Extent of Dorsolateral Prefrontal Cortex Plasticity and Its Association With Working Memory in Patients With Alzheimer Disease

Abstract

The extent of dorsolateral prefrontal cortex (DLPFC) plasticity in Alzheimer disease (AD) and its association with working memory are not known. To determine whether participants with AD had impaired DLPFC plasticity compared with healthy control participants, to compare working memory between participants with AD and controls, and to determine whether DLPFC plasticity was associated with working memory. This cross-sectional study included 32 participants with AD who were 65 years or older and met diagnostic criteria for dementia due to probable AD with a score of at least 17 on the Mini-Mental State Examination and 16 age-matched control participants. Participants were recruited from a university teaching hospital from May 2013 to October 2016. Plasticity of the DLPFC measured as potentiation of cortical-evoked activity using paired associative stimulation (a combination of peripheral nerve electrical stimulation and transcranial magnetic stimulation) combined with electroencephalography. Working memory was assessed with the n-back task (1- and 2-back) and measured using the A' statistic. Among the 32 participants with AD, 17 were women and 15 were men (mean [SD] age, 76.3 [6.3] years); among the 16 controls, 8 were men and 8 were women (mean [SD] age, 76.4 [5.1] years). Participants with AD had impaired DLPFC plasticity (mean [SD] potentiation, 1.18 [0.25]) compared with controls (mean [SD] potentiation, 1.40 [0.35]; F1,44 = 5.90; P = .02; between-group comparison, Cohen d = 0.77; P = .01). Participants with AD also had impaired performances on the 1-back condition (mean [SD] A' = 0.47 [0.30]) compared with controls (mean [SD] A' = 0.96 [0.01]; Cohen d = 1.86; P < .001), with similar findings for participants with AD on the 2-back condition (mean [SD] A' = 0.29 [0.2]) compared with controls (mean [SD], A' = 0.85 [0.18]; Cohen d = 2.83; P < .001). Plasticity of DLPFC was positively associated with working memory performance on the 1-back A' (parameter estimate B [SE] = 0.32 [0.13]; standardized β = 0.29; P = .02) and 2-back A' (B [SE] = 0.43 [0.15]; β = 0.39; P = .006) across both groups after controlling for age, education, and attention. This study demonstrated impaired in vivo DLPFC plasticity in patients with AD. The findings support the use of DLPFC plasticity as a measure of DLPFC function and a potential treatment target to enhance DLPFC function and working memory in patients with AD.