Prevalence and Clinical Determinants of Cognitive Impairment in Diverse Patients with Parkinson's Disease

Technical Summary: Prevalence and Clinical Determinants of Cognitive Impairment in Diverse Patients with Parkinson Disease

Problem Statement
Cognitive impairment is a major complication of Parkinson disease (PD), significantly impacting quality of life and caregiver burden. While the prevalence of cognitive decline in PD is estimated to be high, existing data is limited by a lack of diversity in study populations (predominantly White, higher-income groups) and reliance on Level I diagnostic criteria based on screening scales (e.g., MoCA). These screening tools are known to be susceptible to cultural, linguistic, and educational biases, potentially leading to misclassification or overestimation of impairment in diverse cohorts. Furthermore, the specific clinical determinants—vascular, motor, and nonmotor factors—associated with cognitive impairment in ethnically and socioeconomically diverse populations remain understudied.

Methodology
This study utilized a cross-sectional, case-only analysis of the East London Parkinson disease (ELPD) project, conducted at the Royal London Hospital between July 2022 and July 2025.

• Participants: The analysis included 223 patients with clinically diagnosed PD (median disease duration 4.0 years). The cohort was highly diverse: 43.9% White, 38.6% South Asian, and 17.5% "Other" (including Black, Middle Eastern, and mixed ethnicity).
• Outcome Definition: The primary outcome was Parkinson disease cognitive impairment (PDCI), defined as either mild cognitive impairment (PDMCI) or Parkinson disease dementia (PDD). Crucially, diagnoses were not based solely on screening scores but were established via expert multidisciplinary team (MDT) consensus. The MDT reviewed demographic data, native language, education, clinical history, comorbidities, medication, MoCA scores, Clinical Dementia Rating (CDR), depression/anxiety scores, and neuroimaging (available for 76% of cases).
• Exposures: The study examined vascular risk factors (hypercholesterolemia, hypertension, type 2 diabetes, smoking), motor severity (MDS-UPDRS Part III), and nonmotor features (daytime somnolence, sleep disturbance, RBD, olfaction, hearing loss), alongside socioeconomic deprivation (Index of Multiple Deprivation).
• Statistical Analysis: Univariate and multivariable logistic regression models were employed. Multivariable models were adjusted for age, gender, ethnicity, years of education, disease duration, and depression scores. Missing data (<5%) was imputed; variables with >5% missingness were excluded from multivariable modeling.

Key Contributions

• Robust Diagnostic Approach: The study moves beyond Level I screening criteria by utilizing a gold-standard, expert MDT consensus approach that accounts for cultural and educational biases, providing a more accurate prevalence estimate in a diverse real-world setting.
• Diverse Cohort: It addresses the gap in literature regarding ethnically and socioeconomically diverse PD populations, specifically focusing on a multi-ethnic cohort in East London.
• Differentiation of Determinants: The analysis disentangles the associations of vascular, motor, and nonmotor factors with cognitive impairment, specifically testing whether ethnic disparities persist after adjusting for these clinical and socioeconomic variables.

Results

• Prevalence: Among the 223 participants, 50.2% (112/223) were classified as having PDCI. This included 27.8% (62/223) with PDD and 22.4% (50/223) with PDMCI.
• Ethnicity: South Asian ethnicity was strongly associated with PDCI. In univariate analysis, the odds ratio (OR) was 2.30 (95% CI, 1.32–4.00). After adjusting for age, gender, education, disease duration, and depression, the association strengthened (OR, 3.60; 95% CI, 1.68–7.69). This association remained robust even after further adjustment for vascular risk factors, motor severity, and socioeconomic deprivation.
• Vascular and Lifestyle Factors: In the adjusted model, smoking was independently associated with increased odds of PDCI (OR, 3.62; 95% CI, 1.56–8.41). Hypercholesterolemia showed a weaker but significant association (OR, 2.03). Hypertension and type 2 diabetes were not independently associated.
• Motor and Nonmotor Factors: Greater motor severity (MDS-UPDRS Part III) and higher daytime somnolence (Epworth Sleepiness Scale) were independently associated with PDCI.
• Sub-analyses (PDD): For the PDD subgroup specifically, South Asian ethnicity remained associated (OR, 3.02), though this was attenuated when socioeconomic deprivation was added to the model. Lower olfaction scores were specifically linked to PDD but not PDCI in the broader model.

Significance and Claims
The paper claims that cognitive impairment is highly prevalent (affecting over half) in a diverse PD cohort at a median of four years post-diagnosis. The study highlights that South Asian individuals have approximately threefold higher odds of cognitive impairment compared to White individuals, a disparity that persists even after controlling for vascular risk factors, motor severity, and depression.

The authors suggest that while biological mechanisms (such as vascular risk or APOE status) may contribute, the attenuation of the South Asian ethnicity association with PDD upon adjustment for socioeconomic deprivation indicates that socioeconomic factors and structural inequities likely play a significant role in cognitive disparities. The study concludes that the use of expert consensus rather than screening tools alone mitigates cultural bias, revealing a high burden of impairment that requires culturally sensitive assessment methods. The findings underscore the need for targeted, methodologically rigorous cognitive assessments in underrepresented populations, though the cross-sectional design precludes causal inference regarding the temporal relationship between risk factors and cognitive decline.