJ (2012) Relationship between genetic risk factors and markers for Alzheimer’s disease pathology. Biomark Med 6: 477–495.
  10. 10. McKhann G, Drachman D, Folstein M, Katzman R, Price D, et al. (1984) Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 34: 939–944.
  11. 11. Bertram L, McQueen MB, Mullin K, Blacker D, Tanzi RE (2007) Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nat Genet 39: 17–23.
  12. 12. Strittmatter WJ, Saunders AM, Schmechel D, Pericak-Vance M, Enghild J, et al. (1993) Apolipoprotein E: high-avidity binding to beta-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci U S A 90: 1977–1981.
  13. 13. Farrer LA, Cupples LA, Haines JL, Hyman B, Kukull WA, et al. (1997) Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta Analysis Consortium. JAMA 278: 1349–1356.
  14. 14. Schellenberg GD, Montine TJ (2012) The genetics and neuropathology of Alzheimer’s disease. Acta Neuropathol.
  15. 15. Olgiati P, Politis AM, Papadimitriou GN, De Ronchi D, Serretti A (2011) Genetics of late-onset Alzheimer’s disease: update from the alzgene database and analysis of shared pathways. Int J Alzheimers Dis 2011: 832379.
  16. 16. Hollingworth P, Harold D, Sims R, Gerrish A, Lambert JC, et al. (2011) Common variants at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33 and CD2AP are associated with Alzheimer’s disease. Nat Genet 43: 429–435.
  17. 17. Seshadri S, Fitzpatrick AL, Ikram MA, DeStefano AL, Gudnason V, et al. (2010) Genome-wide analysis of genetic loci associated with Alzheimer disease. JAMA 303: 1832–1840.
  18. 18. Naj AC, Jun G, Beecham GW, Wang LS, Vardarajan BN, et al. (2011) Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer’s disease. Nat Genet 43: 436–441.
  19. 19. Han MR, Schellenberg GD, Wang LS (2010) Genome-wide association reveals genetic effects on human Abeta42 and tau protein levels in cerebrospinal fluids: a case control study. BMC Neurol 10: 90.
  20. 20. Kim S, Swaminathan S, Shen L, Risacher SL, Nho K, et al. (2011) Genome-wide association study of CSF biomarkers Abeta1–42, t-tau, and p-tau181p in the ADNI cohort. Neurology 76: 69–79.
  21. 21. Kramer PL, Xu H, Woltjer RL, Westaway SK, Clark D, et al. (2011) Alzheimer disease pathology in cognitively healthy elderly: a genome-wide study. Neurobiol Aging 32: 2113–2122.
  22. 22. Iivonen S, Corder E, Lehtovirta M, Helisalmi S, Mannermaa A, et al. (2004) Polymorphisms in the CYP19 gene confer increased risk for Alzheimer disease. Neurology 62: 1170–1176.
  23. 23. Kauwe JS, Wang J, Mayo K, Morris JC, Fagan AM, et al. (2009) Alzheimer’s disease risk variants show association with cerebrospinal fluid amyloid beta. Neurogenetics 10: 13–17.
  24. 24. Helisalmi S, Vepsalainen S, Koivisto AM, Mannermaa A, Iivonen S, et al. (2006) Association of CYP46 intron 2 polymorphism in Finnish Alzheimer’s disease samples and a global scale summary. J Neurol Neurosurg Psychiatry 77: 421–422.
  25. 25. Johansson A, Katzov H, Zetterberg H, Feuk L, Johansson B, et al. (2004) Variants of CYP46A1 may interact with age and APOE to influence CSF Abeta42 levels in Alzheimer’s disease. Hum Genet 114: 581–587.
  26. 26. Vepsalainen S, Helisalmi S, Mannermaa A, Pirttila T, Soininen H, et al. (2009) Combined risk effects of IDE and NEP gene variants on Alzheimer disease. J Neurol Neurosurg Psychiatry 80: 1268–1270.
  27. 27. Vepsalainen S, Parkinson M, Helisalmi S, Mannermaa A, Soininen H, et al. (2007) Insulin-degrading enzyme is genetically associated with Alzheimer’s disease in the Finnish population. J Med Genet 44: 606–608.
  28. 28. Blomqvist ME, Chalmers K, Andreasen N, Bogdanovic N, Wilcock GK, et al. (2005) Sequence variants of IDE are associated with the extent of beta-amyloid deposition in the Alzheimer’s disease brain. Neurobiol Aging 26: 795–802.
  29. 29. Lambert JC, Heath S, Even G, Campion D, Sleegers K, et al. (2009) Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer’s disease. Nat Genet 41: 1094–1099.
  30. 30. Lambert JC, Zelenika D, Hiltunen M, Chouraki V, Combarros O, et al.. (2011) Evidence of the association of BIN1 and PICALM with the AD risk in contrasting European populations. Neurobiol Aging 32: 756 e711–755.
  31. 31. Vepsalainen S, Castren E, Helisalmi S, Iivonen S, Mannermaa A, et al. (2005) Genetic analysis of BDNF and TrkB gene polymorphisms in Alzheimer’s disease. J Neurol 252: 423–428.
  32. 32. Purcell S, Cherny SS, Sham PC (2003) Genetic Power Calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics 19: 149–150.
  33. 33. Lenth RV (2006–9) Java Applets for Power and Sample Size.
  34. 34. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society, Series B 57: 289–300.
  35. 35. Folstein MF, Folstein SE, McHugh PR (1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12: 189–198.
  36. 36. Jun G, Vardarajan BN, Buros J, Yu CE, Hawk MV, et al.. (2012) Comprehensive Search for Alzheimer Disease Susceptibility Loci in the APOE Region. Arch Neurol: 1–10.
  37. 37. Kauwe JS, Cruchaga C, Bertelsen S, Mayo K, Latu W, et al. (2010) Validating predicted biological effects of Alzheimer’s disease associated SNPs using CSF biomarker levels. J Alzheimers Dis 21: 833–842.
  38. 38. Schott JM (2011) Using CSF biomarkers to replicate genetic associations in Alzheimer’s disease. Neurobiol Aging.
  39. 39. Schjeide BM, Schnack C, Lambert JC, Lill CM, Kirchheiner J, et al. (2011) The role of clusterin, complement receptor 1, and phosphatidylinositol binding clathrin assembly protein in Alzheimer disease risk and cerebrospinal fluid biomarker levels. Arch Gen Psychiatry 68: 207–213.
  40. 40. Rosenthal SL, Wang X, Demirci FY, Barmada MM, Ganguli M, et al. (2012) Beta-Amyloid Toxicity Modifier Genes and the Risk of Alzheimer’s Disease. Am J Neurodegener Dis 1: 191–198.
  41. 41. Schmidt C, Wolff M, von Ahsen N, Zerr I (2012) Alzheimer’s disease: genetic polymorphisms and rate of decline. Dement Geriatr Cogn Disord 33: 84–89.
  42. 42. Karch CM, Jeng AT, Goate AM (2012) Calcium phosphatase calcineurin influences tau metabolism. Neurobiol Aging.
  43. 43. Reese LC, Taglialatela G (2011) A role for calcineurin in Alzheimer’s disease. Curr Neuropharmacol 9: 685–692.
  44. 44. Reitz C, Cheng R, Rogaeva E, Lee JH, Tokuhiro S, et al. (2011) Meta-analysis of the association between variants in SORL1 and Alzheimer disease. Arch Neurol 68: 99–106.
  45. 45. Alexopoulos P, Guo LH, Kratzer M, Westerteicher C, Kurz A, et al. (2011) Impact of SORL1 single nucleotide polymorphisms on Alzheimer’s disease cerebrospinal fluid markers. Dement Geriatr Cogn Disord 32: 164–170.
  46. 46. Kolsch H, Jessen F, Wiltfang J, Lewczuk P, Dichgans M, et al. (2008) Influence of SORL1 gene variants: association with CSF amyloid-beta products in probable Alzheimer’s disease. Neurosci Lett 440: 68–71.
  47. 47. Olgiati P, Politis A, Albani D, Rodilossi S, Polito L, et al. (2012) Association of SORL1 alleles with late-onset Alzheimer’s disease. findings from the GIGAS_LOAD study and mega-analysis. Curr Alzheimer Res 9: 491–499.
  48. 48. Cuenco KT, Lunetta KL, Baldwin CT, McKee AC, Guo J, et al. (2008) Association of distinct variants in SORL1 with cerebrovascular and neurodegenerative changes related to Alzheimer disease. Arch Neurol 65: 1640–1648.
  49. 49. Kauwe JS, Cruchaga C, Mayo K, Fenoglio C, Bertelsen S, et al. (2008) Variation in MAPT is associated with cerebrospinal fluid tau levels in the presence of amyloid-beta deposition. Proc Natl Acad Sci U S A 105: 8050–8054.