BackgroundGait and cognition are closely associated in Parkinson's disease (PD), with specific cognitive domains being associated with different motor symptoms. By identifying gait parameters affected by cognition, clinicians can develop targeted interventions that address cognitive impairment, improve gait, and reduce the risk of injury in PD patients. Research questionWhat gait parameters are affected by cognition in PD patients during dual-task walking, and how are these parameters related to cognitive function as measured by the Montreal Cognitive Assessment (MoCA)? Methods36 patients with available gait data and cognitive assessments were enrolled. Gait data of usual and dual-task walking sessions were recorded using lightweight wireless wearable sensors attached to trunk, lower, and upper extremities. Dual-task costs were calculated from usual and dual-task measures. Statistical analysis included non-parametric tests, Wilcoxon signed-rank test, Spearman’s correlation, and stepwise linear regression models. ResultsWalking speed, cadence, asymmetry in arm swing (ASA), between arms’ amplitude symmetry (BAS), average stride time, and jerk of the acceleration movement of the legs were found to be affected during the dual-task walking session (P<0.05). Spearman’s correlation showed significant correlations between MoCA scores and ASA (ρ=-0.469, P=0.036) and BAS (ρ=-0.448, P=0.036) costs. Stepwise linear regression models found that MoCA scores were significant predictors of BAS and ASA costs (P<0.05). SignificanceThis study found a significant association between global cognitive ability and several gait parameters costs under cognitive load caused by dual-task walking in PD patients. The study identified the gait parameters that were affected by cognitive load and found that MoCA scores were significant predictors of those gait parameters. Identifying gait parameters affected by cognition can lead to more targeted interventions for improving gait and reducing injury risk in PD patients.