• Dylan M Williams, Ida K Karlsson, Nancy L Pedersen, and Sara Hägg.
• From the Department of Medical Epidemiology & Biostatistics (D.M.W., I.K.K., N.L.P., S.H.), Karolinska Institutet, Stockholm, Sweden; and Department of Psychology (N.L.P.), University of Southern California, Los Angeles. Dylan.williams@ki.se.
• Neurology. 2018 Jan 23; 90 (4): e291-e297.

ObjectiveTo examine whether genetically predicted variation in circulating insulin-like growth factor 1 (IGF1) or its binding protein, IGFBP3, are associated with risk of Alzheimer disease (AD), using a mendelian randomization study design.MethodsWe first examined disease risk by genotypes of 9 insulin-like growth factor (IGF)-related single nucleotide polymorphisms (SNPs) using published summary genome-wide association statistics from the International Genomics of Alzheimer's Project (IGAP; n = 17,008 cases; 37,154 controls). We then assessed whether any SNP-disease results replicated in an independent sample derived from the Swedish Twin Registry (n = 984 cases; 10,304 controls).ResultsMeta-analyses of SNP-AD results did not suggest that variation in IGF1, IGFBP3, or the molar ratio of these affect AD risk. Only one SNP appeared to affect AD risk in IGAP data. This variant is located in the gene FOXO3, implicated in human longevity. In a meta-analysis of both IGAP and secondary data, the odds ratio of AD per FOXO3 risk allele was 1.04 (95% confidence interval 1.01-1.08; p = 0.008).ConclusionsThese findings suggest that circulating IGF1 and IGFBP3 are not important determinants of AD risk. FOXO3 function may influence AD development via pathways that are independent of IGF signaling (i.e., pleiotropic actions).Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology.

19 keywords...

hide…