War Exposure and DNA Methylation in Syrian Refugee Children and Adolescents

Abstract

Importance: Exposure to war is associated with poor mental health outcomes. Adverse and traumatic experiences can lead to long-lasting DNA methylation changes, potentially mediating the link between adversity and mental health. To date, limited studies have investigated the impact of war on DNA methylation in children or adolescents, hampering our understanding of the biological impact of war exposure.

Objective: To identify salivary DNA methylation differences associated with war exposure in refugee children and adolescents.

Design, setting, and participants: This cohort study included Syrian refugee children and adolescents, and their primary caregiver were recruited from tented settlements in Lebanon. Data collection was carried out in 2 waves, 1 year apart, from October 2017 to January 2018 and October 2018 to January 2019. Children and their caregiver were interviewed, and children provided saliva samples for DNA extraction. Data analysis was conducted in 2022, 2023, and 2024.

Exposure: War exposure assessed by interviewing children and their caregiver using the War Events Questionnaire.

Main outcomes and measures: Salivary DNA methylation levels were assayed with the Infinium MethylationEPIC BeadChip (Illumina). Epigenetic aging acceleration was estimated using a set of preexisting epigenetic aging clocks. A literature search was conducted to identify previously reported DNA methylation correlates of childhood trauma.

Results: The study population included 1507 children and adolescents (mean [SD] age, 11.3 [2.4] years; age range, 6-19 years; 793 female [52.6%]). A total of 1449 children provided saliva samples for DNA extraction in year 1, and 872 children provided samples in year 2. Children who reported war events had a number of differentially methylated sites and regions. Enrichment analyses indicated an enrichment of gene sets associated with transmembrane transport, neurotransmission, and intracellular movement in genes that exhibited differential methylation. Sex-stratified analyses found a number of sex-specific DNA methylation differences associated with war exposure. Only 2 of 258 (0.8%) previously reported trauma-associated DNA methylation sites were associated with war exposure (B = -0.004; 95% CI, -0.005 to -0.003; Bonferroni P = .04 and B = -0.005; 95% CI, -0.006 to -0.004; Bonferroni P = .03). Any war exposure or bombardment was nominally associated with decreased epigenetic age using the Horvath multitissue clock (B = -0.39; 95% CI, -0.63 to -0.14; P = .007 and B = -0.42; 95% CI, -0.73 to -0.11; P = .002).

Conclusions and relevance: In this cohort of Syrian refugee children and adolescents, war exposure was associated with a small number of distinct differences in salivary DNA methylation.

Figure 1.. Manhattan Plots of the Epigenome-Wide Association Studies for Violence in the Home, Any War Exposure, Total War Exposure, and Other Forms of Violence

The orange dashed line indicates the Bonferroni-corrected P value threshold. Genome-wide significant differentially methylated probes are in orange and the only differentially methylated region is indicated by the blue vertical dotted line.

Figure 2.. Top 10 Gene Sets Associated With War Exposure

Each gene set is labeled with the description and war exposure measure with which it is associated. P values have been false discovery rate (FDR)–corrected. Bar length represents the number of genes within that gene set. mRNA indicates messenger RNA.

Figure 3.. Volcano Plot of the Regression B Values Against P Values for the Replication Analyses

Dashed lines represent the nominal (P = .05) and Bonferroni-corrected P value thresholds. Cytosine-guanine dinucleotide sites with P < .001 have been labeled.

Figure 4.. Results for the Association Between Epigenetic Age Acceleration and War Exposure

Dots represent regression coefficients and error bars represent 95% confidence intervals. Negative estimates indicate slower age acceleration and positive estimates indicate faster age acceleration. Where P < .05, the P value is given. PACE indicates DunedinPACE or Dunedin Pace of Aging Calculated From the Epigenome; PedBE, Pediatric-Buccal-Epigenetic.