Rs16923760, which associates with AD risk in a Caribbean-Hispanic series (OR = 0.74) [6] , is in LD with rs4746648 (D’ = 1.0, r2 = 0.01) that has nominally significant AD risk association in our combined cohort (p = 0.02, OR = 0.86).
Though they did not achieve significance in this replication cohort (p = 0.55), when Cohorts 1 and 2 were jointly assessed (3,166 LOAD and 3,261 controls), the global MLG association was significant (p = 0.036), with three MLGs achieving nominal significance (p<0.05) and five others that are suggestive (0.05<p<0.2).
There were two other probes for APP (ILMN_1653283 and ILMN_2404065), both of which showed significant positive correlations in the cerebellum and temporal cortex, except ILMN_2404065, which was not significant (p = 0.21) but had a positive trend in the temporal cortex (β = 0.04).