Air pollution exposure in pregnancy can cause molecular level alterations that might influence later disease susceptibility.
Objectives
We investigated DNA methylation (DNAm) and telomere length (TL) in the cord blood in relation to gestational PM10 exposure and explored potential gestational windows of susceptibility.
Methods
Cord blood epigenome-wide DNAm (N = 384) and TL (N = 500) were measured in children of the Italian birth cohort Piccolipiù, using the Infinium Methylation EPIC BeadChip and qPCR, respectively. PM10 daily exposure levels, based on maternal residential address, were estimated for different gestational periods using models based on satellite data. Epigenome-wide analysis to identify differentially methylated probes (DMPs) and regions (DMRs) was conducted, followed by a pathway analysis and replication analysis in an second Piccolipiù dataset. Distributed lag models (DLMs) using weekly exposures were used to study the association of PM10 exposure across pregnancy with telomere length, as well as with the DMPs that showed robust associations.
Results
Gestational PM10 exposure was associated with the DNA methylation of more than 250 unique DMPs, most of them identified in early gestation, and 1 DMR. Out of 151 DMPs available in the replication dataset, ten DMPs showed robust associations: eight were associated with exposure during early gestation and 2 with exposure during the whole pregnancy. These exposure windows were supported by the DLM analysis. The PM10 exposure between 15th and 20th gestational week seem to be associated with shorter telomeres at birth, while exposure between 24th and 29th was associated with longer telomeres.
Discussion
The early pregnancy period is a potential critical window during which PM10 exposure can influence cord blood DNA methylation. The results from the TL analysis were consistent with previous findings and merit further exploration in future studies. The study underlines the importance of considering gestational windows outside of the predefined trimesters that may not always overlap with biologically relevant windows of exposure.
