Methylation Marks to Inform Association between Early-Life Air Pollution Exposures and Child Body Mass Index: An Analysis Based on A Priori Selected Pathways

摘要

Background: Early-life exposure to air pollution has been suggested to influence childhood growth and Body Mass Index (BMI). DNA methylation changes might explain part of this effect. Yet, methylome-wide studies with population sizes under 1000 generally suffer from lack of power. Using a priori information to reduce data dimension is an option to face this challenge. Aim: To identify genes whose methylation level is possibly implied in air pollution effect on child BMI, with a 'Meet in the middle' approach, using a priori information from functional pathways. Methods: Among 1173 mother-infant pairs from Helix cohorts, exposures to 4 outdoor air pollutants were assessed during pregnancy and at age 7-9 years (total 16 exposures). Genome-wide DNA methylation levels in peripheral blood at 7-9 years was measured using HumanMethylation450 BeadChip, with quality control, batch and cell types adjustments. We a priori considered only 2276 CpGs from genes belonging to 16 growth or obesity-relevant pathways according to KEGG pathway database. Stabilized Elasticnet regression (ENet) was used to independently assess associations of DNA methylation with exposures and BMI. We identified exposures sharing associated CpG sites with BMI and assessed their direct association with BMI using ENet. Results: Exposures were associated with 358 CpG loci; 12 exposures were associated with CpGs also associated with BMI (total, 22 CpGs). One of these exposures, PM10 level the day before blood collection, was associated with BMI. This exposure was associated with 62 CpGs, 2 of which being also associated with BMI, on PIK3CA (thyroid hormone signaling pathway) and ELOVL3 genes (fatty acid elongation pathway). Conclusion: Our hypothesis-generating approach identified 358 CpGs from BMI-relevant pathways associated with air pollutants, including 22 related to child BMI. These CpGs provide hints as to which genes may be implied in mechanisms underlying air pollution effects on child growth.