Inhibitory P300 subprocesses and neural compensation in genetic risk for Alzheimer's disease: The case for temporal-spatial principal component analysis.

The P300 event-related potential (ERP) is widely investigated in cognitive neuroscience, including related to aging, with smaller amplitudes and delayed latency consistently reported in Alzheimer's disease (AD). Given that AD-related neurological changes begin years before symptom onset, ERPs in asymptomatic elders with AD risk may characterize early changes. ERPs are seldom studied in this population. Yet, healthy carriers of apolipoprotein-E (APOE) ε4 have evidenced delayed P300 latencies, while P300 amplitude differences are seldom found. However, despite its frequent study, the specific cognitive processes reflected by P300 remain unclear. We propose that these challenges are due to the relatively long P300 window, which likely encompasses multiple underlying subprocesses that overlap in time. Temporal-spatial principal component analysis (tsPCA) maintains the high temporal resolution of EEG and is better suited to isolate processes that overlap in time. Thus, we interrogated APOE ε4 differences in P300 activity during successful stop-signal inhibitory control in healthy, cognitively intact older adults (25 ε4-, 20 ε4+), using both conventional ERP metrics (i.e., mean and peak amplitude) and P300 tsPCA factors. P300 amplitudes did not differ by ε4 using conventional metrics. tsPCA revealed two P300 factors in each ε4 group: first, a Posterior P300 (attention allocation) factor, and second, a relatively Anterior P300 (performance monitoring, evaluating, and updating) factor. tsPCA uniquely revealed greater activity in ε4+ vs. ε4- in Anterior P300. ε4 groups had comparable task performance, suggesting that greater P300 activity in ε4+ likely reflects neural compensation for ε4-related deficits, thereby enabling the maintenance of good task performance.