Quantitative analysis of motor performance in epilepsy patients treated with valproate
Article Outline
- Abstract
- 1. Introduction
- 2. Subjects and methods
- 3. Results
- 4. Discussion
- 5. Conclusion
- Conflict of interest statement
- Acknowledgments
- References
- Copyright
Abstract
Purpose
The aim of our study was to detect objective signs of deterioration of motor performance in epilepsy patients treated with chronic valproate therapy.
Methods
We examined 14 controls and 15 epilepsy patients receiving chronic valproate monotherapy, who had no subjective complaints related to motor function. Regularity and maximum frequency of repetitive hand and finger movements, and simple reaction time were measured. Intensity and frequency characteristics of resting and postural tremor were assessed using accelerometry. Data were statistically evaluated.
Results
Repetitive hand and finger movements were significantly more irregular and the maximum frequency of repetitive movements was significantly lower in the valproate group than in controls. Resting tremor peak frequency and motor reaction time of the two cohorts did not differ.
Conclusions
This is the first study, which quantitatively assesses motor performance of patients with epilepsy on chronic valproate therapy. The results suggest significant irregularity of repetitive hand movements and finger tapping even in patients with no motor complaints. Objective methods might help to recognize valproate-induced motor performance deterioration.
Keywords: Repetitive hand movements, Finger tapping, Tremor, Epilepsy, Valproate
1. Introduction
Valproate is one of the most frequently prescribed antiepileptic drugs worldwide.1 The number of patients treated with valproate is growing since it has been shown to be effective in the management of bipolar illness,2 and migraine3 as well.
A well-known side effect of valproate is tremor, which might affect as many as 45% of patients.1 Some studies suggest that valproate might also induce other motor deficits, like decreased alternate motion rates, rigidity, abnormalities of posture and gait, with an incidence of about 5%.4, 5, 6, 7 The early motor manifestations may not be evident unless particularly sought, and the association with valproate may escape notice due to insidious onset.8
While valproate-induced tremor has already been investigated using objective methods9, 10, 11, 12 the quantitative measurement of motor performance in patients with chronic valproate medication has not been performed yet.
Valproate-induced motor symptoms are reversible on withdrawal of the drug, however the recovery might be slow.6, 8 The reversible nature of the alterations emphasizes the importance of early recognition, since discontinuation of the drug might be necessary.
In the present study, we performed complex quantitative assessment of motor performance of adults with epilepsy, who were on chronic valproate monotherapy for at least 2 years, and had no clinically noticeable motor symptoms. We measured the regularity and maximum speed of alternating hand and repetitive finger movements, the simple motor reaction time and the resting and postural tremor. Our aim was to detect subclinical changes of motor performance in valproate-treated patients, which might be used for early detection of valproate-related motor dysfunction.
2. Subjects and methods
2.1. Subjects
We examined 14 healthy control subjects, who were all employees of our Department, and had no history of any neurological disorder. Controls did not take any medication regularly.
Fifteen epilepsy patients receiving valproate monotherapy were selected from the Epilepsy Outpatient Service of our Department. Inclusion criteria were the following: (1) The patient has primarily generalized epilepsy with tonic–clonic seizures only, treated with controlled-release valproate monotherapy for a minimum of 2 years, and he/she is not taking other medication regularly. (2) He/she does not have any complaint regarding tremor or motor abnormality. (3) The patient does not have any clinically detectable motor symptom (upper motor neuron signs, rigidity, tremor, gait disturbance or impaired alternating movements) on neurological examination. (4) The number of seizures per year is less than three. (5) The last seizure occurred at least 2 months before the motor performance assessment. (6) The serum valproate level does not exceed 100
mg/l measured within 1 month prior motor testing. (7) The patient has a negative CT/MRI examination performed within 2 years. (8) The Mini Mental State Examination (MMSE) score is greater than 28. (9) The hematological panel, serum TSH, ammonia and vitamin B12 levels are normal.
Clinical characteristics of the groups are shown in Table 1. There was no significant difference between the mean age of the groups. All subjects were right handed.
Table 1. Clinical characteristics of the control and valproate-treated groups.
| Parameter | Control | Valproate |
|---|---|---|
| Number of subjects | 14 | 15 |
| Age (year) mean±SD | 36.3±13.4 | 40.7±15.4 |
| Range | 25–65 | 21–65 |
| Gender (female/male) | 7/7 | 6/9 |
| Epilepsy type | – | 15 IGE-GTCS |
| Median daily dosage (mg) | – | 900 (300–1500) |
| Serum drug level (mg/l) mean±SD | – | 78.4±11.76 |
| Range | 62–98 | |
| Duration of treatment (year) mean±SD | – | 5.2±4.6 |
| Range | 2–13 | |
| Median seizure frequency (seizures/year) | – | 1 |
| Range | 0–3 |
Subjects were asked to refrain from taking sleeping pills, and drinking coffee and/or alcohol 24h prior to examination, since these are known to affect tremor, motor coordination and alertness.13, 14, 15, 16 Measurements were carried out within 2h taking the regular morning valproate medication.
The study was performed according to the Declaration of Helsinki. The experimental procedure was approved by the Local Ethical Committee. All individuals gave written informed consent.
2.2. Recording methods
Tremor parameters and regularity of repetitive movements were measured using a computerized test system (CATSYS 2000, Danish Product Development Ltd., Snekkersten, Denmark, www.catsys.dk).17, 18, 19 Patients were seated comfortably in a soundproof room. Measurements were carried out on the right and left side consecutively.
Regularity and maximum frequency of repetitive finger and alternating hand movements were examined using a touch sensitive drum.
For finger tapping measurements, subjects hit the drum with the index finger while the wrist was supported.
Alternating hand movements were investigated with the drum in hand. Subjects were instructed to perform pronation/supination movements for 10s, keeping precise pace with the 2.5Hz acoustic signal generated by the computer. The time offset (ms) between the signal and the subject's beat was measured.
Regularity of repetitive movements was quantified by the standard deviation of the time-offset values (tap-to-cue variability, ms). A subject, who is able to maintain the constant frequency of 2.5Hz, has a tap-to-cue variability of 30±20ms for finger and 40±20ms for hand pronation/supination movement according to the normal dataset of CATSYS.19
To define the maximum frequency (Hz) of repetitive movements subjects were instructed to perform finger tapping or hand pronation/supination paced by gradually accelerating acoustic signal (from 1.6 to 7.5Hz). The touch-sensitive period around each signal, in which the subject's action was recorded, was 250ms at stimulus frequency of 1.6Hz, and it gradually decreased to 50ms at stimulus frequency of 7.5Hz. If the latency of the hit was longer than the touch-sensitive period, the action was not accepted. To compensate for random errors, subjects were allowed to miss one hit if the next two were acceptable. The frequency of hand or finger movements at the time of the last legal hit determined the maximum frequency.
Simple motor reaction time (ms) was measured in a traditional stimulus-response test using a handle switch. The computer gave random auditory signals to which the subjects had to press the handle switch with the thumb. Reaction times shorter than 0.1s or longer than 0.5 were excluded. Average reaction time of ten trials was calculated.
Tremor was recorded for 32.8s in two different positions: (1) at rest the forearm and the hand was fully supported on the arm-rest of the chair; (2) in postural position the arm and hand were held against gravity in an outstretched, horizontal, prone position.
Tremor was measured using a biaxial micro-accelerometer (weight: 10.5g, sensitivity: >0.3m/s2), fixed on the third metacarpal bone, 2cm proximal to the metacarpophalangeal joint. Accelerometry signals of the two axes were digitized at 128Hz, and combined by root sum of squares. Tremor parameters were derived from Fourier power spectra. Because of the inter-individual tremor intensity differences, for frequency power analysis the power spectrum of each subject was normalized.
Tremor intensity (TI, m/s2) was calculated as the root-mean-square of acceleration.
Peak frequency (Hz), at which the tremor had the highest power in the spectrum, was determined from the normalized power spectra.
Frequency dispersion (FD, Hz) reflects the regularity of tremor. It is defined as the half width of the frequency band centered around the peak frequency, which contains 68% of the total power. Frequency dispersion of the physiological tremor is broad (3–4Hz), while it is reduced (0.5–1Hz) in parkinsonian or essential tremor.20, 21
2.3. Statistical analysis
Our principal experimental hypothesis was that regularity of alternating hand movements and finger tapping as well as physiological tremor were affected in patients treated with VPA.
This hypothesis was examined by non-parametric tests because data were not normally distributed (Kolmogorov–Smirnov test). Statistical analysis was performed using the Statistica software package (Statsoft Inc., 6.0 version, Tulsa, OK, USA).
Regularity data of motor performance of the two groups were analysed using the Mann–Whitney test.
To assess the tremor power distribution in the 0–15Hz spectrum of the two groups, the power of each 1Hz wide frequency band (0–0.99, 1–1.99, 2–2.99, etc.) was compared using the Mann–Whitney test.
To investigate changes of resting and postural tremor parameters within one group we used the Wilcoxon-matched-pairs test.
The relationship between duration of medication and various parameters of hand and finger movements was examined by regression analysis.
Level of significance for all statistics was set at p<0.05. Statistical analysis showed no significant difference between the right and left sides. Data of the right hand are presented.
3. Results
3.1. Motor performance
Individual data of motor performance parameters for controls and patients are presented in Table 2a, Table 2b, respectively. Data of the control group were similar to those reported for normal population.19
Table 2a. Individual data of motor parameters in the control group.
| Controls | PS | Tap | MF/PS | MF/Tap | RT |
|---|---|---|---|---|---|
| 1 | 19 | 15 | 7.1 | 6.9 | 259 |
| 2 | 25 | 16 | 5.7 | 7.5 | 273 |
| 3 | 30 | 19 | 5.1 | 7.3 | 220 |
| 4 | 19 | 35 | 6.1 | 6.8 | 283 |
| 5 | 36 | 36 | 5.2 | 7.2 | 228 |
| 6 | 28 | 23 | 5.6 | 7.5 | 259 |
| 7 | 27 | 50 | 7.5 | 7.5 | 224 |
| 8 | 25 | 32 | 6.0 | 6.8 | 194 |
| 9 | 27 | 33 | 6.9 | 6.9 | 233 |
| 10 | 21 | 32 | 5.4 | 7.5 | 206 |
| 11 | 22 | 33 | 6.1 | 5.4 | 219 |
| 12 | 45 | 56 | 4.7 | 7.5 | 240 |
| 13 | 50 | 13 | 4.4 | 7.1 | 236 |
| 14 | 17 | 40 | 6.6 | 7.5 | 239 |
| Av±SD | 27.93±9.7* | 30.93±12.8* | 5.89±0.9* | 7.1±0.6* | 236.64±24.9 |
*p |
Table 2b. Individual data of motor parameters in the valproate group.
| Patients | PS | Tap | MF/PS | MF/Tap | RT |
|---|---|---|---|---|---|
| 1 | 68 | 32 | 3.8 | 7.5 | 226 |
| 2 | 109 | 50 | 4.9 | 7.5 | 247 |
| 3 | 33 | 71 | 6.7 | 7.5 | 208 |
| 4 | 105 | 73 | 5.9 | 6.2 | 188 |
| 5 | 97 | 45 | 6.2 | 7.5 | 229 |
| 6 | 23 | 46 | 6.7 | 5.6 | 225 |
| 7 | 139 | 101 | 3.7 | 4.8 | 228 |
| 8 | 82 | 54 | 4.8 | 4.6 | 206 |
| 9 | 108 | 65 | 4.7 | 6.0 | 286 |
| 10 | 22 | 54 | 4.6 | 6.4 | 239 |
| 11 | 47 | 54 | 5.2 | 6.4 | 208 |
| 12 | 96 | 102 | 2.6 | 3.8 | 317 |
| 13 | 68 | 87 | 3.2 | 5.3 | 208 |
| 14 | 55 | 75 | 4.3 | 7.5 | 206 |
| 15 | 71 | 120 | 3.0 | 6.2 | 258 |
| Av±SD | 74.87±34.6* | 68.60±24.8* | 4.69±1.3* | 6.19±1.2* | 231.93±34.0 |
*p |
The tap-to-cue variability of the valproate-group was significantly greater for both pronation/supination movements and finger tapping compared to controls (Table 2a, Table 2b).
The individual tap-to-cue variability was longer than the mean+2SD of the control group in the pronation/supination test in 11 patients (73%) and in the finger tapping test in 8 patients (53%). There was only one value greater than normal in the control group in the pronation/supination and in the tapping test (Subject No. 13 and 12, respectively) (shaded cells in Table 2a, Table 2b). These findings show that valproate-treated patients performed alternating repetitive hand movements and finger tapping with significantly less regularity compared to controls.
There was no significant correlation between the duration of valproate treatment and the severity of either the hand or the finger movement irregularity (pronation/supination: r2=0.275, p=0.267, finger tapping: r2=0.071, p=0.435). There was also no correlation between serum level of valproate and motor performance deterioration (pronation/supination: r2=0.087, p=0.428, finger tapping: r2=0.0047, p=0.562).
The maximum frequency of pronation/supination movements and finger tapping was significantly lower in valproate patients compared to controls (Table 2a, Table 2b). The individual maximum frequency of hand and finger movements was lower than the mean−2SD of the control group in 5 valproate-treated patients (33.3%). There was only one value lower than normal in the control group in the maximum tapping frequency (Subject No. 11, shaded values in Table 2a, Table 2b).
There was no significant correlation between the duration of valproate treatment and the maximum frequency of either the hand or the finger movement (maximum frequency of pronation/supination: r2=0.001, p=0.506, maximum frequency of finger tapping: r2=0.053, p=0.366).
The average reaction time of the two groups was statistically similar (Table 2a, Table 2b).
3.2. Tremor parameters
Data of tremor parameters are presented in Table 3. The results of the control group were similar to those reported for normal population.19
Table 3. Tremor data of the control and the valproate groups.
| Parameter | Position | Control | Valproate |
|---|---|---|---|
| Tremor intensity (m/s2) | Rest | 0.08±0.04 | 0.09±0.06 |
| Postural | 0.17±0.09 | 0.21±0.08 | |
| Peak frequency (Hz) | Rest | 8.92±1.38 | 7.73±1.98 |
| Postural | 7.23±1.50 | 6.92±1.43 | |
| Frequency dispersion (Hz) | Rest | 2.50±0.87 | 2.77±0.89 |
| Postural | 3.33±0.71 | 3.32±0.80 | |
Tremor intensity of VPA-patients was similar to controls both at rest and in postural position (Table 3). In both groups tremor intensity was significantly higher in postural compared to resting position, which is in accordance with data of the normal physiological tremor.18
The peak frequency of rest tremor was lower in the valproate-group, but the difference did not reach significance (Table 3). The analysis of power distribution showed significantly higher power in the 2–2.99Hz, 3–3.99Hz, 5–5.99Hz and 6–6.99Hz bands in the valproate-group compared to controls (Fig. 1). Power values in the higher frequency bands (from 8 to 15Hz) were similar in the two cohorts.
Fig. 1.
Normalized average tremor power (mean
±SE) of the control and valproate-treated groups. The tremor power was significantly higher in the low frequency bands of the power spectrum in the valproate group compared to controls. There was no statistically significant difference in the higher frequency range (7–15Hz, data not shown).
The peak frequency and the power spectrum of postural tremor were similar in the two groups (Table 3). Peak frequency was significantly lower in postural position compared to rest in both groups, in accordance with normal physiological tremor data.22
Frequency dispersion of rest and postural tremor was similar in the two groups (Table 3). Postural tremor had significantly greater frequency dispersion value compared to resting tremor in both groups, similarly to what can be observed in the normal population.
4. Discussion
We report for the first time the results of quantitative analysis of motor performance in patients with epilepsy treated with chronic valproate monotherapy. We found that repetitive finger and hand movements were significantly more irregular and the maximum frequency of movements was significantly lower in valproate-treated patients than in controls. This deterioration of motor performance was detected not only at the group level; individual values of more than 70% of valproate patients fell outside the normal range (i.e. mean±2SD of the controls).
Irregularity and slowing of repetitive movements were probably not related to valproate-induced sedation, since the control and the patient groups performed similarly well in the simple motor reaction time test. Motor reaction time is a measure of akinesia,23 and our data suggest that valproate-treated patients were not affected in this respect.
Tremor analysis, in accordance with the inclusion criteria, showed normal intensity of both rest and postural tremor in both groups. Chronic valproate administration caused a non-significant decrease of peak frequency of rest tremor, which was due to an augmentation of tremor power in the lower frequency ranges (between 2 and 6.99Hz).
Our data prove that the motor performance of the majority of valproate-treated patients might already be impaired even when their tremor still shows physiological characteristics. This suggests that tremor is probably not the first appearing motor adverse effect of valproate treatment. Quantitative methods might reveal higher incidence of valproate related motor disturbances than it is considered now.
We did not find correlation between duration of treatment and severity of motor deficit or decrease of tremor peak frequency. There was also no correlation between serum levels of valproate and degree of motor performance deterioration or tremor parameters. These results imply that the development of valproate-related motor symptoms might be influenced by individual predisposition, but its nature is not yet identified.
The analysis of follow-up results was not the aim of the present study. However, because one of the patients (Patient No. 12, Table 2b) had pathological values in all subtests of the quantitative assessment, we performed a control examination 3 months later. We noted further deterioration of motor performance, and the patient complained of hand tremor. To avoid further decline valproate treatment was replaced. Six months later the control quantitative test showed significant improvement, however the results were still below normal.
One drawback of our study is that the scientifically most robust protocol, i.e. the challenge–dechallenge–rechallenge design regarding valproate treatment could not be applied, most importantly, because of the withdrawal of the drug might lead to seizure exacerbation. Another reason is that although valproate-induced motor symptoms are reversible but the recovery might take several months. Therefore, this trial design would have been difficult to implement in practice.
Since patients with epilepsy may differ from normal controls in a number of ways, regardless of antiepileptic drug therapy, a more valid control group might have been patients taking other antiepileptic drugs. This approach however also seems problematic since most antiepileptic drugs (phenytoin, benzodiazepines, carbamazepine, topiramate, gabapentine, and lamotrigine) are known to cause tremor and ataxia.24
We did not intend to prove that valproate induces motor adverse reactions, since this is already a well-established fact.25 Instead, we investigated epilepsy patients who were on chronic valproate therapy and were free of motor complaints in order to demonstrate subclinical signs of motor performance deterioration. Our results suggest that in the majority of valproate-treated patients significant irregularity of alternating hand and repetitive finger movements could be detected using quantitative methods. This phenomenon most probably is not related to the epilepsy disorder itself, since discontinuation of the drug leads to considerable improvement.
5. Conclusion
We propose that quantitative methods might be used for objective evaluation and monitoring the motor performance status of patients on chronic valproate therapy. In case of progressive deterioration, discontinuation of the drug might be initiated.
Conflict of interest statement
None of the authors has any conflict of interest to disclose.
Acknowledgments
We thank Dr. A. Balogh for permission to report three of his patients, and Dr. G. Hábel for her help during the recordings.
AK was supported by the Hungarian Scientific Research Fund (K49269).
References
- . Pharmacological and therapeutic properties of valproate: a summary after 35 years of clinical experience. CNS Drugs. 2002;16:695–714
- Antiepileptic drugs in mood-disordered patients. Epilepsia. 2005;46:38–44
- Evidence-based medicine in migraine prevention. Expert Review of Neurotherapeutics. 2005;5:333–341
- Reversible parkinsonism and cognitive impairment with chronic valproate use. Neurology. 1996;47:626–635
- . Prospective evaluation of parkinsonism and tremor in patients treated with valproate. Parkinsonism and Related Disorders. 1999;5:67–68
- . Reversible parkinsonism with normal beta-CIT-SPECT in patients exposed to sodium valproate. Neurology. 2004;62:1435–1437
- . The frequency of reversible parkinsonism and cognitive decline associated with valproate treatment: a study of 364 patients with different types of epilepsy. Epilepsia. 2006;47:2183–2185
- . Valproate-induced Parkinsonism in epilepsy patients. Movement Disorders. 2007;22:130–133
- . Tremor due to sodium valproate. Neurology. 1979;29:1177–1180
- . Valproate tremors. Neurology. 1982;32:428–432
- . Clinical and surface EMG characteristics of valproate induced tremors. Electromyography and Clinical Neurophysiology. 2005;45:177–182
- Computerized tremor analysis of valproate-induced tremor: a comparative study of controlled-release versus conventional valproate. Epilepsia. 2005;46:320–323
- . The effect of alcohol on physiological tremor. Experimental Physiology. 1994;79:273–276
- . A possible mechanism for the beneficial effect of ethanol in essential tremor. European Journal of Neurology. 2008;15:697–705
- . Dissociations between motor timing, motor coordination, and time perception after the administration of alcohol or caffeine. Psychopharmacology. 2009;202:719–729
- . Effects of benzodiazepines on psychomotor performance. British Journal of Clinical Pharmacology. 1979;7:61S–67S
- . Sensitivity and specificity of a portable system measuring postural tremor. Neurotoxicology and Teratology. 1997;19:95–104
- . Evaluation of electronic equipment for quantitative registration of tremor. Acta Neurologica Scandinavica. 1998;97:36–40
- . Standardization of a neuromotor test battery: the CATSYS system. Neurotoxicology. 2000;21:725–735
- . Asymmetry of tremor intensity and frequency in Parkinson's disease and essential tremor. Parkinsonism & Related Disorders. 2006;12:49–55
- . Objective monitoring of tremor and bradykinesia during DBS surgery for Parkinson disease. Neurology. 2008;70:1244–1249
- . Effects of aging on the regularity of physiological tremor. Journal of Neurophysiology. 2005;93:3064–3074
- . Bradykinesia akinesia in coordination test (BRAIN TEST): an objective computerised assessment of upper limb motor function. Journal of Neurology, Neurosurgery and Psychiatry. 1999;67:624–629
- In: Engel J, Pedley TA editor. Epilepsy: a comprehensive textbook. second ed.. Wolters Kluwer, Lippincott Williams & Wilkins; 2007;p. 1431–1747
- Movement disorders in patients taking anticonvulsants. Journal of Neurology, Neurosurgery and Psychiatry. 2007;78:147–151
PII: S1059-1311(10)00014-2
doi:10.1016/j.seizure.2010.01.013
© 2010 British Epilepsy Association. Published by Elsevier Inc. All rights reserved.
