Richard, Christian

AbstractBackgroundEvent‐related‐potentials (ERPs) during cognitive tasks are sensitive measures of cognitive decline [1]‐[2] and could quantitatively measure the trajectory of disease progression or response to clinical interventions. In this longitudinal study, we aimed to compare ERPs between individuals with mild cognitive impairment (MCI) and controls, and to demonstrate the test‐retest reliability of ERP measures to assess their potential utility for future observational studies and treatment trials.MethodsParticipants included individuals with mild cognitive impairment (MCI, n=31) and healthy controls stratified by age: HC2 (n=46, 40‐60 yrs.), HC3 (n=56, 60‐90 yrs.). ERPs during attention, image recognition, and verbal memory tasks were recorded using a wireless EEG to determine amplitudes and latencies of the P200 and late positive potential (LPP) ERP components in response to target (memorized) and non‐target (novel) stimuli. Differences between MCI and HC3 groups were characterized at the initial visit, and longitudinal changes in ERP components were determined at 1‐year follow‐up for both groups. Reliability of ERP measures was demonstrated by computing intraclass correlation coefficient (ICC) measures of test‐retest variability in all healthy controls (HC2 and HC3) from initial visit to 1‐year follow‐up.ResultsAt the initial visit, the MCI group demonstrated reduced LPP amplitude (Hedges‐g=0.8, p=10‐3) and delayed P200‐latency (Hedges‐g=0.89, p=10‐4) when compared to HC3. Reliability in healthy controls was high for LPP (0.85 and 0.75 for target and non‐target stimuli, respectively) and moderate for P200‐latency (0.67 and 0.53 for target and non‐target stimuli, respectively), without longitudinal changes after 1‐year follow‐up. Individuals with MCI demonstrated increased P200‐latency (Hedges‐g= 1.05, p=0.02, df=8).ConclusionsIndividuals with MCI demonstrated quantifiable memory ERP deficits at baseline with detectable change in the P200 component at 1‐year follow‐up, demonstrating that ERPs are sensitive measures for quantifying mild cognitive impairment. Test‐retest reliability of these measures was good or moderate after 1‐year, supporting their feasibility for longitudinal monitoring of patients during disease progression or quantifying the cognitive impact of potential interventions.

AbstractBackgroundResting state electroencephalography (rsEEG) is a neurophysiologic Alzheimer disease (AD) and alpha‐synucleinopathy (dementia with Lewy body (DLB) and Parkinson disease (PDD)) biomarker [1]–[5]. Background slowing (increased low frequency power, dominant rhythm (DR) slowing) is frequent in AD and DLB, more so in DLB. We sought to identify distinguishing rsEEG signatures between DLB and AD.MethodParticipants included dementia patients [AD (n=26), DLB (n=12), PD (PDD; n=5)] and controls (HC; n=56). 20‐channel rsEEG was collected in eyes‐closed/eyes‐open conditions (5‐min each). Power spectral densities and coherences were measured using established methods [6]. Individual rsEEG power (delta, 1‐3 Hz; theta, 3‐7 Hz; alpha, 8‐13 Hz), theta/alpha ratio (TAR), dominant rhythm (DR), and DR frequency domain prominence were measured and averaged across groups, and effect sizes were determined and compared using Hedges g.ResultAD, DLB, and PDD groups exhibited increased low frequency (delta‐theta, 1‐5 Hz) power and TAR compared to controls (p<4x10‐4), with higher effect sizes in the DLB and PDD groups compared to AD. DR frequency was also decreased in all patient groups (DLB>PDD>AD) compared to controls (p=3x10‐9, p=3x10‐5, p=4x10‐4).AD alpha power and DR frequency domain prominence were reduced compared to controls and PDD/DLB groups (p<0.003). Both AD patients and controls had comparable posterior power distribution predominance of DR, while DLB power of DR was more anteriorly distributed. Compared to controls, delta coherence was higher in AD, but lower in DLB. Delta power was positively correlated with cognitive fluctuation (CAF) in DLB (r(8)=+0.67, p=0.03), while negatively correlated with MMSE in AD (r(24)=‐0.64, p=10‐5).ConclusionDLB and AD had both common and distinct rsEEG signatures. Both had DR slowing and background slowing that was more severe in DLB than AD. Distinctive DLB findings included reduced posterior DR spatial dominance and decreased delta coherence. Conversely, distinct AD features were increased delta coherence, decreased alpha power, and decreased DR frequency prominence. Finally, the DLB and PDD groups had a similar magnitude of DR slowing, consistent with posteriorly dominant functional impairment in synucleinopathy reflected by corresponding regional rsEEG slowing, although future longitudinal prospective sequential rsEEG analyses are necessary to confirm these findings.