Long-term levodopa (L-DOPA) therapy is known to effectively relieve motor symptoms in Parkinson''s disease (PD); however, its effects on non-motor symptoms including cognition, hallucinations and affective disorders are inconsistent, and can adversely affect certain aspects of cognitive function. Here, we investigated the non-motor symptoms of PD patients to evaluate the risk factors associated with Parkinson''s disease dementia (PD-dementia). The ameliorating effects of an Gynostemma pentaphyllum ethanol extract (GP-EX) and gypenosides (GPS) on memory deficits in the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)- lesioned mouse model of PD were also examined to develop phytonutrients to serve as adjuvants for PD-dementia treatment. First, 88 PD patients taking levodopa consecutively from 2011 to 2015 were enrolled in this retrospective study. Predictor variables were compared between patients with and without PD-dementia. Of the 88 PD patients, 30 patients (34.1%) developed PD-dementia, of which 21 (70%) developed dementia symptoms within 6 years post PD onset. The average daily dose of levodopa was higher in the PD-dementia group than in the non-dementia group, and the differences between the initial and average daily doses were statistically significant in the PD-dementia group (P = 0.003). Age and daily dose at enrollment were positive independent variables to predict the risk of dementia development in PD patients. Next, the MPTP-lesioned mice administered L-DOPA (25 mg/kg) exhibited deficits in memory, including habit learning and spatial memory. However, GP-EX treatment improved the decreases in retention latency time in the passive avoidance test, number of tyrosine hydroxylase-immunopositive cells, and dopamine levels of the substantia nigra-striatum. GP-EX also reduced the increase in retention transfer latency time in the elevated plus-maze test as well as expression of the N-methyl-D-aspartate (NMDA) receptor, and reversed the reduction of phosphorylation of extracellular signal-regulated kinase and cyclic AMP-response element binding protein in the hippocampus in the same models. GPS also showed effects similar to GP-EX. Taken together, these results suggest that the levodopa dose may be associated with the risk of cognitive impairment in PD. Moreover, GP-EX has protective effects on habit learning memory deficits by activation of dopaminergic neurons and on spatial memory deficits by modulation of the NMDA receptor-mediated signaling system in the MPTP-lesioned mice treated with L-DOPA. Thus, the administration of adjuvant phytonutrients should be considered during levodopa therapy to reduce cognitive impairment in PD patients. Furthermore, GP-EX may serve as a medicinal phytonutrient adjuvant for treating cognitive impairment in PD patients receiving levodopa.