Environment International (Jan 2025)

Association of exposure to second-hand smoke during childhood with blood DNA methylation

  • Marta Cosin-Tomas,
  • Thanh Hoang,
  • Cancan Qi,
  • Giulietta S. Monasso,
  • Ryan Langdon,
  • Simon Kebede Merid,
  • Lucinda Calas,
  • Paula de Prado-Bert,
  • Rebecca Richmond,
  • Vincent VW Jaddoe,
  • Liesbeth Duijts,
  • John Wright,
  • Isabella Annesi-Maesano,
  • Regina Grazuleviciene,
  • Marianna Karachaliou,
  • Gerard H. Koppelman,
  • Erik Melén,
  • Olena Gruzieva,
  • Martine Vrijheid,
  • Paul Yousefi,
  • Janine F. Felix,
  • Stephanie J. London,
  • Mariona Bustamante

Journal volume & issue
Vol. 195
p. 109204

Abstract

Read online

Introduction: By recent estimates, 40% of children worldwide are exposed to second-hand smoke (SHS), which has been associated with adverse health outcomes. While numerous studies have linked maternal smoking during pregnancy (MSDP) to widespread differences in child blood DNA methylation (DNAm), research specifically examining postnatal SHS exposure remains sparse. To address this gap, we conducted epigenome-wide meta-analyses to identify associations of postnatal SHS and child blood DNAm. Methods: Six cohorts from the Pregnancy And Childhood Epigenetics (PACE) Consortium (total N = 2,695), with SHS data and child blood DNAm (aged 7–9 years) measured with the Illumina 450K array were included in the meta-analysis. Linear regression models adjusted for covariates were fitted to examine the association between the number of household smokers in postnatal life (0, 1, 2+) and child blood DNAm. Sensitivity models without adjusting for MSDP and restricted to mothers who did not smoke during pregnancy were evaluated. Results: Our analysis revealed significant associations (False Discovery Rate < 0.05) between household postnatal SHS exposure and DNAm at 11 CpGs in exposed children. Nine CpGs were mapped to genes (MYO1G, FAM184B, CTDSPL2, LTBP3, PDE10A, and FIBCD1), while 2 CpGs were located in open sea regions. Notably, all except 2 CpGs (mapped to FIBCD1 and CTDSPL2) have previously been linked to either personal smoking habits or in utero exposure to smoking. The models restricted to non-smoking mothers provided similar results. Importantly, several of these CpGs and their associated genes are implicated in conditions exacerbated by or directly linked to SHS. Conclusions: Our findings highlight the potential biological effects of SHS on blood DNAm. These findings support further research on epigenetic factors mediating deleterious effects of SHS on child health and call for public health policies aimed at reducing exposure, particularly in environments where children are present.

Keywords