Finally, it is worth noting that we did not attempt to characterize the SNPs that comprised each GRS nor did we create biologically informed GRS by selecting variants that related to a specific neurotransmitter pathway (for example, dopamine variant profile). Given the high false discovery rates typically associated with ascribing a functional direction to such variants of purported biological importance (except, perhaps, the alcohol dehydrogenase variants) and our currently limited understanding of the etiology of AD, we employed a more conservative and agnostic approach of utilizing genome-wide data. Future studies that contrast such genome-wide PRS with biologically informed risk scores may be valuable in the construction of the architecture underlying the polygenicity associated with complex traits such as AO-AD. Nonetheless, our approach precludes any mechanistic interpretation of the polygenicity represented by each GRS.
