Deep learning-based physical exercise assessment of older adults using single-camera videos.

Lewy body disease (LBD) encompasses a disease spectrum that includes Parkinson's disease (PD) and dementia with Lewy bodies (DLB). PD is the fastest growing neurological disorder globally and DLB accounts for approximately 3%–7% of diagnosed dementias, highlighting both as major public health priorities [, ]. Both conditions share an overlapping prodromal phase, where neurodegeneration is occurring long before a diagnostic clinical syndrome [, , ]. In this phase, it is generally not possible to distinguish PD from DLB or other LBDs. The prodrome is highly heterogeneous, with variability in the rate of disease progression, age of onset, and the pattern and severity of symptoms [, , ]. Understanding this phase remains challenging, in part due to the difficulty of identifying and recruiting patients to research trials before manifestation of overt clinical symptoms. Studies exploring heterogeneity in prodromal LBD have either been hypothesis‐led, testing predefined subgroups, or data‐driven approaches, using unsupervised clustering [, ]. Across numerous recent imaging, clinical marker, and postmortem studies using both methodologies, isolated rapid eye movement sleep behavior disorder (iRBD) emerges as a recurring phenotype across the PD‐DLB spectrum, with iRBD‐positive and iRBD‐negative prodromal presentations commonly described [, , , , ]. Recent PPMI analyses suggest that baseline non‐motor symptom profiles in prodromal cohorts can also have prognostic value, predicting both time to PD diagnosis and motor phenotype at diagnosis []. However, detailed characterization of between‐group clinical differences at the prodromal stage itself remains limited. This study aims to characterize early clinical distinctions between two enriched prodromal cohorts: an iRBD‐positive group with polysomnography‐confirmed iRBD and an iRBD‐negative group defined by hyposmia with abnormal Dopamine Transporter Single Photon Emission Computed Tomography (DAT‐SPECT). Both iRBD and hyposmia with DAT deficit are established prodromal markers associated with a high risk of conversion to LBD. As demonstrated in the Parkinson Associated Risk Syndrome study, 67% of hyposmic individuals with DAT deficit converted to PD within 4 years, with high conversion rates confirmed at long‐term follow‐up and in the largest multicentre iRBD study to date, the phenoconversion rate to an overt neurodegenerative syndrome was 6.3% per year, reaching 73.5% at 12 years [, , ].