Journal of Affective Disorders 57 (2000) 8393 www.elsevier.

com / locate / jad

Research report

Motor and cognitive aspects of motor retardation in depression

Michael P. Caligiuri*, Joel Ellwanger
Department of Psychiatry (0603), Movement Disorders Laboratory, University of California at San Diego, La Jolla, CA 92093, USA Received 22 December 1998; accepted 22 March 1999

Abstract Background: Motor retardation is a common feature of major depressive disorder having potential prognostic and etiopathological signicance. According to DSM-IV, depressed patients who meet criteria for psychomotor retardation, must exhibit motor slowing of sufcient severity to be observed by others. However, overt presentations of motor slowing cannot distinguish slowness due to cognitive factors from slowness due to neuromotor disturbances. Methods: We examined cognitive and neuromotor aspects of motor slowing in 36 depressed patients to test the hypothesis that a signicant proportion of patients exhibit motor programming disturbances in addition to psychomotor impairment. A novel instrumental technique was used to assess motor programming in terms of the subjects ability to program movement velocity as a function of movement distance. A traditional psychomotor battery was combined with an instrumental measure of reaction time to assess the cognitive aspects of motor retardation. Results: The depressed patients exhibited signicant impairment on the velocity scaling measure and longer reaction times compared with nondepressed controls. Approximately 40% of the patients demonstrated abnormal psychomotor function as measured by the traditional battery; whereas over 60% exhibited some form of motor slowing as measured by the instruments. Approximately 40% of the patients exhibited parkinsonian-like motor programming decits. A ve-factor model consisting of motor measures predicted diagnosis among bipolar and unipolar depressed patients with 100% accuracy. Limitations: The ability of motor measures to discriminate bipolar from unipolar patients must be viewed with caution considering the relatively small sample size of bipolar patients. Conclusions: These ndings suggest that a subgroup of depressed patients exhibit motor retardation that is behaviorally similar to parkinsonian bradykinesia and may stem from a similar disruption within the basal ganglia. 2000 Elsevier Science B.V. All rights reserved.
Keywords: Depression; Motor retardation; Psychomotor impairment

1. Introduction Current diagnostic criteria for major depression include the presence of psychomotor retardation
*Corresponding author. Tel.: 1 1-619-552-8585, ext. 2163. E-mail address: mcaligiuri@ucsd.edu (M.P. Caligiuri)

which, according to DSM-IV, . . . must be severe enough to be observed by others (American Psychiatric Association, 1994; p. 321). Most features of motor retardation are easily observed. These consist of reduced speed, slow speaking rate, delayed motor initiation, body immobility, loss of facial expression, and postural abnormalities (Parker et al., 1993).

0165-0327 / 00 / $ see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0165-0327( 99 )00068-3

84

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93

Parker et al. (1993) suggested a trunk and branch analogy as a means of modeling motor disturbances in depression. In this model, a truncal psychic component arborises into retarded and agitated branches. While the model accounts for three potentially overlapping factors, it ignores the possibility that the root, which gives rise to the truncal component, may contain elements shared by other neuropsychiatric disorders, such as Parkinsons disease (PD). Restricting the criteria for motor disturbances in depression to those that are observable, as suggested by DSM-IV, minimizes the likelihood that the neurobiology of depression can be fully understood. It is noteworthy that the motor signs which characterize motor retardation are indistinguishable from characteristics of idiopathic PD. This has led some investigators to hypothesize the presence of a common neurobiological mechanism for motor retardation in depression and parkinsonism (Sachdev and Aniss, 1994), while others have found marked differences in the motor behavior of PD and depressed patients (Fleminger, 1992). The evidence for a mechanistic overlap between depression and parkinsonism is strong. Depression has been reported as a feature in PD (Rogers et al., 1987; Tom and Cummings, 1988; Tandberg et al., 1997) and psychosis (Prosser et al., 1987), and it is possible that some aspects of motor retardation may actually signal the presence of specic neuromotor pathology or a medication side effect (Casey, 1994). Overall, the overt signs associated with slowness, such as reduced movement velocities, increased reaction times, and increased movement times are similar in depression and parkinsonism. In a previous study, (Caligiuri et al., 1998a), we described a novel approach for separating motor slowing into two components: a basic neuromotor component derived from movement velocity and a higher level cognitive or attentional component derived from a measure of reaction time. The purpose of this dichotomy is to distinguish patients whose slowness is driven by neuromotor factors, such as medication side effects, from those whose slowness is related to cognitive factors. Using wrist movement velocity, we measured the subjects ability to program (or scale) movement velocity in anticipation of changing target distances. The idea

that neuromotor slowing would manifest as impaired velocity scaling stems from extensive research on the mechanism of parkinsonian bradykinesia (Draper and Johns, 1964; Berardelli et al., 1986; Warabi et al., 1986; Benecke, 1989; Hufschmidt and Lucking, 1995). Our ndings conrmed the presence of a motor programming disorder in all of the Parkinsonian patients studied and in nearly one-third of the psychosis patients. In a follow-up study of motor performance of the depressed and nondepressed psychosis patients (Caligiuri et al., 1998b), the nondepressed patients exhibited normal psychomotor performance, but scored in the abnormal range on measures of motor programming. Conversely, depressed patients exhibited abnormal psychomotor performance, but scored within normal limits on measures of motor programming, even after controlling for medication and age. These ndings suggest that motor slowing may have a cognitive basis in the presence of depression and a neuromotor basis in the absence of depression. In the present study, we used the velocity scaling measure together with a battery of traditional psychomotor measures to test the hypothesis that a signicant proportion of depressed patients exhibit motor programming disturbances. Disturbances in the programming of movement velocity have been thought of as contributing to the bradykinesia seen in patients with Parkinsons disease (Hallett and Khoshbin, 1980; Hallet, 1985). Impaired motor programming in depression would suggest that for these patients, their motor retardation may stem from disturbed dopamine neurotransmission within the basal ganglia. Alternatively, motor slowing in the absence of a velocity programming decit would suggest a more cognitive origin of the slowness in depression.

2. Methods

2.1. Patients
Thirty-six depressed outpatients (12 women and 24 men) and 22 nondepressed healthy controls (nine women and 13 men) were studied. The patients had a mean (S.D.) age of 45.1 (10.7) years and the controls

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93

85

had a mean age of 46.7 (16.6) years. Thirty-one of the patients met DSM-IV criteria for major depressive disorder; whereas ve patients met DSM-IV criteria for bipolar II disorder. Patients were recruited from the VA San Diego Healthcare System Clinical Research Center within the Department of Psychiatry. Ten patients were currently enrolled in cognitive behavioral therapy, while the remaining 26 patients were enrolled into ongoing clinical trials. The control subjects were recruited from psychiatry service staff and medical center volunteers. For all patients, the assessments described herein were made during their initial visit of their respective trials. Because the majority of the patients were enrolling into pharmaceutical efcacy trials, most patients were unmedicated at the time of motor assessment. Only 11 of the 36 patients were on antidepressants at the time of assessment. Three patients were receiving antipsychotic medication and two patients were on mood stabilizers. Severity of depression was rated using the 17-item Hamilton Rating Scale for Depression (HRSD; Hamilton, 1967) for all patients enrolled into pharmaceutical trials and the Beck Depression Inventory (BDI; Beck, 1972) for patients enrolled in behavioral therapy. On the basis of established cut-points for these instruments (Knesevich et al., 1977; Beck et al., 1988), two patients exhibited mild, 22 moderate, and 12 severe depression.

2.2. Instrumental assessment of motor programming and slowing

Following the clinical assessment, patients who met DSM-IV criteria for major depressive disorder or bipolar disorder were provided a description of the study. Those who signed informed consent to participate in research were referred to the Movement Disorders Laboratory for motor assessments. Motor assessments were conducted within 1 week of the clinical depression ratings. Reassessments of clinical symptoms at the time of the motor assessments were deemed unnecessary because of this short interval. Three measures were obtained using a single instrument: velocity scaling, reaction time, and peak instantaneous velocity. A platform instrumented with

a rotation sensor attached to a handle (electrogoniometer) was used to record wrist rotation. A line drawing of the instrumental set-up is shown in Fig. 1. The analog output from the electrogoniometer was digitized at 100 samples / s (12-bit) and custom software was used to convert the digitized signal to a large cursor displayed on the monitor. The position of the cursor along the horizontal plane was calibrated in degrees of rotation. In addition to the rotation-based cursor, target boxes were displayed on the monitor, also along the horizontal plane and were located at 25 or 458 left of midline for right wrist exion, and right of midline for left wrist exion. Subjects sat facing a computer monitor while holding the electrogoniometer handle with one hand. Subjects were instructed to ex the wrist as quickly and as accurately as possible when a target box appeared on the screen causing the cursor to move horizontally toward one of two target boxes. The target remained displayed for 2 s with an interstimulus interval of 2 s. Two blocks of trials or movements, consisting of 16 trials for each of two randomly presented target locations were administered for each hand, for a total of 64 movements. For velocity scaling, the peak instantaneous velocities and angular distances from each of the 32 trials for a given hand were used to compute the difference scores for 25 and 408 targets. A velocity scaling score was derived from the equation: VS 5 (V 40 2 V 25) /(D 40 2 D 25), where V and D represent velocity and angular distance, respectively, for the 40 and 258 targets. Reaction time was measured by obtaining the time from stimulus onset to movement onset in milliseconds. Peak instantaneous velocity was measured by computing the rst derivative of rotation (angular velocity). The reliability of the procedure was established in a previous study (Caligiuri et al., 1998a). In that study, 10 psychosis patients were retested within the same session with the dominant hand. The reliability coefcient for the slope was 0.98 suggesting very high repeatability for this measure. This measure may be considered an example of open loop motor control, insofar as the rapid wrist movements were ballistic in nature (Schmidt, 1982). Accordingly, analyses were based on the initial peak velocity rather than average velocity. Feedback-derived corrections would appear

86

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93

Fig. 1. Line drawing of the subjectapparatus interface showing the wrist electrogoniometer, amplier and computer monitor. Subject sits facing the monitor and responds to target cues presented on the screen. Flexion movements of the wrist cause the subjects cursor (circle) to move to target (box). See text for data acquisition and analysis details.

as multiple velocity peaks which were not incorporated into the measure of velocity scaling.

bol substitution test (DSST), Trails A, and Trails B. These tests were administered and scored by a trained staff member.

2.3. Assessment of psychomotor function 2.4. Statistical analyses

Three psychomotor assessments were administered for the purpose of comparing performance on traditional measures of psychomotor retardation with performance on the instrumental measures. The psychomotor measures are part of the comprehensive HalsteadReitan battery and included the digit symIndependent t -tests were used to test differences between patients and controls on all instrumental motor measures once the group data were found to satisfy all assumptions for a parametric statistic. Pearson correlational analyses were used to examine

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93

87

the relationships between severity of depression and performance on motor measures. Lastly, the effects of medication on motor variables were examined using t -tests by comparing patients on versus off any of three classes of medication: antipsychotics, antidepressants, and mood stabilizers.

3. Results

possibility that their motor slowing may have its origin in faulty motor programming. Four of the ve patients (80%) with bipolar disorder exhibited abnormal scores on the velocity scaling measure compared with 10 of the 31 (32.2%) patients with unipolar depression ( x 2 5 4.13; P 5 0.04). For reaction time, none of the bipolar patients met criteria for abnormality; whereas 11 of the 31 patients (35.5%) with unipolar depression exhibited abnormally long reaction times.

3.1. Instrumental measures of motor slowing 3.2. Psychomotor disturbances

Table 1 shows the descriptive statistics for patients and controls for the instrumental motor measures. Results of independent t -tests indicated signicant differences between depressed patients and nondepressed healthy controls on all of the instrumental motor measures. Depressed patients exhibited signicantly longer reaction times, lower velocities, and lower scores on the velocity scaling measure. Additional analyses were performed on the individual data to estimate the prevalence of impaired motor function. Patients were considered to exhibit abnormality if their scores exceeded cut-points of 1.5 S.D. above the normal control mean for reaction time and 1.5 S.D. below the normal control mean for peak velocity, and velocity scaling. Twenty-two of the 36 patients (61.1%) exhibited abnormalities on at least one instrumental measure based on these cut-points. The proportions of patients meeting this criteria for specic abnormalities of reaction time, velocity scaling, and peak velocity were 30.5% (11 patients), 40% (14 patients) and 27.8% (10 patients), respectively. Nine patients (25.7%) exhibited abnormal velocity scaling in addition to abnormalities on at least one other instrumental measure, suggesting the Regarding performance on the psychomotor battery, the depressed patients as a group did not exhibit impaired performance relative to standard published age and education corrected norms (Heaton et al., 1986, 1991). Four patients were unable to complete the psychomotor battery for various reasons. The raw scores for these measures were converted to t -scores for the purpose of correcting for age and education. The mean corrected scores (and standard deviations) for the remaining 32 patients for Trails A, Trails B, and the DSST were 48.0 (11.2), 48.8 (9.7) and 49.9 (10.0), respectively, which did not differ from the published normal mean value of 50.0 (10.0). However, individually, nine patients (28.1%) met standard criteria for impaired performance on Trails A, seven patients (21.8%) met standard criteria for impaired performance on Trails B, and eight patients (25%) met standard criteria for impaired performance on the DSST. Overall 14 patients (43.7%) exhibited impairment on at least one of the three psychomotor tests. Interestingly, a similar proportion of bipolar versus unipolar depressed patients exhibited abnormal psychomotor performance on at

Table 1 Means (and standard deviations) for the instrumental measures of motor slowing for 22 non-depressed healthy controls and 36 depressed patients Group Controls Patients t -Test P value Reaction time (ms) 366.07 (105.91) 464.62 (88.27) 2 3.82 0.0003 Velocity at 258 (deg / s) 146.77 (59.10) 110.90 (53.23) 2.38 0.02 Velocity at 408 (deg / s) 198.91 (71.95) 147.02 (66.36) 2.79 0.007 Velocity scaling (deg / s / deg) 4.27 (1.93) 2.67 (1.49) 3.19 0.002

88

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93

least one of the three measures with two of the ve bipolar patients (40%) and 12 of the 27 unipolar patients (44%) demonstrating a disturbance.

3.3. Relationships between motor slowing and clinical factors

Correlational analyses were performed to examine the relationship between severity of motor impairment and clinically relevant variables including severity of depression, medication status, and age. DSST score and peak velocity to 408 targets were signicantly correlated with HRSD with coefcients of r 5 2 0.61 (P 5 0.002) and r 5 2 0.44 (P 5 0.033), respectively. These ndings indicate that as depression severity increased, performance on DSST worsened and movement velocities decreased. Eight patients who were rated at least mild on the motor retardation item of the HRSD were compared on the ve measures of motor slowing with 18 patients who were rated as not exhibiting motor retardation. While none of the comparisons reached statistical signicance, patients exhibiting clinically rated motor retardation had lower scores for both the Trails A (41.4 vs. 48.7) and DSST (45.7 vs. 50.1), slightly longer reaction times (482 vs. 474 ms), lower peak velocities (132 vs. 164 deg / s), and lower velocity scaling scores (2.06 vs. 3.00 deg / s / s) than patients rated normal on the motor retardation item of the HRSD. Seven of the eight patients with observer-rated motor retardation exhibited abnormal velocities, three of the eight exhibited abnormal reaction times, and four of the eight exhibited abnormal velocity scaling based on the cut-points described above.

Table 2 shows the results of analyses of the effects of medication status on selected instrumental and psychomotor variables. Patients on antidepressants were not less depressed than patients off antidepressants. Of the 11 medicated patients, 27% met criteria for severe depression; whereas 36% of the 25 unmedicated patients met criteria for severe depression ( x 2 5 0.46; P . 0.10). The ndings for the effects of antipsychotics and mood stabilizers on motor slowing are difcult to interpret because of the small sample sizes. However, the results for antidepressants indicate a benecial effect on the DSST and average reaction time. Patients on antidepressants scored signicantly higher on the DSST than patients off antidepressants (t 5 2 2.33; P 5 0.026). Similarly, patients on antidepressants exhibited shorter reaction times than patients off antidepressants (t 5 2.60; P 5 0.013). The question of whether the motor measures offer any diagnostic discriminability was addressed using logistic regression. The results of a ve-factor model for predicting diagnosis (bipolar disorder versus unipolar depression) are shown in Table 3. The ve-factor model accurately classied 100% of the patients into their respective diagnoses (F 5 9.91; P , 0.0001). Removing Trails A and DSST from the model reduced the overall accuracy to 91.4%, classifying 80% of the bipolar disorder patients and 93.3% of the unipolar depressed patients. Positive b values in Table 3 indicate higher scores; whereas negative b values indicate lower scores for one group versus the other. The results indicate that patients with bipolar disorder had higher scores on Trails A, lower scores on the DSST, higher peak velocities, shorter reaction times, and lower scores

Table 2 Means (and S.D.) for the instrumental (reaction time, RT) and non-instrumental (trails A and digit symbol substitution test, DSST) measures of psychomotor function that were found to be inuenced by medication status Antipsychotics On (n 5 3) Trails A DSST RT 64.0** (19.8) 55.0 (26.8) 377.7* (68.1) Off (n 5 33) 47.1 (10.2) 49.5 (9.1) 474.6 (86.9) Antidepressants On (n 5 11) 52.9 (12.7) 56.1** (10.6) 411.2** (54.7) Off (n 5 25) 46.1 (10.2) 47.6 (8.9) 488.1 (90.7) Mood stabilizers On (n 5 2) 53.0 (15.5) 47.0 (11.3) 467.6 (143.7) Off (n 5 34) 47.8 (11.2) 50.1 (10.3) 466.2 (88.1)

*Different from off medication at P , 0.10; **different from off medication at P , 0.05.

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93 Table 3 Results of the logistic regression for classifying diagnosis as either bipolar disorder or unipolar depression a Factor Trails A DSST Peak velocity (408 ) Reaction time Velocity scaling

89

b
2 0.498 0.557 2 0.306 0.434 0.536

t -ratio 2 3.12 3.50 2 2.48 3.54 4.44

P value 0.004 0.002 0.020 0.002 0.0002

a Positive b values indicate higher scores; whereas negative b values indicate lower scores for unipolar depressives. This vefactor model classied patients into their respective diagnoses with 100% accuracy.

on the velocity scaling measure compared with unipolar depressed patients.

4. Discussion The results of the present study are consistent with the notion that motor retardation is an essential feature of major depressive disorder. As a group, the patients in the present study exhibited longer reaction times and lower peak velocities compared with healthy control subjects. Over 40% of the depressed exhibited mild impairment on traditional measures of psychomotor function. Similar results have been previously reported for reaction times (see Sobin and Sackeim, 1997, for review). In Section 1 we emphasized that sensitive instrumental measures of motor slowing may allow detection of motor system abnormalities that are not observed clinically. This is exactly what was observed in the present study in which 22% of the patients were rated as having motor retardation based on the HRSD compared with 60% based on instrumental measures. There were no signicant relationships between scores on the retardation item on the HRSD and abnormalities as measured by the present technologies; however, seven of the eight patients who were impaired on the HRSD motor retardation item also exhibited abnormalities on at least one instrumental measure and six patients exhibited abnormalities on two of the three instrumental measures. The absence of a signicant statistical relationship between the HRSD motor retardation item and performance on the instrumental

battery may be explained by a small sample size and insufcient statistical power. A novel nding of the present study was that many patients exhibited decits in the programming of movement velocity. As many as 40% of the patients were unable to increase movement velocity as target distance increased. For simple ballistic movements, the amplitude of the initial agonist muscle burst governs movement velocity (Freund and Budingen, 1978). On the basis of disturbances in the generation and timing of a normal muscle ring pattern in patients with Parkinsons disease, investigators have argued that parkinsonian bradykinesia is primarily a decit in the timing or programming of movement velocity (Draper and Johns, 1964; Hallett and Khoshbin, 1980; Berardelli et al., 1986; Warabi et al., 1986; Benecke, 1989; Hufschmidt and Lucking, 1995). The present ndings support the concept that motor slowing in depressive disorders and the bradykinesia of parkinsonism may stem from a common underlying mechanism, at least for a subgroup of patients. Such a mechanism could involve reduced mesocorticolimbic and prefrontal dopamine neurotransmission (Price et al., 1978; Javoy-Agid and Agid, 1980; Scatton et al., 1982). Support for a dopaminergic role in depression comes from several arenas. Lauterbach et al. (1997) noted that pallidothalamic lesions were associated with secondary depression in patients with cerebral vascular accidents localized to subcortical areas. On the basis of an extensive review of the clinical and neuropharmacological literature, Brown and Gershon (1993) concluded that dopamine deciency was responsible for the psychomotor retardation and diminished motivation observed in depression. Pharmacological ndings indicate that depressed patients have been reported to exhibit reductions in the dopamine metabolite, homovanillic acid, relative to control subjects (Post et al., 1980) and inhibited prolactin release in response to apomorphine (reective of abnormal upregulation of postsynaptic dopamine receptors) (Jimerson and Post, 1984). Magnetic resonance imaging studies also support the notion that the basal ganglia are involved in the pathogenesis of depression. Husain et al. (1991) found that patients meeting DSM-III criteria for major depression demonstrated diminished putamenal volumes. Overall, these ndings suggest that dopamine

90

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93

may play a key role in the expression of motor retardation in depression. The present nding that a subset of unipolar depressed patients, and nearly all of the bipolar depressed patients, exhibited a pattern of motor slowing consistent with a parkinsonian-like motor programming decit provides further support for a dopaminergic hypothesis in depression. The results of the present study may offer insight into which aspects of psychomotor slowing are related to state changes and which may be considered trait characteristics of depression. It has been argued that subtypes of depression may be distinguished on the basis of psychomotor disturbances with melancholic type demonstrating more pervasive psychomotor disturbances (Sobin and Sackeim, 1997). The present study was unable to differentiate patients on the basis of clinical characteristics to support or weaken this notion. However, correlational analyses between motor signs and clinical symptoms indicated that some aspects of motor slowing were related to severity of depression. In particular, we found that movement velocity decreased with increasing severity of depression and performance on the DSST worsened with increasing severity of depression. These ndings are partially consistent with the literature on the relationship between mood symptoms and psychomotor behavior. For example, in a study of 20 melancholic patients, Moffoot et al. (1994) found that reaction time and movement speed tracked changes in depression, whereas performance on the DSST did not track symptoms over a 24-h period. The differences between this and the present study could be explained by differences in study design. A second set of results from the present study also supports the notion that performance on temporal measures of motor function, such as reaction time, may be related to state of depression. While we found that patients on antidepressants scored signicantly higher on the DSST and exhibited shorter reaction times than patients off antidepressants, the cross-sectional design of the present study limits the interpretation of this nding. Benecial effects of antidepressants on reaction time have been reported previously (vanHoof et al., 1993; Sabbe et al., 1997). vanHoof et al. (1993) studied six patients prior to and following 23 months of treatment with conventional tricyclic antidepressants. Their changes in

HRSD scores correlated signicantly with changes in reaction time. Similarly, Sabbe et al. (1997) followed 22 patients through 6 weeks of treatment with uoxetine and found that antidepressants imparted a signicant reduction in the reaction time but not movement time. The ndings of the present study are consistent with these previous studies; however, in the absence of longitudinal data, we cannot determine whether antidepressants which reduce the severity of depression impart a parallel reduction in the severity of the motor impairment or whether the antidepressants improve motor performance independent of change in mood symptoms. The distinction between bipolar and unipolar depression on the basis of a battery of motor measures supports the idea that some aspects of motor slowing may be trait characteristics in depression. Despite the very small sample of patients and the preliminary nature of these ndings we found that nearly all of the bipolar depressed patients demonstrated abnormalities on the velocity scaling measure compared with less than one-third of the unipolar depressed patients. Conversely, none of the bipolar depressed patients demonstrated abnormalities on the reaction time measure compared with over one-third of the unipolar depressed patients. These ndings suggest that motor retardation associated with bipolar disorder more closely resembles parkinsonian bradykinesia compared with motor retardation in unipolar depression. This is consistent with previous ndings by Sachdev and Aniss (1994) on similarities between retarded melancholia and Parkinsons disease and the factor analyses of Parker et al. (1993). Motor retardation can present with a variety of neuropsychiatric conditions including dementia (Nebes et al., 1998), Parkinsons disease (Sachdev and Aniss, 1994), schizophrenia (Prosser et al., 1987), and depression (Sobin and Sackeim, 1997). The observable characteristics of motor slowing are virtually indistinguishable across the spectrum of neuropsychiatric disorders contributing to the hypothesis that there is a common unied motor retardation syndrome (Widlocher, 1983; Gunther et al., 1988; Bermanzohn and Siris, 1992). An alternative hypothesis pursued in this study, is that motor retardation may have different etiologies and pathophysiologies in at least some of the neuro-

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93

91

psychiatric conditions in which it is observed. Findings from this and previous studies (Caligiuri et al., 1998a,b) suggest that in psychosis or bipolar depression, motor retardation may stem from a neurobiological mechanism similar to parkinsonian bradykinesia; whereas in unipolar depression motor retardation appears to have both motor and cognitive bases. Clinical factors which predispose a patient to exhibit either the cognitive, motor, or cognitive motor form of motor retardation remain unclear. On the basis of the present ndings, we propose that motor slowing may have a distinct neurobiological basis (neuromotor retardation) in some forms of depression, while in others it has a generalized cognitive basis (psychomotor retardation). We found marked heterogeneity in motor performance among unipolar depressed patients. Several factors could explain this, including variability in prescribed medication, age, illness severity, and illness duration. For example, among unipolar depressed patients, illness severity explained 36% of the variability in DSST performance, 23% of the variability in peak velocity, and 18% of the variability in velocity scaling. Despite this heterogeneity, the motor tasks offered meaningful diagnostic or differential diagnostic information. A ve-factor model consisting of Trails A, DSST, peak velocity, reaction time, and velocity scaling correctly classied 100% of the patients as unipolar or bipolar depressed. Thus, in cases of diagnostic uncertainty when a patient presents depression and motor slowing, the pattern of performance on a motor battery which includes psychomotor and neuromotor measures may be helpful in differentiating bipolar from unipolar depression. Insofar as the management strategies may differ for bipolar and unipolar depression, any information which increases diagnostic precision could lead to more effective treatment.

the bipolar depressed patients exhibited a fundamental disturbance in motor programming and that this disturbance may remain hidden from the clinical presentation of motor retardation. The existence of a parkinsonian-like bradykinesia masquerading as motor retardation in patients with depression has implications for the potential pharmacological management of depression. While there is some evidence to suggest that motor retardation predicts a favorable response to pharmacotherapy (Ranelli and Miller, 1981) and ECT (Buchan et al., 1992), the ndings are mixed (see Sobin and Sackeim, 1997). Perhaps the advantage of motor retardation as a prognosticator of favorable treatment outcome is carried largely by a subgroup of patients who exhibit a fundamental neuromotor disturbance.

Acknowledgements This research was supported in part from USPHS grants (P30-MH496771, T32-MH19934), the UCSD Mental Health Clinical Research Center (MH30914), and the Department of Veteran Affairs VISN-22 MIRECC. The authors wish to thank the following individuals for their assistance on this research: Dr Chris Gillin, Dr John McQuaid, Kathy Resovsky, Lorraine Goyette, and Grace Shiao.

References
American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, 4th ed., APA Press, Washington, DC. Beck, A.T., 1972. Measuring depression: The depression inventory. In: Williams, T.A., Katz, M.M., Shields, J.A. (Eds.), Recent Advances in the Psychobiology of Depressive Illness, US Government Printing Ofce, Washington, DC, pp. 299 302. Beck, A.T., Steer, R.A., Garbin, M.G., 1988. Psychometric properties of the Beck Depression Inventory: twenty-ve years of evaluation. Clin. Psychol. Rev. 8, 77100. Benecke, R., 1989. The pathophysiology of Parkinsons Disease. In: Quinn, N.P., Jenner, P.G. (Eds.), Disorders of Movement, Academic Press, San Diego, pp. 5972. Berardelli, A., Dick, J.P.R., Rothwell, J.C., Day, B.L., Marsden, C.D., 1986. Scaling of the size of the rst agonist EMG burst during rapid wrist movements in patients with Parkinsons disease. J. Neurol. Neurosurg. Psychiatry 49, 12731279. Bermanzohn, P.C., Siris, S.G., 1992. Akinesia: a syndrome

5. Conclusions Patients may meet DSM-IV criteria for major depressive disorder without exhibiting overt signs of motor retardation. DSM-IV explicitly dened psychomotor retardation as an observable phenomena. The results from the present study suggest that over 30% of the unipolar depressed patients and 80% of

92

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93 dopamine agonists in depression. In: Post, R.M., Ballenger, J.C. (Eds.), Neurobiology of Mood Disorders, Williams and Wilkins, Baltimore, MD. Javoy-Agid, F., Agid, Y., 1980. Is the mesolimbic dopaminergic system involved in Parkinsons disease? Neurology 30, 1326 1331. Knesevich, J., Biggs, J., Clayton, P., Ziegler, V., 1977. Validity of the Hamilton Rating Scale for Depression. Br. J. Psychiatr. 131, 4952. Lauterbach, E.C., Jackson, J.G., Price, S.T., Wilson, A.N., Kirsh, A.D., Dever, G.E.A., 1997. Clinical, motor, and biological correlates of depressive disorders after focal subcortical lesions. J. Neuropsychiatr. Clin. Neurosci. 9, 259266. Moffott, A.P.R., OCarroll, R.E., Dick, S., Ebmeier, K.P., Goodwin, G.M., 1994. Diurnal variation of mood and neuropsychological function in major depression with melancholia. J. Affect. Disord. 32, 257269. Nebes, R.D., Halligan, E.M., Rosen, J., Reynolds, C.F., 1998. Cognitive and motor slowing in Alzheimers disease and geriatric depression. J. Int. Neuropsychol. Soc. 4, 426434. Parker, G., Hadzi-Pavlovic, D., Brodaty, H., Boyce, P., Mitchell, P., Wilhelm, K., Hickie, I., Eyers, K., 1993. Psychomotor disturbances in depression: dening the constructs. J. Affect. Disord. 27, 255265. Post, R.M., Ballenger, J.C., Goodwin, F.K., 1980. Cerebrospinal uid studies of neurotransmitter function in manic and depressive illness. In: Wood, J.C. (Ed.), Neurobiology of Cerebrospinal Fluid, Plenum Press, New York. Price, K.S., Farley, I.J., Hornykiewicz, O., 1978. Neurochemistry of Parkinsons disease: relation between striatal and limbic dopamine. In: Roberts, J., Woodruff, G.N., Iversen, L.L. (Eds.), Dopamine. Advances in Biochemical Psychopharmacology, Vol. 19, Raven Press, New York, pp. 293300. Prosser, E.S., Csernansky, J.G., Kaplan, J., Thiemann, S., Becker, T.J., Hollister, L.E., 1987. Depression, parkinsonian symptoms, and negative symptoms in schizophrenics treated with neuroleptics. J. Nerv. Ment. Dis. 175, 100105. Ranelli, C.J., Miller, E.R., 1981. Behavioral predictors of amitriptyline response in depression. Am. J. Psychiatry 138, 30 34. Rogers, D., Lees, A.J., Smith, E., Trimble, M., Stern, G.M., 1987. Bradyphrenia in Parkinsons disease and psychomotor retardation in depressive illness: an experimental study. Brain 110, 761776. Sabbe, B., vanHoof, J., Hulstijn, W., Zitman, F., 1997. Depressive retardation and treatment with uoxetine: assessment of the motor component. J. Affect. Disord. 43, 5361. Sachdev, P., Aniss, A.M., 1994. Slowness of movement in melancholic depression. Biol. Psychiatry 35, 253262. Scatton, B., Rouquier, L., Javoy-Agid, F., Agid, Y., 1982. Dopamine deciency in the cerebral cortex in Parkinsons disease. Neurology 32, 10391040. Schmidt, R.A., 1982. In: Motor Control and Learning, Human Kinetics Publishers, Champaign, IL. Sobin, C., Sackeim, H.A., 1997. Psychomotor symptoms of depression. Am. J. Psychiatry 154, 417. Tandberg, E., Larsen, J.P., Aarsland, D., Laake, K., Cummings,

common to parkinsonism, retarded depression, and negative symptoms of schizophrenia. Compr. Psychiatry 33, 221232. Brown, A.S., Gershon, S., 1993. Dopamine and depression. J. Neural Transm. 91, 75109. Buchan, H., Johnstone, E., McPherson, K., Palmer, R.L., Crow, T.J., Brandon, S., 1992. Who benets from electroconvulsive therapy? Combined results from the Leicester and Northwick Park trials. Br. J. Psychiatry 160, 355359. Caligiuri, M.P., Lohr, J.B., Ruck, R.K., 1998a. Scaling of movement velocity: a measure of neuromotor retardation in patients with psychopathology. Psychophysiology 35, 431437. Caligiuri, M.P., Lohr, J.B., Ellwanger, J., 1998b. Effects of depression on motor slowing. Biol. Psychiatry Abstr. 43, 86. Casey, D.E., 1994. Motor and mental aspects of acute extrapyramidal syndromes. Acta Psychiatr. Scand. 89 (Suppl. 380), 14 20. Draper, I.T., Johns, R.J., 1964. The disordered movement in parkinsonism and the effect of drug treatment. Bull. Johns Hopkins Hosp. 115, 465480. Fleminger, S., 1992. Control of simultaneous movements distinguishes depressive motor retardation from Parkinsons disease and neuroleptic parkinsonism. Brain 115, 14591480. Freund, H.J., Budingen, H.J., 1978. The relationship between speed and amplitude of the fastest voluntary contractions of human arm muscles. Exp. Brain Res. 31, 112. Gunther, W., Gunther, R., Streck, P., Romig, H., Rodel, A., 1988. Psychomotor disturbances in psychiatric patients as a possible basis for new attempts at differential diagnosis and therapy. III Cross validation study on depressed patients: the psychotic motor syndrome as a possible state marker for endogenous depression. Eur. Arch. Psychiatr. Neurol. Sci. 237, 6573. Hallett, M., 1985. Quantitative assessment of motor deciency in Parkinsons disease: ballistic movements. In: Delwaide, P.J., Agnoli, A. (Eds.), Clinical Neurophysiology in Parkinsonism, Elseviers, Amsterdam, pp. 7582. Hallett, M., Khoshbin, C., 1980. A physiological mechanism of bradykinesia. Brain 103, 301314. Hamilton, M., 1967. Development of a rating scale for primary depressive illness. Br. J. Soc. Clin. Psychiatry 6, 278296. Heaton, R.K., Grant, I., Matthews, C.G., 1986. Differences in neuropsychological test performance associated with age, education, and sex. In: Grant, I., Adams, K.M. (Eds.), Neuropsychological Assessment in Neuropsychiatric Disorders: Clinical Methods and Empirical Findings, Oxford Press, New York, pp. 100120. Heaton, R.K., Grant, I., Matthews, C.G., 1991. Comprehensive Norms for Expanded HalsteadReitan Battery: Demographic Corrections, Research Findings, and Clinical Application, Psychological Assessment Resources, Odessa, FL. Hufschmidt, A., Lucking, C.H., 1995. Abnormalities of tracking behavior in Parkinsons disease. Mov. Disord. 10, 267276. Husain, M.M., McDonald, W.M., Doraiswamy, P.M., Figiel, G.S., Na, P., Escalona, R., Boyko, O.B., Nemeroff, C.B., Krishnan, R.R., 1991. A magnetic resonance imaging study of putamen nuclei in major depression. Psychiatr. Res. Neuroimaging 40, 9599. Jimerson, D.C., Post, R.M., 1984. Psychomotor stimulants and

M.P. Caligiuri, J. Ellwanger / Journal of Affective Disorders 57 (2000) 83 93 J.L., 1997. Risk factors for depression in Parkinsons disease. Arch. Neurol. 54, 625630. Tom, T., Cummings, J.L., 1988. Depression in Parkinsons disease. Pharmacological characteristics and treatment. Drugs Aging 12, 5574. vanHoof, J.J.M., Hulstijn, W., vanMier, H., Pagen, M., 1993. Figure drawing and psychomotor retardation: preliminary report. J. Affect. Disord. 29, 263266.

93

Warabi, T., Noda, H., Yanagisawa, N., Tashiro, K., Shindo, R., 1986. Changes in sensorimotor function associated with the degree of bradykinesia of Parkinsons disease. Brain 109, 12091224. Widlocher, D., 1983. Retardation: A basic emotional response. In: Davis, J.M., Maas, J.W. (Eds.), The Affective Disorders, American Psychiatric Press, Washington, DC, pp. 165181.