CoChat
Sign Up
Cited by 33Epigenetics and DNA MethylationDementia and Cognitive Impairment ResearchAlzheimer's disease research and treatments

Distinct sex-specific DNA methylation differences in Alzheimer’s disease

Open access
C. Silva, Tiago, Zhang, Wei, Young, Juan I., Gomez, Lissette, Schmidt, Michael A., Varma, Achintya, Chen, X. Steven, Martin, Eden R. +1 more·2022·Alzheimer's Research & Therapy
DOI 10.1186/s13195-022-01070-zPMC PMC9479371
Open this paper in CoChat — free accountOpen PDF

Abstract

Background: Sex is increasingly recognized as a significant factor contributing to the biological and clinical heterogeneity in AD. There is also growing evidence for the prominent role of DNA methylation (DNAm) in Alzheimer’s disease (AD). Methods: We studied sex-specific DNA methylation differences in the blood samples of AD subjects compared to cognitively normal subjects, by performing sex-specific meta-analyses of two large blood-based epigenome-wide association studies (ADNI and AIBL), which included DNA methylation data for a total of 1284 whole blood samples (632 females and 652 males). Within each dataset, we used two complementary analytical strategies, a sex-stratified analysis that examined methylation to AD associations in male and female samples separately, and a methylation-by-sex interaction analysis that compared the magnitude of these associations between different sexes. After adjusting for age, estimated immune cell type proportions, batch effects, and correcting for inflation, the inverse-variance fixed-effects meta-analysis model was used to identify the most consistent DNAm differences across datasets. In addition, we also evaluated the performance of the sex-specific methylation-based risk prediction models for AD diagnosis using an independent external dataset. Results: In the sex-stratified analysis, we identified 2 CpGs, mapped to the PRRC2A and RPS8 genes, significantly associated with AD in females at a 5% false discovery rate, and an additional 25 significant CpGs (21 in females, 4 in males) at P- value < 1×10 −5 . In methylation-by-sex interaction analysis, we identified 5 significant CpGs at P- value < 10 −5 . Out-of-sample validations using the AddNeuroMed dataset showed in females, the best logistic prediction model included age, estimated immune cell-type proportions, and methylation risk scores (MRS) computed from 9 of the 23 CpGs identified in AD vs. CN analysis that are also available in AddNeuroMed dataset (AUC = 0.74, 95% CI: 0.65–0.83). In males, the best logistic prediction model included only age and MRS computed from 2 of the 5 CpGs identified in methylation-by-sex interaction analysis that are also available in the AddNeuroMed dataset (AUC = 0.70, 95% CI: 0.56–0.82). Conclusions: Overall, our results show that the DNA methylation differences in AD are largely distinct between males and females. Our best-performing sex-specific methylation-based prediction model in females performed better than that for males and additionally included estimated cell-type proportions. The significant discriminatory classification of AD samples with our methylation-based prediction models demonstrates that sex-specific DNA methylation could be a predictive biomarker for AD. As sex is a strong factor underlying phenotypic variability in AD, the results of our study are particularly relevant for a better understanding of the epigenetic architecture that underlie AD and for promoting precision medicine in AD.

AI snapshot

Written from the paper’s full text.

TL;DR

This study performed sex-specific blood DNA methylation meta-analyses in ADNI and AIBL and found largely distinct AD-associated methylation differences in females versus males. Sex-specific methylation risk scores validated in AddNeuroMed predicted AD better in females (AUC = 0.74) than males (AUC = 0.70).

Key findings

  • In sex-stratified meta-analysis of 1284 blood samples (632 females, 652 males), 2 CpGs reached 5% FDR in females, mapped to PRRC2A and RPS8; no CpGs reached 5% FDR in males. At P < 1×10−5, an additional 25 CpGs were significant (21 in females, 4 in males).
  • Methylation-by-sex interaction analysis identified 5 CpGs at P < 1×10−5, mapped to MYO19, ESRRB, APLNR, and intergenic regions; there was no overlap with sex-stratified significant CpGs.
  • Region-based analysis identified 41 DMRs in females and 24 DMRs in males at 5% Sidak-corrected P-value, with no overlap between female and male significant CpGs or DMRs.
  • Out-of-sample validation in AddNeuroMed (107 females, 64 males) showed the best female model included age, immune cell-type proportions, and MRS from 9 CpGs (AUC = 0.74, 95% CI: 0.65–0.83); the best male model included age and MRS from 2 interaction CpGs (AUC = 0.70, 95% CI: 0.56–0.82).

Methods

  • Sex-specific meta-analyses of two blood EWAS, ADNI and AIBL, using Illumina EPIC arrays; external validation used AddNeuroMed (Illumina 450k).
  • Within each dataset, logistic regression or logistic mixed-effects models tested methylation–AD associations separately by sex and methylation-by-sex interaction, adjusting for age, methylation plate/batch, and estimated immune cell-type proportions; bacon correction addressed inflation.
  • Inverse-variance fixed-effects meta-analysis combined datasets; CpGs at FDR < 5% were significant, with P < 1×10−5 used as a suggestive threshold. Methylation risk scores were trained in AIBL and tested in AddNeuroMed.

Limitations

  • Stated: bulk whole blood contains mixed cell types, potentially reducing power; the study used a liberal P < 1×10−5 threshold; comb-p DMRs may have inflated false-positive rates; MCI subjects were excluded; testing on 450k arrays limited MRS CpGs compared with EPIC training data; education was unavailable in AIBL and AddNeuroMed; and methylation may reflect both cause and consequence of AD.
  • Stated: the validation dataset had more female than male samples, which may partly explain better female prediction performance.
  • Evident from the text: the sex-specific meta-analysis was described as underpowered given modest sample sizes, and no CpGs reached 5% FDR in males or in cross-tissue meta-analysis.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1186/s13195-022-01070-z

Cites

99 of 99 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

57 of 61 cited sources checked · 33 supported · 3 flagged · accuracy not yet measured

  • Study design and conductown data

    The sex-specific meta-analysis included 1284 whole blood samples (632 females and 652 males) from ADNI and AIBL.

    “included DNA methylation data for a total of 1284 whole blood samples (632 females and 652 males).”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [38] Vasanthakumar A, Davis JW, Idler K, Waring JF, Asque E, Riley-Gillis B, et al. Harnessing peripheral DNA methylation differences in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) to reveal novel biomarkers of disease. Clin Epigenetics. 2020;12:84. source

      Methods > Subjects reports samples from 653 ADNI participants, and Methods > Sample identity checks says 1,707 samples were analyzed after quality control—not 1,284. Methods > Statistical analysis also distinguishes whole blood from buffy coat. The cited paper does not describe the claimed sex-specific ADNI–AIBL meta-analysis or its female–male counts. Page numbers are unavailable in our copy. (LLM, conf=0.88)

      Checked by a reasoning model — may be wrong.

      Experimental — accuracy not yet measured
    • cites [24] Nabais MF, Laws SM, Lin T, Vallerga CL, Armstrong NJ, Blair IP, et al. Meta-analysis of genome-wide DNA methylation identifies shared associations across neurodegenerative disorders. Genome Biol. 2021;22:90. source

      The Results (Study design; page unavailable) identify AIBL and ADNI as cohorts with whole-blood DNA methylation data, and the Methods (Cohorts description; page unavailable) describe their methylation samples. Our copy does not report a sex-specific meta-analysis of 1,284 samples or the stated breakdown of 632 females and 652 males (searched those numbers and the analysis description). (LLM, conf=0.83)

      Experimental — accuracy not yet measured
  • Findingown data

    The diminishing B cells in AD is more pronounced in females.

    “our findings revealed that the diminishing B cells in AD is more pronounced in females”
    Quote in the paperSubstantial claim3 of 4 sources checked
    • cites [68] Busse M, Michler E, von Hoff F, Dobrowolny H, Hartig R, Frodl T, et al. Alterations in the peripheral immune system in dementia. J Alzheimers Dis. 2017;58:1303–13. source
    • cites [69] Xiong LL, Xue LL, Du RL, Niu RZ, Chen L, Chen J, et al. Single-cell RNA sequencing reveals B cell-related molecular biomarkers for Alzheimer’s disease. Exp Mol Med. 2021;53:1888–901. source

      The Results section (‘Differential gene expression profiles in B cells’; page unavailable) reports fewer B cells in AD patients than in controls. It does not report a female–male comparison, so the claim that the decrease is more pronounced in females is not supported by this source. The citing sentence describes that comparison as the citing authors’ own finding. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites [70] Pellicano M, Bulati M, Buffa S, Barbagallo M, Di Prima A, Misiano G, et al. Systemic immune responses in Alzheimer’s disease: in vitro mononuclear cell activation and cytokine production. J Alzheimers Dis. 2010;21:181–92. source

      The Lymphocyte subsets section reports fewer circulating B cells in AD patients than in healthy controls, but not a larger decrease in females. The PBMC activation section states that the results were not affected by gender, contrary to the claimed female difference. Page numbers are unavailable in our copy. (LLM, conf=0.86)

      Checked by a reasoning model — may be wrong.

      Experimental — accuracy not yet measured
    • cites [71] Paranjpe MD, Belonwu S, Wang JK, Oskotsky T, Gupta A, Taubes A, et al. Sex-Specific cross tissue meta-analysis identifies immune dysregulation in women with Alzheimer’s disease. Front Aging Neurosci. 2021;13:735611. source

      The Results section on cell-type deconvolution reports decreased memory B cells in females with AD and no significant immune-cell proportion differences in males. But it also reports increased naïve B cells in females with AD, so the source does not support a general decrease in B cells. (LLM, conf=0.87)

      Experimental — accuracy not yet measured
  • Premiseown data

    A prior meta-analysis of more than 1000 prefrontal cortex brain samples identified sex-specific AD neuropathology-associated DNA methylation differences.

    “we recently meta-analyzed more than 1000 prefrontal cortex brain samples and identified a number of sex-specific AD neuropathology-associated DNAm differences”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [36] Zhang L, Young JI, Gomez L, Silva TC, Schmidt MA, Cai J, et al. Sex-specific DNA methylation differences in Alzheimer’s disease pathology. Acta Neuropathol Commun. 2021;9:77. source
  • Premisecitation

    Those sex-specific AD neuropathology-associated DNA methylation differences were enriched in integrin activation in females and complement activation in males.

    “which were enriched in divergent biological processes such as integrin activation in females and complement activation in males [36].”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [36] Zhang L, Young JI, Gomez L, Silva TC, Schmidt MA, Cai J, et al. Sex-specific DNA methylation differences in Alzheimer’s disease pathology. Acta Neuropathol Commun. 2021;9:77. source
  • Premisecitation

    Sex is a strong factor underlying phenotypic variability in AD.

    “As sex is a strong factor underlying phenotypic variability in AD”
    • Also stated in: § Introduction · plain
      “Sex has increasingly been recognized as an important factor contributing to the phenotypic heterogeneity in AD [4–8].”
    Quote in the paperSubstantial claim3 of 5 sources checked
    • cites [4] Ferretti MT, Iulita MF, Cavedo E, Chiesa PA, Schumacher Dimech A, Santuccione Chadha A, et al. Sex differences in Alzheimer disease - the gateway to precision medicine. Nat Rev Neurol. 2018;14:457–69. source
    • cites [5] Mazure CM, Swendsen J. Sex differences in Alzheimer’s disease and other dementias. Lancet Neurol. 2016;15:451–2. source
    • cites [6] Nebel RA, Aggarwal NT, Barnes LL, Gallagher A, Goldstein JM, Kantarci K, et al. Understanding the impact of sex and gender in Alzheimer’s disease: A call to action. Alzheimers Dement. 2018;14:1171–83. source
    • cites [7] Hampel H, Vergallo A, Giorgi FS, Kim SH, Depypere H, Graziani M, et al. Alzheimer Precision Medicine I: Precision medicine and drug development in Alzheimer’s disease: the importance of sexual dimorphism and patient stratification. Front Neuroendocrinol. 2018;50:31–51. source
    • cites [8] Lin KA, Doraiswamy PM. When Mars versus Venus is not a cliche: gender differences in the neurobiology of Alzheimer’s disease. Front Neurol. 2014;5:288. source
  • Classified as probably not a substantial claim. 21
  • May combine two claims. 19

  • Classified as probably not a substantial claim.

    Study design and conductown data

    The sex-specific methylation-based prediction models were evaluated for AD diagnosis using an independent external dataset.

    “In addition, we also evaluated the performance of the sex-specific methylation-based risk prediction models for AD diagnosis using an independent external dataset.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [28] Roubroeks JAY, Smith AR, Smith RG, Pishva E, Ibrahim Z, Sattlecker M, et al. An epigenome-wide association study of Alzheimer’s disease blood highlights robust DNA hypermethylation in the HOXB6 gene. Neurobiol Aging. 2020;95:26–45. source

      The Discussion (page unavailable) explicitly says the study did not replicate its findings in an independent cohort; its HOXB6 validation used the same samples. The Methods describe adjusting for sex in methylation analyses, not evaluating sex-specific prediction models in an external dataset. (LLM, conf=0.88)

      Checked by a reasoning model — may be wrong.

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    The external validation dataset AddNeuroMed included 171 whole blood samples (107 females and 64 males).

    “The external validation samples included 171 whole blood samples (107 females and 64 males) from the AddNeuroMed study [28] (GEO accession: GSE144858).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [28] Roubroeks JAY, Smith AR, Smith RG, Pishva E, Ibrahim Z, Sattlecker M, et al. An epigenome-wide association study of Alzheimer’s disease blood highlights robust DNA hypermethylation in the HOXB6 gene. Neurobiol Aging. 2020;95:26–45. source

      The Methods (Subjects) identify blood samples from AddNeuroMed, and Methods (DNA methylation analysis) gives the cited GEO accession, GSE144858. The paper describes 284 samples and does not identify an external-validation subset of 171 with 107 females and 64 males; those figures are not supported by our copy. (LLM, conf=0.86)

      Experimental — accuracy not yet measured

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductcitation

    The meta-analysis was expected to be underpowered given modest sex-specific sample sizes, so CpGs at P < 1×10−5 were also prioritized.

    “Based on our experiences and previous studies in the analysis of EWAS measured in blood [22, 38, 47], we expected our meta-analysis to be underpowered, given the modest sample sizes of the sex-specific analyses. Therefore, we also prioritized CpGs with suggestive significance at the pre-specified significance threshold of P-value < 1×10−5.”
    Quote in the paperSubstantial claim3 of 3 sources checked
    • cites [22] Silva TC, Young JI, Zhang L, Gomez L, Schmidt MA, Varma A, et al. Cross-tissue meta-analysis of blood and brain epigenome-wide association studies in Alzheimer’s disease. Nat Commun. 2022; In Press. source

      The Discussion reports that only five CpGs reached 5% FDR in the paper’s blood-sample meta-analysis, so the authors examined 45 more at P < 10⁻⁵; it also discusses ways to improve the power of blood-based methylation studies. The paper does not attribute limited power to modest sex-specific sample sizes. In the citing sentence, that sex-specific rationale and the decision to prioritize CpGs describe the citing authors’ own analysis, rather than a finding of this source. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites [38] Vasanthakumar A, Davis JW, Idler K, Waring JF, Asque E, Riley-Gillis B, et al. Harnessing peripheral DNA methylation differences in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) to reveal novel biomarkers of disease. Clin Epigenetics. 2020;12:84. source
    • cites [47] Starnawska A, Tan Q, McGue M, Mors O, Borglum AD, Christensen K, et al. Epigenome-wide association study of cognitive functioning in middle-aged monozygotic twins. Front Aging Neurosci. 2017;9:413. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Alterations in DNA methylation levels are involved in many diseases, including AD.

    “Alterations of DNA methylation levels have been shown to be involved in many diseases, including AD [19–24].”
    Quote in the paperNot a substantial claim6 of 6 sources checked
    • cites [19] De Jager PL, Srivastava G, Lunnon K, Burgess J, Schalkwyk LC, Yu L, et al. Alzheimer’s disease: early alterations in brain DNA methylation at ANK1, BIN1, RHBDF2 and other loci. Nat Neurosci. 2014;17:1156–63. source

      The Discussion reports replicated associations between altered DNA methylation and AD pathology, supporting the AD portion of the claim. It says the changes suggest a role in AD onset but explicitly cautions that the study cannot establish whether they contribute to the disease or are an early consequence of it. Thus, 'shown to be involved' is stronger than the source establishes. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
    • cites [20] Lunnon K, Smith R, Hannon E, De Jager PL, Srivastava G, Volta M, et al. Methylomic profiling implicates cortical deregulation of ANK1 in Alzheimer’s disease. Nat Neurosci. 2014;17:1164–70. source
    • cites [21] Zhang L, Silva TC, Young JI, Gomez L, Schmidt MA, Hamilton-Nelson KL, et al. Epigenome-wide meta-analysis of DNA methylation differences in prefrontal cortex implicates the immune processes in Alzheimer’s disease. Nat Commun. 2020;11:6114. source

      The Results report DNA methylation differences in people with pathological AD, and the Discussion describes their potential roles in AD. The paper does not establish the broader claim about many diseases; its findings concern AD. It also reports associations rather than demonstrating that the methylation changes cause disease. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites [22] Silva TC, Young JI, Zhang L, Gomez L, Schmidt MA, Varma A, et al. Cross-tissue meta-analysis of blood and brain epigenome-wide association studies in Alzheimer’s disease. Nat Commun. 2022; In Press. source

      The introduction says DNA methylation plays an important role in AD, and the Discussion reports methylation differences between AD subjects and cognitively normal subjects. The paper does not establish the broader claim that methylation alterations are involved in many diseases. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites [23] Smith RG, Hannon E, De Jager PL, Chibnik L, Lott SJ, Condliffe D, et al. Elevated DNA methylation across a 48-kb region spanning the HOXA gene cluster is associated with Alzheimer’s disease neuropathology. Alzheimers Dement. 2018;14:1580–8. source

      The Introduction and Discussion (page numbers unavailable) report altered DNA methylation associated with AD neuropathology. The paper also mentions differential methylation in Down syndrome, but it does not establish the broader claim that methylation alterations are involved in many diseases; its AD finding is an association, not proof of causation. (LLM, conf=0.83)

      Experimental — accuracy not yet measured
    • cites [24] Nabais MF, Laws SM, Lin T, Vallerga CL, Armstrong NJ, Blair IP, et al. Meta-analysis of genome-wide DNA methylation identifies shared associations across neurodegenerative disorders. Genome Biol. 2021;22:90. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    DNA methylation is susceptible to environmental factors that modify disease risk.

    “and it is susceptible to environmental factors that modify the risk of diseases [18].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [18] Kubota T, Miyake K, Hirasawa T. Epigenetic understanding of gene-environment interactions in psychiatric disorders: a new concept of clinical genetics. Clin Epigenetics. 2012;4:1. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    DNA methylation regulates gene expression without altering DNA sequence.

    “DNA methylation (DNAm) is an epigenetic mechanism that regulates gene expression without altering DNA sequence”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [18] Kubota T, Miyake K, Hirasawa T. Epigenetic understanding of gene-environment interactions in psychiatric disorders: a new concept of clinical genetics. Clin Epigenetics. 2012;4:1. source

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    Bacon correction was applied to obtain corrected effect sizes, standard errors, and P-values for each cohort.

    “genomic correction using the bacon method [45], as implemented in the bacon R package, was applied to obtain bacon-corrected effect sizes, standard errors, and P-values for each cohort.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [45] van Iterson M, van Zwet EW, Consortium B, Heijmans BT. Controlling bias and inflation in epigenome- and transcriptome-wide association studies using the empirical null distribution. Genome Biol. 2017;18:19. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Dysregulated DNA methylation is an early feature of AD neuropathology in the brain.

    “dysregulated DNAm has been observed as an early feature of AD neuropathology in the brain [19]”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [19] De Jager PL, Srivastava G, Lunnon K, Burgess J, Schalkwyk LC, Yu L, et al. Alzheimer’s disease: early alterations in brain DNA methylation at ANK1, BIN1, RHBDF2 and other loci. Nat Neurosci. 2014;17:1156–63. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    These results are consistent with the previous observation that about 45% of DNA methylation sites targeted by the Illumina 450k array are influenced by genetic variants in the blood.

    “These results are consistent with the previous observation that a substantial proportion (about 45%) of the DNA methylation sites targeted by the Illumina 450k array are influenced by genetic variants in the blood [58].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [58] Min JL, Hemani G, Hannon E, Dekkers KF, Castillo-Fernandez J, Luijk R, et al. Genomic and phenotypic insights from an atlas of genetic effects on DNA methylation. Nat Genet. 2021;53:1311–21. source

  • May combine two claims.

    Interpretationcitation

    The limited commonality between genetic and epigenetic loci in AD could be due to low power in EWAS and/or GWAS but could also reflect relatively independent roles of genetic variants and DNA methylation in influencing AD susceptibility.

    “The limited commonality between genetic and epigenetic loci in AD could be due to the low power in EWAS and/or GWAS but could also reflect the relatively independent roles of genetic variants and DNA methylation in influencing AD susceptibility [65, 66].”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [65] Klein HU, Bennett DA, De Jager PL. The epigenome in Alzheimer’s disease: current state and approaches for a new path to gene discovery and understanding disease mechanism. Acta Neuropathol. 2016;132:503–14. source

      In “First generation of epigenomic studies in AD” (page unavailable), the review says validated AD epigenomic associations do not appear to be driven by genetic variation and describes genetic and methylation associations at BIN1 as independently influencing risk. It discusses sample size and statistical power, but does not identify low power in EWAS or GWAS as an explanation for limited overlap between loci. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites [66] Chibnik LB, Yu L, Eaton ML, Srivastava G, Schneider JA, Kellis M, et al. Alzheimer’s loci: epigenetic associations and interaction with genetic factors. Ann Clin Transl Neurol. 2015;2:636–47. source

      The Introduction and Abstract Interpretation say that genetic variation and DNA methylation may have largely independent roles in AD pathology and susceptibility. The Introduction also says stringent multiple-comparison criteria limited an earlier methylome-wide analysis. However, the paper does not establish the claimed limited commonality between genetic and epigenetic loci or specifically attribute it to low power in EWAS and/or GWAS. No page numbers are available in our copy. (LLM, conf=0.85)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    The comb-p software may have an inflated false-positive rate, so the resulting DMRs need to be interpreted cautiously.

    “However, because the comb-p software may have an inflated false-positive rate [95], the resulting DMRs need to be interpreted cautiously.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [95] Suderman M, Staley JR, French R, Arathimos R, Simpkin A, Tilling K. dmrff: identifying differentially methylated regions efficiently with power and control. bioRxiv. 2018. https://doi.org/10.1101/508556. source

  • May combine two claims.

    Premisecitation

    The second most significant CpG is located on the RPS8 gene, which encodes a ribosomal protein and was recently found to be significantly down-regulated in blood samples of AD patients.

    “The second most significant CpG is located on the RPS8 gene, which encodes a ribosomal protein and was recently found to be significantly down-regulated in blood samples of AD patients [77].”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [77] Shigemizu D, Mori T, Akiyama S, Higaki S, Watanabe H, Sakurai T, et al. Identification of potential blood biomarkers for early diagnosis of Alzheimer’s disease through RNA sequencing analysis. Alzheimers Res Ther. 2020;12:87. source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation

    This result is consistent with the previous observation that deficiency in PRRC2A reduces the oligodendroglia population in the brain and induces hypomyelination, which leads to impaired locomotive and cognitive functions.

    “This result is consistent with the previous observation that deficiency in PRRC2A reduces the oligodendroglia population in the brain and induces hypomyelination, which leads to an impaired locomotive and cognitive functions [74–76].”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites [74] Wu R, Li A, Sun B, Sun JG, Zhang J, Zhang T, et al. A novel m(6)A reader Prrc2a controls oligodendroglial specification and myelination. Cell Res. 2019;29:23–41. source
    • cites [75] Chen X, Wang F, Gan J, Zhang Z, Liang X, Li T, et al. Myelin deficits caused by Olig2 deficiency lead to cognitive dysfunction and increase vulnerability to social withdrawal in adult mice. Neurosci Bull. 2020;36:419–26. source

      The Discussion and Results report that Olig2 deletion in oligodendroglial cells caused hypomyelination and cognitive deficits in young adult mice. They do not establish those effects for PRRC2A deficiency or a reduced oligodendroglia population. The Results also report normal locomotor activity in the Olig2 knockout mice, rather than impaired locomotion (page numbers unavailable). (LLM, conf=0.90)

      Experimental — accuracy not yet measured
    • cites [76] Sehmbi M, Rowley CD, Minuzzi L, Kapczinski F, Steiner M, Sassi RB, et al. Association of intracortical myelin and cognitive function in bipolar I disorder. Acta Psychiatr Scand. 2018;138:62–72. source

  • May combine two claims.

    Findingown data

    Before bacon correction, P-values from ADNI logistic mixed-effects models were overly conservative and P-values from AIBL logistic regression models were overly liberal.

    “These estimated inflation factors showed that the P-values of the logistic mixed effects models based on analytical formula in the analysis of ADNI dataset is overly conservative, while P-values of the logistic regression models in the analysis of AIBL dataset is overly liberal.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The best-performing prediction model for females included age, immune cell type proportions, and MRS, while the best-performing prediction model for males included only age and MRS.

    “while the best-performing prediction model for females included age, immune cell type proportions, and MRS, the best-performing prediction model for males included only age and MRS.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The most significant CpG in females is located near the promoter region of the PRRC2A gene, with significant hypermethylation in AD subjects.

    “In sex-stratified analysis, the most significant CpG in females is located near the promoter region of the PRRC2A gene, with significant hypermethylation in AD subjects (Table 2).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In males, the most predictive model included MRS and age (AUC = 0.70, 95% CI: 0.56-0.82), significantly more predictive than a random classifier (P-value = 5.62×10−3).

    “In males, the most predictive model include MRS and age (AUC = 0.70, 95% CI: 0.56–0.82), significantly more predictive than a random classifier (P-value = 5.62×10−3).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In females, the most predictive model included MRS, age, and estimated immune cell-type proportions (AUC = 0.74, 95% CI: 0.65-0.83), significantly more predictive than a random classifier (P-value = 8.42×10−6).

    “In females, the most predictive model include MRS, age, and estimated immune cell-type proportions (AUC = 0.74, 95% CI: 0.65–0.83), significantly more predictive than a random classifier (P-value = 8.42×10−6).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The prediction performance of the best-performing sex-specific logistic regression models in males and females remained very similar when the ADNI dataset was additionally included as a training dataset, or when CpGs from AD-associated DMRs and/or significant CpGs in cross-tissue analyses were included in the computation of MRS.

    “The results indicated the prediction performance of these models in males and females remained very similar when the ADNI dataset was additionally included as a training dataset in the development of the logistic regression models, or when CpGs from AD-associated DMRs and/or significant CpGs in cross-tissue analyses are also included in the computation of MRS”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    All 27 sex-specific CpGs remained highly significant after additionally adjusting for smoking score, with meta-analysis P-values ranging from 2.59 × 10−5 to 5.83 × 10−8.

    “The results of our expanded logistic regression model that additionally included smoking score showed all 27 sex-specific CpGs (Supplementary Table 2) remained highly significant, with meta-analysis P-values ranging from 2.59 × 10−5 to 5.83 × 10−8 (Supplementary Table 15)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    An additional 21 CpGs in females and 4 CpGs in males reached P-value < 1×10−5.

    “An additional 21 CpGs in females and 4 CpGs in males reached P-value < 1×10−5 (Supplementary Table 2).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    A logistic regression model on all three datasets showed a significant B cell type proportion × sex interaction (P-value = 0.017), indicating associations between B-cell type proportions and AD were significantly different between males and females.

    “The results showed a significant B cell type proportion × sex interaction (P-value = 0.017), indicating the associations between B-cell type proportions and AD were significantly different between males and females.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    After bacon correction, estimated bias was near zero and inflation factors were approximately one in both sexes and cohorts.

    “Indeed, after bacon correction, for males, the estimated bias is − 2.47×10−4 and − 8.64×10−6, and the estimated inflation factors were λ = 1.03 and 1.05, and λ.bacon = 1.00 and 1.00, for the ADNI and AIBL datasets, respectively.”
    • Also stated in: § Inflation assessment and correction · plain
      “For females, the estimated bias is 2.85 × 10−3 and 4.75 × 10−5, and the estimated inflation factors were λ = 0.98 and 1.05, and λ.bacon = 1.00 and 1.00 for the ADNI and AIBL cohorts, respectively.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    The sex-specific meta-analyses adjusted for age, estimated immune cell-type proportions, batch effects, and genomic inflation.

    “After adjusting for age, estimated immune cell type proportions, batch effects, and correcting for inflation, the inverse-variance fixed-effects meta-analysis model was used to identify the most consistent DNAm differences across datasets.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Estimated sex status from DNA methylation agreed with recorded sex for all samples.

    “We also estimated sex status using the estimateSex() function in wateRmelon R package, which agreed with the recorded sex for all the samples.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    The interaction analysis identifies CpGs with large differences in sex-specific effect estimates that are in different directions, but these effects might not have reached the P < 1×10−5 significance threshold in sex-stratified analysis.

    “the interaction analysis identifies CpGs with large differences in sex-specific effect estimates that are in different directions, but these effects might not have reached the P < 1×10−5 significance threshold in sex-stratified analysis.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The greater number of differentially expressed genes in brain samples compared to blood samples could be due to the larger sample size of brain samples available (502 brain samples vs. 265 ADNI blood samples).

    “The greater number of differentially expressed genes in brain samples compared to blood samples could be due to the larger sample size of brain samples available (502 brain samples in the meta-analysis of GSE33000 and GSE44772 vs. 265 ADNI blood samples).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    MCI subjects were not considered in this study because of considerable heterogeneity among MCI subjects, with subjects converting to AD at different trajectories.

    “Third, we did not consider MCI subjects in this study because there is considerable heterogeneity among MCI subjects, with subjects converting to AD at different trajectories.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: There is considerable heterogeneity among MCI subjects, with subjects converting to AD at different trajectories.

  • May combine two claims.

    Interpretationown data

    In addition to implicating sex-specificity for methylation differences in genes previously known in AD (e.g., PM20D1), the authors also nominated additional differentially methylated genes that might be associated with AD (e.g., PRRC2A and MCCC1).

    “In addition to implicating sex-specificity for methylation differences in genes previously known in AD (e.g., PM20D1), we also nominated additional differentially methylated genes that might be associated with AD (e.g., PRRC2A and MCCC1).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    The use of bulk whole blood samples may have introduced substantial variability and reduced the power of the study.

    “First, we analyzed methylation data generated from bulk whole blood samples, which contain a complex mixture of cell types, and might have introduced substantial variability in the samples and reduced the power of our study.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationown data

    The higher prediction accuracy of immune cell type proportions in females might be due to a greater change in AD-associated B cell type proportions in females.

    “which might be due to a greater change in AD-associated B cell type proportions in females (Supplementary Fig. 6).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationreasoning

    Taken together, these results demonstrated that the sex-specific meta-analyses are consistent with recent literature in brain research.

    “Taken together, these results demonstrated our sex-specific meta-analyses are consistent with recent literature in brain research.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    A more liberal significance threshold (P-value < 10−5) was pre-defined for the meta-analysis to select a small number of loci, which were then further prioritized using integrative analyses.

    “Second, based on our previous experiences with the analysis of blood samples in AD, we pre-defined a more liberal significance threshold (i.e., P-value < 10−5) for our meta-analysis to select a small number of loci, which were then further prioritized using integrative analyses.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    Estimated cell-type proportions were included as covariate variables in all analyses to reduce confounding effects due to different cell types.

    “To reduce confounding effects due to different cell types, we included estimated cell-type proportions as covariate variables in all our analyses.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    The methylation-based prediction model could be further improved because DNA methylation samples in the testing dataset (AddNeuroMed) were measured by 450k arrays, different from the EPIC arrays used by AIBL, so only the subset of significant CpGs that mapped to both array types was included in the computation of MRS.

    “Fourth, the methylation-based prediction model could be further improved. Because DNA methylation samples in the testing dataset (AddNeuroMed) were measured by 450k arrays, which are different from the EPIC arrays used by the AIBL study, we only included the subset of significant CpGs that mapped to both types of arrays in the computation of MRS.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    The temporal relationship between DNA methylation sampling and AD diagnosis is uncertain in all three publicly available datasets, so the identified associations do not necessarily reflect causal relationships.

    “Finally, we cannot be certain that the DNA methylation samples were obtained prior to AD diagnosis in all three publicly available datasets; therefore, the DNA methylation may represent both cause and consequence of AD, so the associations we identified do not necessarily reflect causal relationships.”
    Quote in the paperNot a substantial claim
36 of 36

  • The paper relies on this without support.

    Sex-specific DNA methylation differences in blood samples of AD subjects are not well characterized.

    “However, currently, sex-specific DNAm differences in the blood samples of AD subjects, which can be accessed relatively easily in living individuals, are still not well characterized”

  • The paper relies on this without support.

    The feasibility of using sex-specific blood DNA methylation differences as biomarkers for AD remains undetermined.

    “and the feasibility of their use as biomarkers for AD remains to be determined.”

  • The paper relies on this without support.

    Blood samples can be accessed relatively easily in living individuals.

    “which can be accessed relatively easily in living individuals”

  • The paper relies on this without support.

    There is considerable heterogeneity among MCI subjects, with subjects converting to AD at different trajectories.

    “because there is considerable heterogeneity among MCI subjects, with subjects converting to AD at different trajectories.”

  • The paper relies on this without support.

    Advanced aging is the strongest factor for AD.

    “As advanced aging is the strongest factor for AD, we also compared our results with those from Yusipov et al. [94] and McCartney et al. [35]”
5 of 5

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Conflicting directions of female cross-tissue methylation changes

    Major

    Table 5 says all but 5 of the 25 significant female CpGs changed in the same direction in brain and blood. The Results instead say 20 of the 25 changed in opposite directions, reversing the conclusion about agreement between tissues.

  • Cross-tissue meta-analysis credited with 31 rather than 19 CpGs

    Minor

    The Results attribute 13 female and 6 male CpGs-19 total-to the cross-tissue meta-analysis. The Discussion credits that meta-analysis with 31 CpGs, which is the total prioritized by both cross-tissue approaches together.

2 of 2

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
C. Silva, T., Zhang, W., Young, J. I., Gomez, L., Schmidt, M. A., Varma, A., Chen, X. S., Martin, E. R., & Wang, L.
Affiliations
University of Miami (US)
Corresponding author
Wang, Lily
ORCID
Venue
Alzheimer's Research & Therapy
Publisher
Springer Science and Business Media LLC
Published
2022-09-15
Volume
14
Issue
1
Pages
133-133
ISSN
1758-9193 (ISSN-L)
DOI
10.1186/s13195-022-01070-z
PMC
PMC9479371
OpenAlex
W4296034468
Semantic Scholar
098b324d9b27eaa416ce02f67b292b35b50e7ca2
Cited by
33
References
98
Citations by year
2026: 6 · 2025: 11 · 2024: 9 · 2023: 6 · 2022: 1
FWCI
1.86
Type
journal article
Language
en
Open access
Licence
Locations
  • Alzheimer s Research & Therapy (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • Rush University
  • National Institutes of Health (1rf1ag061127-01, 5u01ag024904-10, R01 AG15819, R01 AG062634, R01AG30146, P30AG10161, R21AG060459, R01 AG17917)
  • National Institute on Aging (U01-AG-024904, P30-AG 10161, U01 AG32984, R01 AG30146, R01‐AG17917, R01 AG15819, R01-AG36836, R01AG036836, R01-AG030146, R01 AG017917, U01AG046152, R21AG060459, R01AG061127, U01AG032984, R01AG062634, RF1 AG061127, P30AG010161, R01-AG015819)
  • National Institute of Neurological Disorders and Stroke (RF1NS128145)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Epigenetics and DNA Methylation
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
  • Health Sciences › Medicine › Physiology › Alzheimer's disease research and treatments
Keywords
Alzheimer's disease, DNA methylation, sex differences, Alzheimer's disease diagnosis, methylation risk score, immune cell proportions, sex-stratified analysis, precision medicine, biomarkers, ADNI, AIBL
MeSH terms
Alzheimer Disease, Female, Humans, Male, Sex Characteristics, CpG Islands, DNA Methylation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-18.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-18.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review