Signs of Central Nervous System Dysfunction in Obsessive-Compulsive Disorder Eric Hollander, MD; Erica Schiffman, MD; Barry Cohen, MS; Maria A. Rivera-Stein; Wilma Rosen, PhD; Jack M. Gorman, MD; Abby J. Fyer, MD; Laszlo Papp, MD; Michael R. Liebowitz, MD \s=b\ Obsessive-compulsive disorder (OCD) has been linked to altered neurological function following head trauma, encephalitis, abnormal birth events, and Gilles de la Tourette's syndrome.
Abnormalities in
computed tomographic scans, electroencephalograms, positron emission tomographic scans, and evoked potentials have been described in this disorder, but are neither
consistent nor pathognomonic of OCD. Neurological soft signs nonlocalizing signs of deviant performance on a motor or sensory test where no other sign of a neurological lesion is present. We studied 41 medication-free patients with OCD who are
met DSM-III-R criteria, as well as 20 normal controls, matched for age, sex, and handedness, on 20 individual tasks that involved fine motor coordination, involuntary movements, and sensory
and visuospatial function. There were significantly more signs of central nervous system dysfunction in the OCD group, as shown by abnormalities in fine motor coordination, involuntary and mirror movements, and visuospatial function. An excess of findings on the left side of the body and abnormalities of cube drawing may suggest right hemispheric dysfunction in a subgroup of patients with OCD. Soft signs correlated with a severity of obsessions. There was also a correlation between abnormalities in visual memory and recognition on neuropsychological testing and total soft signs. These findings provided additional evidence for a neurological deficit in some patients with OCD. However, further comparisons with other psychiatric populations are needed to determine whether these findings are unique to OCD or are a property of other psychiatric disorders as well. (Arch Gen Psychiatry. 1990;47:27-32)
disorder to has been be far more common than and neuroof the chemical research. However, to our knowledge, there have been few systematic studies of neurological findings and no clear attempt to delineate neurological subgroups, and the relationship between biochemical and neurological abnormali¬ ties is unresolved.
(OCD), recently reported previously believed,1 Obsessi v e-compul s i v e object increasing psychopharmacological
Accepted for publication December 13,1988. From the Department of Psychiatry, Columbia University College of Physicians and Surgeons, New York, NY. Reprint requests to Department of Psychiatry, Columbia University College of Physicians and Surgeons, 722 W 168th St, New York, NY 10032 (Dr Hollander).
Reports suggestive of a neurological cause in OCD include the following findings: onset of OCD following head trauma2 or von Economo's encephalitis3; a high incidence of other neuro¬ logical premorbid illness4; an association of "bizarre cases" of OCD with birth trauma5; the increased incidence in males of childhood-onset OCD6; abnormalities in electroencephalo¬ graphic,' somatosensory, and auditory evoked potentials8,9 and in the ventricular-brain ratio of computed tomographic scans10; a decreased caudate volume on volumetric computed tomo¬ graphic scans11; abnormalities on positron emission tomo¬ graphic scans12; close familial links with Gilles de la Tourette's syndrome18; comorbidity with diabetes insipidus14; abnormali¬ ties in neuropsychological testing10,15; improvement of OCD symptoms following prefrontal leukotomy and cingulotomy16; and increased stereotypical movements following electrical stimulation of the cingulum." However, while suggestive of neurological dysfunction, these findings are neither consistent nor pathognomonic. For example, not all patients with OCD have electroencephalo¬ graphic abnormalities, and when present, the electroenceph¬ alographic abnormalities are not consistent. Two studies did not replicate an abnormal ventricular-brain ratio in OCD.11,18 There are conflicting reports of neuropsychological tests in OCD, with some failing to find diffuse or frontal impairment.18 Finally, stereotypical movements produced by selective elec¬ trical stimulation have only a vague resemblance to obsessions and compulsions. Neurological soft signs are nonlocalizing deviant perfor¬ mances on a motor or sensory test where no other sign of a focal neurological disorder is present. Abnormalities may in¬ clude disorders of coordination, involuntary movements, and sensory signs. Previous studies have documented a link be¬ tween neurological soft signs and psychiatric illnesses, such as childhood asocial schizophrenia and emotionally unstable character disorder,19 as well as childhood hyperactivity and minimal brain dysfunction.20,21 These were not nonspecific find¬ ings, however, since several psychiatric groups, including nonasocial schizophrenics and patients with affective disorders and other character disorders, did not show them.19 In a prospective study, early soft signs at the age of 7 years were associated with the development of anxiety disorders in adolescence.22
Two reports of neurological examination
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
findings in child-
Our neurological soft sign battery includes items from two previous studies19,22 plus visuospatial tasks reported to be relevant to OCD. Detailed descriptions of the individual tests have been previously
Table 1 .—Subject Characteristics*
Subject Group OCD
Normal
No.
41
Mean±SD age, y (range) Sex, No. (%) M
33.9±8.1
20 32.4±10.7
(18-53)
Obsessions
.54
12 8
(60) (40)
0.87 .94
33
(80)
15 5
(75) (25)
1.53
8 (20) 12.9 ± 9.0
.13
(1 -38)
(YBOCS) 13.9 ±3.2
Compulsions
11.4 ±4.7
Total
25.3 ±5.6
Mean±SDHDRS
0.61
(21-60)
(61) (39)
Dominance, No. (%)
Mean ± SD duration of illness (range), y Mean ± SD OCD severity score
Coordination is assessed for smoothness and accuracy of the follow¬
16
25
R
provided.19
8.7 ±5.9
severity score *OCD indicates obsessive-compulsive disorder; YBOCS, Yale-Brown Ob¬ sessive-Compulsive Scale; HDRS, Hamilton Depression Rating Scale.
hood-onset OCD were also suggestive of neurological dysfunc¬ tion. In a small study of seven adolescents with OCD, neurodevelopmental examinations yielded a high frequency of ageinappropriate synkinesias and lateralization of neurodevelopmental deficits to the left side of the body.10 A nonblind, uncontrolled study using a clinical neurological ex¬ amination in childhood and adolescent subjects with OCD reported that a majority of patients had abnormal neurological findings.22 These included choreiform movements, nonspecific neurodevelopmental signs, left hemisyndrome, and miscella¬ neous findings. The patients with left-hemisyndrome OCD were also impaired on spatial tests, such as the Money's Road Map and Stylus Maze Learning tests. To our knowledge, we report the first comprehensive, controlled study of neurologi¬ cal soft signs in adult patients with OCD.
ing fine motor movements: finger to finger, rapid alternating move¬ ments of the upper and lower extremities, mirror movements (of the opposite finger and thumb), hopping, toe walking, and heel walking. Two tongue twisters are used to assess speech. Involuntary movements are tested on standing in the Romberg position, with the first 20 seconds assessing station and motor persis¬ tence. For the next 20 seconds, assessment of athetosis, chorea, tremor (resting and intention), and abnormal posturing is conducted. Sensory function is assessed for astereognosis (recognition of a penny, nickel, dime, and quarter), agraphesthesia (recognition of numbers written on the palmar surface of the index finger), position sense, and direction of cutaneous kinesthesia (recognition of a direc¬ tion of a line drawn on the palmar surface of the index finger). Visuospatial function is assessed with the face-hand test, evalua¬ tion of right-left confusion on self and examiner, and the drawing of a
cube. Dominance is determined for each subject on the eye, ear, and upper and lower extremities. This is assessed by asking subjects which eye they would use when looking through a microscope or telescope, which hand they would write with and throw a ball with, which ear they would listen on the telephone with, and which leg they would kick a ball with. Each item is rated as normal or abnormal, and a score for total soft signs, right- and left-sided soft signs, each of the four categories (coordination, abnormal movements, sensory findings, and visuospa¬ tial function), and 20 individual items is reported for each subject. Thirty subjects with OCD also underwent neuropsychological test¬ ing on a blind basis with the Benton Visual Retention Test and the Matching Familiar Figures Test, which have been described previ¬ ously.24 The Benton Visual Retention Test measures the ability to reproduce geometric forms from memory after a 10-second exposure. The Matching Familiar Figures Test requires making fine visual discriminations in the context of pattern recognition. Each of the 12 stimuli remains in view and is matched from multiple choice (five foils and one target), where the foils differ in small detail. The subject continues to choose until the correct match is selected. Dependent variables are the mean latency to the first response and the mean number of errors. DATA ANALYSIS
Soft signs are not normally distributed in the two populations studied, and many normal controls have no soft signs. Thus, the OCD group (n 41) was compared with the control group (n 20) for the mean number of abnormal soft signs by nonparametric Mann-Whit¬ ney [/tests. Comparisons are reported for total soft signs, right- and left-sided signs, the four categories, and the 20 individual items. Significance levels are conservatively two-tailed (P=.05). In addi¬ tion, the two groups were compared for the number of subjects with any soft sign abnormalities by 2 analysis. Receiver operating characteristic analysis is used, post hoc, to determine the number of soft signs that minimizes combined errors for sensitivity and specificity, as previously described.25 The first half of the subjects studied was compared with the second half of the subjects studied by cross-tabulation (log-linear) analysis to assess consistency of our findings in the original and replication groups, as =
SUBJECTS AND METHODS
Subjects included 41 adult outpatients between the ages of 18 and 53 years (mean ± SD age, 33.9 ±8.1 years) (25 male and 16 female) who met DSM-III-R criteria for OCD (Table 1). Duration of illness ranged from 1 to 38 years (mean± SD duration, 12.9±9.0 years). All patients had been medication free for at least 2 weeks, and they had
focal neurological disorder or major medical illness. Severity of OCD symptoms was assessed with the Yale-Brown Obsessive-Com¬ pulsive Scale. Severity of depression was assessed with the Hamilton Depression Rating Scale. Normal controls (n 20) ranging in age from 21 to 60 years (mean ± SD age, 32.4 ± 10.7 years) (12 male and 8 female) were screened with a diagnostic interview by a research psychiatrist and found free of any DSM-III-R diagnosis or neurologi¬ cal or medical illnesses. All subjects were administered an examination for neurological soft signs by a trained psychiatrist (E.H.) or neurologist (E.S.) "blind" to the subjects' diagnoses, or administered by a trained psychiatrist (E. S. ), videotaped, and independently rated by a trained neurologist (E.H.) blind to the diagnosis. To maintain the study blind, subjects were instructed not to discuss anything that might reveal their diagnostic status to the rater. Each rater rated a similar percentage of patients and controls (E.H.: 11 patients and 5 controls) (E.S.: 30 patients and 15 controls). Interrater reliability was estab¬ lished (see the "Results" section). The soft sign examination involves four categories (fine motor coordination, involuntary movements, sensory function, and visuospatial tasks) and a total of 20 individual tasks. no
=
=
discussed below. The effects of severity of OCD symptoms and depression on total soft signs were assessed with Spearman's correlations. Spearman's correlations of the number of soft sign abnormalities with the number of neuropsychological abnormalities on the Benton Visual Retention and Matching Familiar Figures tests were conducted. RESULTS
Sample Characteristics The OCD and normal control groups did not significantly differ in age, sex, or number of subjects with left-handed dominance (Table 1). The mean ± SD OCD severity scores for the patients on the YaleBrown Obsessive-Compulsive Scale for obsessions was 13.9 ± 3.3, for
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
Table 2. —Statistical
Summary Neurological Soft Signs* No. of Subjects With Abnormality
Mean Abnormalities OCD
Normal
OCD
Normal
Group (N 41)
Subjects (N 20)
zt
Group (N 41)
Subjects (N 20) =
xt
5.31
1.40
4.47
.00001
12.05
0.80
3.13
.0017
39 33
11
2.31
10
4.63
3.00
0.60
4.74
.00001
38
20.23
2.97
0.75
3.34
.0008
34
7.56
22
=
Side(s) Total§ R§ L§ Coordination§ Finger-finger§ Finger-nose
=
ft
=
P* .0005 .03
.0000 .006
1.12
0.45
2.07
.03
3.38
.06
0.24
0.00
1.78
.07
1.80
.17
0.05 0.35
0.00
0.69
.48
Finger-thumb
0.00 2.44
0.24
1.97 1.97
.048
RAM
0.00 0.00
.048
2.44
.11 .11
Mirror§ Hop
0.78
0.15
2.90
7.54
.006
0.02
0.00
1.00
NS
NS
Toe
0.02
0.00 0.00
0.000
1.00
NS
Heel
0.05
0.00
0.69
.48
0.00
Speech Movements§
0.41
0.15
1.62
.10
14
2.44
0.51
2.21
.026
13
4.02
Station
0.05 0.46 0.85
0.10 0.0
0.69
.48
0.10
2.30
.02
13
4.01
.04
0.50
1.79
19
2.92
0.32
0.20
1.66 1.30
.07 .09
.08 .19
.19
15
Heel-shin
Choreoathetoid§ Sensory Astereog Agraphes
0.51
0.30
Position
0.02
DCK
0.00
0.0 0.0 0.05 0.0
3.65
Face-hand
0.78 0.22
R-L confusion
0.146
0.05
0.63
Visuospatial§
0.69 0.00
21
.003
1.0
NS 1.0 .11 .045 1.0
0.00
1.72 1.03
.48
.30
0.00
1.0
1.0
NS
NS
.10
11.55 1.28
.0007 .25
.52
0.01
.0003
22
.88
0.00 3.50 .0005 18 0.43 9.52 .002 drawing§ OCD indicates obsessive-compulsive disorder: NS, not significant; Astereog, astereognosis; Agraphes, agraphesthesia; RAM, rapid alternating movements; Station, station and motor persistence; and DCK, direction of cutaneous kinesthesia. Right- vs left-sided findings: normal subjects, 0.98 and =.32; OCD 2.97 and .003. Friedman analysis of variance (2 2) for right- vs left-sided findings: OCD group vs normal subjects, 2 2.68 and .26. group, tBy Mann-Whitney U test (two-tailed). t;By 2 analysis (two-tailed). §Significant findings.
Cube
=
=
=
=
compulsions was 11.4 ± 4.7, and for total OCD severity was 25.3 ± 5.6. This is a similar level of severity compared with other OCD studies that report Yale-Brown Obsessive-Compulsive Scale scores. The mean ± SD Hamilton Depression Rating Scale score for the patients was 8.7 ±5.9.
=
replication group of 22 subjects with OCD and 10 controls by using 2 analysis. There were no differences between the initial and replication groups on the presence of soft signs for subjects with OCD ( 2 1.8, P=.17), or normal controls ( 2 0.2, P=.65). Log-linear analysis revealed no significant three-way interaction among the study, diagnosis, and number of soft signs ( 2 1.57, P=.210, df= 1). The only significant relationship was between the diagnosis and soft signs, irrespective of the study. The findings were consistent, with patients with OCD having more abnormal soft signs than normal subjects whether or not they were in the original or replication groups. =
=
=
Interrater Reliability Interrater reliability scores for the two raters on a single occasion obtained by simultaneous ratings of eight subjects. Raters alternated, with one examining and the other watching. Interrater reliability scores for each item range from a correlation of .58 to .95 (Pearson's r*s). This includes total soft signs (r= .95, Ps.OOOl), rightsided soft signs (r=.72, Ps.014), left-sided soft signs (r=.89, Ps.001), coordination (r=.58, P=s.052), abnormal movements (r= .61, P*s.041), sensory abnormalities (r .93, P=s.0001), finger-tofinger movements (r=.67, P=s.025), speech abnormalities (r=.75, Pss.01), cube drawing (r=1.0, Pe.0001), and mirror movements were
=
(r= 1.0, «.0001).
Consistency of Findings To cross-validate our soft sign findings prospectively, we compared initial group of 19 subjects with OCD and 10 controls with our
our
Soft Sign
Findings
Patients with OCD and normal controls were compared for total soft signs, right- and left-sided soft signs, the four categories of 12sts (coordination, abnormal movements, sensory findings, and visuo¬ spatial function), and the 20 individual items. Significant findings are reported and the number of subjects with abnormal findings on each of the individual items are listed on Table 2. Patients with OCD had significantly more total soft signs (mean 5.31) compared with normal subjects (mean 1.40). There was a significant increase in abnormalities of fine motor coordination in the patients with OCD compared with controls. This included finger-to-finger, tongue twister, and mirror movement abnor¬ malities. There was also an increased number of abnormal involuntary =
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
=
10
10-
9·· 8-
9-
7-
7-
8c
6
o
Ü
5·-
o
4..
o
6· 54·. 3-
3·2-
2-
il 0-
01
23456789 10 11 12 13
0
25
1
2
No. of Abnormal Soft Signs
3 4
5 6 7
8 9 10 11 12 13
H-
25
No. of Abnormal Soft Signs
Left, Distribution of total abnormal neurological soft signs in patients (n 41) with obsessive-compulsive disorder Is represented. Number of total abnormal soft signs range from 0 to 25. Number of patients with obsessivecompulsive disorder who manifest each number of total soft signs range from 0 to 9. Right, Distribution of total abnormal neurological soft signs In normal controls (n 20) Is represented. Number of total abnormal soft signs =
=
range from 0 to 5. Number of normal controls who manifest each number of total soft signs range from 0 to 9. movements and
visuospatial findings, such
as
cube-drawing
abnormalities in the patients with OCD compared with controls.
There was no difference between the two groups in sensory abnormalities. There were a similar number of right- and left-sided findings in the control group. However, in the patients with OCD, there was a significant increase in left-sided findings compared with right-sided findings (Table 2). However, further analysis with 2x2 Friedman nonparametric analysis of variance did not reveal a significant interac¬ tion effect of laterality with the diagnosis.
The Figure demonstrates the frequency distributions for the num¬ ber of total soft signs for the patients and controls. The normal controls all had five or fewer soft signs. Fifteen (36.5%) of the patients with OCD had six or more soft signs. Receiver operating characteristic analysis suggests that a cutoff of total soft signs of three or more yields the minimum number of combined errors of sensitivity and specificity in distinguishing sub¬ jects with OCD and controls. Setting this cutoff correctly identified 83% of patients and 75% of controls. A cutoff of four or more correctly identified 61% of patients and 85% of controls, whereas a more strin¬ gent cutoff of five or more would enhance specificity to 95% at the cost oflowering sensitivity to 44%. Thus, for case finding, a cutoff of three is best, whereas if high specificity is needed, as with biological studies, a cutoff of five is best. The OCD and control groups were compared with 2 analysis for the number of subjects with at least one soft sign abnormality in total soft signs, the four categories, and the 20 individual items (Table 2). There were significantly more patients with OCD who had abnormalities in the following comparisons. Of the 41 patients with OCD, 39 patients had at least one total soft sign abnormality, 38 had left-sided abnor¬ malities, and 33 had right-sided abnormalities. Three of the four categories differed, with 34 patients with OCD having coordination abnormalities, 22 with visuospatial abnormalities, and 13 with invol¬ untary movements. The groups only marginally differed in sensory abnormalities. In the individual items, 22 patients with OCD had finger-to-finger fine motor coordination abnormalities, 20 patients had mirror movements, 18 had cube-drawing abnormalities, and 13 had choreoathetoid movements. Correlations With Other Clinical Factors The effects of age, sex, handedness, duration of illness, and pres¬ of childhood or adult onset on total and individual neurological soft signs were assessed with Spearman's correlations. No significant correlations were found. The effects of severity of depression and OCD symptoms on total soft signs were assessed with Spearman's correlations. Severity of .79). While compul¬ depression did not influence soft signs (r .04, sions (r=.01, P=.93) and total Yale-Brown Obsessive-Compulsive ence
=
=
Scale severity (r .10, P= .55) did not correlate with soft signs, the severity of obsessions did significantly correlate with total soft signs (r=.37,P=.02). =
Correlation With Neuropsychological Tests The total number of neurological soft signs was correlated by Spearman's analysis with the number of errors on the Benton Visual Retention Test (r= .49, P= .006), a measure of visual memory reten¬ tion, but not on the Matching Familiar Figures Test (r .24, =. 196), a test of visual recognition and matching. =
COMMENT
This study has reported abnormalities in neurological soft signs suggestive of central nervous system dysfunction in an adult OCD population that was closely matched with normal
controls on age, sex, and handedness. Results of our neurological soft sign examination are con¬ sistent between raters. The least consistent items do not account for the findings. In addition, there is good prospective cross-validation, with the first half of the subjects examined having similar findings as the second half. Previous studies have demonstrated high intrarater and interrater reliability over time for many neurological soft sign items.19,22,24 One extensive reliability study of soft signs in adolescents sug¬ gested that certain motor and sensory tasks can be measured at the same degree of reliability reported for numerous other medical tests.26 Furthermore, soft sign abnormalities ap¬ peared to be stable for 10 years in a follow-up study.27 We have reported significant abnormalities in total soft signs, fine motor coordination, involuntary movements, and visuospatial function in adult patients with OCD compared with normal controls. While there seemed to be an increased number of left-sided abnormalities in subjects with OCD compared with normal controls, there was no significant in¬ teraction between the diagnosis and laterality, which does not support theories of laterality in OCD. However, the in¬ creased number of findings on the left side of the body and abnormalities of cube drawing in some patients with OCD may suggest right hemispheric dysfunction in a subgroup of patients with OCD. Perhaps the presence of two or fewer soft signs represents background noise, whereas the presence of
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
a distinct subgroup of neuropsychiatrie impairment. Setting a cutoff of three or more soft signs best distinguishes the OCD group from normal controls, by minimizing the sum of falsenegative and false-positive errors. Soft signs significantly correlated with the severity of ob¬ sessions, but not compulsions. In most patients with OCD, obsessions are a primary feature, and compulsions develop either in response to the obsessions, or at a later stage of the disorder. Some patients with OCD only suffer from obses¬ sions. In our sample, the more severe the obsessions, the greater the number of soft signs. Soft signs did not correlate with the severity of depression. This supports the notion that soft signs are not reflective of psychiatric illness in general, but may be more specifically related to obsessionality. Patients with OCD had significantly more signs of central nervous system dysfunction than normal controls. Thirtynine of the 41 patients with OCD had at least one abnormal soft sign. Abnormalities in fine motor coordination in the OCD group, a nonspecific finding, occurred in 34 patients and may be suggestive of diffuse brain dysfunction. Thirty-eight of the subjects with OCD had left-sided soft signs, and 18 patients with OCD showed disturbances in cube drawing. In addition, those patients with OCD who had in¬ creased soft signs had neuropsychological abnormalities on visual memory. These findings may suggest right hemispheric dysfunction in the OCD group, but are not definitive. Involuntary movements in 13 patients with OCD and mirror movements in 21 patients with OCD might be suggestive of basal ganglia disturbance. This would be consistent with posi¬ tron emission tomographic scan findings of altered metabo¬
three
or more
patients with
soft signs represents
true
lism of the basal ganglia,12 as well as an association of OCD with Sydenham's chorea.4,28 A descriptive comparison of these results with previous reports in other psychiatric populations is of interest. Quitkin et al19 studied schizophrenics with premorbid asociality, schizophrenics of mixed subgroups, patients with emotionally unstable character disorder, subjects with other character disorders, and patients with affective disorders. Compared with these other psychiatric groups, patients with OCD have more involuntary movements and abnormal agraphesthesia, and slightly more mirror movement, speech, and rapid alter¬ nating movement abnormalities. These other groups were not tested for abnormalities in cube drawing. Asocial schizo¬ phrenics appear to have more abnormalities on the face-hand test and hopping. A greater percent of patients with OCD in our sample had soft sign abnormalities (95%) compared with children rated as hyperactive at the age of 4 years (13%). " To date, most neurobiological studies of OCD have centered on the pharmacology and biochemistry of this disorder. Sero¬ tonergic involvement in OCD is supported by pharmacological treatment studies with serotonin reuptake blockers, such as
clomipramine, fluoxetine, and fluvoxamine,29 cerebrospinal fluid studies of 5-hydroxyindoleacetic acid,30 peripheral mea¬ sures of platelet 5-hydroxytryptamine levels,31 and biological challenge studies with oral m-chlorophenylpiperazine, which is a selective serotonin agonist reported to cause exacerbation of OCD symptoms.32,33 Serotonergic pathways ascend from nuclei in the brain stem through the median forebrain bundle and project diffusely to cortical and motor systems. Structural brain abnormalities at various levels could result in functional dysregulation of the serotonergic system. Further attempts to localize structural abnormalities in OCD have utilized imag¬ ing and neuropsychological approaches. Imaging studies with positron emission tomographic scans12 and volumetric comput¬ ed tomographic scans11 have implicated abnormalities in the basal ganglia, an area involved in the production of abnormal and choreiform movements, as well as the orbital frontal gyri. Neuropsychological studies of OCD also have reported visuo¬ spatial abnormalities consistent with right hemisphere dys¬ function,18,24,34 as well as frontal abnormalities.15,24 Additional work is needed to integrate fully structural findings with functional abnormalities in OCD. Further studies should ad¬ dress whether there are separate neurological and biochemi¬ cal subgroups of OCD, or whether neuropsychiatrie and bio¬ chemical abnormalities cluster and are interrelated in OCD. Abnormalities in coordination, abnormal movements, and vi¬ suospatial function in a subgroup of patients with OCD are consistent with other neurobiological findings in OCD. The results of this study have confirmed and expanded evidence for neurological abnormalities in at least a subgroup of patients with OCD and have confirmed the usefulness of assessing soft signs in adult anxiety disorder populations. This is exploratory work that confirms the observation that there is clumsiness in OCD and points to some intrinsic organic diathe¬ sis that other investigators have noticed. However, this is only the first step. While some other psychiatric conditions have been reported to have subtle neurological soft signs, others have not. It is not clear to what extent soft signs may be a property of psychiatric disorders in general, and to what extent they are unique to OCD. The next step would be to compare patients with OCD on these measures to other psy¬ chiatric samples. Further studies should also address whether soft signs support the distinction of OCD from other anxiety subtypes (ie, social phobia, panic disorder, etc). Additionally, soft signs may help to clarify issues of heterogeneity within OCD with regard to familial background, diagnostic subtypes (ie, washers, checkers, and pure obsessionals), treatment response, origin, and pathophysiology. This study was supported in part by grants MH-30906 and MH-37592 and Research Scientist Development Award MH-00750 (Dr Hollander) from the National Institute of Mental Health, Bethesda, Md. The authors gratefully acknowledge the helpful advice of Donald F. Klein, MD, the statistical consultation with Donald Ross, PhD, and the assistance of Concetta DeGria, MS.
References 1. Myers JK, Weissman MM, Tischler GL, Holzer LE III, Leaf PJ, Orvaschel H, Anthony JC, Boyd JH, Burke JD, Kramer M, Stoltzman R. Six-month prevalence of psychiatric disorders in three communities. Arch Gen Psychiatry.
1984;41:959-967. 2. McKeon J, McGuffin P, Robinson P. Obsessive-compulsive neurosis following head injury: a report of 4 cases. Br J Psychiatry. 1984;144:190\x=req-\
192. 3. Schilder P. The organic background of obsessions and compulsions. Am J
Psychiatry. 1938;94:1397.
4. Grimshaw L. Obsessional disorder and neurological illness. J Neurol Neurosurg Psychiatry.1964;27:229-231. 5. Capstick N, Seldrup U. Obsessional states: at study in the relationship between abnormalities occurring at birth and subsequent development of obsessional symptoms. Acta Psychiatr Scand. 1977;56:427-439. 6. Rapoport J, Elkins R, Langer DH, Sceery W, Buchsbaum MS, Gillin JC, Murphy DL, Zahn TP, Lake R, Ludlow C, Mendelson W. Childhood obsessive
compulsive disorder. Am J Psychiatry. 1981;138:1545-1554. 7. Pacella BL, Polatin P, Nagle SH. Clinical and EEG studies in obsessive compulsive states. Am J Psychiatry. 1944;100:830-838. 8. Shagass C, Roemer RA, Straumanis JJ, Josiassen J. Distinctive somatosensory evoked potential features in obsessive compulsive disorder. Biol Psychiatry. 1984;19:1507-1524. 9. Ciesielksi HR, Beech HR, Gordon PK. Some electrophysiological observations in obsessional states. Br J Psychiatry. 1981;138:479-484. 10. Behar D, Rapoport JL, Berg C, Denckla MB, Mann L, Cox C, Fidio P, Zahn T, Wolfman MG. Computerized tomography and neuropsychological test measures in adolescents with obsessive-compulsive disorder. Am J Psychiatry. 1984;141:363-369. 11. Luxenberg JS, Swedo SE, Flament MF, Friedland RP, Rapoport J, Rapoport SI. Neuroanatomical abnormalities in obsessive-compulsive disorder detected with quantitative X-ray computed tomography. Am J Psychiatry. 1988;145:1089-1093.
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
12. Baxter LR, Phelps ME, Mazziotta JC, Guze BH, Schwartz JM, Selin LE. Local cerebral glucose metabolic rates in obsessive-compulsive disorder: a comparison with rates in unipolar depression and in normal controls. Arch Gen
Psychiatry. 1987;44:211-218. 13. Pauls DL, Towbin KE, Leckman JF, Zahner GE, Cohen DJ. Gilles de la Tourette's and obsessive-compulsive disorder. Arch Gen Psychiatry. 1986;43:1180-1182. 14. Barton R. Diabetes insipidus and obsessional neurosis: a syndrome. Lancet. 1965;1:133-135. 15. Flor-Henry P, Yeudall LT, Koles ZJ, Howarth BG. Neuropsychological and power spectral EEG investigations of the obsessive-compulsive syndrome. Biol Psychiatry. 1979;14:119-130. 16. Lewin W. Selective leukotomy: a review. In: Laitinin LV, Livingston K, eds. Surgical Approaches in Psychiatry. Lancaster, England: Medical & Technical Publishing Co; 1973:69-73. 17. Talairach J, Bancard J, Geiver S, Bordas-Fener A, Bonis A, Szikla G, Rusu M. The cingulate gyrus and human behavior. Electroencephalogr Clin Neurophysiol. 1973;34:45-52. 18. Insel T, Donnelly E, Lalakea M, Alterman IS, Murphy DL. Neurological and neuropsychological studies of patients with obsessive-compulsive disorder.
Biol Psychiatry. 1983;18:741-751. 19. Quitkin FM, Rifkin A, Klein DF. Neurologic soft signs in schizophrenia and character disorders: organicity in schizophrenia with premorbid asociality and emotionally unstable character disorders. Arch Gen Psychiatry.
1976;33:845-853.
20. Nichols PL. Minimal brain dysfunction and soft signs: the collaborative perinatal project. In: Tupper D, ed. Soft Neurological Signs. New York, NY:
Grune & Stratton; 1987:179-199.
21. Denckla MB: Minimal brain dysfunction. In: Chall J, Mirsky AF, Relage KJ, eds. Education and the Brain. Chicago, Ill: University of Chicago Press;
1978:223-268. 22. Shaffer D, Schonfeld IS, O'Connor PA, Stokman C, Trautman P, Shafer S, Ng SC. Neurological soft signs and their relationship to psychiatric disorder and intelligence in childhood and adolescence. Arch Gen Psychiatry.
1985;42:342-351.
23. Denckla MB.
Neurological examination. In: Rapoport JL, ed.
Obse iv
Compulsive Disorder in Children and Adolescents. Washington, Psychiatric Press Inc; 1988:107-118.
DC: Ameri-
can
24. Rosen WG, Hollander E, Stannick V, Liebowitz MR. Test performance variables in obsessive-compulsive disorder. J Clin Exp Neuropsychol.
1988;10:73.
25. Swets JA. Measuring the accuracy of diagnostic systems. Science. 1988;24:1285-1293. 26. Stokman CJ, Shafer SQ, Shaffer D, Ng SC, O'Connor PA, Wolff RM. Assessment of neurological soft signs in adolescents: reliability studies. Dev Med Child Neurol. 1986;28:428-439. 27. Shafer SQ, Stokman CJ, Shaffer D, Ng SC, O'Connor PA, Schonfeld IS. Ten-year consistency in neurological test performance of children without focal neurological deficit. Dev Med Child Neurol. 1986;28:417-427. 28. Swedo SE, Rapoport JL, Cheslow DL, Leonard HL, Ayoub EM, Hosier DM, Wald E. Increased incidence of obsessive-compulsive symptoms in patients with Sydenham's chorea. Am J Psychiatry. 1989;146:246-249. 29. Hollander E, Liebowitz MR, Gorman JM. Anxiety disorders. In: Talbott JA, Hales RE, Yudofsky, SS, eds. American Psychiatric Press Textbook of Psychiatry. Washington, DC: American Psychiatric Press Inc; 1988:443\x=req-\ 493. 30. Thoren P, Asberg M, Bertilsson L, Mellstr\l=o"\mB, Sjoqvist F, Tr\l=a"\skman L. Clomipramine treatment of obsessive-compulsive disorder, II: biochemical
aspects. Arch Gen Psychiatry. 1980;37:1289-1294. 31. Flament MF, Rapoport JL, Murphy DL, Lake CR, Berg CJ. Biochemical
changes during clomipramine treatment of childhood obsessive-compulsive disorder. Arch Gen Psychiatry. 1987;44:219-225. 32. Zohar J, Mueller EA, Insel TR, Zohar-Kadovch RC, Murphy DL. Serotonergic responsivity in obsessive-compulsive disorder: comparison of patients and healthy controls. Arch Gen Psychiatry. 1987;44:946-951. 33. Hollander E, Fay M, Cohen B, Campess R, Gosman JM, Liebowitz MR. Serotonergic and noradrenergic sensitivity in obsessive-compulsive disorder: behavioral findings. Am J Psychiatry. 1988;145:1015-1017. 34. Diamond BM. Albrecht W, Borison RL. Neuropsychology of obsessive compulsive disorders. In: Forty-Third Annual Convention and Scientific Program, the Society of Biological Psychiatry: May 4-8,1988; Montreal, Canada.
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
Abnormalities in
computed tomographic scans, electroencephalograms, positron emission tomographic scans, and evoked potentials have been described in this disorder, but are neither
consistent nor pathognomonic of OCD. Neurological soft signs nonlocalizing signs of deviant performance on a motor or sensory test where no other sign of a neurological lesion is present. We studied 41 medication-free patients with OCD who are
met DSM-III-R criteria, as well as 20 normal controls, matched for age, sex, and handedness, on 20 individual tasks that involved fine motor coordination, involuntary movements, and sensory
and visuospatial function. There were significantly more signs of central nervous system dysfunction in the OCD group, as shown by abnormalities in fine motor coordination, involuntary and mirror movements, and visuospatial function. An excess of findings on the left side of the body and abnormalities of cube drawing may suggest right hemispheric dysfunction in a subgroup of patients with OCD. Soft signs correlated with a severity of obsessions. There was also a correlation between abnormalities in visual memory and recognition on neuropsychological testing and total soft signs. These findings provided additional evidence for a neurological deficit in some patients with OCD. However, further comparisons with other psychiatric populations are needed to determine whether these findings are unique to OCD or are a property of other psychiatric disorders as well. (Arch Gen Psychiatry. 1990;47:27-32)
disorder to has been be far more common than and neuroof the chemical research. However, to our knowledge, there have been few systematic studies of neurological findings and no clear attempt to delineate neurological subgroups, and the relationship between biochemical and neurological abnormali¬ ties is unresolved.
(OCD), recently reported previously believed,1 Obsessi v e-compul s i v e object increasing psychopharmacological
Accepted for publication December 13,1988. From the Department of Psychiatry, Columbia University College of Physicians and Surgeons, New York, NY. Reprint requests to Department of Psychiatry, Columbia University College of Physicians and Surgeons, 722 W 168th St, New York, NY 10032 (Dr Hollander).
Reports suggestive of a neurological cause in OCD include the following findings: onset of OCD following head trauma2 or von Economo's encephalitis3; a high incidence of other neuro¬ logical premorbid illness4; an association of "bizarre cases" of OCD with birth trauma5; the increased incidence in males of childhood-onset OCD6; abnormalities in electroencephalo¬ graphic,' somatosensory, and auditory evoked potentials8,9 and in the ventricular-brain ratio of computed tomographic scans10; a decreased caudate volume on volumetric computed tomo¬ graphic scans11; abnormalities on positron emission tomo¬ graphic scans12; close familial links with Gilles de la Tourette's syndrome18; comorbidity with diabetes insipidus14; abnormali¬ ties in neuropsychological testing10,15; improvement of OCD symptoms following prefrontal leukotomy and cingulotomy16; and increased stereotypical movements following electrical stimulation of the cingulum." However, while suggestive of neurological dysfunction, these findings are neither consistent nor pathognomonic. For example, not all patients with OCD have electroencephalo¬ graphic abnormalities, and when present, the electroenceph¬ alographic abnormalities are not consistent. Two studies did not replicate an abnormal ventricular-brain ratio in OCD.11,18 There are conflicting reports of neuropsychological tests in OCD, with some failing to find diffuse or frontal impairment.18 Finally, stereotypical movements produced by selective elec¬ trical stimulation have only a vague resemblance to obsessions and compulsions. Neurological soft signs are nonlocalizing deviant perfor¬ mances on a motor or sensory test where no other sign of a focal neurological disorder is present. Abnormalities may in¬ clude disorders of coordination, involuntary movements, and sensory signs. Previous studies have documented a link be¬ tween neurological soft signs and psychiatric illnesses, such as childhood asocial schizophrenia and emotionally unstable character disorder,19 as well as childhood hyperactivity and minimal brain dysfunction.20,21 These were not nonspecific find¬ ings, however, since several psychiatric groups, including nonasocial schizophrenics and patients with affective disorders and other character disorders, did not show them.19 In a prospective study, early soft signs at the age of 7 years were associated with the development of anxiety disorders in adolescence.22
Two reports of neurological examination
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
findings in child-
Our neurological soft sign battery includes items from two previous studies19,22 plus visuospatial tasks reported to be relevant to OCD. Detailed descriptions of the individual tests have been previously
Table 1 .—Subject Characteristics*
Subject Group OCD
Normal
No.
41
Mean±SD age, y (range) Sex, No. (%) M
33.9±8.1
20 32.4±10.7
(18-53)
Obsessions
.54
12 8
(60) (40)
0.87 .94
33
(80)
15 5
(75) (25)
1.53
8 (20) 12.9 ± 9.0
.13
(1 -38)
(YBOCS) 13.9 ±3.2
Compulsions
11.4 ±4.7
Total
25.3 ±5.6
Mean±SDHDRS
0.61
(21-60)
(61) (39)
Dominance, No. (%)
Mean ± SD duration of illness (range), y Mean ± SD OCD severity score
Coordination is assessed for smoothness and accuracy of the follow¬
16
25
R
provided.19
8.7 ±5.9
severity score *OCD indicates obsessive-compulsive disorder; YBOCS, Yale-Brown Ob¬ sessive-Compulsive Scale; HDRS, Hamilton Depression Rating Scale.
hood-onset OCD were also suggestive of neurological dysfunc¬ tion. In a small study of seven adolescents with OCD, neurodevelopmental examinations yielded a high frequency of ageinappropriate synkinesias and lateralization of neurodevelopmental deficits to the left side of the body.10 A nonblind, uncontrolled study using a clinical neurological ex¬ amination in childhood and adolescent subjects with OCD reported that a majority of patients had abnormal neurological findings.22 These included choreiform movements, nonspecific neurodevelopmental signs, left hemisyndrome, and miscella¬ neous findings. The patients with left-hemisyndrome OCD were also impaired on spatial tests, such as the Money's Road Map and Stylus Maze Learning tests. To our knowledge, we report the first comprehensive, controlled study of neurologi¬ cal soft signs in adult patients with OCD.
ing fine motor movements: finger to finger, rapid alternating move¬ ments of the upper and lower extremities, mirror movements (of the opposite finger and thumb), hopping, toe walking, and heel walking. Two tongue twisters are used to assess speech. Involuntary movements are tested on standing in the Romberg position, with the first 20 seconds assessing station and motor persis¬ tence. For the next 20 seconds, assessment of athetosis, chorea, tremor (resting and intention), and abnormal posturing is conducted. Sensory function is assessed for astereognosis (recognition of a penny, nickel, dime, and quarter), agraphesthesia (recognition of numbers written on the palmar surface of the index finger), position sense, and direction of cutaneous kinesthesia (recognition of a direc¬ tion of a line drawn on the palmar surface of the index finger). Visuospatial function is assessed with the face-hand test, evalua¬ tion of right-left confusion on self and examiner, and the drawing of a
cube. Dominance is determined for each subject on the eye, ear, and upper and lower extremities. This is assessed by asking subjects which eye they would use when looking through a microscope or telescope, which hand they would write with and throw a ball with, which ear they would listen on the telephone with, and which leg they would kick a ball with. Each item is rated as normal or abnormal, and a score for total soft signs, right- and left-sided soft signs, each of the four categories (coordination, abnormal movements, sensory findings, and visuospa¬ tial function), and 20 individual items is reported for each subject. Thirty subjects with OCD also underwent neuropsychological test¬ ing on a blind basis with the Benton Visual Retention Test and the Matching Familiar Figures Test, which have been described previ¬ ously.24 The Benton Visual Retention Test measures the ability to reproduce geometric forms from memory after a 10-second exposure. The Matching Familiar Figures Test requires making fine visual discriminations in the context of pattern recognition. Each of the 12 stimuli remains in view and is matched from multiple choice (five foils and one target), where the foils differ in small detail. The subject continues to choose until the correct match is selected. Dependent variables are the mean latency to the first response and the mean number of errors. DATA ANALYSIS
Soft signs are not normally distributed in the two populations studied, and many normal controls have no soft signs. Thus, the OCD group (n 41) was compared with the control group (n 20) for the mean number of abnormal soft signs by nonparametric Mann-Whit¬ ney [/tests. Comparisons are reported for total soft signs, right- and left-sided signs, the four categories, and the 20 individual items. Significance levels are conservatively two-tailed (P=.05). In addi¬ tion, the two groups were compared for the number of subjects with any soft sign abnormalities by 2 analysis. Receiver operating characteristic analysis is used, post hoc, to determine the number of soft signs that minimizes combined errors for sensitivity and specificity, as previously described.25 The first half of the subjects studied was compared with the second half of the subjects studied by cross-tabulation (log-linear) analysis to assess consistency of our findings in the original and replication groups, as =
SUBJECTS AND METHODS
Subjects included 41 adult outpatients between the ages of 18 and 53 years (mean ± SD age, 33.9 ±8.1 years) (25 male and 16 female) who met DSM-III-R criteria for OCD (Table 1). Duration of illness ranged from 1 to 38 years (mean± SD duration, 12.9±9.0 years). All patients had been medication free for at least 2 weeks, and they had
focal neurological disorder or major medical illness. Severity of OCD symptoms was assessed with the Yale-Brown Obsessive-Com¬ pulsive Scale. Severity of depression was assessed with the Hamilton Depression Rating Scale. Normal controls (n 20) ranging in age from 21 to 60 years (mean ± SD age, 32.4 ± 10.7 years) (12 male and 8 female) were screened with a diagnostic interview by a research psychiatrist and found free of any DSM-III-R diagnosis or neurologi¬ cal or medical illnesses. All subjects were administered an examination for neurological soft signs by a trained psychiatrist (E.H.) or neurologist (E.S.) "blind" to the subjects' diagnoses, or administered by a trained psychiatrist (E. S. ), videotaped, and independently rated by a trained neurologist (E.H.) blind to the diagnosis. To maintain the study blind, subjects were instructed not to discuss anything that might reveal their diagnostic status to the rater. Each rater rated a similar percentage of patients and controls (E.H.: 11 patients and 5 controls) (E.S.: 30 patients and 15 controls). Interrater reliability was estab¬ lished (see the "Results" section). The soft sign examination involves four categories (fine motor coordination, involuntary movements, sensory function, and visuospatial tasks) and a total of 20 individual tasks. no
=
=
discussed below. The effects of severity of OCD symptoms and depression on total soft signs were assessed with Spearman's correlations. Spearman's correlations of the number of soft sign abnormalities with the number of neuropsychological abnormalities on the Benton Visual Retention and Matching Familiar Figures tests were conducted. RESULTS
Sample Characteristics The OCD and normal control groups did not significantly differ in age, sex, or number of subjects with left-handed dominance (Table 1). The mean ± SD OCD severity scores for the patients on the YaleBrown Obsessive-Compulsive Scale for obsessions was 13.9 ± 3.3, for
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
Table 2. —Statistical
Summary Neurological Soft Signs* No. of Subjects With Abnormality
Mean Abnormalities OCD
Normal
OCD
Normal
Group (N 41)
Subjects (N 20)
zt
Group (N 41)
Subjects (N 20) =
xt
5.31
1.40
4.47
.00001
12.05
0.80
3.13
.0017
39 33
11
2.31
10
4.63
3.00
0.60
4.74
.00001
38
20.23
2.97
0.75
3.34
.0008
34
7.56
22
=
Side(s) Total§ R§ L§ Coordination§ Finger-finger§ Finger-nose
=
ft
=
P* .0005 .03
.0000 .006
1.12
0.45
2.07
.03
3.38
.06
0.24
0.00
1.78
.07
1.80
.17
0.05 0.35
0.00
0.69
.48
Finger-thumb
0.00 2.44
0.24
1.97 1.97
.048
RAM
0.00 0.00
.048
2.44
.11 .11
Mirror§ Hop
0.78
0.15
2.90
7.54
.006
0.02
0.00
1.00
NS
NS
Toe
0.02
0.00 0.00
0.000
1.00
NS
Heel
0.05
0.00
0.69
.48
0.00
Speech Movements§
0.41
0.15
1.62
.10
14
2.44
0.51
2.21
.026
13
4.02
Station
0.05 0.46 0.85
0.10 0.0
0.69
.48
0.10
2.30
.02
13
4.01
.04
0.50
1.79
19
2.92
0.32
0.20
1.66 1.30
.07 .09
.08 .19
.19
15
Heel-shin
Choreoathetoid§ Sensory Astereog Agraphes
0.51
0.30
Position
0.02
DCK
0.00
0.0 0.0 0.05 0.0
3.65
Face-hand
0.78 0.22
R-L confusion
0.146
0.05
0.63
Visuospatial§
0.69 0.00
21
.003
1.0
NS 1.0 .11 .045 1.0
0.00
1.72 1.03
.48
.30
0.00
1.0
1.0
NS
NS
.10
11.55 1.28
.0007 .25
.52
0.01
.0003
22
.88
0.00 3.50 .0005 18 0.43 9.52 .002 drawing§ OCD indicates obsessive-compulsive disorder: NS, not significant; Astereog, astereognosis; Agraphes, agraphesthesia; RAM, rapid alternating movements; Station, station and motor persistence; and DCK, direction of cutaneous kinesthesia. Right- vs left-sided findings: normal subjects, 0.98 and =.32; OCD 2.97 and .003. Friedman analysis of variance (2 2) for right- vs left-sided findings: OCD group vs normal subjects, 2 2.68 and .26. group, tBy Mann-Whitney U test (two-tailed). t;By 2 analysis (two-tailed). §Significant findings.
Cube
=
=
=
=
compulsions was 11.4 ± 4.7, and for total OCD severity was 25.3 ± 5.6. This is a similar level of severity compared with other OCD studies that report Yale-Brown Obsessive-Compulsive Scale scores. The mean ± SD Hamilton Depression Rating Scale score for the patients was 8.7 ±5.9.
=
replication group of 22 subjects with OCD and 10 controls by using 2 analysis. There were no differences between the initial and replication groups on the presence of soft signs for subjects with OCD ( 2 1.8, P=.17), or normal controls ( 2 0.2, P=.65). Log-linear analysis revealed no significant three-way interaction among the study, diagnosis, and number of soft signs ( 2 1.57, P=.210, df= 1). The only significant relationship was between the diagnosis and soft signs, irrespective of the study. The findings were consistent, with patients with OCD having more abnormal soft signs than normal subjects whether or not they were in the original or replication groups. =
=
=
Interrater Reliability Interrater reliability scores for the two raters on a single occasion obtained by simultaneous ratings of eight subjects. Raters alternated, with one examining and the other watching. Interrater reliability scores for each item range from a correlation of .58 to .95 (Pearson's r*s). This includes total soft signs (r= .95, Ps.OOOl), rightsided soft signs (r=.72, Ps.014), left-sided soft signs (r=.89, Ps.001), coordination (r=.58, P=s.052), abnormal movements (r= .61, P*s.041), sensory abnormalities (r .93, P=s.0001), finger-tofinger movements (r=.67, P=s.025), speech abnormalities (r=.75, Pss.01), cube drawing (r=1.0, Pe.0001), and mirror movements were
=
(r= 1.0, «.0001).
Consistency of Findings To cross-validate our soft sign findings prospectively, we compared initial group of 19 subjects with OCD and 10 controls with our
our
Soft Sign
Findings
Patients with OCD and normal controls were compared for total soft signs, right- and left-sided soft signs, the four categories of 12sts (coordination, abnormal movements, sensory findings, and visuo¬ spatial function), and the 20 individual items. Significant findings are reported and the number of subjects with abnormal findings on each of the individual items are listed on Table 2. Patients with OCD had significantly more total soft signs (mean 5.31) compared with normal subjects (mean 1.40). There was a significant increase in abnormalities of fine motor coordination in the patients with OCD compared with controls. This included finger-to-finger, tongue twister, and mirror movement abnor¬ malities. There was also an increased number of abnormal involuntary =
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
=
10
10-
9·· 8-
9-
7-
7-
8c
6
o
Ü
5·-
o
4..
o
6· 54·. 3-
3·2-
2-
il 0-
01
23456789 10 11 12 13
0
25
1
2
No. of Abnormal Soft Signs
3 4
5 6 7
8 9 10 11 12 13
H-
25
No. of Abnormal Soft Signs
Left, Distribution of total abnormal neurological soft signs in patients (n 41) with obsessive-compulsive disorder Is represented. Number of total abnormal soft signs range from 0 to 25. Number of patients with obsessivecompulsive disorder who manifest each number of total soft signs range from 0 to 9. Right, Distribution of total abnormal neurological soft signs In normal controls (n 20) Is represented. Number of total abnormal soft signs =
=
range from 0 to 5. Number of normal controls who manifest each number of total soft signs range from 0 to 9. movements and
visuospatial findings, such
as
cube-drawing
abnormalities in the patients with OCD compared with controls.
There was no difference between the two groups in sensory abnormalities. There were a similar number of right- and left-sided findings in the control group. However, in the patients with OCD, there was a significant increase in left-sided findings compared with right-sided findings (Table 2). However, further analysis with 2x2 Friedman nonparametric analysis of variance did not reveal a significant interac¬ tion effect of laterality with the diagnosis.
The Figure demonstrates the frequency distributions for the num¬ ber of total soft signs for the patients and controls. The normal controls all had five or fewer soft signs. Fifteen (36.5%) of the patients with OCD had six or more soft signs. Receiver operating characteristic analysis suggests that a cutoff of total soft signs of three or more yields the minimum number of combined errors of sensitivity and specificity in distinguishing sub¬ jects with OCD and controls. Setting this cutoff correctly identified 83% of patients and 75% of controls. A cutoff of four or more correctly identified 61% of patients and 85% of controls, whereas a more strin¬ gent cutoff of five or more would enhance specificity to 95% at the cost oflowering sensitivity to 44%. Thus, for case finding, a cutoff of three is best, whereas if high specificity is needed, as with biological studies, a cutoff of five is best. The OCD and control groups were compared with 2 analysis for the number of subjects with at least one soft sign abnormality in total soft signs, the four categories, and the 20 individual items (Table 2). There were significantly more patients with OCD who had abnormalities in the following comparisons. Of the 41 patients with OCD, 39 patients had at least one total soft sign abnormality, 38 had left-sided abnor¬ malities, and 33 had right-sided abnormalities. Three of the four categories differed, with 34 patients with OCD having coordination abnormalities, 22 with visuospatial abnormalities, and 13 with invol¬ untary movements. The groups only marginally differed in sensory abnormalities. In the individual items, 22 patients with OCD had finger-to-finger fine motor coordination abnormalities, 20 patients had mirror movements, 18 had cube-drawing abnormalities, and 13 had choreoathetoid movements. Correlations With Other Clinical Factors The effects of age, sex, handedness, duration of illness, and pres¬ of childhood or adult onset on total and individual neurological soft signs were assessed with Spearman's correlations. No significant correlations were found. The effects of severity of depression and OCD symptoms on total soft signs were assessed with Spearman's correlations. Severity of .79). While compul¬ depression did not influence soft signs (r .04, sions (r=.01, P=.93) and total Yale-Brown Obsessive-Compulsive ence
=
=
Scale severity (r .10, P= .55) did not correlate with soft signs, the severity of obsessions did significantly correlate with total soft signs (r=.37,P=.02). =
Correlation With Neuropsychological Tests The total number of neurological soft signs was correlated by Spearman's analysis with the number of errors on the Benton Visual Retention Test (r= .49, P= .006), a measure of visual memory reten¬ tion, but not on the Matching Familiar Figures Test (r .24, =. 196), a test of visual recognition and matching. =
COMMENT
This study has reported abnormalities in neurological soft signs suggestive of central nervous system dysfunction in an adult OCD population that was closely matched with normal
controls on age, sex, and handedness. Results of our neurological soft sign examination are con¬ sistent between raters. The least consistent items do not account for the findings. In addition, there is good prospective cross-validation, with the first half of the subjects examined having similar findings as the second half. Previous studies have demonstrated high intrarater and interrater reliability over time for many neurological soft sign items.19,22,24 One extensive reliability study of soft signs in adolescents sug¬ gested that certain motor and sensory tasks can be measured at the same degree of reliability reported for numerous other medical tests.26 Furthermore, soft sign abnormalities ap¬ peared to be stable for 10 years in a follow-up study.27 We have reported significant abnormalities in total soft signs, fine motor coordination, involuntary movements, and visuospatial function in adult patients with OCD compared with normal controls. While there seemed to be an increased number of left-sided abnormalities in subjects with OCD compared with normal controls, there was no significant in¬ teraction between the diagnosis and laterality, which does not support theories of laterality in OCD. However, the in¬ creased number of findings on the left side of the body and abnormalities of cube drawing in some patients with OCD may suggest right hemispheric dysfunction in a subgroup of patients with OCD. Perhaps the presence of two or fewer soft signs represents background noise, whereas the presence of
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
a distinct subgroup of neuropsychiatrie impairment. Setting a cutoff of three or more soft signs best distinguishes the OCD group from normal controls, by minimizing the sum of falsenegative and false-positive errors. Soft signs significantly correlated with the severity of ob¬ sessions, but not compulsions. In most patients with OCD, obsessions are a primary feature, and compulsions develop either in response to the obsessions, or at a later stage of the disorder. Some patients with OCD only suffer from obses¬ sions. In our sample, the more severe the obsessions, the greater the number of soft signs. Soft signs did not correlate with the severity of depression. This supports the notion that soft signs are not reflective of psychiatric illness in general, but may be more specifically related to obsessionality. Patients with OCD had significantly more signs of central nervous system dysfunction than normal controls. Thirtynine of the 41 patients with OCD had at least one abnormal soft sign. Abnormalities in fine motor coordination in the OCD group, a nonspecific finding, occurred in 34 patients and may be suggestive of diffuse brain dysfunction. Thirty-eight of the subjects with OCD had left-sided soft signs, and 18 patients with OCD showed disturbances in cube drawing. In addition, those patients with OCD who had in¬ creased soft signs had neuropsychological abnormalities on visual memory. These findings may suggest right hemispheric dysfunction in the OCD group, but are not definitive. Involuntary movements in 13 patients with OCD and mirror movements in 21 patients with OCD might be suggestive of basal ganglia disturbance. This would be consistent with posi¬ tron emission tomographic scan findings of altered metabo¬
three
or more
patients with
soft signs represents
true
lism of the basal ganglia,12 as well as an association of OCD with Sydenham's chorea.4,28 A descriptive comparison of these results with previous reports in other psychiatric populations is of interest. Quitkin et al19 studied schizophrenics with premorbid asociality, schizophrenics of mixed subgroups, patients with emotionally unstable character disorder, subjects with other character disorders, and patients with affective disorders. Compared with these other psychiatric groups, patients with OCD have more involuntary movements and abnormal agraphesthesia, and slightly more mirror movement, speech, and rapid alter¬ nating movement abnormalities. These other groups were not tested for abnormalities in cube drawing. Asocial schizo¬ phrenics appear to have more abnormalities on the face-hand test and hopping. A greater percent of patients with OCD in our sample had soft sign abnormalities (95%) compared with children rated as hyperactive at the age of 4 years (13%). " To date, most neurobiological studies of OCD have centered on the pharmacology and biochemistry of this disorder. Sero¬ tonergic involvement in OCD is supported by pharmacological treatment studies with serotonin reuptake blockers, such as
clomipramine, fluoxetine, and fluvoxamine,29 cerebrospinal fluid studies of 5-hydroxyindoleacetic acid,30 peripheral mea¬ sures of platelet 5-hydroxytryptamine levels,31 and biological challenge studies with oral m-chlorophenylpiperazine, which is a selective serotonin agonist reported to cause exacerbation of OCD symptoms.32,33 Serotonergic pathways ascend from nuclei in the brain stem through the median forebrain bundle and project diffusely to cortical and motor systems. Structural brain abnormalities at various levels could result in functional dysregulation of the serotonergic system. Further attempts to localize structural abnormalities in OCD have utilized imag¬ ing and neuropsychological approaches. Imaging studies with positron emission tomographic scans12 and volumetric comput¬ ed tomographic scans11 have implicated abnormalities in the basal ganglia, an area involved in the production of abnormal and choreiform movements, as well as the orbital frontal gyri. Neuropsychological studies of OCD also have reported visuo¬ spatial abnormalities consistent with right hemisphere dys¬ function,18,24,34 as well as frontal abnormalities.15,24 Additional work is needed to integrate fully structural findings with functional abnormalities in OCD. Further studies should ad¬ dress whether there are separate neurological and biochemi¬ cal subgroups of OCD, or whether neuropsychiatrie and bio¬ chemical abnormalities cluster and are interrelated in OCD. Abnormalities in coordination, abnormal movements, and vi¬ suospatial function in a subgroup of patients with OCD are consistent with other neurobiological findings in OCD. The results of this study have confirmed and expanded evidence for neurological abnormalities in at least a subgroup of patients with OCD and have confirmed the usefulness of assessing soft signs in adult anxiety disorder populations. This is exploratory work that confirms the observation that there is clumsiness in OCD and points to some intrinsic organic diathe¬ sis that other investigators have noticed. However, this is only the first step. While some other psychiatric conditions have been reported to have subtle neurological soft signs, others have not. It is not clear to what extent soft signs may be a property of psychiatric disorders in general, and to what extent they are unique to OCD. The next step would be to compare patients with OCD on these measures to other psy¬ chiatric samples. Further studies should also address whether soft signs support the distinction of OCD from other anxiety subtypes (ie, social phobia, panic disorder, etc). Additionally, soft signs may help to clarify issues of heterogeneity within OCD with regard to familial background, diagnostic subtypes (ie, washers, checkers, and pure obsessionals), treatment response, origin, and pathophysiology. This study was supported in part by grants MH-30906 and MH-37592 and Research Scientist Development Award MH-00750 (Dr Hollander) from the National Institute of Mental Health, Bethesda, Md. The authors gratefully acknowledge the helpful advice of Donald F. Klein, MD, the statistical consultation with Donald Ross, PhD, and the assistance of Concetta DeGria, MS.
References 1. Myers JK, Weissman MM, Tischler GL, Holzer LE III, Leaf PJ, Orvaschel H, Anthony JC, Boyd JH, Burke JD, Kramer M, Stoltzman R. Six-month prevalence of psychiatric disorders in three communities. Arch Gen Psychiatry.
1984;41:959-967. 2. McKeon J, McGuffin P, Robinson P. Obsessive-compulsive neurosis following head injury: a report of 4 cases. Br J Psychiatry. 1984;144:190\x=req-\
192. 3. Schilder P. The organic background of obsessions and compulsions. Am J
Psychiatry. 1938;94:1397.
4. Grimshaw L. Obsessional disorder and neurological illness. J Neurol Neurosurg Psychiatry.1964;27:229-231. 5. Capstick N, Seldrup U. Obsessional states: at study in the relationship between abnormalities occurring at birth and subsequent development of obsessional symptoms. Acta Psychiatr Scand. 1977;56:427-439. 6. Rapoport J, Elkins R, Langer DH, Sceery W, Buchsbaum MS, Gillin JC, Murphy DL, Zahn TP, Lake R, Ludlow C, Mendelson W. Childhood obsessive
compulsive disorder. Am J Psychiatry. 1981;138:1545-1554. 7. Pacella BL, Polatin P, Nagle SH. Clinical and EEG studies in obsessive compulsive states. Am J Psychiatry. 1944;100:830-838. 8. Shagass C, Roemer RA, Straumanis JJ, Josiassen J. Distinctive somatosensory evoked potential features in obsessive compulsive disorder. Biol Psychiatry. 1984;19:1507-1524. 9. Ciesielksi HR, Beech HR, Gordon PK. Some electrophysiological observations in obsessional states. Br J Psychiatry. 1981;138:479-484. 10. Behar D, Rapoport JL, Berg C, Denckla MB, Mann L, Cox C, Fidio P, Zahn T, Wolfman MG. Computerized tomography and neuropsychological test measures in adolescents with obsessive-compulsive disorder. Am J Psychiatry. 1984;141:363-369. 11. Luxenberg JS, Swedo SE, Flament MF, Friedland RP, Rapoport J, Rapoport SI. Neuroanatomical abnormalities in obsessive-compulsive disorder detected with quantitative X-ray computed tomography. Am J Psychiatry. 1988;145:1089-1093.
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
12. Baxter LR, Phelps ME, Mazziotta JC, Guze BH, Schwartz JM, Selin LE. Local cerebral glucose metabolic rates in obsessive-compulsive disorder: a comparison with rates in unipolar depression and in normal controls. Arch Gen
Psychiatry. 1987;44:211-218. 13. Pauls DL, Towbin KE, Leckman JF, Zahner GE, Cohen DJ. Gilles de la Tourette's and obsessive-compulsive disorder. Arch Gen Psychiatry. 1986;43:1180-1182. 14. Barton R. Diabetes insipidus and obsessional neurosis: a syndrome. Lancet. 1965;1:133-135. 15. Flor-Henry P, Yeudall LT, Koles ZJ, Howarth BG. Neuropsychological and power spectral EEG investigations of the obsessive-compulsive syndrome. Biol Psychiatry. 1979;14:119-130. 16. Lewin W. Selective leukotomy: a review. In: Laitinin LV, Livingston K, eds. Surgical Approaches in Psychiatry. Lancaster, England: Medical & Technical Publishing Co; 1973:69-73. 17. Talairach J, Bancard J, Geiver S, Bordas-Fener A, Bonis A, Szikla G, Rusu M. The cingulate gyrus and human behavior. Electroencephalogr Clin Neurophysiol. 1973;34:45-52. 18. Insel T, Donnelly E, Lalakea M, Alterman IS, Murphy DL. Neurological and neuropsychological studies of patients with obsessive-compulsive disorder.
Biol Psychiatry. 1983;18:741-751. 19. Quitkin FM, Rifkin A, Klein DF. Neurologic soft signs in schizophrenia and character disorders: organicity in schizophrenia with premorbid asociality and emotionally unstable character disorders. Arch Gen Psychiatry.
1976;33:845-853.
20. Nichols PL. Minimal brain dysfunction and soft signs: the collaborative perinatal project. In: Tupper D, ed. Soft Neurological Signs. New York, NY:
Grune & Stratton; 1987:179-199.
21. Denckla MB: Minimal brain dysfunction. In: Chall J, Mirsky AF, Relage KJ, eds. Education and the Brain. Chicago, Ill: University of Chicago Press;
1978:223-268. 22. Shaffer D, Schonfeld IS, O'Connor PA, Stokman C, Trautman P, Shafer S, Ng SC. Neurological soft signs and their relationship to psychiatric disorder and intelligence in childhood and adolescence. Arch Gen Psychiatry.
1985;42:342-351.
23. Denckla MB.
Neurological examination. In: Rapoport JL, ed.
Obse iv
Compulsive Disorder in Children and Adolescents. Washington, Psychiatric Press Inc; 1988:107-118.
DC: Ameri-
can
24. Rosen WG, Hollander E, Stannick V, Liebowitz MR. Test performance variables in obsessive-compulsive disorder. J Clin Exp Neuropsychol.
1988;10:73.
25. Swets JA. Measuring the accuracy of diagnostic systems. Science. 1988;24:1285-1293. 26. Stokman CJ, Shafer SQ, Shaffer D, Ng SC, O'Connor PA, Wolff RM. Assessment of neurological soft signs in adolescents: reliability studies. Dev Med Child Neurol. 1986;28:428-439. 27. Shafer SQ, Stokman CJ, Shaffer D, Ng SC, O'Connor PA, Schonfeld IS. Ten-year consistency in neurological test performance of children without focal neurological deficit. Dev Med Child Neurol. 1986;28:417-427. 28. Swedo SE, Rapoport JL, Cheslow DL, Leonard HL, Ayoub EM, Hosier DM, Wald E. Increased incidence of obsessive-compulsive symptoms in patients with Sydenham's chorea. Am J Psychiatry. 1989;146:246-249. 29. Hollander E, Liebowitz MR, Gorman JM. Anxiety disorders. In: Talbott JA, Hales RE, Yudofsky, SS, eds. American Psychiatric Press Textbook of Psychiatry. Washington, DC: American Psychiatric Press Inc; 1988:443\x=req-\ 493. 30. Thoren P, Asberg M, Bertilsson L, Mellstr\l=o"\mB, Sjoqvist F, Tr\l=a"\skman L. Clomipramine treatment of obsessive-compulsive disorder, II: biochemical
aspects. Arch Gen Psychiatry. 1980;37:1289-1294. 31. Flament MF, Rapoport JL, Murphy DL, Lake CR, Berg CJ. Biochemical
changes during clomipramine treatment of childhood obsessive-compulsive disorder. Arch Gen Psychiatry. 1987;44:219-225. 32. Zohar J, Mueller EA, Insel TR, Zohar-Kadovch RC, Murphy DL. Serotonergic responsivity in obsessive-compulsive disorder: comparison of patients and healthy controls. Arch Gen Psychiatry. 1987;44:946-951. 33. Hollander E, Fay M, Cohen B, Campess R, Gosman JM, Liebowitz MR. Serotonergic and noradrenergic sensitivity in obsessive-compulsive disorder: behavioral findings. Am J Psychiatry. 1988;145:1015-1017. 34. Diamond BM. Albrecht W, Borison RL. Neuropsychology of obsessive compulsive disorders. In: Forty-Third Annual Convention and Scientific Program, the Society of Biological Psychiatry: May 4-8,1988; Montreal, Canada.
Downloaded From: http://archpsyc.jamanetwork.com/ by a La Trobe University User on 02/08/2016
