Impaired Novelty Processing in Apathy After Subcortical Stroke
Background and Purpose— Apathy is associated with decreased novelty-seeking behaviors and is a prevailing behavioral symptom after stroke affecting cortical and subcortical regions. We studied the relationship between apathetic state after subcortical stroke and neural orienting response to novel events using an event-related evoked potential (ERP) technique.
Methods— Twenty-nine patients with subcortical ischemic stroke were grouped according to whether they showed apathy or not. We analyzed apathy state scaled by the modified Starkstein apathy score and auditory P3 ERP components evoked by task-relevant target stimuli (target P3) and task-irrelevant novel stimuli (novelty P3).
Results— The apathetic group showed a significantly lower score of verbal fluency test and global cognitive function test compared with the nonapathetic group. The novelty P3 latency was significantly prolonged, and its amplitude was reduced over the frontal site in the apathy group. The apathy scale was correlated with the novelty P3 latency and amplitude at the frontal site. The target P3 measures were related to global cognitive function.
Conclusions— The present study suggests that apathy after subcortical stroke is associated with impaired neural processing of novel events within the frontal–subcortical system and that the novelty P3 is a useful physiological measure for assessing apathy after stroke.
Apathy is often observed after stroke and is defined as reduced motivation and lack of initiative and exploration. Apathy state after stroke prevents patients from engaging in rehabilitation programs, resulting in delay of physical and social recovery. The frontal cortex is a neural structure responsible for apathy state after stroke.1,2 In addition, subcortical structures including caudate nucleus, thalamus, and basal ganglia are also reported to be related to apathy.3–5 Thus, the dysfunction of a frontal–subcortical system could be involved in apathy, but the underlying mechanism remains uncertain in its relationship with subcortical stroke lesions.
Decreased novelty-seeking behaviors could be associated with apathy state resulting from reduced interest in environments. Although the ability to respond to novel events in environments has been studied using event-related evoked potential (ERP) in humans,6 electrophysiological studies for poststroke apathy have been conducted scarcely.7 A novelty P3 is known to be generated by a salient stimulus that occasionally occurs in a train of frequent background stimuli.8 We demonstrated abnormalities in novelty processing in patients with vascular dementia and Parkinson’s disease, in which the abnormality of novelty P3 was correlated with impaired frontal lobe functions.9,10 A single seminal study demonstrated that the degree of apathy was correlated with the reduction of novelty P3 amplitude in apathetic patients with prefrontal cortical lesions.7 Here we studied a direct relationship between apathy and subcortical stroke lesion by using a similar ERP technique to gain insights into pathophysiology of apathy after subcortical stroke.
Clinical data suggest that apathy is often associated with cognitive impairments, but they can be dissociable. Apathy is related to hypofunction of dopaminergic system,11 whereas dementia is more linked to impairments of acetylcholine system. This suggests that differentiation of apathy and cognitive impairments would be clinically useful for future treatment of these behavioral abnormalities. ERP studies in dementia and Parkinson’s disease suggested the dissociation of target and novelty P3.9,10 Thus, this study sought a possibility that 2 types of P3 measures can be associated differentially with apathy and global cognitive functions.
Subjects and Methods
Twenty-nine patients (21 men and 8 women; mean age 71.7±9.4 [SD] years) participated in this study after giving informed consent approved by the institutional review committee. The criterion for patient selection was that they had at least 1 episode of stroke and their lesions were detected using T1 and T2 images in MRI. All subjects had only subcortical cerebral infarctions. Information about individual subcortical lesion was listed in Table 1. We excluded patients with dementia that was assessed using the modified Diagnostic Statistical Manual IV criteria. None of the patients had been diagnosed previously with primary psychiatric disorders.
To quantify apathetic state, we adopted a Japanese version of the apathy scale reported by Starkstein et al12,13 The apathy scale consisted of 14 questions concerning spontaneity, initiation, emotionality, activity level, and interest in hobbies. This scale was used in a self-assessment style. An answer to each question was scored in 4 grades, 0 to 3, and the total score was used for the analysis. We classified the patients into 2 groups according to the score: nonapathetic group (apathy score <16 points) and apathetic group (apathy ≥16 points). The cut-off point was determined on the basis of our previous results obtained from Japanese stroke patients.14 The apathy score was 8.5±4.4 for the nonapathetic group (7 men and 6 women, 72.9±9.6 years old) and 22.4±5.2 for the apathetic group (11 men and 5 women, 70.6±9.4 years old). There were no significant differences in age, length of education, male/female ratio, lesion location, lesion size, and lesion number between 2 groups (Table 1).
The global cognitive function was evaluated by the revised version of the Hasegawa dementia rating scale (HDS-R).15 The test assessed range of information, short-term memory, digit span, and verbal fluency. The maximum score for this scale is 30, and no subjects scored <18. We also evaluated verbal fluency in a separate session in which patients were requested to list as many vegetable names as possible within 1 minute. In addition, depression state was assessed using a self-rating depression scale (SDS).16
The experiment was conducted with the subjects seated comfortably in a sound-attenuated room. Before the experiment, binaural audiometric threshold was determined at 1000 Hz for each subject (threshold range ≈5 to 20 dB). All stimuli were presented binaurally through stereo headphones at 60 dB above each individual’s threshold. Auditory ERPs were elicited using a 3-stimulus oddball paradigm that consisted of a series of computer-generated tone bursts 100 ms in duration. A total of 400 stimuli, including frequent 1000 Hz tones (65%), infrequent 2000 Hz tones (20%), and novel sounds (15%), were delivered randomly with an interstimulus interval of ≈1.0 to 1.3 s. The novel sounds consisted of unique environmental sounds that had been recorded digitally from cartoon movies (dog barking, chime, etc.). To maintain the novelty of stimuli, each novel sound was used only once for each subject. The subjects were instructed to press a button with the right thumb when they detected the 2000-Hz tone and were not warned of the novel sounds. A 1-minute practice session without novel sounds was given before the experiment.
The EEG was recorded from antigen (Ag)/AgCl electrodes placed at 15 scalp locations (Fpz, F3, Fz, F4, T3, C3, Cz, C4, T4, T5, P3, Pz, P4, T6, and Oz) on the basis of the 10-20 system and below the left eye referred to linked earlobes. The electrode impedance was kept at <5 kΩ. The EEG was amplified (bandpass 0.05 to 100 Hz/s), digitized (250 Hz/channel), and stored on a hard disk for offline analysis. The averaging epoch was 1024 ms, including 200 ms of prestimulus baseline. Individual trials with excessive muscle activity or blinking (>80 μV peak-to-peak amplitude) were excluded. Target trials with a reaction time (RT) of <150 ms or >1000 ms were also excluded from averaging. For averaging, 216±12 trials for standard stimuli, 63±6 for target stimuli, and 45±6 for novel stimuli were included.
We measured P1, N1, and P2 components to standard stimuli and P3 components to target and novel stimuli (Figure 1). The P1, P2, and P3 were defined as the largest positive peak in 30- to 60-, 150- to 270-, and 300- to 600-ms range poststimulus, respectively. The N1 is the largest negative peak in 75- to 150-ms range poststimulus. The peak amplitude of each component was measured relative to the 200-ms prestimulus baseline.
Group comparisons were performed using the Wilcoxon (W) signed rank test for neuropsychological data and correct response rate during the ERP experiment. RT data and ERP measures were subjected to ANOVA with RT and each ERP measure as dependent factors and groups as group factor. Topographical maps were obtained for target and novelty P3 using a spline-interpolated method. To analyze their topographies, repeated-measures ANOVA was performed on midline 5 electrodes (Fpz, Fz, Cz, Pz, Oz) with the Greenhouse–Geisser adjustment to the degree of freedom to correct violations of the assumption of sphericity after amplitude normalization within each subject.17 Post hoc analysis was done with the Bonferroni–Dunn procedure. Correlation analysis was performed to examine the relationship between neuropsychological scores and ERP measures using the Spearman rank correlation method. The level of significance was set at P<0.05.
Neuropsychological and Behavioral Data
The HDS-R score was lower for the apathetic group compared with the nonapathetic group (W=175.0; P<0.005; Table 2). The verbal fluency score was also lower for the apathetic group (W=189.0; P<0.05). There was no significant difference in the SDS between the apathetic and nonapathetic groups (40.3±8.8 versus 34.2±11.2). The incidence of depression (SDS>50) was also the same between the 2 groups (15.4% versus 12.5%). The behavioral performance was comparable between them: RT to targets was 507.6±122.6 ms and 511.5±108.6 ms, and correct response rate was 82.3±23.3% and 84.1±19.0% for the apathetic and nonapathetic group, respectively.
Figure 1 demonstrates grand-average ERPs to novel and target stimuli. There were no group differences in either latency or amplitude in P1, N1, and P2 components. The novelty P3 latency at Cz was increased in the apathetic group compared with the nonapathetic group (407.0±59.6 ms versus 356.6±32.5 ms; F1,27=7.45; P<0.02). However, there was no significant difference in the target P3 latency at Pz between the apathetic and nonapathetic groups (377.3±46.8 ms versus 356.3±43.5 ms). As seen in Table 3, there were no differences in the novelty and target P3 amplitude between 2 groups except for Fz site, in which there was significant reduction of the novelty P3 amplitude (F1,27=7.68; P<0.01) and marginal reduction of the target P3 amplitude for the apathetic group (F1,27=2.96; P<0.1).
We compared scalp topographies of the target and novelty P3 between 2 groups (Figure 2). The target P3 was distributed maximally over the parietal site in both groups. This was shown by a significant main electrode effect (F4,108=30.0; ε=0.69; P<0.0001) without an interaction of group x electrode. In contrast, there was a significant interaction of group x electrode in the novelty P3 (F4,108=3.12; ε=0.66; P<0.05). Post hoc analysis indicated that the novelty P3 amplitude was maximal at Cz for the nonapathy group (electrode effect; F4,48=22.8; ε=0.78; P<0.0001), whereas it distributed maximally at Pz for the apathy group (electrode effect; F4,60=23.0; ε=0.46; P<0.0001).
We further examined the relationship between P3 measures and neuropsychological data. The novelty P3 latency at Cz was well correlated with apathy scale (Spearman ρ=0.63; P<0.001; Figure 3). The target P3 latency at Pz was marginally correlated with apathy scale (ρ=0.32; P<0.1). The HDS-R score was correlated negatively with the target P3 latency (ρ=−0.52; P<0.005) and was marginally correlated negatively with the novelty P3 latency (ρ=−0.31; P<0.1). These results suggest that measures of novelty and target P3 latencies could reflect a dissociation of apathy and general intelligence. For example, 2 nonapathetic patients who had relatively lower HDS-R score (<24 points) showed prolonged target P3 latency (402.0±8.5 ms) with shorter novelty P3 latency (358.0±36.8 ms), whereas 2 apathetic patients with relatively higher HDS-R score (>27 points) showed prolonged novelty P3 latency (430.0±25.5 ms) with shorter target P3 latency (360.0±5.66 ms). The correlation between verbal fluency score and either target or novelty P3 latency was not significant. The novelty P3 amplitude at Fz was correlated with apathy scale (ρ=−0.52; P<0.005) and verbal fluency (ρ=0.61; P<0.001), but these relationships were not significant in other scalp sites. The target P3 amplitude at Pz did not show significant correlation with neuropsychological measures. We determined a reliable threshold of novelty P3 latency for the risk of apathy by calculating an odds ratio. The maximal odds ratio was 9.3 (95% CI, ≈0.97 to 90.0; P=0.053) when the threshold level of novelty P3 latency was set at 395 ms.
The present study provides 3 main findings on neurophysiological measures related to apathetic state after subcortical stroke. The apathetic group showed: (1) prolonged latency of the novelty P3; (2) reduced novelty P3 amplitude over the frontal site, resulting in posterior shift of the scalp topography; and (3) correlations between the changes in novelty P3 measures and degree of apathy state. The modulations of P3 measures could not be explained by the demographic differences or lesion characteristics between the 2 groups. The novelty P3 measures were better correlated to apathy than the target P3 measures. Reduced amplitude and prolonged latency of novelty P3 have been reported in patients with frontal lobe lesion who experienced mild apathy.7 Our patients did not have direct damage to the frontal lobe but showed similar delay of novelty P3 latency with localized amplitude reduction at the frontal site.
Evidence from patients with focal lesions or implanted electrodes suggests involvement of distributed prefrontal–hippocampal and multimodal association cortices in novelty detection.18–20 The present findings are consistent with recent neuroimaging data21,22 and support a notion that involvement of structures supportive of frontal function is a requirement for apathy.2 Cerebral blood flow study revealed that apathetic patients showed significantly reduced blood flow in the right dorsolateral frontal and left frontotemporal regions.14 The present study provides further physiological evidence that dysfunction of the frontal–subcortical system is involved in apathy after subcortical stroke.
Although no patients were diagnosed with dementia in the present study, apathy was associated with decreased cognitive functions. Novelty P3 abnormality has been demonstrated in patients with vascular dementia associated with subcortical pathology.10 Their cognitive impairments are often associated with attentional and affective disorders, including apathy, emotional lability, and depression.23 This association is not surprising because the limbic–cortical network is implicated in memory function as well as attentional and emotional function. In addition, the present data suggest that apathy is not always serving as a mediating factor for cognitive function impairments. The present study showed differential modulations of target and novelty P3. The target P3 latency was more related to global intelligence measure impairments, whereas the novelty P3 latency was associated with apathy score. Because cognitive impairments and apathy could be dissociable in some cases,2 the present study suggests that the novelty P3 might provide useful information when therapeutic interventions are considered.
The paradigm for recording a novelty P3 does not require any behavioral output to a novel stimulus itself. This feature of automaticity is clearly suitable for assessing subjects with a lack of psychic initiative to environments. Anatomical study has limitation regarding investigation of vascular cognitive impairments because some patients show a symptom yet other patients do not, despite similar lesion location. Additional physiological index such as ERP measures could be useful for assessing functional impairments in poststroke behavioral disorders. One limitation of the present study is in the apathy assessment method. Although we modified the original questionnaire suitable for Japanese people,13 the assessment was basically performed with a self-reported style. So there was a possibility that severely apathetic patients might have been underestimated with this battery. Development of a new objective battery for assessing different types of apathy would contribute greatly to understanding the pathophysiology of poststroke apathy.
The present study provided physiological evidence that apathy after subcortical stroke is linked tightly to dysfunction of the frontal–subcortical system and suggests that recording the novelty P3 is a useful clinical tool for objective and qualitative assessments of poststroke apathy patients.
This work was supported by a grant from the Japanese Ministry of Education, Science, Sports, and Culture to S.Yamaguchi (11670626) and S.K. (10670586).
- Received January 29, 2004.
- Revision received May 5, 2004.
- Accepted May 14, 2004.
Daffner KR, Mesulam MM, Holcomb PJ, Calvo V, Acar D, Chabrerie A, Kikinis R, Jolesz FA, Rentz DM, Scinto LF. Disruption of attention to novel events after frontal lobe injury in humans. J Neurol Neurosurg Psychiatry. 2000; 68: 18–24.
Stuss DT, Reekum RV, Murphy KJ. Differentiation of states and causes of apathy. In: Borod JC, ed. The Neuropsychology of Emotion. New York, NY: Oxford University Press; 2000: 340–363.
Laplane D, Baulac M, Widlocher D, Dubois B. Pure psychic akinesia with bilateral lesions of basal ganglia. J Neurol Neurosurg Psychiatry. 1984; 47: 377–385.
Bhatia KP, Marsden CD. The behavioral and motor consequences of focal lesions of the basal ganglia in man. Brain. 1994; 117: 859–876.
Engelborghs S, Marien P, Pickut BA, Verstraeten S, De Deyn PP. Loss of psychic self-activation after paramedian bithalamic infarction. Stroke. 2000; 31: 1762–1765.
Daffner KR, Mesulam MM, Scinto LF, Acar D, Calvo V, Faust R, Chabrerie A, Kennedy B, Holcomb P. The central role of the prefrontal cortex in directing attention to novel events. Brain. 2000; 123: 927–939.
Starkstein SE, Fedoroff JP, Price TR, Leiguarda R, Robinson RG. Apathy following cerebrovascular lesions. Stroke. 1993; 24: 1625–1630.
Okada K, Kobayashi S, Aoki K, Suyama N, Yamaguchi S. Assessment of motivational loss in poststroke patients using the Japanese version of Starkstein’s apathy scale. Jpn J Stroke. 1998; 20: 318–327. In Japanese.
Okada K, Kobayashi S, Yamagata S, Takahashi K, Yamaguchi S. Poststroke apathy and regional cerebral blood flow. Stroke. 1997; 28: 2437–2441.
Kato S, Shimogaki H, Onodera A, Ueda H, Oikawa K, Ikeda K, Kosaka A, Imai Y, Hasegawa K. Development of the revised version of Hasegawa’s dementia scale (HDS-R). Jpn J Geriatr Psychiatry. 1991; 2: 1339–1347. In Japanese.
Yamaguchi S, Knight RT. Anterior and posterior association cortex contributions to the somatosensory P300. J Neurosci. 1991; 11: 2039–2054.