Examining a telecognitive assessment‐derived Preclinical Alzheimer’s Cognitive Composite, the tPACC, for cognitive migration over one year in community‐dwelling older adults

Abstract

AbstractBackgroundADRD clinical trials often target preclinical disease, where hope of averting the pathologic cascade and eventual cognitive decline remains highest. The preclinical Alzheimer’s cognitive composite (PACC5) was developed for in‐person administration to capture subtle cognitive decline. Early cognitive fluctuations may inform multidirectional disease progression, defined as cognitive migration. It is desirable to have a harmonized composite measurement derived from in‐person and remote assessments for detecting cognitive migration.MethodWe examined in‐person data from 417 adults (70.6±8.0y) from the Wake Forest ADRC Clinical Core, who received MRI, clinical evaluation and cognitive testing. The primary outcome was cognitive migration status determined by global CDR (0 and 0.5) increase/decrease between baseline (BL) and follow‐up (mean difference = 13.9 months): CDR‐0 Stable (51% maintained CDR‐0), CDR‐0.5 Stable (27% maintained CDR‐0.5), Reverter+ (13% migrated from CDR‐0.5 to CDR‐0) and Migrant‐ (9% migrated from CDR‐0 to CDR‐0.5). In‐person PACC5 was calculated (RAVLT Delayed Recall, Digit Symbol Coding [DSC], semantic fluency, Craft Story Delayed Verbatim, total MMSE) using BL participants with global CDR = 0 as reference. tPACC used measurements available at both in‐person and remote visits with minor modifications from PACC5: Montreal Cognitive Assessment (MoCA) replaced MMSE and DSC not included. We performed intraclass correlation coefficient (ICC) analysis and generated Bland‐Altman plots for BL tPACC and PACC5. We also evaluated associations between tPACC and neuroimaging.ResultParticipant characteristics and group differences in MRI markers are in Table 1. Bland‐Altman plots show differences and mean of tPACC and PACC5 (dashed line: ±1.96 SD, solid line: mean difference/bias). 92% of values lie within limit of agreement, indicating good agreement (overall ICC = .960, 95%CI = [.952, .967]), between tPACC and PACC5 in each group (Figure 1). At BL, there was a significant absolute agreement between in‐person tPACC and PACC5 in all groups (Figure 1). tPACC showed significant associations (Table 2) with Temporal Meta ROI thickness, hippocampal volume as % head size (ICV) and white matter hyperintensities (logWMH controlled for ICV) overall and in some cognitive migration groups.ConclusionThere is generally good agreement between BL tPACC and PACC5 across cognitive migration groups. BL tPACC may predict BL neuroimaging based on cognitive migration status. Longitudinal tPACC examination is warranted.