Parkinson's disease (PD) impairs internally generated (IG) movement through dysfunction of striato-thalamo-cortical circuits, while externally guided (EG) movements may rely more on relatively preserved cerebello-thalamo-cortical pathways. The partnered dance form adapted tango provides a naturalistic context to compare IG and EG movement strategies: Leaders predominantly self-initiate movement, whereas Followers depend on partner-delivered sensory cues. The investigators tested whether these role-specific movement strategies would differentially influence clinical outcomes and resting-state functional connectivity in people with mild-to-moderate PD.
People with mild-to-moderate PD were randomized to one of three conditions: adapted tango as Leaders, adapted tango as Followers, or a non-dance health education Control. The adapted tango interventions were delivered over a 12-week period. The primary comparisons examined clinical motor outcomes and resting-state connectivity changes associated with Leader versus Follower training, and relative to Control.
During therapeutic adapted tango classes, participants assigned to the Leader role self-initiated direction, timing, and amplitude of movement. Followers used auditory, tactile, and proprioceptive cues from their partner to guide movement. This intentional contrast was used to model internally guided (Leader) and externally guided (Follower) motor control strategies within the same partnered activity.
Both dance groups showed significant clinical improvement. Specifically, Leaders and Followers each demonstrated reduced overall disease severity and motor impairment after the 12-week intervention, whereas the Control group did not show a corresponding significant pattern of clinical change. When Leaders and Followers were pooled into a Dance group, the pooled group exhibited overall clinical and motor improvement and enhanced lower-limb motor outcomes relative to Control.
Connectivity analyses indicated partly distinct neural routes associated with the Leader and Follower interventions. Leaders displayed relative network stability with a directional shift toward lower connectivity values on cerebellar-sensorimotor connections; these connectivity changes resembled connectivity patterns observed in healthy older adults for connections that are typically strengthened in PD. In contrast, Followers exhibited increased connectivity across sensorimotor and caudate-motor circuits, a pattern interpreted as consistent with cue-supported action updating. The Control arm showed some overlapping connectivity increases in motor-network regions of interest but did not show comparable clinical improvements, suggesting that connectivity increases by themselves were not sufficient to produce clinical benefit in this study.
The authors interpret the combined clinical and imaging results to indicate that adapted tango can improve motor outcomes in PD via complementary movement-weighted mechanisms. Leader training—emphasizing self-initiation—may promote neural normalization, evidenced by shifts in cerebellar-sensorimotor connectivity toward values seen in healthy older adults. Follower training—emphasizing cue-driven movement—may engage distributed cue-to-action networks, reflected in increased connectivity across sensorimotor and caudate-motor circuits. These divergent connectivity signatures offer a mechanistic account for how IG and EG strategies can each support motor gains in PD.
The clinical trial is registered as NCT02457832. Ethical approval was provided by the IRB of Emory University School of Medicine and the Research and Development Committee of the Atlanta VA Health Care System. The authors report that some underlying data are subject to access restrictions and are housed in VA-secure environments (VINCI and a VA secured server) under data-use agreements; aggregated analysis and result files are available via the authors' GitHub repository as indicated in the source document.
This report is a medRxiv preprint and has not undergone peer review. The authors note data access restrictions and describe storage within secured VA infrastructures. The preprint status means results require independent peer review and replication before being used to inform clinical practice.