Skip to Content
You are currently on the new version of our website. Access the old version .
EpigenomesEpigenomes
  • Article
  • Open Access

16 November 2026

18 Pages

Set-Level Cross-Tissue and Longitudinal Transportability of Discovery-Defined DNA Methylation Regions from Social Anxiety Disorder: A Cross-Cohort Analysis

1
Psychiatry Research Center, Beijing Huilongguan Hospital, Capital Medical University, Beijing 100096, China
2
Beijing Huilongguan Hospital, Peking University HuiLongGuan Clinical Medical School, Beijing 100096, China
Epigenomes2026, 10(4), 74;https://doi.org/10.3390/epigenomes10040074 
(registering DOI)
This article belongs to the Special Issue DNA Methylation Markers in Health and Disease

Abstract

Background: Epigenome-wide association studies have reported candidate DNA methylation signals for anxiety disorders, but their stability across tissues and developmental periods remains uncertain. This study evaluated locus-specific and set-level transportability of adult whole-blood methylation signals associated with social anxiety disorder to longitudinal saliva measurements from a youth cohort characterized by a broader anxiety-related phenotype. Methods: The discovery cohort comprised 143 adults with social anxiety disorder or control status profiled using the Illumina Methylation EPIC array. The external cohort comprised 94 HumanMethylation450 saliva samples from 47 participants assessed in 2008 and 2013. Discovery analyses included CpG-wise and regional association testing, co-methylation analysis, and nested elastic-net modelling. A locked set of discovery differentially methylated regions (DMRs) was evaluated using wave-specific regional tests, directional concordance, covariate-adjusted permutation tests, cross-cohort effect correlations, matched genomic windows, and participant-blocked longitudinal models. Results: No individual CpG reached false-discovery-rate significance, whereas 130 DMRs met the harmonic-mean false-discovery-rate criterion. Ninety-seven DMRs contained at least two eligible saliva probes. In 2008, 61/97 regions were directionally concordant with discovery (62.9%; exact p = 0.0072; adjusted p = 0.014). Covariate-adjusted permutation tests, discovery–external effect correlations, and matched-window analyses also supported transportability of the DMR set. No corresponding set-level evidence was detected in 2013, and no individual external DMR survived multiple-testing correction at either wave. The locked 20-CpG classifier yielded a mean nested-cross-validation AUC of 0.596 and external AUCs of 0.735 and 0.455 in 2008 and 2013, respectively. Conclusions: Discovery-defined blood DNA methylation regions associated with adult social anxiety disorder showed set-level transportability to the earlier saliva assessment of a broader anxiety-related phenotype, without corrected locus-specific replication. This evidence was not sustained at follow-up, supporting assessment-specific ensemble transportability rather than a stable or clinically validated methylation biomarker.

Article Metrics

Citations

Article Access Statistics

Article metric data becomes available approximately 24 hours after publication online.