Gait training with intelligent treadmill for Parkinson’s disease (GaIT-PD)

Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterised by degeneration of dopaminergic neurons in the substantia nigra, resulting in bradykinesia, rigidity, tremor, and marked gait disturbances. (Bloem et al., 2021; Tolosa et al., 2006). Spatiotemporal gait alterations, including reduced step length, increased cadence, decreased gait speed, and prolonged double support time, represent key factors contributing to impaired mobility and reduced functional independence in PD. (Abbruzzese et al., 2016; Tomlinson et al., 2014). In PD, gait automaticity is reduced due to impairments in motor learning and sensory feedback-based control, resulting in compensatory reliance on anticipatory (feedforward) mechanisms. (Nieuwboer et al., 2009; Olson et al., 2019). These deficits become especially apparent under dual-task (DT) conditions, where performing a cognitive task (such as counting or serial subtractions) while walking further disrupts gait spatiotemporal parameters (Kelly et al., 2012; Rochester et al., 2005). DT gait assessment, therefore, serves as a sensitive marker of impaired automatic walking. Recent research indicates that both single- and DT gait tests can identify early locomotor deficits in PD and distinguish patients from age-matched healthy controls, implying their utility in diagnosis and prognosis value (Caronni et al., 2025; Zhang et al., 2022). In parallel, a growing body of evidence suggests that DT gait is also a valuable measure for examining the effects of targeted rehabilitation interventions. Importantly, two training modalities involving cognitive components operate through different mechanisms: DT training and biofeedback (BF) training. DT training combines gait with an additional cognitive task (e.g., counting, serial subtraction, verbal fluency). In this approach, the same cognitive functions assessed during DT testing, divided attention, working memory, and executive control, are directly targeted. Systematic reviews and randomised controlled trials demonstrate that this method can enhance DT performance, reduce DT costs, and improve gait speed, step length, and stability in individuals with PD (Beck et al., 2018; Johansson et al., 2022; Sarasso et al., 2024; Wollesen et al., 2021; Yang et al., 2019; Zheng et al., 2021). BF training, by contrast, does not involve performing a separate cognitive task; instead, it requires a high level of sustained attention and concentration. Patients must continuously monitor visual or auditory feedback delivered in real-time and adapt their gait parameters accordingly. BF mainly stimulates attentional rather than executive cognitive processes. Studies using treadmill-based BF report significant improvements in step length, gait symmetry, gait regularity, and postural stability (Carpinella et al., 2017; Ginis et al., 2016; Huang et al., 2006; McMaster et al., 2022). Unlike open-loop rhythmic cueing, BF functions in a closed-loop mode, allowing for immediate motor adjustments and supporting sensorimotor learning even when basal ganglia dysfunction is present (Muthukrishnan et al., 2019). Within this framework, a DT gait test that includes standardised and progressively challenging cognitive conditions can be a particularly useful tool to systematically evaluate how cognitive load influences gait and to assess the impact of interventions that depend on attentional demands during walking. This is especially pertinent for protocols requiring continuous monitoring and voluntary adjustment of gait parameters, such as treadmill training with BF, where individuals must modify step length, cadence, or symmetry in response to real-time visual or auditory cues feedback (Carpinella et al., 2017; McMaster et al., 2022). Research involving treadmill training, cognitive tasks, Biofeedback or augmented reality has shown improvements in DT gait performance, gait speed, and stability, as well as a decrease in fall risk. This confirms that the DT gait test can serve both as a rehabilitative stimulus and a sensitive clinical outcome measure (Kim et al., 2022; Mylius et al., 2021; Schaeffer et al., 2019; Vieira et al., 2014). Further evidence suggests that exercise therapy is an effective method for enhancing gait and reducing motor decline in PD, especially when training is intensive, personalised, and started early (Frazzitta et al., 2009; van der Kolk et al., 2019). Taken together, these findings indicate that a DT gait test with varying cognitive difficulty might be an ideal framework for assessing the effectiveness of treadmill-based BF gait training. This training combines aerobic and attentional elements and aims to improve both locomotor automaticity and the capacity to handle concurrent cognitive tasks. This approach aligns with the literature highlighting the importance of exercise intensity and task complexity in fostering meaningful motor and cognitive adaptations (MacInnis and Gibala, 2017; Zheng et al., 2021). Aims of the study The overall aim of this work is to develop evidence-based assessment tools and rehabilitation interventions to improve spatiotemporal gait parameters in PD, with a particular focus on the role of DT performance and treadmill training with BF. The specific objectives are: Develop and validate a standardised ST gait test with three levels of increasing cognitive difficulty (serial subtractions: -1, -3, -7). This test aims to quantify in a reproducible and clinically meaningful way how increasing cognitive load affects gait spatiotemporal parameters and to measure DT costs across conditions, providing a sensitive tool to evaluate gait automaticity and its changes in response to rehabilitation. Examine how cognitive load affects gait in early-stage PD compared with age-matched healthy controls, to identify between-group differences in locomotor parameters and DT costs, determining the extent to which reduced gait automaticity is already present in early disease, and assessing whether DT gait may serve as a sensitive marker of functional impairment. To assess the effectiveness of an aerobic treadmill training protocol with BF in enhancing gait parameters and aerobic fitness in individuals with PD, this analysis will examine its effects on gait speed, step length, and locomotor stability under both single- and DT conditions. Additionally, it will explore the relationship between exercise intensity (measured via oxygen consumption, V̇O₂) and improvements in aerobic fitness.