Meta-analysis of epigenome-wide association studies of cognitive abilities.
Marioni RE., McRae AF., Bressler J., Colicino E., Hannon E., Li S., Prada D., Smith JA., Trevisi L., Tsai P-C., Vojinovic D., Simino J., Levy D., Liu C., Mendelson M., Satizabal CL., Yang Q., Jhun MA., Kardia SLR., Zhao W., Bandinelli S., Ferrucci L., Hernandez DG., Singleton AB., Harris SE., Starr JM., Kiel DP., McLean RR., Just AC., Schwartz J., Spiro A., Vokonas P., Amin N., Ikram MA., Uitterlinden AG., van Meurs JBJ., Spector TD., Steves C., Baccarelli AA., Bell JT., van Duijn CM., Fornage M., Hsu Y-H., Mill J., Mosley TH., Seshadri S., Deary IJ.
Cognitive functions are important correlates of health outcomes across the life-course. Individual differences in cognitive functions are partly heritable. Epigenetic modifications, such as DNA methylation, are susceptible to both genetic and environmental factors and may provide insights into individual differences in cognitive functions. Epigenome-wide meta-analyses for blood-based DNA methylation levels at ~420,000 CpG sites were performed for seven measures of cognitive functioning using data from 11 cohorts. CpGs that passed a Bonferroni correction, adjusting for the number of CpGs and cognitive tests, were assessed for: longitudinal change; being under genetic control (methylation QTLs); and associations with brain health (structural MRI), brain methylation and Alzheimer's disease pathology. Across the seven measures of cognitive functioning (meta-analysis n range: 2557-6809), there were epigenome-wide significant (P -8) associations for global cognitive function (cg21450381, P = 1.6 × 10-8), and phonemic verbal fluency (cg12507869, P = 2.5 × 10-9). The CpGs are located in an intergenic region on chromosome 12 and the INPP5A gene on chromosome 10, respectively. Both probes have moderate correlations (~0.4) with brain methylation in Brodmann area 20 (ventral temporal cortex). Neither probe showed evidence of longitudinal change in late-life or associations with white matter brain MRI measures in one cohort with these data. A methylation QTL analysis suggested that rs113565688 was a cis methylation QTL for cg12507869 (P = 5 × 10-5 and 4 × 10-13 in two lookup cohorts). We demonstrate a link between blood-based DNA methylation and measures of phonemic verbal fluency and global cognitive ability. Further research is warranted to understand the mechanisms linking genomic regulatory changes with cognitive function to health and disease.