- Email: [email protected]
Perceptual alternation in obsessive compulsive disorder — implications for a role of the cortico-striatal circuitry in mediating awareness
Behavioural Brain Research 111 (2000) 61 – 69 www.elsevier.com/locate/bbr
Research report
Perceptual alternation in obsessive compulsive disorder — implications for a role of the cortico-striatal circuitry in mediating awareness Chiang-shan Ray Li a,b,c,* , Mon-chu Chen a,1, Yong-yi Yang a, Hsueh-ling Chang a , Chia-yih Liu a, Seng Shen a, Ching-yen Chen a a
Department of Psychiatry, Chang Gung Memorial Hospital, 5 Fu-Hsing Street, Kwei-shan, 333 Tao-yuan, Taiwan Medical Research Center, Chang Gung Memorial Hospital, 5 Fu-Hsing Street, Kwei-shan, 333 Tao-yuan, Taiwan c Department of Physiology, Chang Gung Uni6ersity, 259 Wen-Hwa First Road, Kwei-shan, 333 Tao-yuan, Taiwan
b
Received 17 September 1999; received in revised form 17 December 1999; accepted 19 December 1999
Abstract Mounting evidence suggests that obsessive compulsive disorder (OCD) results from functional aberrations of the fronto-striatal circuitry. However, empirical studies of the behavioral manifestations of OCD have been relatively lacking. The present study employs a behavioral task that allows a quantitative measure of how alternative percepts are formed from one moment to another, a process mimicking the brain state in which different thoughts and imageries compete for access to awareness. Eighteen patients with OCD, 12 with generalized anxiety disorder, and 18 normal subjects participated in the experiment, in which they viewed one of the three Schro¨der staircases and responded by pressing a key to each perceptual reversal. The results demonstrate that the patients with OCD have a higher perceptual alternation rate than the normal controls. Moreover, the frequency of perceptual alternation is significantly correlated with the Yale – Brown obsessive compulsive and the Hamilton anxiety scores. The increase in the frequency of perceptual reversals cannot easily be accounted for by learning or by different patterns of eye fixations on the task. These results provide further evidence that an impairment of the inhibitory function of the cortico-striatal circuitry might underlie the etiology of OCD. The implications of the results for a general role of the cortico-striatal circuitry in mediating awareness are discussed. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Obsessive compulsive disorder; Anxiety disorder; Perceptual alternation; Cortico-striatal; Fronto-striatal; Basal ganglia; Striatum; Gating
1. Introduction Patients with obsessive compulsive disorder (OCD) experience tremendous anxiety caused by recurrent and persistent thoughts that are deemed intrusive and inappropriate. The patients may also attempt to suppress or ‘undo’ such thoughts and impulses with other mental acts or repetitive behaviors. For example, an individual might experience recurring thoughts that he had not * Corresponding author. Tel./fax: +886-3-3274435. E-mail address: [email protected] (C.-s. Ray Li) 1 Present address: Department of Psychology, Carnegie Mellon University, Pittsburgh, PA 15213, USA.
properly locked the door when he left home for work. The ideation would force him to return to check the door repeatedly, until his anxiety resulting from the uncertainty resolved. Though a causal relationship has yet to be established, evidence has accumulated that OCD is the result of a functional deficit in the cortico-striatal circuitry. Functional imaging studies, for instance, have shown that the metabolic activities in the orbitofrontal cortex, the cingulate cortex and the caudate nucleus are significantly higher in OCD patients than in normal subjects [2,3,8,34,36,40,48,49,56] (see also [52] for a review). The elevated metabolic activities normalize in parallel with
0166-4328/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 4 3 2 8 ( 0 0 ) 0 0 1 4 0 - 6
62
C. R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
symptomatic remission when patients are treated with serotonin-specific reuptake inhibitors (SSRIs) [29,38]. Other studies have demonstrated that the activities of the orbitofrontal cortex and the basal ganglia are significantly correlated in patients with OCD, and the correlation is disrupted after treatment with SSRI or behavioral therapy [4,53]. Further evidence for abnormal cortico-striatal functions has come from patients who have acquired obsessive compulsive disorder secondary to a lesion in the frontal cortex and/or basal ganglia [9,31]. Moreover, it has been well documented that children with Sydenham’s chorea have an increased tendency to have concomitant obsessive/compulsive symptoms [55]. It has been proposed that reverberation or ‘self-driving’ of neural activities in the cortical-striatal-thalamiccortical loop leads to the obsessive compulsive symptoms through a disinhibitory mechanism [4,30]. The hypothesis is consistent with the idea that the basal ganglia are involved in gating incoming sensory and outgoing motor signals [13,51]. Results obtained from electrophysiological studies have provided evidence for this hypothesis. For example, visual and auditory evoked potentials have been shown to be decreased in latency in patients with OCD [5,17,18,47,54,61]. Results obtained in studies of rats have shown that electrical stimulation of the caudate nucleus prolongs the latency of reaching movements to acoustic and visual targets [39]. Moreover, studies in behaving primates have demonstrated that striatal neurons habituate to repeated visual stimulation, as though the irrelevant stimuli are actively ‘gated out’ in the processing of afferent information [14]. Taken together, these findings suggest that the intrusive thoughts observed in OCD might result from a deficit in the cortico-striatal circuitry in gating ‘unwanted’ neural signals. On the other hand, since the ego-dystonic thoughts and imageries represent internal states of the brain and are relatively refractory to experimental manipulation, few studies have tried to characterize the behavioral manifestations of OCD in a more systematic and quantifiable manner. It would be of particular interest to understand the neural operations that lead to the recurrent occupation of awareness by intrusive ideas in the patients with OCD. As an effort to fill this gap, the present study employed a behavioral task that mimicked the process in which different imageries, thoughts and motor sets rivalled for access to awareness. Using a Schro¨der staircase, which produced a bistable percept, we characterized the frequency of perceptual alternation and perceptual dwell time in the patients and compared the results to those obtained in normal controls and patients with Generalized Anxiety Disorder (GAD). We hypothesized that the hyperactivation of the disinhibitory processes in the cortico-striatal circuitry would facilitate the involuntary perceptual rever-
sals in patients with OCD. A preliminary report of part of the study appeared previously in abstract form [16].
2. Subjects and methods Eighteen patients diagnosed with Obsessive Compulsive Disorder (OCD, Diagnostic and Statistical Manual of Mental Disorders, 4th edn., DSM-IV, American Psychiatric Association, 1994) and 12 with Generalized Anxiety Disorder (GAD, DSM-IV) at the psychiatric outpatient service of the Chang Gung Memorial Hospital participated in the experiments. They were recruited in the study over a period of 8 months. Those with psychotic symptoms or symptoms and signs that fulfilled the criteria of Major Depressive Disorder (DSMIV) at the time of the experiment were not included in the study. Eighteen normal subjects were also recruited to participate in the experiments, most of whom were nurses, students of occupation therapy and other hospital personnel. OCD and GAD patients, and the normal subject group averaged 28, 32 and 25 years of age and 12.1, 12.7 and 14.7 years of education, respectively. All OCD patients had both obsessive and compulsive symptoms, predominantly pathological doubt, washing, checking and ordering. None of them had experienced motor tics. At the time of the experiment, the duration of illness for the OCD and GAD group of patients were from 2 to 96 months, and from 6 to 60 months, respectively. During the course of illness, the patients with OCD were treated with an antidepressant (Fluoxetine or Paroxetine 20–40 mg/day or Trazodone 100– 200 mg/day) and benzodiazepines (lorazepam 0.5–2.0 mg, alprazolam, 0.5–1.5 mg or clonazepam 0.5 –1 mg/ day). The patients with GAD were mostly treated with benzodiazepines (lorazepam 0.5–2.5 mg, alprazolam, 0.5–1.5 mg or clonazepam 0.5–2.0 mg/day). All subjects denied having major systemic illness, history of head injury or use of substance other than tobacco and alcohol on a social basis. The Yale–Brown Obsessive Compulsive Scale and Hamilton Anxiety Scale were administered to each subject prior to the experiment, to assess the severity of obsessive compulsive and anxiety symptoms [22,23,25]. Each participant was given a diagnostic interview of 30–60 min after the experiment. Informed consent was obtained from all subjects before the study was conducted. The experiment was usually carried out in a room, in which auditory noise was reduced. Visual stimuli, consisting of a red fixation target (0.2° of visual angle in diameter) and a Schro¨der staircase (approximately 3 by 5° across, Fig. 1A), were displayed on a computer monitor situated at eye level and 40 cm in front of the subject. None of the participants reported having seen the Schro¨der staircases prior to the experiment. After they recognized the Schro¨der figure as a staircase, the
C. R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
subjects were questioned about whether they sometimes felt that the staircase flipped upside down. All subjects reported seeing an inverted staircase after a variable period of time, typically within 5 min. The experiment proceeded by defining for the subjects ‘flipping’ as the perceptual experience that an upright staircase switched into an inverted one or vice versa (Fig. 1A,B). The subjects were then given 10 test trials before the experiment started, in which they would view the stimulus and respond to the perceptual flipping by pressing the space bar on the key board. Each trial started with
Fig. 1. The visual stimuli and behavioral paradigm. (A) The visual stimuli consisted of three Schro¨der figures, each of a slightly different perspective. The figure would produce alternating percept of an upright (U) and an inverted (I) staircase upon prolonged viewing. (B) The subjects pressed a key to signal alternation (x) between the two different percepts during a 30-s trial interval. The frequency of perceptual alternation and the perceptual dwell time (Td) associated with each percept could be analyzed off-line. (C) In the reaction time control experiment, two unambiguous staircases resembling the Schro¨der figure alternated in presentation at a random time interval between 1.5 and 6 s and the subjects pressed a key to signal the alternation. Each trial also lasted for 30 s. Table 1 Reaction time (control experiment) and perceptual dwell timesa
RT in control experiment (ms) Perceptual DT, upright (s) Perceptual DT, inverted (s) a
GAD
OCD
Normal
704 (93)
710 (131)
694 (248)
4.1 (3.2)
3.9 (3.5)
5.6 (4.8)
3.5 (2.8)
2.7 (2.1)
3.2 (2.4)
RT, reaction time; DT, dwell time; S.D. in parentheses.
63
the red fixation target at the center of the monitor. One of the three different Schro¨der staircases appeared and was centered on the fixation target, approximately 600 ms after the onset of fixation. The subject pressed the key to signal the perception of an upright staircase and the beginning of the trial proper. The subjects were instructed to fixate on the red target and to do their best not to blink during fixation. They were also instructed to hold onto a particular percept when one was formed, instead of making any effort to switch between percepts. Three Schro¨der staircases were used in the experiment, each containing a slightly different perspective (Fig. 1A). One of the staircases was presented for 30 s in each trial. Each staircase was repeated 10 times according to a pseudorandom schedule, totalling 30 trials in an experiment. The key presses were recorded by the computer. The frequency of perceptual alternation and the perceptual ‘dwell’ time (i.e. the time associated with each of the two percepts) were computed off-line for analysis. The last dwell time was not included in the analyses as it involved termination of the trial. The subjects were given a 30-s break every 10 trials, during which they were reminded to fixate on the red target and not to blink when the experiment resumed. In case of a response mistake, the subject clicked a ‘stop’ button on the monitor to terminate the trial, and a new trial would replace the aborted one at the end of the cycle. For a subset of subjects, we employed a video-based eye tracker (Eyelink system, SR Research, Toronto) to monitor the eye position (at 250 Hz and with a spatial resolution of 0.1°) while subjects performed on the task. The precision of fixation and the number of saccadic eye movements were evaluated off-line for analysis. A reaction time control experiment involved two staircases disambiguated from one of the Schro¨der’s, alternating in presentation at a random time interval between 1.5 and 6 s (Fig. 1C). The subjects pressed the key in response to the alternation and reaction time was computed for each key press for a total of 15 trials, each lasting for 30 s.
3. Results The data of the reaction time control experiment are summarized in Table 1. The results demonstrated that there was no significant difference in the average reaction time among the three groups (mean, OCD= 710 ms, GAD = 704 ms; normal = 694 ms; P\ 0.5, oneway analysis of variance, ANOVA). The perceptual dwell time for the upright and inverted staircase, respectively, for the three groups of subjects are also shown in Table 1. The average dwell time was longer for the upright (mean, 4.67 s) than for the inverted
64
C. R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
Fig. 2. The average numbers of perceptual alternation (y-axis) for the OCD, GAD, and normal subjects during the 30-s trial interval. The data were pooled from all three staircases for each group of subjects. (A) The average reversal rates plotted against time. Each data point represents an average across all subjects of the mean reversal rates of three successive trials in an experiment. (B) The average reversal rates for the three groups of subjects. Post-hoc analysis showed that the OCD subjects had a higher alternation rate than the normal controls (PB 0.001), but not the GAD patients (P\ 0.09). There was also no significant difference between the OCD and the GAD groups (P \ 0.2). See text for further explanation.
(mean, 3.08 s) staircase, though the difference did not reach statistical significance (P =0.09, ANOVA using Generalized Linear Models). Nor did the difference vary with the group (P \ 0.3). We next examined the frequency of perceptual alternation. The OCD, GAD and the normal groups of subjects each had an average of 9.3, 7.1 and 4.3 perceptual reversals during the 30-s time interval. Fig. 2B demonstrates the average frequency of perceptual alternation for the three groups of subjects. Analysis of variance using generalized linear models showed that the alternation rate significantly varied among the three groups (P B 0.002), but not with the three different staircases (P \0.2). Nor was there a significant interaction between the two factors (P \0.2). Post-hoc analyses showed that OCD subjects had a significantly higher alternation rate than normal controls (P B 0.001, Bonferroni test), but not the GAD subjects (P \ 0.09). There was also no significant difference between the OCD and the GAD group (P \0.2). Fig. 2A plots the frequency of reversals against the time in the experi-
ment, averaged each for the OCD, GAD and normal subjects, to examine the temporal aspect of perceptual alternation. Each data point represents the average of three successive trials, obtained from all subjects in each group. It could be seen that the frequency of reversals increased as the subject continued on the task and the temporal course of change appeared to be similar for all three groups of subjects. OCD and GAD patients and the normal controls scored an average of 21.9, 15.7 and 6.2 on the Hamilton anxiety (PB 0.001, one-way ANOVA) and 20.2, 1.7 and 0.1 on the Yale–Brown obsessive compulsive scales (PB 0.001), respectively. To examine whether the frequency of perceptual reversals is related to the severity of symptomatic manifestations, Fig. 3 plots the alternation rates against the Hamilton anxiety scores (right panels) and the Yale–Brown obsessive compulsive scores (left panels) each for the three groups of subjects. The results showed that the alternation rate was significantly correlated with both the anxiety (r 2 = 0.64, PB 0.001) and obsessive compulsive (r 2 = 0.52, PB0.001) scores for the patients with OCD, but not for the patients with GAD (P\ 0.6 for OC score; P\0.8 for anxiety score) or normal controls (P\ 0.4 for both OC and anxiety scores).
Fig. 3. The relationship between the frequency of perceptual alternation and the Yale – Brown obsessive compulsive (YBOC, left panels) and the Hamilton anxiety (HA, right panels) scores. The y-axis in each panel represents the average number of perceptual alternation in the 30-s trial interval. Note that the YBOC scores in panel A and panel B are not in the same scale as in panel C. It could be seen that the frequency of perceptual alternation is linearly correlated both with the HA and the YBOC scores for the patients with OCD, but not for the patients with GAD or the normal controls. See text for statistics and further explanation.
C.R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
65
Fig. 4. The eye traces of one patient with OCD (lower panel) and one normal subject (upper panel) during one trial on the task. Saccadic eye movements, blinks, and perceptual reversals were marked by circles, crosses, and diamonds, respectively. The x- and y-values of the eye position were displayed in the panel, each centered on zero. It could be seen that most of the eye fixations were confined within a window of 1 by 1° from the central fixation. It could also be seen that perceptual reversals did not appear to be temporally correlated with the occurrence of the saccadic eye movements, for either the OCD or the normal subject.
We next examined for a relationship between the frequency of perceptual alternation and the duration of illness at the time when the patients were tested. The results showed that the reversal rate did not appear to be linearly correlated with the duration of illness for either patient group (P \0.2, for OCD and P \ 0.4, for GAD). For the subjects (six for each group) who performed the task with the eye tracker, we analyzed their eye movements on the task. The occurrence of a saccadic eye movement was detected by preset velocity and motion criteria (velocity\30°/s and motion \0.1°). The results showed that the OCD, GAD and the normal subjects each made an average of 30.1, 24.2 and 19.3 saccades during the 30-s trial period, which significantly differed from one another (P B 0.03, one-way ANOVA). Since the frequency of perceptual reversals also differed among the three groups of subjects, we examined for a possible correlation between saccadic eye movements and perceptual reversals. The results showed that the individual alternation rate was not significantly correlated with the average number of saccadic eye movements for OCD (P \ 0.5, r 2 =0.10), GAD (P\0.2, r 2 = 0.34) or normal (P \ 0.1, r 2 =0.21) subjects. Further examination showed that perceptual alternation did not appear to bear any temporal rela-
tionship to the occurrence of saccadic eye movements. Fig. 4 shows the eye traces of one OCD and one normal control subject concurrently with the timing of perceptual alternation. The perceptual reversal occurred at times both when saccades were made and when fixations were stable for a relatively longer period of time. Similar results were obtained for the other subjects.
4. Discussion The results of the current study showed that the normal subjects had an average of 4.3 reversals during a 30-s trial period, comparable to those obtained in previous experiments involving perceptual alternation in viewing another multi-stable visual stimulus, the Necker cube [12,60]. Moreover, we replicated the finding that the frequency of perceptual reversals increased over a period of several minutes [12,60]. That the perceptual dwell time was longer for the upright staircase than for the inverted was also consistent with the results obtained in earlier experiments [27]. On the other hand, we were not able to find any difference in the frequency of perceptual alternation among the three staircases of different perspectives, as found in other
66
C.R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
studies [26,27]. The latter inconsistency perhaps resulted from the fact that the subjects in our experiment were instructed to fixate on the center target, and thus were not free to explore the stimulus to take advantage of the different perspective inherent in the stimuli. The main result obtained in the study is that patients with OCD have a higher alternation rate than the normal subjects. The results could not easily be accounted for by the effect of medications. The reasons are twofold. First, benzodiazepines have been shown to slow down visual information processing [43], a finding not commensurate with faster perceptual alternation. Second, the results obtained in the reaction time control experiment also ruled out an effect of medications on visual motor processing. Since a greater number of patients with OCD were taking SSRIs, however, further studies would be required to disambiguate an effect of the antidepressants from that of the disease process. The possibility would also have to be excluded that the patients with OCD adopted a particular visual scanning strategy to facilitate perceptual reversals. Examination of the eye traces superimposed on the staircase did not reveal any systematic pattern of eye movements for either patient group or the normal subjects. Most of the fixations centered on the red target and saccadic eye movements brought the fixation away from and back to the center target. On the other hand, subjects with OCD demonstrated a higher number of saccadic eye movements on the task, a result consistent with the finding of frequent saccadic intrusions during smooth pursuit obtained in Pallanti et al. [41]. The lack of correlation between the number of saccades and the frequency of perceptual reversals suggests that the increased rates of saccadic intrusions and perceptual reversals, rather than causally linked, perhaps are related to different aspects of the underlying etiology of OCD. However, these data were collected from a relatively small number of subjects and might not warrant any strong conclusions. Another potential confound in the experiment is that a significant component of learning is involved in perceptual alternation, as suggested by an increase of the reversal rate with trial number. The training effect, however, is unlikely to compromise the main results obtained in the study. None of the subjects had seen an ambiguous figure prior to the experiment and the average ‘learning’ curves (Fig. 2A) did not appear to differ among the three different groups of subjects. Finally, even though the subjects were instructed not to make effort to switch between percepts, it might not be impossible that the pathological doubt commonly observed in patients with OCD led to a behavioral outcome of faster perceptual alternation. Experiments employing a less arbitrary task would resolve this uncertainty. Other than an increase in reversal rate, we demonstrated that the frequency of perceptual reversal was
significantly correlated with the Yale–Brown obsessive compulsive and the Hamilton anxiety scores in patients with OCD. There were no such correlations for the patients with GAD or the normal subjects. The correlation between the reversal rate and the Yale–Brown obsessive compulsive score suggests that faster perceptual alternation reflects neural operations directly related to the underlying pathology of OCD. One might surmise that the functional aberration is perhaps more of a ‘state’ than ‘trait’ in the patients with OCD, since the reversal rate is also significantly correlated with the Hamilton anxiety score. On the other hand, although evidence exists that anxiety state shares common neural correlates with obsession and compulsion [11,20,56], patients with GAD do not show a significant increase of alternation rate (over that of the normal subjects) or a correlation thereof with anxiety score. This suggests that the altered performance in the patients with OCD does not simply result from an anxiety state or an effect of anxiety on the perceptual judgements in this behavioral task. The different results obtained with the OCD and GAD subjects suggest that, though anxiety is commonly involved in both conditions, the two disorders involve different neural mechanisms. The results are consistent with findings of a functional imaging study investigating the neural correlates of binocular rivalry [33], a process likely involving the same neural substrates as the current multistable phenomenon [32]. These authors showed that the occipitotemporal visual areas were activated during each of the two perceptual states, whereas the frontoparietal area mediated transitions of awareness between the two percepts [33]. Though the activities of the basal ganglia were not specifically examined in the latter study, it would not be inconceivable that, in relaying signals in and out of most cortical areas, this structure also played an active role in mediating perceptual alternation [13,37]. Along with the imaging experiments demonstrating functional hyperactivity of the frontal cortex and basal ganglia in OCD, the results obtained in the present study provide further evidence that the fronto-striatal circuitry plays a pivotal role in gating the access of sensory information to perceptual awareness. In general, such hyperactivity of the cortico-striatal circuitry would lead to a failure in preventing inadvertent neural signals from accessing awareness and motor output systems. Further support for this hypothesis has come from OCD patients who demonstrated a higher percentage of response suppression failures and higher error rates in an anti-saccade task, in which eye movements were made in a direction opposite to that of the target [46,59]. Swerdlow et al. demonstrated that OCD subjects were impaired in prepulse inhibition, which suggested deficient sensorimotor gating in the striatal circuitry [57]. The functional aberration of the disinhibitory circuitry has also been observed in the cogni-
C.R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
tive domain. Martinot et al. demonstrated that OCD subjects made more errors in a Stroop task, in which they were required to name the ink color of an incongruous color word [35]. Another study described a failure of OCD subjects, but not of patients with other anxiety disorders, in showing any negative priming effect in a repetition priming experiment [19]. Taken overall, the results obtained in the present along with the previous studies suggest a mechanism of how hyperactivity of the cortico-striatal circuitry might be implicated in the etiology of OCD. The predominant thoughts (obsessions about cleanliness and order, for example) reverberate through the disinhibited, ‘leaky’ basal ganglia and activate the brain area(s) in charge of moment-to-moment awareness. It is interesting to note that a similar mechanism could be postulated to account for the occurrence of motor tics in the Gilles de la Tourette Syndrome (TS), which lately has been suggested to share similar pathology to OCD [21,42]. In TS, the sporadic discharges in the motor cortical areas are not properly controlled through the striatal circuitry and result in a constant urge for movements. Similarly, ‘sensory tics,’ such as crawling and tingling sensation over a particular body part, could develop out of a similar mechanism in patients with OCD and TS [10,15]. As has been noted, the phenomenon of perseveration observed in OCD poses a paradox [30,58]. On one hand, the patients with OCD are impaired in a myriad of cognitive and behavioral tasks involving executive control and set shifting, that are poorly performed by individuals with frontal lobe damage [7,24,28,35,44] (see Refs. [1,6,50] however, for different results). On the other hand, OCD subjects were characterized in functional imaging studies to have increased activity in the frontal striatal circuitry. It is possible that both hypoand hyperactivity of the frontal cortex produce apparent set-shifting deficits, but for very different reasons [58]. The perseveration observed in subjects with frontal cortex damage presumably results from a genuine setshifting deficit, as observed in the perseverative errors in the Wisconsin Card Sorting Test. The frontal patients are not able to generate alternative strategies given changing contextual requirements. It is interesting to note that subjects with frontal cortex damage demonstrated impaired perceptual shifting in recognizing ambiguous figures without explicit external prompting [45]. On the other hand, the seemingly perseverative patterns of behaviors observed in OCD subjects perhaps to a larger extent reflect the ontogenetic predominance of certain evolutionary endowments. Common OCD themes involve violence and contamination, which for good survival reasons would entrain actions until threats and concerns about the existence of threats are surpassed. These concerns are normally kept under control and are active only in appropriate behavioral
67
contexts. In OCD, the hyperactivation of the frontostriatal circuitry leads to malfunctioning of the inhibitory mechanism, rendering it impossible to keep these concerns in check and resulting in their constant access to the structure in charge of awareness. The results obtained in the current study provide a clue to resolve this seeming discrepancy regarding perseveration observed in OCD and frontal cortical damage. It warrants further studies to investigate the neural mechanism of how the cortical striatal circuitry allows perceptual and cognitive flexibility.
Acknowledgements We thank Sho-fen Lin and Ling-yuan Kao for technical assistance and Dr Ken Grieve and two anonymous reviewers for valuable comments on the manuscript. This study was supported by a grant from the Chang Gung Memorial Hospital to Chiang-shan Ray Li.
References [1] Abbruzzese M, Ferri S, Scarone S. The selective breakdown of frontal functions in patients with obsessive-compulsive disorder and in patients with schizophrenia: A double dissociation experimental finding. Neuropsychologia 1997;35:907 – 12. [2] Baxter LR, Phelps ME, Mazziotta JC, et al. Local cerebral glucose metabolic rates in obsessive-compulsive disorder. Arch Gen Psychiatry 1987;44:211 – 8. [3] Baxter LR, Schwartz JM, Mazziotta JC, et al. Cerebral glucose metabolic rates in nondepressed patients with obsessive-compulsive disorder. Am J Psychiatry 1988;145:1560– 3. [4] Baxter LR, Schwartz JM, Bergman KS, et al. Caudate glucose metabolic rate changes with both drug and behavior therapy for obsessive-compulsive disorder. Arch Gen Psychiatry 1992;49:681 – 9. [5] Beech HR, Ciesielski KT, Gordon PK. Further observation of evoked potentials in obsessional patients. Br J Psychiatry 1983;142:605 – 9. [6] Beers SR, Rosenberg DR, Dick EL, et al. Neuropsychological study of frontal lobe function in psychotropic-naı¨ve children with obsessive-compulsive disorder. Am J Psychiatry 1999;156:777 – 9. [7] Behar D, Rapoport JL, Berg CJ, et al. Computerized tomography and neuropsychological test measures in adolescents with obsessive-compulsive disorder. Am J Psychiatry 1984;41:363–9. [8] Benkelfat C, Nordahl TE, Semple WE, et al. Local cerebral glucose metabolic rates in obsessive-compulsive disorder. Arch Gen Psychiatry 1990;47:840 – 8. [9] Berthier ML, Kulisevsky J, Gironell A, Heras JA. Obsessivecompulsive disorder associated with brain lesions: Clinical phenomenology, cognitive function, and anatomic correlates. Neurology 1996;47:353 – 61. [10] Bliss J, Cohen DJ, Freedman DX, et al. Sensory experiences of Gilles de la Tourette syndrome. Arch Gen Psychiatry 1980;37:1343 – 7. [11] Breiter HC, Rauch SL, Kwong KK, et al. Functional magnetic resonance imaging of symptom provocation in obsessive-compulsive disorder. Arch Gen Psychiatry 1996;53:595 – 606.
68
C. R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
[12] Brown KT. Rate of apparent change in a dynamic ambiguous figure as a function of observation-time. Percept Motor Skills 1954;67:358 – 71. [13] Brown LL, Schneider JS, Lidsky TI. Sensory and cognitive functions of the basal ganglia. Curr Opin Neurobiol 1997;7:157 – 63. [14] Caan W, Perrett DI, Rolls ET. Responses of striatal neurons in the behaving monkey. 2. Visual processing in the caudal neostriatum. Brain Res 1984;290:53–65. [15] Chee KY, Sachdev PA. A controlled study of sensory tics in Gilles de la Tourette syndrome and obsessive-compulsive disorder using a structured interview. J Neurol Neurosurg Psychiatry 1997;62:188 – 92. [16] Chen M, Li C-SR, Yang Y-Y, et al. Perceptual alternation in obsessive compulsive disorder — Implications for the functions of the frontostriatal circuitry (Abstract). 3rd International Conference on Cognitive and Neural Systems, Boston, MA, May 1999. [17] Ciesielski KT, Beech HR, Gordon PK. Some electrophysiological observations in obsessional states. Br J Psychiatry 1981;138:479 – 84. [18] Drake ME, Hietter SA, Padamadan H, et al. Auditory evoked potentials in Gilles de la Tourette syndrome. Clin Electrophysiol 1992;23:19 – 23. [19] Enright SJ, Beech AR. Reduced cognitive inhibition in obsessivecompulsive disorder. Br J Clin Psychol 1993;32:67–74. [20] Fredrikson M, Fischer H, Wik G. Cerebral blood flow during anxiety provocation. J Clin Psychiatry 1997;58(Suppl. 16):16 – 21. [21] Goldsmith T, Shapira MA, Phillips KA, McElroy SL. Conceptual foundations of obsessive-compulsive spectrum disorders. In: Swinson RP, Antony MM, Rachman S, Richter MA, editors. Obsessive-Compulsive Disorder, Theory, Research and Treatment. New York: Guilford, 1998:397–425. [22] Goodman WK, Price LH, Rasmussen SA, et al. The Yale – Brown obsessive compulsive scale. I. Development, use, and reliability. Arch Gen Psychiatry 1989;46:1006–11. [23] Goodman WK, Price LH, Rasmussen SA, et al. The Yale – Brown obsessive compulsive scale. II. Validity. Arch Gen Psychiatry 1989;46:1012 – 6. [24] Gross-Isseroff R, Sasson Y, Voet H, et al. Alternation learning in obsessive-compulsive disorder. Biol Psychiatry 1996;39:733 – 8. [25] Hamilton M. The assessment of anxiety states by rating. Br J Med Psychol 1959;32:50–5. [26] Harris JP. The Schro¨der staircase: A new perspective. Percept Psychophys 1979;26:312–8. [27] Harris JP, Phillipson OT. Chlorpromazine reduces the perceptual ambiguity of a reversible visual figure. Neuropharmacology 1981;20:1337 – 8. [28] Head EK, Bolton D, Hymas N. Deficit in cognitive shifting ability in patients with obsessive compulsive disorders. Biol Psychiatry 1989;25:929–37. [29] Hoehn-Saric R, Pearlson GD, Harris GJ, et al. Effects of fluoxetine on regional cerebral blood flow in obsessive-compulsive patients. Am J Psychiatry 1991;148:1243–5. [30] Insel TR. Toward a neuroanatomy of obsessive-compulsive disorder. Arch Gen Psychiatry 1992;49:739–44. [31] Laplane D, Levasseur M, Pillon B, et al. Obsessive-compulsive and other behavioral changes with bilateral basal ganglia lesions. Brain 1989;112:699 –725. [32] Logothetis NK, Leopold DA, Sheinberg DL. What is rivaling during binocular rivalry? Nature 1996;380:621–4. [33] Lumer ED, Friston KJ, Rees G. Neural correlates of perceptual rivalry in the human brain. Science (Wash DC) 1998;280:1930– 4. [34] Machlin A, Harris GJ, Pearlson GD, et al. Elevated medial-frontal cerebral block flow in obsessive-compulsive patients. A SPECT study. Am J Psychiatry 1991;148:1240–2.
[35] Martinot JL, Allilaire JF, Mazoyer BM, et al. Obsessive-compulsive disorder: a clinical, neuropsychological and positron emission tomography study. Acta Psychiatr Scand 1990;82:233–42. [36] McGuire PK, Bench CJ, Frith CD, et al. Functional anatomy of obsessive-compulsive phenomena. Br J Psychiatry 1994;164:459– 68. [37] Middleton FA, Strick PL. The temporal lobe is a target of output from the basal ganglia. Proc Natl Acad Sci USA 1996;93:8683 – 7. [38] Molina V, Montz R, Perez-Castejon MJ, et al. Cerebral perfusion, electrical activity and effects of serotonergic treatment in Obsessive-Compulsive disorder. A preliminary study. Neuropsychobiology 1995;32:139 – 48. [39] Moroz VM, Bures J. Caudate stimulation prolongs latency of acoustically and visually signalled reaching in rats. Brain Res 1983;259:298 – 300. [40] Nordahl TE, Benkelfat C, Semple WE, et al. Cerebral glucose metabolic rates in obsessive compulsive disorder. Neuropsychopharmacology 1989;2:23 – 8. [41] Pallanti S, Grecu LM, Gangemi PF, et al. Smooth-pursuit eye movement and saccadic intrusions in obsessive-compulsive disorder. Biol Psychiatry 1996;40:1164 – 72. [42] Palumbo D, Maughan A, Kurlan R. Tourette syndrome is only one of several causes of a developmental basal ganglia syndrome. Arch Neurol 1997;54:475 – 83. [43] Pang E, Fowler B. Discriminating the effects of triazolam on stimulus and response processing by means of reaction time and P300 latency. Psychopharmacology 1994;115:509 – 15. [44] Purcell R, Maruff P, Kyrios M, Pantelis C. Cognitive deficits in obsessive-compulsive disorder on tests of frontal-striatal function. Biol Psychiatry 1998;43:348 – 57. [45] Ricci C, Blundo C. Perception of ambiguous figures after focal brain lesions. Neuropsychologia 1990;28:1163 – 73. [46] Rosenberg DR, Averbach DH, O’Hearn KM. Oculomotor response inhibition abnormalities in pediatric obsessive-compulsive disorder. Arch Gen Psychiatry 1997;54:831 – 8. [47] Savage CR, Weilburg JB, Duffy FH, et al. Low-level sensory processing in obsessive-compulsive disorder: an evoked potential study. Biol Psychiatry 1994;35:247 – 52. [48] Sawle GV, Hymas NF, Lees AJ, Frackowiak RSJ. Obsessional slowness. Brain 1991;114:2191 – 202. [49] Saxena S, Brody AL, Schwartz JM, Baxter LR. Neuroimaging and frontal-subcortical circuitry in obsessive-compulsive disorder. Br J Psychiatry 1998;173(Suppl. 35):26 – 37. [50] Schmidtke K, Schorb A, Winkelmann G, Hohagen F. Cognitive frontal lobe dysfunction in obsessive-compulsive disorder. Biol Psychiatry 1998;43:666 – 73. [51] Schneider JS. Basal ganglia role in behavior: Importance of sensory gating and its relevance to psychiatry. Biol Psychiatry 1984;19:1693 – 710. [52] Schwartz JM. Neuroanatomical aspects of cognitive-behavioural therapy response in obsessive-compulsive disorder. Br J Psychiatry 1998;173(Suppl. 35):38 – 44. [53] Schwartz JM, Stoessel PW, Baxter LR Jr, et al. Systematic changes in cerebral glucose metabolic rate after successful behavior modification treatment of obsessive-compulsive disorder. Arch Gen Psychiatry 1996;53:109 – 13. [54] Shagass C, Roemer RA, Straumanis JJ, Josiassen RC. Evoked potentials in obsessive-compulsive disorder. Adv Biol Psychiatry 1984;15:69 – 75. [55] Swedo SE, Leonard HL, Schapiro MB, et al. Sydenham’s chorea: Physical and psychological symptoms of St Vitus dance. Pediatrics 1993;91:706 – 13. [56] Swedo SE, Schapiro MB, Grady CL, et al. Cerebral glucose metabolism in childhood-onset obsessive-compulsive disorder. Arch Gen Psychiatry 1989;46:518 – 23.
C.R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69 [57] Swerdlow NR, Benbow CH, Zisook S, et al. A preliminary assessment of sensorimotor gating in patients with obsessive compulsive disorder. Biol Psychiatry 1993;33:298–301. [58] Tallis F. The neuropsychology of obsessive-compulsive disorder: A review and consideration of clinical implications. Br J Clin Psychol 1997;36:3 – 20. [59] Tien AY, Pearlson GD, Machlin SR. Oculomotor performance in
.
69
obsessive-compulsive disorder. Am J Psychiatry 1992;149:641–6. [60] Toppino TC, Long GM. Selective adaptation with reversible figures: don’t change that channel. Percept Psychophys 1987;42:37 – 48. [61] Towey J, Bruder G, Hollander ET, et al. Endogenous event-related potentials in obsessive-compulsive disorder. Biol Psychiatry 1990;28:92 – 8.
Research report
Perceptual alternation in obsessive compulsive disorder — implications for a role of the cortico-striatal circuitry in mediating awareness Chiang-shan Ray Li a,b,c,* , Mon-chu Chen a,1, Yong-yi Yang a, Hsueh-ling Chang a , Chia-yih Liu a, Seng Shen a, Ching-yen Chen a a
Department of Psychiatry, Chang Gung Memorial Hospital, 5 Fu-Hsing Street, Kwei-shan, 333 Tao-yuan, Taiwan Medical Research Center, Chang Gung Memorial Hospital, 5 Fu-Hsing Street, Kwei-shan, 333 Tao-yuan, Taiwan c Department of Physiology, Chang Gung Uni6ersity, 259 Wen-Hwa First Road, Kwei-shan, 333 Tao-yuan, Taiwan
b
Received 17 September 1999; received in revised form 17 December 1999; accepted 19 December 1999
Abstract Mounting evidence suggests that obsessive compulsive disorder (OCD) results from functional aberrations of the fronto-striatal circuitry. However, empirical studies of the behavioral manifestations of OCD have been relatively lacking. The present study employs a behavioral task that allows a quantitative measure of how alternative percepts are formed from one moment to another, a process mimicking the brain state in which different thoughts and imageries compete for access to awareness. Eighteen patients with OCD, 12 with generalized anxiety disorder, and 18 normal subjects participated in the experiment, in which they viewed one of the three Schro¨der staircases and responded by pressing a key to each perceptual reversal. The results demonstrate that the patients with OCD have a higher perceptual alternation rate than the normal controls. Moreover, the frequency of perceptual alternation is significantly correlated with the Yale – Brown obsessive compulsive and the Hamilton anxiety scores. The increase in the frequency of perceptual reversals cannot easily be accounted for by learning or by different patterns of eye fixations on the task. These results provide further evidence that an impairment of the inhibitory function of the cortico-striatal circuitry might underlie the etiology of OCD. The implications of the results for a general role of the cortico-striatal circuitry in mediating awareness are discussed. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Obsessive compulsive disorder; Anxiety disorder; Perceptual alternation; Cortico-striatal; Fronto-striatal; Basal ganglia; Striatum; Gating
1. Introduction Patients with obsessive compulsive disorder (OCD) experience tremendous anxiety caused by recurrent and persistent thoughts that are deemed intrusive and inappropriate. The patients may also attempt to suppress or ‘undo’ such thoughts and impulses with other mental acts or repetitive behaviors. For example, an individual might experience recurring thoughts that he had not * Corresponding author. Tel./fax: +886-3-3274435. E-mail address: [email protected] (C.-s. Ray Li) 1 Present address: Department of Psychology, Carnegie Mellon University, Pittsburgh, PA 15213, USA.
properly locked the door when he left home for work. The ideation would force him to return to check the door repeatedly, until his anxiety resulting from the uncertainty resolved. Though a causal relationship has yet to be established, evidence has accumulated that OCD is the result of a functional deficit in the cortico-striatal circuitry. Functional imaging studies, for instance, have shown that the metabolic activities in the orbitofrontal cortex, the cingulate cortex and the caudate nucleus are significantly higher in OCD patients than in normal subjects [2,3,8,34,36,40,48,49,56] (see also [52] for a review). The elevated metabolic activities normalize in parallel with
0166-4328/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 4 3 2 8 ( 0 0 ) 0 0 1 4 0 - 6
62
C. R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
symptomatic remission when patients are treated with serotonin-specific reuptake inhibitors (SSRIs) [29,38]. Other studies have demonstrated that the activities of the orbitofrontal cortex and the basal ganglia are significantly correlated in patients with OCD, and the correlation is disrupted after treatment with SSRI or behavioral therapy [4,53]. Further evidence for abnormal cortico-striatal functions has come from patients who have acquired obsessive compulsive disorder secondary to a lesion in the frontal cortex and/or basal ganglia [9,31]. Moreover, it has been well documented that children with Sydenham’s chorea have an increased tendency to have concomitant obsessive/compulsive symptoms [55]. It has been proposed that reverberation or ‘self-driving’ of neural activities in the cortical-striatal-thalamiccortical loop leads to the obsessive compulsive symptoms through a disinhibitory mechanism [4,30]. The hypothesis is consistent with the idea that the basal ganglia are involved in gating incoming sensory and outgoing motor signals [13,51]. Results obtained from electrophysiological studies have provided evidence for this hypothesis. For example, visual and auditory evoked potentials have been shown to be decreased in latency in patients with OCD [5,17,18,47,54,61]. Results obtained in studies of rats have shown that electrical stimulation of the caudate nucleus prolongs the latency of reaching movements to acoustic and visual targets [39]. Moreover, studies in behaving primates have demonstrated that striatal neurons habituate to repeated visual stimulation, as though the irrelevant stimuli are actively ‘gated out’ in the processing of afferent information [14]. Taken together, these findings suggest that the intrusive thoughts observed in OCD might result from a deficit in the cortico-striatal circuitry in gating ‘unwanted’ neural signals. On the other hand, since the ego-dystonic thoughts and imageries represent internal states of the brain and are relatively refractory to experimental manipulation, few studies have tried to characterize the behavioral manifestations of OCD in a more systematic and quantifiable manner. It would be of particular interest to understand the neural operations that lead to the recurrent occupation of awareness by intrusive ideas in the patients with OCD. As an effort to fill this gap, the present study employed a behavioral task that mimicked the process in which different imageries, thoughts and motor sets rivalled for access to awareness. Using a Schro¨der staircase, which produced a bistable percept, we characterized the frequency of perceptual alternation and perceptual dwell time in the patients and compared the results to those obtained in normal controls and patients with Generalized Anxiety Disorder (GAD). We hypothesized that the hyperactivation of the disinhibitory processes in the cortico-striatal circuitry would facilitate the involuntary perceptual rever-
sals in patients with OCD. A preliminary report of part of the study appeared previously in abstract form [16].
2. Subjects and methods Eighteen patients diagnosed with Obsessive Compulsive Disorder (OCD, Diagnostic and Statistical Manual of Mental Disorders, 4th edn., DSM-IV, American Psychiatric Association, 1994) and 12 with Generalized Anxiety Disorder (GAD, DSM-IV) at the psychiatric outpatient service of the Chang Gung Memorial Hospital participated in the experiments. They were recruited in the study over a period of 8 months. Those with psychotic symptoms or symptoms and signs that fulfilled the criteria of Major Depressive Disorder (DSMIV) at the time of the experiment were not included in the study. Eighteen normal subjects were also recruited to participate in the experiments, most of whom were nurses, students of occupation therapy and other hospital personnel. OCD and GAD patients, and the normal subject group averaged 28, 32 and 25 years of age and 12.1, 12.7 and 14.7 years of education, respectively. All OCD patients had both obsessive and compulsive symptoms, predominantly pathological doubt, washing, checking and ordering. None of them had experienced motor tics. At the time of the experiment, the duration of illness for the OCD and GAD group of patients were from 2 to 96 months, and from 6 to 60 months, respectively. During the course of illness, the patients with OCD were treated with an antidepressant (Fluoxetine or Paroxetine 20–40 mg/day or Trazodone 100– 200 mg/day) and benzodiazepines (lorazepam 0.5–2.0 mg, alprazolam, 0.5–1.5 mg or clonazepam 0.5 –1 mg/ day). The patients with GAD were mostly treated with benzodiazepines (lorazepam 0.5–2.5 mg, alprazolam, 0.5–1.5 mg or clonazepam 0.5–2.0 mg/day). All subjects denied having major systemic illness, history of head injury or use of substance other than tobacco and alcohol on a social basis. The Yale–Brown Obsessive Compulsive Scale and Hamilton Anxiety Scale were administered to each subject prior to the experiment, to assess the severity of obsessive compulsive and anxiety symptoms [22,23,25]. Each participant was given a diagnostic interview of 30–60 min after the experiment. Informed consent was obtained from all subjects before the study was conducted. The experiment was usually carried out in a room, in which auditory noise was reduced. Visual stimuli, consisting of a red fixation target (0.2° of visual angle in diameter) and a Schro¨der staircase (approximately 3 by 5° across, Fig. 1A), were displayed on a computer monitor situated at eye level and 40 cm in front of the subject. None of the participants reported having seen the Schro¨der staircases prior to the experiment. After they recognized the Schro¨der figure as a staircase, the
C. R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
subjects were questioned about whether they sometimes felt that the staircase flipped upside down. All subjects reported seeing an inverted staircase after a variable period of time, typically within 5 min. The experiment proceeded by defining for the subjects ‘flipping’ as the perceptual experience that an upright staircase switched into an inverted one or vice versa (Fig. 1A,B). The subjects were then given 10 test trials before the experiment started, in which they would view the stimulus and respond to the perceptual flipping by pressing the space bar on the key board. Each trial started with
Fig. 1. The visual stimuli and behavioral paradigm. (A) The visual stimuli consisted of three Schro¨der figures, each of a slightly different perspective. The figure would produce alternating percept of an upright (U) and an inverted (I) staircase upon prolonged viewing. (B) The subjects pressed a key to signal alternation (x) between the two different percepts during a 30-s trial interval. The frequency of perceptual alternation and the perceptual dwell time (Td) associated with each percept could be analyzed off-line. (C) In the reaction time control experiment, two unambiguous staircases resembling the Schro¨der figure alternated in presentation at a random time interval between 1.5 and 6 s and the subjects pressed a key to signal the alternation. Each trial also lasted for 30 s. Table 1 Reaction time (control experiment) and perceptual dwell timesa
RT in control experiment (ms) Perceptual DT, upright (s) Perceptual DT, inverted (s) a
GAD
OCD
Normal
704 (93)
710 (131)
694 (248)
4.1 (3.2)
3.9 (3.5)
5.6 (4.8)
3.5 (2.8)
2.7 (2.1)
3.2 (2.4)
RT, reaction time; DT, dwell time; S.D. in parentheses.
63
the red fixation target at the center of the monitor. One of the three different Schro¨der staircases appeared and was centered on the fixation target, approximately 600 ms after the onset of fixation. The subject pressed the key to signal the perception of an upright staircase and the beginning of the trial proper. The subjects were instructed to fixate on the red target and to do their best not to blink during fixation. They were also instructed to hold onto a particular percept when one was formed, instead of making any effort to switch between percepts. Three Schro¨der staircases were used in the experiment, each containing a slightly different perspective (Fig. 1A). One of the staircases was presented for 30 s in each trial. Each staircase was repeated 10 times according to a pseudorandom schedule, totalling 30 trials in an experiment. The key presses were recorded by the computer. The frequency of perceptual alternation and the perceptual ‘dwell’ time (i.e. the time associated with each of the two percepts) were computed off-line for analysis. The last dwell time was not included in the analyses as it involved termination of the trial. The subjects were given a 30-s break every 10 trials, during which they were reminded to fixate on the red target and not to blink when the experiment resumed. In case of a response mistake, the subject clicked a ‘stop’ button on the monitor to terminate the trial, and a new trial would replace the aborted one at the end of the cycle. For a subset of subjects, we employed a video-based eye tracker (Eyelink system, SR Research, Toronto) to monitor the eye position (at 250 Hz and with a spatial resolution of 0.1°) while subjects performed on the task. The precision of fixation and the number of saccadic eye movements were evaluated off-line for analysis. A reaction time control experiment involved two staircases disambiguated from one of the Schro¨der’s, alternating in presentation at a random time interval between 1.5 and 6 s (Fig. 1C). The subjects pressed the key in response to the alternation and reaction time was computed for each key press for a total of 15 trials, each lasting for 30 s.
3. Results The data of the reaction time control experiment are summarized in Table 1. The results demonstrated that there was no significant difference in the average reaction time among the three groups (mean, OCD= 710 ms, GAD = 704 ms; normal = 694 ms; P\ 0.5, oneway analysis of variance, ANOVA). The perceptual dwell time for the upright and inverted staircase, respectively, for the three groups of subjects are also shown in Table 1. The average dwell time was longer for the upright (mean, 4.67 s) than for the inverted
64
C. R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
Fig. 2. The average numbers of perceptual alternation (y-axis) for the OCD, GAD, and normal subjects during the 30-s trial interval. The data were pooled from all three staircases for each group of subjects. (A) The average reversal rates plotted against time. Each data point represents an average across all subjects of the mean reversal rates of three successive trials in an experiment. (B) The average reversal rates for the three groups of subjects. Post-hoc analysis showed that the OCD subjects had a higher alternation rate than the normal controls (PB 0.001), but not the GAD patients (P\ 0.09). There was also no significant difference between the OCD and the GAD groups (P \ 0.2). See text for further explanation.
(mean, 3.08 s) staircase, though the difference did not reach statistical significance (P =0.09, ANOVA using Generalized Linear Models). Nor did the difference vary with the group (P \ 0.3). We next examined the frequency of perceptual alternation. The OCD, GAD and the normal groups of subjects each had an average of 9.3, 7.1 and 4.3 perceptual reversals during the 30-s time interval. Fig. 2B demonstrates the average frequency of perceptual alternation for the three groups of subjects. Analysis of variance using generalized linear models showed that the alternation rate significantly varied among the three groups (P B 0.002), but not with the three different staircases (P \0.2). Nor was there a significant interaction between the two factors (P \0.2). Post-hoc analyses showed that OCD subjects had a significantly higher alternation rate than normal controls (P B 0.001, Bonferroni test), but not the GAD subjects (P \ 0.09). There was also no significant difference between the OCD and the GAD group (P \0.2). Fig. 2A plots the frequency of reversals against the time in the experi-
ment, averaged each for the OCD, GAD and normal subjects, to examine the temporal aspect of perceptual alternation. Each data point represents the average of three successive trials, obtained from all subjects in each group. It could be seen that the frequency of reversals increased as the subject continued on the task and the temporal course of change appeared to be similar for all three groups of subjects. OCD and GAD patients and the normal controls scored an average of 21.9, 15.7 and 6.2 on the Hamilton anxiety (PB 0.001, one-way ANOVA) and 20.2, 1.7 and 0.1 on the Yale–Brown obsessive compulsive scales (PB 0.001), respectively. To examine whether the frequency of perceptual reversals is related to the severity of symptomatic manifestations, Fig. 3 plots the alternation rates against the Hamilton anxiety scores (right panels) and the Yale–Brown obsessive compulsive scores (left panels) each for the three groups of subjects. The results showed that the alternation rate was significantly correlated with both the anxiety (r 2 = 0.64, PB 0.001) and obsessive compulsive (r 2 = 0.52, PB0.001) scores for the patients with OCD, but not for the patients with GAD (P\ 0.6 for OC score; P\0.8 for anxiety score) or normal controls (P\ 0.4 for both OC and anxiety scores).
Fig. 3. The relationship between the frequency of perceptual alternation and the Yale – Brown obsessive compulsive (YBOC, left panels) and the Hamilton anxiety (HA, right panels) scores. The y-axis in each panel represents the average number of perceptual alternation in the 30-s trial interval. Note that the YBOC scores in panel A and panel B are not in the same scale as in panel C. It could be seen that the frequency of perceptual alternation is linearly correlated both with the HA and the YBOC scores for the patients with OCD, but not for the patients with GAD or the normal controls. See text for statistics and further explanation.
C.R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
65
Fig. 4. The eye traces of one patient with OCD (lower panel) and one normal subject (upper panel) during one trial on the task. Saccadic eye movements, blinks, and perceptual reversals were marked by circles, crosses, and diamonds, respectively. The x- and y-values of the eye position were displayed in the panel, each centered on zero. It could be seen that most of the eye fixations were confined within a window of 1 by 1° from the central fixation. It could also be seen that perceptual reversals did not appear to be temporally correlated with the occurrence of the saccadic eye movements, for either the OCD or the normal subject.
We next examined for a relationship between the frequency of perceptual alternation and the duration of illness at the time when the patients were tested. The results showed that the reversal rate did not appear to be linearly correlated with the duration of illness for either patient group (P \0.2, for OCD and P \ 0.4, for GAD). For the subjects (six for each group) who performed the task with the eye tracker, we analyzed their eye movements on the task. The occurrence of a saccadic eye movement was detected by preset velocity and motion criteria (velocity\30°/s and motion \0.1°). The results showed that the OCD, GAD and the normal subjects each made an average of 30.1, 24.2 and 19.3 saccades during the 30-s trial period, which significantly differed from one another (P B 0.03, one-way ANOVA). Since the frequency of perceptual reversals also differed among the three groups of subjects, we examined for a possible correlation between saccadic eye movements and perceptual reversals. The results showed that the individual alternation rate was not significantly correlated with the average number of saccadic eye movements for OCD (P \ 0.5, r 2 =0.10), GAD (P\0.2, r 2 = 0.34) or normal (P \ 0.1, r 2 =0.21) subjects. Further examination showed that perceptual alternation did not appear to bear any temporal rela-
tionship to the occurrence of saccadic eye movements. Fig. 4 shows the eye traces of one OCD and one normal control subject concurrently with the timing of perceptual alternation. The perceptual reversal occurred at times both when saccades were made and when fixations were stable for a relatively longer period of time. Similar results were obtained for the other subjects.
4. Discussion The results of the current study showed that the normal subjects had an average of 4.3 reversals during a 30-s trial period, comparable to those obtained in previous experiments involving perceptual alternation in viewing another multi-stable visual stimulus, the Necker cube [12,60]. Moreover, we replicated the finding that the frequency of perceptual reversals increased over a period of several minutes [12,60]. That the perceptual dwell time was longer for the upright staircase than for the inverted was also consistent with the results obtained in earlier experiments [27]. On the other hand, we were not able to find any difference in the frequency of perceptual alternation among the three staircases of different perspectives, as found in other
66
C.R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
studies [26,27]. The latter inconsistency perhaps resulted from the fact that the subjects in our experiment were instructed to fixate on the center target, and thus were not free to explore the stimulus to take advantage of the different perspective inherent in the stimuli. The main result obtained in the study is that patients with OCD have a higher alternation rate than the normal subjects. The results could not easily be accounted for by the effect of medications. The reasons are twofold. First, benzodiazepines have been shown to slow down visual information processing [43], a finding not commensurate with faster perceptual alternation. Second, the results obtained in the reaction time control experiment also ruled out an effect of medications on visual motor processing. Since a greater number of patients with OCD were taking SSRIs, however, further studies would be required to disambiguate an effect of the antidepressants from that of the disease process. The possibility would also have to be excluded that the patients with OCD adopted a particular visual scanning strategy to facilitate perceptual reversals. Examination of the eye traces superimposed on the staircase did not reveal any systematic pattern of eye movements for either patient group or the normal subjects. Most of the fixations centered on the red target and saccadic eye movements brought the fixation away from and back to the center target. On the other hand, subjects with OCD demonstrated a higher number of saccadic eye movements on the task, a result consistent with the finding of frequent saccadic intrusions during smooth pursuit obtained in Pallanti et al. [41]. The lack of correlation between the number of saccades and the frequency of perceptual reversals suggests that the increased rates of saccadic intrusions and perceptual reversals, rather than causally linked, perhaps are related to different aspects of the underlying etiology of OCD. However, these data were collected from a relatively small number of subjects and might not warrant any strong conclusions. Another potential confound in the experiment is that a significant component of learning is involved in perceptual alternation, as suggested by an increase of the reversal rate with trial number. The training effect, however, is unlikely to compromise the main results obtained in the study. None of the subjects had seen an ambiguous figure prior to the experiment and the average ‘learning’ curves (Fig. 2A) did not appear to differ among the three different groups of subjects. Finally, even though the subjects were instructed not to make effort to switch between percepts, it might not be impossible that the pathological doubt commonly observed in patients with OCD led to a behavioral outcome of faster perceptual alternation. Experiments employing a less arbitrary task would resolve this uncertainty. Other than an increase in reversal rate, we demonstrated that the frequency of perceptual reversal was
significantly correlated with the Yale–Brown obsessive compulsive and the Hamilton anxiety scores in patients with OCD. There were no such correlations for the patients with GAD or the normal subjects. The correlation between the reversal rate and the Yale–Brown obsessive compulsive score suggests that faster perceptual alternation reflects neural operations directly related to the underlying pathology of OCD. One might surmise that the functional aberration is perhaps more of a ‘state’ than ‘trait’ in the patients with OCD, since the reversal rate is also significantly correlated with the Hamilton anxiety score. On the other hand, although evidence exists that anxiety state shares common neural correlates with obsession and compulsion [11,20,56], patients with GAD do not show a significant increase of alternation rate (over that of the normal subjects) or a correlation thereof with anxiety score. This suggests that the altered performance in the patients with OCD does not simply result from an anxiety state or an effect of anxiety on the perceptual judgements in this behavioral task. The different results obtained with the OCD and GAD subjects suggest that, though anxiety is commonly involved in both conditions, the two disorders involve different neural mechanisms. The results are consistent with findings of a functional imaging study investigating the neural correlates of binocular rivalry [33], a process likely involving the same neural substrates as the current multistable phenomenon [32]. These authors showed that the occipitotemporal visual areas were activated during each of the two perceptual states, whereas the frontoparietal area mediated transitions of awareness between the two percepts [33]. Though the activities of the basal ganglia were not specifically examined in the latter study, it would not be inconceivable that, in relaying signals in and out of most cortical areas, this structure also played an active role in mediating perceptual alternation [13,37]. Along with the imaging experiments demonstrating functional hyperactivity of the frontal cortex and basal ganglia in OCD, the results obtained in the present study provide further evidence that the fronto-striatal circuitry plays a pivotal role in gating the access of sensory information to perceptual awareness. In general, such hyperactivity of the cortico-striatal circuitry would lead to a failure in preventing inadvertent neural signals from accessing awareness and motor output systems. Further support for this hypothesis has come from OCD patients who demonstrated a higher percentage of response suppression failures and higher error rates in an anti-saccade task, in which eye movements were made in a direction opposite to that of the target [46,59]. Swerdlow et al. demonstrated that OCD subjects were impaired in prepulse inhibition, which suggested deficient sensorimotor gating in the striatal circuitry [57]. The functional aberration of the disinhibitory circuitry has also been observed in the cogni-
C.R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
tive domain. Martinot et al. demonstrated that OCD subjects made more errors in a Stroop task, in which they were required to name the ink color of an incongruous color word [35]. Another study described a failure of OCD subjects, but not of patients with other anxiety disorders, in showing any negative priming effect in a repetition priming experiment [19]. Taken overall, the results obtained in the present along with the previous studies suggest a mechanism of how hyperactivity of the cortico-striatal circuitry might be implicated in the etiology of OCD. The predominant thoughts (obsessions about cleanliness and order, for example) reverberate through the disinhibited, ‘leaky’ basal ganglia and activate the brain area(s) in charge of moment-to-moment awareness. It is interesting to note that a similar mechanism could be postulated to account for the occurrence of motor tics in the Gilles de la Tourette Syndrome (TS), which lately has been suggested to share similar pathology to OCD [21,42]. In TS, the sporadic discharges in the motor cortical areas are not properly controlled through the striatal circuitry and result in a constant urge for movements. Similarly, ‘sensory tics,’ such as crawling and tingling sensation over a particular body part, could develop out of a similar mechanism in patients with OCD and TS [10,15]. As has been noted, the phenomenon of perseveration observed in OCD poses a paradox [30,58]. On one hand, the patients with OCD are impaired in a myriad of cognitive and behavioral tasks involving executive control and set shifting, that are poorly performed by individuals with frontal lobe damage [7,24,28,35,44] (see Refs. [1,6,50] however, for different results). On the other hand, OCD subjects were characterized in functional imaging studies to have increased activity in the frontal striatal circuitry. It is possible that both hypoand hyperactivity of the frontal cortex produce apparent set-shifting deficits, but for very different reasons [58]. The perseveration observed in subjects with frontal cortex damage presumably results from a genuine setshifting deficit, as observed in the perseverative errors in the Wisconsin Card Sorting Test. The frontal patients are not able to generate alternative strategies given changing contextual requirements. It is interesting to note that subjects with frontal cortex damage demonstrated impaired perceptual shifting in recognizing ambiguous figures without explicit external prompting [45]. On the other hand, the seemingly perseverative patterns of behaviors observed in OCD subjects perhaps to a larger extent reflect the ontogenetic predominance of certain evolutionary endowments. Common OCD themes involve violence and contamination, which for good survival reasons would entrain actions until threats and concerns about the existence of threats are surpassed. These concerns are normally kept under control and are active only in appropriate behavioral
67
contexts. In OCD, the hyperactivation of the frontostriatal circuitry leads to malfunctioning of the inhibitory mechanism, rendering it impossible to keep these concerns in check and resulting in their constant access to the structure in charge of awareness. The results obtained in the current study provide a clue to resolve this seeming discrepancy regarding perseveration observed in OCD and frontal cortical damage. It warrants further studies to investigate the neural mechanism of how the cortical striatal circuitry allows perceptual and cognitive flexibility.
Acknowledgements We thank Sho-fen Lin and Ling-yuan Kao for technical assistance and Dr Ken Grieve and two anonymous reviewers for valuable comments on the manuscript. This study was supported by a grant from the Chang Gung Memorial Hospital to Chiang-shan Ray Li.
References [1] Abbruzzese M, Ferri S, Scarone S. The selective breakdown of frontal functions in patients with obsessive-compulsive disorder and in patients with schizophrenia: A double dissociation experimental finding. Neuropsychologia 1997;35:907 – 12. [2] Baxter LR, Phelps ME, Mazziotta JC, et al. Local cerebral glucose metabolic rates in obsessive-compulsive disorder. Arch Gen Psychiatry 1987;44:211 – 8. [3] Baxter LR, Schwartz JM, Mazziotta JC, et al. Cerebral glucose metabolic rates in nondepressed patients with obsessive-compulsive disorder. Am J Psychiatry 1988;145:1560– 3. [4] Baxter LR, Schwartz JM, Bergman KS, et al. Caudate glucose metabolic rate changes with both drug and behavior therapy for obsessive-compulsive disorder. Arch Gen Psychiatry 1992;49:681 – 9. [5] Beech HR, Ciesielski KT, Gordon PK. Further observation of evoked potentials in obsessional patients. Br J Psychiatry 1983;142:605 – 9. [6] Beers SR, Rosenberg DR, Dick EL, et al. Neuropsychological study of frontal lobe function in psychotropic-naı¨ve children with obsessive-compulsive disorder. Am J Psychiatry 1999;156:777 – 9. [7] Behar D, Rapoport JL, Berg CJ, et al. Computerized tomography and neuropsychological test measures in adolescents with obsessive-compulsive disorder. Am J Psychiatry 1984;41:363–9. [8] Benkelfat C, Nordahl TE, Semple WE, et al. Local cerebral glucose metabolic rates in obsessive-compulsive disorder. Arch Gen Psychiatry 1990;47:840 – 8. [9] Berthier ML, Kulisevsky J, Gironell A, Heras JA. Obsessivecompulsive disorder associated with brain lesions: Clinical phenomenology, cognitive function, and anatomic correlates. Neurology 1996;47:353 – 61. [10] Bliss J, Cohen DJ, Freedman DX, et al. Sensory experiences of Gilles de la Tourette syndrome. Arch Gen Psychiatry 1980;37:1343 – 7. [11] Breiter HC, Rauch SL, Kwong KK, et al. Functional magnetic resonance imaging of symptom provocation in obsessive-compulsive disorder. Arch Gen Psychiatry 1996;53:595 – 606.
68
C. R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69
[12] Brown KT. Rate of apparent change in a dynamic ambiguous figure as a function of observation-time. Percept Motor Skills 1954;67:358 – 71. [13] Brown LL, Schneider JS, Lidsky TI. Sensory and cognitive functions of the basal ganglia. Curr Opin Neurobiol 1997;7:157 – 63. [14] Caan W, Perrett DI, Rolls ET. Responses of striatal neurons in the behaving monkey. 2. Visual processing in the caudal neostriatum. Brain Res 1984;290:53–65. [15] Chee KY, Sachdev PA. A controlled study of sensory tics in Gilles de la Tourette syndrome and obsessive-compulsive disorder using a structured interview. J Neurol Neurosurg Psychiatry 1997;62:188 – 92. [16] Chen M, Li C-SR, Yang Y-Y, et al. Perceptual alternation in obsessive compulsive disorder — Implications for the functions of the frontostriatal circuitry (Abstract). 3rd International Conference on Cognitive and Neural Systems, Boston, MA, May 1999. [17] Ciesielski KT, Beech HR, Gordon PK. Some electrophysiological observations in obsessional states. Br J Psychiatry 1981;138:479 – 84. [18] Drake ME, Hietter SA, Padamadan H, et al. Auditory evoked potentials in Gilles de la Tourette syndrome. Clin Electrophysiol 1992;23:19 – 23. [19] Enright SJ, Beech AR. Reduced cognitive inhibition in obsessivecompulsive disorder. Br J Clin Psychol 1993;32:67–74. [20] Fredrikson M, Fischer H, Wik G. Cerebral blood flow during anxiety provocation. J Clin Psychiatry 1997;58(Suppl. 16):16 – 21. [21] Goldsmith T, Shapira MA, Phillips KA, McElroy SL. Conceptual foundations of obsessive-compulsive spectrum disorders. In: Swinson RP, Antony MM, Rachman S, Richter MA, editors. Obsessive-Compulsive Disorder, Theory, Research and Treatment. New York: Guilford, 1998:397–425. [22] Goodman WK, Price LH, Rasmussen SA, et al. The Yale – Brown obsessive compulsive scale. I. Development, use, and reliability. Arch Gen Psychiatry 1989;46:1006–11. [23] Goodman WK, Price LH, Rasmussen SA, et al. The Yale – Brown obsessive compulsive scale. II. Validity. Arch Gen Psychiatry 1989;46:1012 – 6. [24] Gross-Isseroff R, Sasson Y, Voet H, et al. Alternation learning in obsessive-compulsive disorder. Biol Psychiatry 1996;39:733 – 8. [25] Hamilton M. The assessment of anxiety states by rating. Br J Med Psychol 1959;32:50–5. [26] Harris JP. The Schro¨der staircase: A new perspective. Percept Psychophys 1979;26:312–8. [27] Harris JP, Phillipson OT. Chlorpromazine reduces the perceptual ambiguity of a reversible visual figure. Neuropharmacology 1981;20:1337 – 8. [28] Head EK, Bolton D, Hymas N. Deficit in cognitive shifting ability in patients with obsessive compulsive disorders. Biol Psychiatry 1989;25:929–37. [29] Hoehn-Saric R, Pearlson GD, Harris GJ, et al. Effects of fluoxetine on regional cerebral blood flow in obsessive-compulsive patients. Am J Psychiatry 1991;148:1243–5. [30] Insel TR. Toward a neuroanatomy of obsessive-compulsive disorder. Arch Gen Psychiatry 1992;49:739–44. [31] Laplane D, Levasseur M, Pillon B, et al. Obsessive-compulsive and other behavioral changes with bilateral basal ganglia lesions. Brain 1989;112:699 –725. [32] Logothetis NK, Leopold DA, Sheinberg DL. What is rivaling during binocular rivalry? Nature 1996;380:621–4. [33] Lumer ED, Friston KJ, Rees G. Neural correlates of perceptual rivalry in the human brain. Science (Wash DC) 1998;280:1930– 4. [34] Machlin A, Harris GJ, Pearlson GD, et al. Elevated medial-frontal cerebral block flow in obsessive-compulsive patients. A SPECT study. Am J Psychiatry 1991;148:1240–2.
[35] Martinot JL, Allilaire JF, Mazoyer BM, et al. Obsessive-compulsive disorder: a clinical, neuropsychological and positron emission tomography study. Acta Psychiatr Scand 1990;82:233–42. [36] McGuire PK, Bench CJ, Frith CD, et al. Functional anatomy of obsessive-compulsive phenomena. Br J Psychiatry 1994;164:459– 68. [37] Middleton FA, Strick PL. The temporal lobe is a target of output from the basal ganglia. Proc Natl Acad Sci USA 1996;93:8683 – 7. [38] Molina V, Montz R, Perez-Castejon MJ, et al. Cerebral perfusion, electrical activity and effects of serotonergic treatment in Obsessive-Compulsive disorder. A preliminary study. Neuropsychobiology 1995;32:139 – 48. [39] Moroz VM, Bures J. Caudate stimulation prolongs latency of acoustically and visually signalled reaching in rats. Brain Res 1983;259:298 – 300. [40] Nordahl TE, Benkelfat C, Semple WE, et al. Cerebral glucose metabolic rates in obsessive compulsive disorder. Neuropsychopharmacology 1989;2:23 – 8. [41] Pallanti S, Grecu LM, Gangemi PF, et al. Smooth-pursuit eye movement and saccadic intrusions in obsessive-compulsive disorder. Biol Psychiatry 1996;40:1164 – 72. [42] Palumbo D, Maughan A, Kurlan R. Tourette syndrome is only one of several causes of a developmental basal ganglia syndrome. Arch Neurol 1997;54:475 – 83. [43] Pang E, Fowler B. Discriminating the effects of triazolam on stimulus and response processing by means of reaction time and P300 latency. Psychopharmacology 1994;115:509 – 15. [44] Purcell R, Maruff P, Kyrios M, Pantelis C. Cognitive deficits in obsessive-compulsive disorder on tests of frontal-striatal function. Biol Psychiatry 1998;43:348 – 57. [45] Ricci C, Blundo C. Perception of ambiguous figures after focal brain lesions. Neuropsychologia 1990;28:1163 – 73. [46] Rosenberg DR, Averbach DH, O’Hearn KM. Oculomotor response inhibition abnormalities in pediatric obsessive-compulsive disorder. Arch Gen Psychiatry 1997;54:831 – 8. [47] Savage CR, Weilburg JB, Duffy FH, et al. Low-level sensory processing in obsessive-compulsive disorder: an evoked potential study. Biol Psychiatry 1994;35:247 – 52. [48] Sawle GV, Hymas NF, Lees AJ, Frackowiak RSJ. Obsessional slowness. Brain 1991;114:2191 – 202. [49] Saxena S, Brody AL, Schwartz JM, Baxter LR. Neuroimaging and frontal-subcortical circuitry in obsessive-compulsive disorder. Br J Psychiatry 1998;173(Suppl. 35):26 – 37. [50] Schmidtke K, Schorb A, Winkelmann G, Hohagen F. Cognitive frontal lobe dysfunction in obsessive-compulsive disorder. Biol Psychiatry 1998;43:666 – 73. [51] Schneider JS. Basal ganglia role in behavior: Importance of sensory gating and its relevance to psychiatry. Biol Psychiatry 1984;19:1693 – 710. [52] Schwartz JM. Neuroanatomical aspects of cognitive-behavioural therapy response in obsessive-compulsive disorder. Br J Psychiatry 1998;173(Suppl. 35):38 – 44. [53] Schwartz JM, Stoessel PW, Baxter LR Jr, et al. Systematic changes in cerebral glucose metabolic rate after successful behavior modification treatment of obsessive-compulsive disorder. Arch Gen Psychiatry 1996;53:109 – 13. [54] Shagass C, Roemer RA, Straumanis JJ, Josiassen RC. Evoked potentials in obsessive-compulsive disorder. Adv Biol Psychiatry 1984;15:69 – 75. [55] Swedo SE, Leonard HL, Schapiro MB, et al. Sydenham’s chorea: Physical and psychological symptoms of St Vitus dance. Pediatrics 1993;91:706 – 13. [56] Swedo SE, Schapiro MB, Grady CL, et al. Cerebral glucose metabolism in childhood-onset obsessive-compulsive disorder. Arch Gen Psychiatry 1989;46:518 – 23.
C.R. Li et al. / Beha6ioural Brain Research 111 (2000) 61–69 [57] Swerdlow NR, Benbow CH, Zisook S, et al. A preliminary assessment of sensorimotor gating in patients with obsessive compulsive disorder. Biol Psychiatry 1993;33:298–301. [58] Tallis F. The neuropsychology of obsessive-compulsive disorder: A review and consideration of clinical implications. Br J Clin Psychol 1997;36:3 – 20. [59] Tien AY, Pearlson GD, Machlin SR. Oculomotor performance in
.
69
obsessive-compulsive disorder. Am J Psychiatry 1992;149:641–6. [60] Toppino TC, Long GM. Selective adaptation with reversible figures: don’t change that channel. Percept Psychophys 1987;42:37 – 48. [61] Towey J, Bruder G, Hollander ET, et al. Endogenous event-related potentials in obsessive-compulsive disorder. Biol Psychiatry 1990;28:92 – 8.