Objective: To assess whether baseline measurements of neuroimaging and biofluid-based biomarkers predict diagnosis of mild cognitive impairment (MCI) and dementia at 5-year follow-up in early Parkinson’s Disease (PD).

Background: Cognitive impairment affects a significant number of individuals with PD, with up to 57% meeting criteria for PD-MCI within 3-5 years of diagnosis [1] and 20% of individuals with PD-MCI converting to dementia over a three-year period [2]. While several risk factors for cognitive impairment in PD have been identified, no single definitive predictor has emerged, and the role of biomarkers in driving cognitive change over time is unclear.

Method: Data from the Parkinson’s Progression Markers Initiative (PPMI) study (n = 256) were used. Cognition at year 5 was defined based on a combination of self/informant-report and performance on a battery of neuropsychological tests (HVLT, BJLOT, SDM, LNS and SFT). Participants were categorised as: no cognitive impairment (n = 182), self-reported cognitive complaint (n = 46), PD-MCI (n = 15) or PD-dementia (n = 13). The ability of baseline clinical, neuroimaging (striatal dopamine transporter (DaT) binding) and biofluid markers (CSF alpha-syn/p-tau/amyloid beta) to predict cognitive classification or cognitive impairment (i.e. presence (n = 28)/absence (n = 228) of either PD-MCI or PD-D) at five-year follow-up was assessed using ordinal or logistic regression, respectively. Receiver operating characteristic (ROC) analysis was also performed.

Results: Lower cognitive function, increased mood dysfunction and reduced striatal DaT binding, as well as alterations in CSF levels of both amyloid beta (decreased) and p-Tau (increased), at baseline predicted both cognitive categorisation and presence of cognitive impairment at year 5 follow-up (Table 1). The models predicted 50.8% and 53.5% of the variance for cognitive categorisation and cognitive impairment, respectively. The ROC curve for cognitive impairment yielded a value of 0.96 (CI = 0.93-0.99) (Figure 1).

Conclusion: Collecting measurements of neuroimaging and biofluid-based biomarkers may help to inform clinical judgment regarding cognitive function trajectory in early PD. This could lead to earlier identification of risk for more personalised management strategies.

Table 1: Regression models.

Figure 1: ROC analysis for cognitive dysfunction.

References: [1] Williams-Gray CH, Foltynie T, Brayne CE, Robbins TW, Barker RA. (2007). Evolution of cognitive dysfunction in an incident Parkinson’s disease cohort. Brain 130:1787-1798; [2] Saredakis D, Collins-Praino LE, Gutteridge DS, Stephan BCM, Keage HAD. (2019). Conversion to MCI and dementia in Parkinson’s disease: A systematic review and meta-analysis. Park Rel Disord 65: 20-31. [3] Hu MTM, Szewczyk-Krolikowski K, Tomlinson P, Nithi K, Rolinksi M, Murray C, Talbot K, Ebmeier KP, Mackey CE, Ben-Shlomo Y. (2014). Predictors of cognitive impairment in an early stage Parkinson’s disease cohort. Mov Disord 29(3): 351-359.

« Back to 2025 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/baseline-cerebrospinal-levels-of-p-tau-and-amyloid-beta-and-striatal-dopamine-transporter-binding-predict-mild-cognitive-impairment-and-dementia-diagnosis-in-parkinsons-disease/