EEG BIS-AD correlates with metrics of cognitive function in healthy subjects and in patients with Alzheimer’s disease

Abstract

Background: EEG BIS® is a widely used monitor of the neurological effects of anesthesia that correlates with cognitive function metrics (e.g., MMSE, memory recall, and sedation level)[1-2]. BIS-AD, a novel revision (0.1) of BIS, was refined to correlate with dementia severity.Objective: To examine the association of BIS-AD and cognitive performance in healthy and demented elderly subjects. Methods: EEG and cognitive assessments (MMSE, ADAScog) were recorded from healthy elderly subjects (NML, CapeCod), patients recently diagnosed with mild AD (MildAD, McLean), and subjects with moderate to severe AD (ModAD, UCLA). Longitudinal assessments were recorded from healthy subjects (q3 mo., ongoing). BIS-AD was calculated from frontotemporal EEG during a three-minute vigilant, eyes-closed period. Partial correlation coefficients accounting for multiple subject visits measured association between BIS-AD and cognitive performance metrics. Results: 365 assessments (324 NML, 30 MildAD, 11 ModAD) were recorded from 111 subjects (84 NML, 16 MildAD, 11 ModAD) (age 79.9 /4.9). BIS-AD correlated with MMSE (R 0.290, p 0.001) and ADAScog (R -0.164, p 0.003), but not with age (i.e., BIS-AD was higher in subjects with better cognitive performance). Although the range of cognitive performance was small in the healthy subjects (MMSE: 28.8 /1.3; ADAScog: 7.4 /3.8), BIS-AD correlated with cognitive performance: MMSE (R 0.107, p 0.05) and ADAScog (R -0.136,p 0.018). At baseline, BIS-AD was higher in healthy subjects who performed better than the median score on both tests (i.e., MMSE 29 and ADAScog 8.1) than those who did not (96 vs. 94, p 0.039). At subsequent visits in healthy subjects, change in BIS-AD from baseline correlated with change in ADAScog (R -0.150 p 0.019) but not MMSE. Conclusions: BIS-AD correlated with cognitive performance in elderly healthy and demented subjects. BIS-AD was significantly higher at baseline in healthy subjects who performed in the upper 50% on both MMSE and ADAScog tests than those who did not. Change in BIS-AD over visits correlated with change in cognitive performance. Follow-up assessments in these subjects will reveal whether low BIS-AD values may be an early marker of dementia onset.