Introduction: Dementia with Lewy Bodies (DLB) or Parkinson’s Disease Dementia (PDD), collectively termed Lewy body dementias (LBD) show heterogenous progression across cognitive, motor and neuropsychiatric symptom domains, yet disease-specific endpoints are lacking. We evaluated whether a composite clinical endpoint using validated scales across different symptom domains could sensitively track disease progression and align with functional and caregiver outcomes. Methods: 116 participants (DLB=72; PDD=44) were assessed at baseline, 3, and 6 months in a cluster-randomized trial comparing usual care versus management informed by an evidence-based toolkit. The Lewy Body Symptom Severity index (LBSS) was constructed by summing rescaled Mini-Mental State Examination, MDS-UPDRS (Part III), Dementia Cognitive Fluctuations Scale, and Neuropsychiatric Inventory 4-item subscore (including hallucinations). Linear mixed-effects models tested change over time. Validity was examined against caregiver Clinical Rating of Change (CRC), Bristol Activities of Daily Living Scale and caregiver Zarit Burden Interview. Results: Over 6 months, LBSS increased significantly (β=0.0307; p=0.0006). Simulation-based power analyses indicated greater statistical efficiency for LBSS than for any individual component. LBSS also detected a significant intervention effect (p=0.0365) not observed with single-domain measures. LBSS correlated with caregiver burden (Zarit; ρ=0.53, p<0.001), functional dependence (Bristol ADL; ρ=0.57, p<0.001), and CRC (ρ=–0.33, p=0.002) permitting derivation of a minimal clinically important difference. Discussion: A simple composite spanning cognition, parkinsonism, cognitive fluctuations, and neuropsychiatric symptoms sensitively detected short-interval progression, with improved statistical efficiency over single-domain measures, and was aligned with functional/caregiver outcomes. These findings support composite endpoints for LBD trials and can inform the design of disease-specific scales.
Matar et al. (Thu,) studied this question.