Abnormalities in Motor Physiology in Bipolar Disorder

A lthough the pathophysiology of bipolar disorder is unknown, many investigators have proposed that it may relate to abnormal functioning of brain circuitry, particularly circuitry involving the basal ganglia. Such basal ganglia circuitry involvement is suggested by imaging studies of bipolar disorder. 1 – 6 Also, bipolar disorder patients have been reported to have a greater risk for developing drug-induced movement disorders, such as Tourette’s, than schizophrenia patients, 7 and the presence of catatonia, a motor disorder, is more likely to occur in the context of bipolar disorder than in schizophrenia or unipolar depression. 8 , 9 Collectively, these observations implicate an abnormality of the dopamine system in the pathophysiology of bipolar disorder.

We considered that if disturbances in basal ganglia functioning are important in bipolar disorder, then it is possible that patients with the condition might show abnormalities in aspects of motor physiology that have been considered to relate to the functioning of these structures. Over the years we have developed quantitative instrumental approaches to measuring abnormalities in motor function that appear to be related to basal ganglia dysfunction. 10 – 12 The two main methods we have developed assess the ability to maintain steady-state force and the ability to scale velocity with distance. Although we have observed that patients with clinically observable movement disorders show abnormalities on these measures (with abnormalities in force steadiness being seen in movement disorders marked by a hyperdopaminergic state, and velocity scaling abnormalities being associated more with a hypodopaminergic condition), 13 we also have found that patients without obvious clinical movement problems may demonstrate abnormalities on these measures. In particular, we have found that patients with schizophrenia may show abnormalities on these measures without evidence of frank tardive dyskinesia or parkinsonism. 14

We hypothesized that patients with bipolar disorder would show abnormalities on these measures of force steadiness and velocity scaling. Further, because of the possible importance of hyperdopaminergia in mania 15 , 16 and hypodopaminergia in depression, 17 , 18 we considered that performance on the force steadiness and velocity scaling measures would correlate with symptoms of mania and depression, respectively. In particular, we hypothesized that the degree of manic symptoms would positively correlate with force steadiness scores, and the degree of depressive symptoms would negatively correlate with velocity scaling scores (lower velocity scaling scores indicate less impairment). A third objective of the study was to explore whether impaired motor performance might relate to current pharmacotherapy.

METHOD

Subjects and Medication Status

We studied 67 individuals (39 men and 28 women) meeting DSM-IV criteria 19 for bipolar disorder and 47 healthy comparison subjects (31 men and 16 women). We recruited these subjects through advertisements placed in the local media and Internet sites for research recruitment. All subjects signed Institutional Review Board-approved consent for voluntary participation in this research. The mean age of the bipolar group was 44.3 (SD=10.9) years and the mean age for the healthy comparison group was 39.9 (SD=13.2) years, which was not significantly different. All patients had been either off medication or on stable doses of medication for at least 1 month prior to undergoing motor assessment. The breakdown of the bipolar disorder patients by medication type was as follows: unmedicated (N=11); mood stabilizers (N=46); antidepressants (N=37); antipsychotics (N=15); and benzodiazepines (N=10). Thirty-nine patients were taking two or more medication types at the time of testing.

Patients were excluded if there was any clinical evidence of neuroleptic-induced side effects such as parkinsonism or tardive dyskinesia, or if there was clinical evidence of tremor related to mood stabilizers, such as lithium or valproate.

Clinical Assessment of Diagnosis and Psychopathology

All patients underwent a Structured Clinical Interview for DSM-IV (SCID) performed by trained personnel. Sixty-three patients met criteria for bipolar type I disorder, and four patients for bipolar type II disorder. Clinical assessments included the 28-item Hamilton Depression Rating Scale (HAM-D) 20 for the severity of depression, and the Young Mania Rating Scale (YMRS) 21 for the severity of mania.

Instrumental Assessment of Motor Function

Force Steadiness

The force steadiness procedure has been in use in our laboratory for over 10 years. 11 , 22 In this procedure, subjects sit facing a computer monitor with one of their hands positioned atop a platform equipped with the load cell (Sensotec, model 31/1426–02, range=1000 g). A load cell mounted onto the platform transduces force applied by flexing the index finger. A target line equivalent to 350 centiNewtons of force and the subject’s force signal are displayed on the monitor. The two waveforms move at a constant speed from left to right across the computer monitor to fill a 20-second window. At the end of each 20-second window, the screen is refreshed.

Subjects were instructed to match the target with their force signal by maintaining steady levels of finger flexion force. Three 20-second trials were obtained, each separated by a 5-second rest. The force signal was sampled at 100 samples/second by a laboratory computer and stored for subsequent analysis. Analysis involved identifying the segment with the greater range in force (determined by computing the force minima and maxima over the medial 80% of each segment) and calculating the waveform mean and standard deviation. Both hands were assessed, and data from the hand with the higher score were used for analysis.

Because many patients taking lithium develop a high frequency tremor 23 which can interfere with performance of the force steadiness task as well as the interpretation of results, we excluded patients with clinical evidence of tremor due to mood stabilizers, as well as low pass-filtered the force waveform to remove high frequency oscillations (>3.0 Hz). The coefficient of variation obtained by dividing the standard deviation of the force waveform by the mean force waveform served as the force steadiness score. Higher scores indicate greater error or instability. Force steadiness error is the direct result of irregular muscle contractions that produce changes in measurable force over time. The force steadiness data acquisition procedure, segmentation, and analysis have been shown to be highly reliable and valid. 24Figure 1 shows examples of force steadiness waveforms from a healthy comparison subject ( Figure 1a ) and a subject with force steadiness impairment ( Figure 1b ).

Velocity Scaling

The ability to scale movements is considered a cardinal function of the basal ganglia. 25 , 26 For example, some patients with parkinsonian hypokinesia execute all movements, regardless of distance or complexity, with a uniform velocity. 27 – 29 Investigators have shown that the slope of the relationship between movement velocity and target distance is sensitive to the extent of bradykinesia. 30 This phenomenon is consistent with recent studies showing that while Parkinson’s disease patients can increase movement velocity with increases in movement distance, 31 the slope of the relationship between peak velocity and target distance is often lower than in healthy comparison subjects. 32

For our measure of velocity scaling, we measured the velocity-distance relationship during simple wrist rotation using a platform equipped with a rotation sensor (electrogoniometer) that was attached to a handle. Rotation of the handle by flexing or extending the wrist produces a continuous signal that is proportional to the angle of rotation. We used custom software to convert the digitized signal (sampled at 100 samples/second) to a large cursor displayed on a computer monitor. The position of the cursor along the horizontal plane was calibrated in degrees of rotation. Along with the cursor, target boxes were displayed on the monitor along a horizontal plane located at 25° or 45°. Targets were placed left of midline for right wrist flexion, and right of midline for left wrist flexion. Subjects were instructed to flex the wrist “as quickly and as accurately as possible” when a target box appears on the screen. Subjects used visual feedback to place their cursor in the target box. The target remains were displayed for 2 seconds with an interstimulus interval of 2 seconds. If the subject needed clarification, instructions were repeated with emphasis on “speed of movement.” Thirty-two 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. Velocity and distance difference scores were obtained by subtracting the respective values for the 25° targets from the 45° targets. The velocity scaling score was derived from the ratio of the velocity difference over the angular distance difference, scaled in degrees/second/degree. Lower scores indicate disturbances in the ability to scale movement velocity with distance. Both hands were assessed, and data from the hand with the lower score were used for analyses.

In previous studies we have determined that test-retest reliability for this measure is very high. 12 , 33Figure 2 shows velocity-displacement relationships for a single subject with normal ( Figure 2a ) and abnormal ( Figure 2b ) velocity scaling. In the healthy sample, there is a significant increase in velocity with increases in angle of wrist rotation. Such increases in velocity are not present in the case with abnormal velocity scaling.

Motor Measure Characteristics

Eighteen psychiatrically and medically stable patients were retested 6 months following their initial assessment to examine measurement reliability. Correlational analyses indicated high levels of repeatability for both the force steadiness (r=0.86, p<0.001) and velocity scaling (r=0.75, p<0.001) measures.

Analyses of the relationships between the force steadiness and velocity scaling measures indicated that performance on one measure was independent of performance on the other. Pearson's product moment correlations were −0.03 for bipolar disorder and −0.17 for healthy comparison subjects. These findings support the specificity of our two motor measures.

There were no effects of gender or age on either the force steadiness or velocity scaling measures. Also, there were no differences on force steadiness or velocity scaling among the patients treated with different medications.

Statistical Methods

Group data were examined for homogeneity of variances and normal distributions prior to undertaking parametric statistical analyses. Group comparisons involving data meeting all assumptions for parametric analyses were subjected to analyses of variance (ANOVA) and post-hoc planned comparisons to test group differences. Separate analyses were performed for force steadiness and velocity scaling data. Group data that did not meet the assumptions for parametric analyses, even after transforming the raw scores, were subjected to Friedman’s ANOVA and post-hoc Mann-Whitney tests. Correlational analyses were used to examine relationships between motor scores and severity of depression and mania. Chi-square tests were used to examine differences in prevalence of motor abnormalities among patients who were on versus patients who were off medication.

Patients were also classified as “normal” or “abnormal” based on the healthy comparison subjects’ 95 ^th percentile score. Patients scoring above 2.26 on the force steadiness measure or below 1.88 on the velocity scaling measure were therefore classified as “abnormal.”

RESULTS

Overall, bipolar disorder patients performed significantly more poorly than healthy comparison subjects. The mean scores for the force steadiness and velocity scaling measures for the healthy comparison group were 1.96 (SD=0.77) and 2.53 (SD=1.35), respectively. The mean scores for the force steadiness and velocity scaling measures for the bipolar disorder patients were 3.46 (SD=3.00) and 1.71 (SD=0.92), respectively. Nonparametric tests for group differences revealed highly significant differences between bipolar disorder patients and healthy comparison subjects for the force steadiness (p<0.00001) and velocity scaling (p<0.001) measures.

For the force steadiness measure, 39 patients (58%) scored within the abnormal range. For the velocity scaling measure, 42 out of 67 patients (63%) scored within the abnormal range. Twenty-five bipolar disorder patients (37%) exhibited both force steadiness and velocity scaling abnormalities, and 56 patients (84%) had abnormal measures on either assessment. Scores on the force steadiness and velocity scaling measures were unrelated to each other. Also, the percentage of patients with abnormalities on both measures (37%) is no greater than would be calculated from a chance relationship between the two measures (58% x 63% = 36%).

Affective State

Patients with abnormal force steadiness scores did not differ from patients with normal force steadiness scores on ratings of depression or mania. Patients exhibiting both force steadiness and velocity scaling abnormalities did not differ from other subgroups on these ratings of mood pathology. There were no significant relationships between severity of mood symptoms based on the HAM-D and YMRS scores and the force steadiness or velocity scaling measures. The Pearson’s r values for the correlations between HAM-D and the motor measures were −0.19 (force steadiness) and −0.09 (velocity scaling), and values between YMRS and the motor measures were −0.06 (force steadiness) and 0.17 (velocity scaling).

Medication Effects

Because of the known influence of certain psychotropic medications on dyskinesia and parkinsonism, we examined whether the motor abnormalities described above might be related to current medication status. The bipolar disorder subjects were divided into those who were being treated with a particular class of psychotropic medications and those who were off these medications. Patients on current pharmacotherapy were separated into five groups, including subjects on antidepressants, mood stabilizers, antipsychotics, and benzodiazepines, and those off medication at the time of testing. We examined each medication group to determine if it was associated with a greater prevalence of instrumentally measured motor abnormalities. Prevalence results indicated no significant differences between the numbers of bipolar disorder subjects meeting criteria for force steadiness or velocity scaling impairment among any of the groups. Prevalence rates for medicated subjects ranged from 50% (for antipsychotics or benzodiazepines) to 62.5% (for antidepressants) for force steadiness impairment, and 50% (for benzodiazepines) to 80% (for antipsychotics) for velocity scaling impairment. Although there were only 11 unmedicated bipolar disorder subjects, seven of the 11 (63.6%) unmedicated bipolar disorder subjects exhibited abnormal force steadiness scores and five (53.5%) exhibited abnormal velocity scaling scores. Thus, the prevalence of instrumentally derived motor abnormalities among unmedicated bipolar disorder subjects was comparable to that of the medicated subjects.

DISCUSSION

We addressed three questions about motor pathophysiology in bipolar disorder. First, we examined whether bipolar disorder patients exhibit disturbances in motor physiology that may relate to abnormalities within the sub-cortical motor circuits. We found that approximately 60% of our bipolar disorder patients exhibited force steadiness or velocity scaling motor abnormalities based on instrumental assessments, with 84% of patients having abnormalities on either measure. Over one-third of the patients exhibited both forms of motor impairment. The second question of this study was to examine whether disturbances in motor physiology in bipolar disorder patients as assessed by force steadiness and velocity scaling were related to an affective state in bipolar disorder. We found no correlation between the motor scores and the severity of manic symptoms (as measured by the YMRS) or depressive symptoms (as measured by the HAM-D). The third question pertained to whether the observed impairment in motor functions was possibly related to current pharmacotherapy. We found no differences in prevalence of impairment between bipolar disorder patients on versus off psychotropic medications. Moreover, unmedicated bipolar disorder patients exhibited similar distributions of force steadiness and velocity scaling impairment as treated patients. The possibility remains that past exposure to certain medications could have contributed to the observed abnormalities in motor function.

The lack of associations between mood state, pharmacotherapy, and force steadiness or velocity scaling performance indicates that these motor impairments may be trait characteristics of bipolar disorder. However, caution is warranted in this interpretation because the present study was cross-sectional and not longitudinal in nature. Nevertheless, it is possible that these measures may be useful as endophenotypes in future genetic studies of bipolar disorder.

With the exception of psychomotor retardation 34 or lithium-induced tremor, 35 few studies have focused specifically on motor abnormalities in bipolar disorder. The few exceptions include studies by Heninger and Kirstein, 36 who reported increased motor activity in manic bipolar disorder patients, and Klein et al., 37 who reported that increased motor activity predicted relapse into mania in bipolar patients. It may be argued that this lack of attention to motor disturbances in bipolar disorder may be related to their relative infrequency in bipolar disorder in comparison to schizophrenia 38 , 39 or major depression. 34 Motor problems in bipolar disorder may also be underreported because they can be subtle and difficult to assess using conventional assessment means. 40

Our findings show that approximately 60% of patients with bipolar disorder demonstrate significant impairments in the ability to maintain steady-state force or the ability to scale velocity with distance; but these two findings appear to be independent, suggesting that two separate physiological mechanisms relating to basal ganglia circuitry may be involved in bipolar disorder. In fact, the putative neural circuits involved in motor dysfunction have been elucidated over the past two decades 41 , 42 and two basic pathways have been identified. Neurotransmission from the motor cortex, through the basal ganglia and thalamus, and back to the sensorimotor cortices has been proposed to follow a direct pathway and an indirect pathway, the latter of which involves the subthalamus. 43 Disturbances within the direct pathway have been suggested to be associated with the reduction of movement or with difficulty initiating movement (hypokinesia), whereas disturbances within the indirect pathway have been proposed to be associated with excessive or unstable movement (hyperkinesia). In this context, our results suggest that velocity scaling abnormalities may possibly relate to dysfunction in the direct pathway, and the force steadiness abnormalities to disturbances in the indirect pathway in bipolar disorder. Although this is highly speculative, these brain regions could serve as a focus in future studies utilizing techniques such as functional magnetic resonance imaging (fMRI).

Our findings suggest that psychopathology and motor dysfunction may share a common pathophysiology. Comorbid mood and motor disorder are common features in neuropsychiatric disorders, such as Parkinson’s and Huntington’s diseases. For example, depression is common among patients with Parkinson’s disease with prevalence estimates ranging from 20% to 45%. 43 , 44 Studies have shown that the onset of depressive symptoms often predates the presentation of motor signs in Parkinson’s disease. 45 This would imply that the mood symptoms may be more closely linked to the disease process than a secondary feature of Parkinson’s disease. Based on a study using positron emission tomography (PET), Remy et al. 46 suggested that depression in Parkinson’s disease might be associated with reduction in dopamine and norepinephrine receptor binding in structures including limbic regions.

Both depression and mania are observed in Huntington’s disease, a prototypic movement disorder of basal ganglia origin. Estimates of the prevalence of a mood disorder in Huntington’s disease range from 22% to 38% for depression 47 and 5% to 10% for mania. 48 PET studies in depressed Huntington’s disease subjects also show pathophysiological correlates (e.g., glucose hypometabolism in the orbitofrontal and inferior parietal regions of the brain) that are similar to those observed in major depression. 49

The co-occurrence of mood symptoms with parkinsonism and choreoathetosis in neurological diseases supports the hypothesis that hypokinetic or hyperkinetic motor abnormalities may coexist with affective disorders because of shared pathophysiology. The present findings of motor problems in bipolar disorder suggest an extension of this hypothesis to include psychiatric disorders as well. Although we did not observe a significant relationship between the severity of mania or depression and severity of motor dysfunction in this cross-sectional study, a longitudinal study might reveal that a fluctuation in affective state is related to a fluctuation in motor control.

In conclusion, the results from this study contribute to the literature by characterizing motor abnormalities in bipolar patients. The use of sensitive quantitative instrumentation of motor function may be helpful in characterizing the brain abnormalities of bipolar disorder. Moreover, better characterization of the motor circuits underlying these abnormalities may help us determine which circuits are disrupted in the modulation of affect and where to target novel treatments.