Alzheimer's Disease Polygenic Risk Score Identifies Mild Cognitive Impairment in Adults in their 50s

Because the pathological process in Alzheimer’s disease (AD) begins decades before onset of dementia, early identification of risk is paramount. We asked whether an AD polygenic risk score (PRS) derived from IGAP data could... [ view full abstract ]

Because the pathological process in Alzheimer’s disease (AD) begins decades before onset of dementia, early identification of risk is paramount. We asked whether an AD polygenic risk score (PRS) derived from IGAP data could identify mild cognitive impairment (MCI)—a transitional phase in the progression from normal cognition to AD—in adults who were only in their 50s. Participants were 1176 white, non-Hispanic community-dwelling men of European ancestry (mean age=56; 89%<60): N=1050 cognitively normal (CN); 83 amnestic MCI (aMCI); 43 non-amnestic MCI (naMCI).  Diagnosis was based on the Jak-Bondi actuarial/neuropsychological approach.  We controlled for non-independence of twins, the first 3 principal components from the SNP data, and factors affecting cognitive function:  age; depression; hypertension; diabetes; head injury.  We tested 6 P-value thresholds (0.05-0.50) for SNPs included in the PRS. After correction for multiple testing, higher PRSs were associated with significantly greater odds of being aMCI than CN (odds ratios [ORs]=1.36-1.43 for thresholds P<0.20-0.50).  For the most significant threshold, the OR=3.22 for the upper versus lower quartile of the PRS distribution.  ORs remained significant after excluding APOE-related SNPs from the PRS.  Diabetes was associated with significantly increased odds of having naMCI (ORs=3.10-3.41 for thresholds P<0.05-0.50).  With a realistic aMCI prevalence of 7%, prediction accuracy for CN versus aMCI was high (AUCs=0.981-0.995).  With cut-offs resulting in high negative and positive predictive values (NPV, PPV), specificity was high and sensitivity was moderate. This threshold may be useful for clinical trials, providing greater certainty that cases are truly cases by substantially reducing false positives.  With cut-offs resulting in high sensitivity and specificity, NPV was high and PPV was moderate.  This threshold may be useful for screening individuals to be referred for further evaluation because people classified as controls would be highly likely to be true controls, whereas aMCI would include increased false positives. Results are consistent with aMCI being more AD-related and naMCI having more vascular/inflammation components.  The results support our neuropsychologically-defined basis for diagnosing MCI, and the use of PRSs in clinical trials aimed at early intervention.  Such efforts are likely to be key factors for more effectively treating or slowing the progression of AD.