Abstract
OBJECTIVE: Cognitive complaints without objective cognitive impairment in Parkinson's Disease, termed Parkinson's Disease-Subjective Cognitive Decline (PD-SCD), have been associated with cognitive decline. However, its progression is heterogeneous, highlighting the need for improved identification of patients at greater risk for deterioration. This cohort study aims to investigate associations between CSF biomarkers and cognitive decline in PD-SCD. METHODS: We included patients from the PPMI cohort with PD-SCD, available baseline CSF beta-amyloid(1-42) (Aβ(42)) and phosphorylated-tau(181) (p-tau(181)), and longitudinal Montreal Cognitive Assessment (MoCA). RESULTS: A total of 80 patients were included, with a mean age of 62.1 years and a median disease duration of 5.6 months at baseline. Five patients were identified with biological AD, who showed a more pronounced decline in MoCA (year 7: β = -4.15, p = 0.009). After excluding AD patients, linear mixed-effects models (LMEM) demonstrated an association between baseline log-transformed Aβ(42) and MoCA progression (β = 2, p = 0.004). Based on these data, we created three CSF-based subgroups: Normal Aβ(42) (n = 50), Low Aβ(42) (n = 25), and Biological AD (n = 5). LMEM predicted greater cognitive decline for Low Aβ(42) versus Normal Aβ(42) (β = -0.161 points-per-year, p = 0.007), and for Biological AD versus Normal Aβ(42) (β = -0.390 points-per-year, p < 0.001). The risk of dementia was increased for Low Aβ(42) (HR = 5.2, p = 0.029) and Biological AD subgroups (HR = 7.7, p = 0.013). INTERPRETATION: In PD-SCD, baseline CSF Aβ(42) is associated with cognitive progression. Furthermore, we identified three CSF biomarker-based cognitive trajectories (Normal Aβ(42), Low Aβ(42), and Biological AD), each characterized by progressively worse cognitive outcomes. These findings could be useful for implementing enrichment strategies in future clinical trials targeting PD-associated cognitive decline.