Auditory evoked potentials (AEPs)
There are several reports regarding AEPs in PD. Gawel et al.  described prolongation of wave V latency, whereas Prasher and Bannister  reported normal AEP findings. This difference may be due to the selection of patients. Tachibana et al.  showed normal latencies in PD patients without dementia. By contrast, PD-D patients revealed a significant prolongation of wave I-V (particularly wave III-V) inter-peak latencies compared with patients without dementia and healthy subjects. These results show that auditory brainstem pathways are involved in PD-D and could be explained by the more widespread involvement of the brainstem pathways by the disease process and/or other concomitant pathologies in such patients. Green and colleagues [26, 27] reported that the P1 (50 ms) of the middle-latency AEP was lacking in 39% of AD and 58% of PD-D patients and suggested that abnormalities of P1 in patients with dementia may be due to cholinergic dysfunction.
Visual evoked potentials (VEPs)
There have been contradictory results concerning the changes in VEPs in patients with PD. Bodis-Wollner and Yahr  reported that the average latency was prolonged but became less prolonged on levodopa therapy, suggesting that the extrapyramidal connections of the visual cortex and the retinal dopaminergic neurons can be affected in PD. Dinner et al.  and Nightingale et al.  did not find prolonged VEP latencies. Calzetti et al.  compared VEP and electro- retinogram data and concluded that changes in VEPs are not entirely dependent on alterations at the retinal level. Okuda et al.  found that prolonged P100 latency was found only in PD-D patients and speculated that dysfunction in the central visual system plays a role in the abnormal pattern of VEPs in PD. In conclusion, the disturbance of both retinal and nigrostriatal dopaminergic systems may contribute to the delayed VEP latency.
Somatosensory evoked potentials (SEPs)
Short-latency SEPs were reported to be normal when recorded from the parietal region regardless of cognitive function [24, 33, 34], although Potolicchio and O’Doherty  found that interwave latencies were prolonged in some patients. In contrast to parietal SEPs, Rossini et al.  described reduced frontal N30 amplitude in PD patients. They suggested that dysfunctions of the thalamo-basal ganglia-frontal cortex circuit or supplementary motor cortex might play a role in the reduction of frontal N30 amplitude.