Predictors of cognitive impairment in Parkinson's disease: a systematic review and meta-analysis of prospective cohort studies

Abstract

Introduction: Cognitive impairment is a debilitating manifestation in Parkinson's disease (PD). We sought to investigate predictors of PD-CI (PD with cognitive impairment).

Methods: We systematically searched PubMed and Cochrane Library for prospective cohort studies and pooled estimates via random-effects models. Primary analyses for all types of cognitive impairments and subgroup analyses by separate outcomes were conducted.

Results: A total of 28,009 studies were identified, of which 57 studies with 31 factors were included in the meta-analysis. In the primary analysis, 13 factors were associated with PD-CI, comprising advanced age [relative risk (RR) = 1.07, 95% confidence interval (CI) = 1.03-1.12], age at onset (RR = 4.43, 95% CI = 1.87-10.54), postural-instability-gait disorder (RR = 3.76, 95% CI = 1.36-10.40), higher Hoehn and Yahr stage (RR = 1.83, 95% CI = 1.35-2.47), higher UPDRS III score (RR = 1.04, 95% CI = 1.01-1.08), rapid eye movement sleep behavior disorder (RR = 3.72, 95% CI = 1.20-11.54), hallucinations (RR = 3.09, 95% CI = 1.61-5.93), orthostatic hypotension (RR = 2.98, 95% CI = 1.41-6.28), anxiety (RR = 2.59, 95% CI = 1.18-5.68), APOE ε2 (RR = 6.47, 95% CI = 1.29-32.53), APOE ε4 (RR = 3.04, 95% CI = 1.88-4.91), electroencephalogram theta power > median (RR = 2.93, 95% CI = 1.61-5.33), and alpha power < median (RR = 1.77, 95% CI = 1.07-2.92). In the subgroup analysis, MAPT H1/H1 genotype increased the risk of dementia in PD. Sixty-four studies were included in the systematic review, of which 12 factors were additionally correlated with PD-CI using single studies.

Conclusions: Advanced age, genetic variation in APOE and MAPT, gait disturbance, motor assessments, non-motor symptoms, and electroencephalogram may be promising predictors for PD-CI.

Keywords: Cognitive impairment; Meta-analysis; Parkinson’s disease; Predictors.