Defining the primary cognitive endpoint is a major decision for Alzheimer’s disease preventive trials. As an example for further trials we present in detail the three-year cognitive decline in the placebo group of MAPT trial, a randomized controlled trial (RCT) using a cognitive composite score (MAPT-PACC). Participants were dementia-free adults 70 years or older, with subjective memory complaints. Our findings as expected showed subjects with older age (>75), higher beta amyloid brain deposition, APOE-ε4 allele carriers, with low RBC DHA+EPA levels and higher CDR level are at higher risk of cognitive decline. The data presented in this paper can be useful for future preventive trials to choose the primary cognitive end point, assess the clinical relevance of cognitive changes and perform sample size calculation for several targeted population eg. ApoE4, amyloid +, oldest old, lower n3-PUFA. We believe that the trial group with CDR 0.5, without being selected by a memory test endpoint is a good target population for AD preventive trials.