A meta-analysis and meta-regression of serial reaction time task performance in Parkinson’s disease.

Abstract

The meta-analysis provides clear support that learning in procedural memory (procedural learning), which underlies implicit sequence learning in the SRT task, is impaired in PD.

Apathy and noradrenaline

“…When controlling for medication effects, studies reveal deficits in learning from trial-by-trial feedback [46], a hallmark of implicit learning [47]. Indeed, a meta-analysis found Parkinson's disease patients to be significantly impaired in implicit learning across 27 studies using the serial reaction time task [48]. While implicit learning is thought to depend on the basal ganglia, explicit, declarative learning relies on the hippocampus and medial temporal lobe [49].…”

“…An [ 11 C] raclopride PET study showed striatal (accumbens) D2 release accompanied learning from discrete feedback in a probabilistic classification task [53 && ]. Further, learning from delayed feedback activates the hippocampus, whereas learning from immediate feedback engages the striatum [48]. A study investigated competing learning mechanisms in Parkinson's disease, with two initial tasks that tested novel tool features (explicit) and novel tool skill (implicit), and a follow-up task that assessed both learning acquisitions 3 weeks later.…”

Feasibility of use of probabilistic reversal learning and serial reaction time tasks in clinical trials of Parkinson's disease

“…A typical PRL task involves participants first learning a stimulus-reward pairing (Phase 1), and later shifting to learn an opposite stimulus-reward pairing (Phase 2). Among individuals with PD, non-medicated participants performed better on PRL than spatial working memory tasks, suggesting PRL is relatively preserved in early PD [1]. By contrast, among individuals with PD treated with DA replacement therapy, learning the PRL task reward contingencies had reversed was impaired but task-set switching on a separate task was enhanced [2].…”

“…These two tasks have been linked to specific DA sensitive frontalsubcortical circuits. Specifically, spatial working memory, including procedural sequence learning, has been associated with the dorsolateral prefrontal cortex (dorsal-striatum-DLPFC circuit, DS-DLPFC) [1,4], whereas probabilistic reversal learning (PRL) has been associated with ventral striatal-orbitofrontal areas (VS-OFC) [4,5].…”

“…Although participants are not told they are seeing a fixed sequence, reaction time decreases as the spatial sequence is learned (implicit learning) [1]. By contrast, PRL involves adapting to changes in stimulus-reward contingencies [8].…”