Visual event-related potential changes in multiple system atrophy: Delayed N2 latency in selective attention to a color task

1. Introduction 

Multiple system atrophy (MSA) is an adult onset neurodegenerative disease characterized by varying degrees of parkinsonism, cerebellar ataxia, and autonomic dysfunction. Although cognitive dysfunction was not considered a main feature of MSA, mild cognitive deficits, particularly in frontal lobe function, had been suspected by neuropsychological tests [1], [2], [3]. Robbins et al. found evidence of impairment in executive function, attention, and working memory [2]. However, motor involvement in MSA may interfere with such neuropsychological testing. This problem can be partly overcome using event-related potentials (ERPs), an index of cognitive function for assessment of cognitive processing, because ERPs are essentially independent of motor influence [4], [5].

Until now, numerous ERP studies have been done to evaluate the cognitive function in various neurodegenerative diseases. Especially much has been reported about the clinical application of P3. In the ERP components, the P3 component is the most well-known endogenous component, as an electrophysiological index of cognitive function in studying cognitive processing. As a matter of course, several reports paid attention to the P3 component, and showed the P3 abnormalities in MSA during visual discrimination tasks. The cognitive impairments in MSA were pointed out from an electrophysiological view [4], [5], [6]. The prolongation of visual P3 latency and the deterioration of P3 amplitude were shown in MSA, although ERP abnormalities in MSA were milder than those in corticobasal degeneration or in progressive supranuclear palsy [7].

Several negative components of ERPs, such as N2, are known to correlate with information processing stages. The N2 component as well as P3 has been given a profound psychological significance, and was useful in clinical application of ERPs [8], [9], [10], [11]. Traditional research on the N2 component focused on stimulus classification [12], [13] and stimulus deviance [14], [15], as determinants of N2 potentials. Recent years have seen an explosion of research on the N2 component of ERPs, a negative wave peaking between 200 and 350ms after stimulus onset. The oddball N2 potential in the visual modality is known to belong to the family of attention-related N2 components, and to consist of an N2a component with posterior scalp distribution and an N2b component with frontocentral scalp distribution [16], [17].

We applied a visual selective attention task to elicit ERPs for MSA patients and age-matched controls and also measured several ERP components earlier than P3, because abnormalities of endogenous ERP components other than P3 have not been described in MSA. We now report N2 abnormalities in MSA during selective attention to color. Our electrophysiological data may contribute to the knowledge of cognitive dysfunction in MSA.

3. Results 

All the normal subjects accurately performed the task. All the MSA patients could perform both selective attention tasks at an error rate not exceeding 5%. Grand average ERP waveforms were obtained from midline electrodes (Fig. 2). Table 1 shows the mean (SD) values of ERP peak latency, interpeak latency, peak amplitude and reaction time in MSA and normal subjects.

View Large Image Download to PowerPoint

Fig. 2 Grand average ERP waveforms in MSA patients and normal control subjects. Upward deflection of the tracing indicates positive activity. The left half shows visual ERPs during the visual selective attention to color task. The solid curves indicate ERPs to rare target color stimuli, and the dotted curves indicate ERPs to frequent non-target color stimuli. The right half shows visual ERPs during the visual selective attention to shape task. The solid curves indicate ERPs to rare target shape stimuli, and the dotted curves indicate ERPs to frequent non-target shape stimuli. N1 (closed circle) is a negative component at Oz occurring between 80 and 180ms after stimulus onset. N2 (opened arrow) is a negative component at Cz occurring between 200 and 400ms after stimulus onset. The P3 component (closed arrow) is the largest positive wave at Pz between 300 and 700ms after stimulus onset.

Table 1. Mean (SD) of peak latencies, interpeak latencies, peak amplitudes, and reaction time in MSA patients and normal subjects

		Normal subjects	MSA patients	Student's t-test (unpaired)
	Selective attention to color task
	N1 latency (Oz)	148±17ms	143±13ms
	N2 latency (Cz)	239±24ms	271±25ms	t=3.16, p=0.0045
	P3 latency (Pz)	396±37ms	429±46ms
	N2 (Cz)–N1 (Oz)	88±28ms	127±25ms	t=3.40, p=0.0030
	P3 (Pz)–N2 (Cz)	157±56ms	154±41ms
	N1 amplitude (Oz)	−10±9μV	−9±9μV
	P3 amplitude (Pz)	17±3μV	16±6μV
	Reaction time	406±48ms	498±148ms
	Selective attention to shape task
	N1 latency (Oz)	154±18ms	153±17ms
	N2 latency (Cz)	267±21ms	286±32ms
	P3 latency (Pz)	409±31ms	442±40ms	t=2.20, p=0.0371
	N2 (Cz)–N1 (Oz)	124±12ms	130±43ms
	P3 (Pz)–N2 (Cz)	148±34ms	163±39ms
	N1 amplitude (Oz)	−12±8μV	−10±8μV
	P3 amplitude (Pz)	18±4μV	17±8μV
	Reaction time	444±53ms	513±124ms

During the visual selective attention to color task, we found the N2 latency delay and the prolongation of interpeak latency from N1 peak to N2 peak in MSA subjects, compared with normal control subjects. The P3 latency delay was found in MSA during a visual selective attention to shape task. We found no significant difference between MSA and normal subjects, for N1 latency at Oz, interpeak latency from N2 peak to P3 peak and reaction time. No significant difference was found between MSA and normal subjects, for N1 amplitude at Oz and P3 amplitude at Pz.

4. Discussion 

Little is known about the ERP changes in MSA, except for abnormalities of the P3 component during a discrimination task. To examine the visual discrimination process in MSA patients more closely, we employed two kinds of selective attention tasks: a sustained visual attention to color task, and a sustained visual attention to shape task. We obtained several important results on ERP measurements by comparing between the MSA group and the age-matched control group. We found a significant prolongation of P3 latency during selective attention to the shape task. We further found a significant prolongation of N2 latency and interpeak latency between N1 and N2, during the selective attention to the color task in the MSA group. However, we did not find any significant prolongation of N2 latency during the selective attention to shape task, and also did not find the significant prolongation P3 latency during color task.

Previously, we studied the ERP components during a visual oddball task in MSA [6]. We applied a visual discrimination task to elicit ERPs, in which three kinds of figures were presented one by one in random sequence, and the subjects were instructed to discriminate shapes of the figures presented. Since all the presented stimuli were of the same color (white), the “color” factor of those stimuli did not exert any influence on the discrimination in this task. We obtained results that showed significantly prolonged P3 latency and preserved N2 latency in MSA [6]. These results of our previous study are similar to our results during the selective attention to shape task which also showed significantly prolonged P3 latency and preserved N2 latency in MSA.

We found a significant prolongation of N2 latency during the selective attention to the color task in the MSA group, although all of our MSA patients executed the color task accurately with a low rate of mistakes, less than 5%. Prolonged N2 latency in the selective attention to color task suggests that MSA patients may have subclinical impairment of color discrimination and that they discriminated or detected the target color stimuli more slowly than normal controls, although, as far as we know, impairment of color discrimination has not been reported in MSA. Abnormal color discrimination has frequently been reported in Parkinson disease (PD) patients [19]. The color impairment in PD is most prominent in the tritan (blue–yellow) axis [20], [21]. The cause of impaired color vision in PD is thought to be related to dopamine depletion, because the abnormality of color vision seen in PD can be reversed by treatment with levodopa and other dopaminergic drugs [22]. Muller et al. reported that the color vision abnormalities in PD did not correlate with dopaminergic nigral degeneration as measured by I123 β-CIT single-photon emission tomography of the dopamine transporter [23]. From these results, they suggested that this visual abnormality seen in PD is extranigral in origin [23]. Some people questioned whether abnormalities of color discrimination in PD are real or just an epiphenomenon related to the motor disability of PD patients. The participation of motor disability in execution of the color discrimination task was denied and the validity of color dysfunction in PD was supported by the additional evidence that abnormalities of the visual-evoked potentials (VEPs) produced by color pattern stimuli are more responsive to levodopa therapy than are those evoked by black-and-white stimuli [24]. PD patients have abnormal VEPs and pattern electroretinograms (ERGs), attributed to dopaminergic transmission deficiency in the visual pathway, probably the retina [25]. VEP abnormalities are not reported in MSA. Sartucci et al. studied the chromatic pattern-reversal ERGs in PD and MSA patients. Their results indicate that both chromatic and achromatic pattern-reversal ERGs are unaffected in MSA, whereas in early PD they are clearly impaired, suggesting different pathogenic retinal mechanisms between MSA and PD [25].

On the basis of our present ERP study, not only PD but also MSA might have some abnormality in color vision. N2 latency during selective attention to the color task showed significant prolongation, whereas N1 latency during the same task was not prolonged in MSA. Generally, the N1 component has been interpreted as an evaluative process for external stimuli and N1 amplitude was related to the degree of the attention to external stimuli. N1 latency is unaffected by discriminative processing [26]. The N2 is thought to reflect a later stage of selective processing [27], [28], and to be an endogenous potential. Interpeak latency from the N1 peak to the N2 peak during selective attention to color task was significantly increased in MSA, while N1 latency showed no difference between the MSA group and the normal controls. Therefore, N2 delay during this color task indicates impaired visual discrimination processing for color and preservation of the evaluative process for visual external stimuli in MSA patients. Although the color processing pathway in humans has not been elucidated completely, it is thought that color information is confined mainly to pathways which convey information from V1, through V2 and V4, to the inferior temporal lobe [29]. Le et al. conducted a functional MRI study in normal human subjects during sustained attention to color task, and showed activation of the inferior temporal gyrus, fusiform gyrus, and lingual gyrus [30] which corresponded to these color processing pathways.

Although our results do not add essential information on pathogenesis or clinical presentation of MSA, and on electrophysiological–clinical correlates, we would like to emphasize, however, that our study is the first which found a significant prolongation of N2 latency during the selective attention to the color task in MSA. Electrophysiological data may facilitate differential diagnosis between MSA and PD. Pattern VEPs and ERGs are unaffected in MSA, whereas in PD they are impaired. Color discrimination is clinically unaffected in MSA, whereas in PD it is impaired. Our most important finding is a prolonged N2 latency in the selective attention to color task. On behavioral observations, no significant difference was observed in an error rate or reaction time between MSA and normal controls. Since ophthalmological function and N1 component as an index of visual cortex activity were normally preserved, we interpreted a prolonged N2 latency as subclinical impairment of color discrimination as an impaired cognitive function caused by structural or neurotransmitter abnormalities in the brain of MSA. An attention-related N2 component has been known as an index of information processing stages. We should further look for additional scientific or clinical benefits of this electrophysiological index in future investigations.

In summary, we compared N2 as well as N1 and P3 in MSA patients with these components in normal controls, by using visual selective attention tasks. While the P3 in MSA was delayed in selective attention to shape, the N2 in MSA was delayed in selective attention to color. N1 was normally preserved. Our electrophysiological results during selective attention to a color task indicate abnormalities of attentional processing of color in MSA.