Abstract B45 Polycyclic aromatic hydrocarbons (PAHs) are widespread organic carcinogenic pollutants. When metabolized in vivo, PAHs form reactive diol epoxides that bind to DNA at guanine residues forming adducts. PAH exposure has also been associated with alterations in genomic cytosine methylation. Both PAH-DNA adducts and altered DNA methylation have been associated with increased cancer risk. We sought to explore the relationship between prenatal PAH exposure, global DNA methylation and PAH-DNA adducts in a New York City birth cohort. Cord blood was collected from deliveries of non-smoking, African Americans and Dominicans, ages 18-35, residing in N. Manhattan and S. Bronx. Personal air monitors measured PAH during their 8th month of pregnancy. We measured benzo[a]pyrene (B[a]P)-DNA adducts (which serve as a proxy for PAH-DNA adducts) in cord blood using high-performance liquid chromatography-fluorescence. We measured genomic DNA methylation in cord blood leukocytes using the MethylampTM global DNA methylation quantification kit. Demographic and epidemiologic risk factors were collected prospectively, including maternal age, ethnicity, parity, and prenatal tobacco smoke exposure, and child’s gender We found that prenatal PAH exposure was associated global DNA hypomethylation such that average methylation among those in the highest quartile of prenatal PAH exposure was 1.17 ng/100 ng total DNA as compared 1.83 ng/100 ng total DNA in the lowest quartile (p < 0.01). Conversely, those with detectable PAH-DNA adducts had higher levels of DNA methylation (1.43 ng/100 ng total DNA) as compared to those with non-detectable adducts (1.16 ng/100 ng total DNA) (p = 0.02). Using multiple variable linear regression, PAH exposure was independently associated with hypomethylation (average difference in methylation among those in the highest vs. lowest quartile of exposure = -0.42, 95% CI: -0.75, -0.09) and the presence of DNA adducts was associated with hypermethyaltion (average difference in methylation among those with detectable vs. non-detectable adducts = 0.28, 95% CI: 0.05, 0.49). Adjusting for potential confounders did not significantly alter these independent associations, indicating that the lowest level of global DNA methylation was found among those with non-detectable adducts and high PAH exposure, and the highest level of DNA methylation was found among those with detectable adducts and low PAH exposure. While the mechanism by which prenatal PAH exposure results in DNA adducts is well understood, the way exposure may impact methylation is not known. These data suggest that in the presence of DNA adducts, PAH exposure is associated with a reduction in methylation, but when adducts are non-detectable, the same PAH exposure is associated with an even larger methylation reduction. It is possible that the presence of adducts may be partially preventing PAH compounds (or metabolites) from accessing the DNA, leading to less of a change in methylation. It is also possible that the shift in global methylation from PAH exposure may prevent adduct formation. These speculations as well as the associations observed in this analysis require confirmation in other settings/populations. Because methylation changes are potentially reversible, it is possible that if these associations are confirmed, methylation may serve as a target for intervention. Citation Information: Cancer Prev Res 2008;1(7 Suppl):B45.