A multimodal training modulates short-afferent inhibition and improves complex walking in a cohort of faller older adults with an increased prevalence of Parkinson's disease

Background: Falls are frequent in Parkinson's disease and aging. Impairments in the cholinergic-mediated attentional supervision of gait may contribute to increased fall risk, especially when obstacles challenge gait. Interventions combining motor-cognitive approaches have been shown to improve motor performance, cognitive skills, and falls number. Here, we hypothesized that an intervention simulating an attention-demanding walking condition could affect not only complex gait performance and fall risk but also short-latency afferent inhibition (SAI), as a marker of cholinergic activity. Methods: Thirty-nine participants at falls risk (24 Parkinson's disease participants and 15 older adults) were recruited in a randomized controlled trial. Participants were assigned to treadmill training or treadmill training with non-immersive virtual reality intervention and trained three times a week for 6 weeks. SAI, a transcranial magnetic stimulation paradigm, was used to assess cholinergic activity. Gait kinematics was measured during usual walking and while negotiating physical obstacles. Transcranial magnetic stimulation and gait assessments were performed pre, post, and 6 months post-intervention. Results: Treadmill training combined with non-immersive virtual reality induced an increase in inhibition of the SAI protocol on cortical excitability, improved obstacle negotiation performance, and induced a reduction of the number of falls compared with treadmill training. Furthermore, the more SAI increased after training, the more the obstacle negotiation performance improved and fall rate decreased. Conclusions: We provide evidence that an innovative rehabilitation approach targeting cognitive components of complex motor actions can induce changes in cortical cholinergic activity, as indexed by SAI, thereby enabling functional gait improvements.