Carpal Tunnel Syndrome Involving Unaffected Limbs of Stroke Patients
Background and Purpose—Stroke-induced hemiparesis involving the arm and hand results in regular, repeated overuse of the opposite hand and wrist. Because repetitive hand and wrist movement is a common cause of carpal tunnel syndrome (CTS), we examined the nonparetic upper limb in stroke patients for evidence of CTS.
Methods—We measured bilaterally sensory nerve conduction velocity (SNCV), motor nerve conduction velocity (MNCV), sensory nerve action potentials (SNAP) at the wrist, palm-to-wrist distal sensory latency (DSL), palm-to-wrist SNAP, compound motor action potentials (CMAP), and distal motor latency (DML) in stroke patients and control subjects. Controls were right-handed, ≥65 years old, lucid, independent in their activities of daily living, and had no disease known to cause CTS. Stroke patients were divided into a functioning hand group (n=61) and a disused hand group (n=71). All patients had hemiplegia.
Results—Tinel’s sign was observed on the nonparetic side in 57.7% of patients with a disused hand and in 31.1% of those with a functioning hand. All electrophysiological indices were significantly more abnormal on the nonparetic side than on the hemiparetic side or in controls. Patients with a disused hand showed greater abnormality on the nonparetic side in SNCV, SNAP, palm-to-wrist DSL, DML, and CMAP than patients with a functioning hand.
Conclusions—Overuse of the nonparetic hand and wrist of the nonparetic side may result in CTS in stroke patients, especially when the paretic hand is not functional. Wrist splinting or other prophylactic treatments beginning soon after stroke might help to prevent CTS.
Hemiparesis constitutes the most common neurological abnormality after stroke, with wide variability of motor deficit severity. The opposite or “unaffected” upper extremity often is considered a reference point, with the assumption that this side has no deficit. However, previous studies have shown various abnormalities in the unaffected hand and arm, such as diminished strength or deficits in grip strength, flexion of fingers and thumb, pinch strength, kinesthesia in the thumb, gross manual dexterity, fine manual dexterity, and motor coordination.1 2 3 Hemiparesis involving the hand and arm results in regular, repeated overuse of the opposite hand and wrist in rough correspondence to the degree of hemiparesis. Because frequent and repetitive use of the hand or wrist is a common cause of carpal tunnel syndrome (CTS),4 5 6 7 8 such hand and wrist overuse may cause CTS in stroke patients. We treated 3 hemiparetic stroke patients who had typical CTS on the opposite side but no apparent systemic diseases known to cause CTS, which was thought to have arisen from overuse. These patients had been using the nonparetic hand exclusively and were mobile without another person’s assistance for 2 to 3.5 years after stroke, with or without assist devices requiring use of the hand. We therefore screened poststroke patients electrophysiologically for subclinical CTS on the “unaffected” side.
Subjects and Methods
Subjects in the stroke group met the following eligibility criteria: unilateral hemiplegia/paresis subsequent to a stroke occurring 2 to 4 years previously, right-handedness before the stroke, age >65 years, residence at home, and independent mobility without assistance from another person but sometimes with assist devices. Exclusion criteria included (1) clinical or electrophysiological evidence of accompanying conditions that could mimic CTS or interfere with its evaluation, such as cervical radiculopathy, proximal median neuropathy, significant polyneuropathy, or marked orthopedic abnormalities; (2) treatment with vitamin B1, B2, or B12, or steroid hormones before or during the study period; (3) systemic diseases known to cause CTS, such as diabetes mellitus, hypothyroidism, rheumatoid arthritis, or chronic renal failure; and (4) typical symptoms and signs of CTS in the nonparetic hand. Hemiparetic hands were classified according to the long-term score of the Scandinavian Stroke Scale9 as functioning (scores 4 to 6, n=61) or disused (scores 0 to 2, n=71).
The comparison group was composed of men and women who were right-handed, ≥65 years old, lucid, independent in their activities of daily living; who had no disease known to cause CTS, no history of frequent, repetitive use of the hand or wrist; and who lacked any upper extremity neurological deficit. The subjects were volunteers drawn from the community, matched by age and sex with the hemiparetic subjects. Fifty-six control subjects with no conditions affecting peripheral nerve evaluation were selected for study.
Neurological Examination of the Median Nerve on the Unaffected Side
Neurological examinations were performed to evaluate the degree or intensity of the following symptoms and signs of CTS in the unaffected hand and wrist: dysesthesia (numbness, tingling, and pain) in the hand; paresthesias and decreased sensation in the median nerve distribution; dry skin or a color change visible over the palm; thenar muscle atrophy; and Tinel’s and Phalen’s signs.
Sensory and Motor Nerve Conduction
Sensory nerve conduction velocity (SNCV) and motor nerve conduction velocity (MNCV) were measured with a Neuropack 2 instrument (Nihon Kohden) in an air-conditioned room at 23°C to 25°C. Skin surface temperature was maintained at 34°C. Median and ulnar nerve conduction studies were performed bilaterally in all subjects. Antidromic sensory responses from the median and ulnar nerve were measured. Ring electrodes were placed on digits II or V, respectively, and SNCV was calculated in both nerves between the wrist and digit II or the wrist and digit V by determining the onset of sensory nerve action potentials (SNAP). Median nerve SNAP were recorded across the wrist with the active recording electrode 14 cm from the stimulator cathode.10 In addition, median sensory conduction from the palm to the wrist was measured over a 7-cm conduction distance.11 Distal sensory latency (DSL) and SNAP at the palm were recorded in the same manner. The voltage used for stimulation was increased until the SNAP reached maximal amplitude in both techniques for median nerve sensory assessment.
MNCV was determined in both median nerves by stimulation at the wrist. Median nerve compound motor action potential (CMAP) and distal motor latency (DML) were measured with the stimulating and recording cathodes 8 cm apart. No needle examination was performed. All patients and volunteers were informed of the nature of the study before witnessed consent was obtained from each participant.
Data are expressed as mean±SD. Group differences involving categorical data were tested by χ2 analyses. The differences in electrophysiological indices between the left and right sides in control subjects and between the affected and unaffected sides in patients were assessed by paired t tests. Student’s t test was used to assess differences in electrophysiological parameters between controls and patients. Because a subclinical compression neuropathy is common in elderly subjects such as the controls in our study, means of electrophysiological indices on the left and right side were calculated for use as control values. Then 1-way ANOVA and Fisher’s protected least significant difference test were used to assess electrophysiological differences among mean values for control, functioning, and disused hands. In performing electrodiagnostic assessment for CTS, we used parametric analyses to define the normal range statistically as values within 2 SD of the mean, assuming a normal distribution in the control population.12 This definition took account of the fact that subclinical CTS exists to a varying extent in the general population. Because a normal distribution in controls was confirmed beforehand, the nerve conduction study was considered to support an electrophysiological diagnosis of CTS when each index was significantly reduced or delayed at least 2 SD below or above the control mean.
Characteristics of Study Subjects
Characteristics of the patient population are summarized in Table 1⇓. Functioning and disused hands were observed in 61 and 71 patients, respectively. No differences were observed between these 2 stroke groups in terms of age, gender, duration of illness, affected side, or type of stroke. The Barthel Index (BI) was significantly lower in the disused hand group than in the functioning hand group. Twenty-seven patients (44%) in the functioning hand group and 27 patients (38%) in the disused hand group were able to move about independently using an assist device such as a cane or wheelchair, with no help from another person. The remaining patients in both groups could ambulate independently without assist devices. No significant difference was evident in relative numbers of patients needing such devices between functioning and disused hand groups.
Symptom and Signs of Carpal Tunnel Syndrome on the Unaffected Side
Dysesthesias were not observed in any patient. As shown in Table 2⇓, paresthesia, dry skin, color change, or mild thenar muscle atrophy occasionally was observed contralaterally to a paretic but functioning or disused hand. No significant difference was seen between functioning and disused groups in the prevalence of these signs. Mildly decreased sensation was observed in a median nerve distribution in 19 patients, more frequently in the disused hand group than in the functioning hand group. Notably, Tinel’s sign was significantly more prevalent in the disused hand group (57.7%) than in the functioning hand group (31.1%).
Sensory and Motor Conduction in the Median Nerve
For electrophysiological indices measured separately on the left and right in controls, left and right sides were averaged and the mean was used as the control value (see Table 3⇓). In control subjects, only slight and insignificant differences were observed in electrophysiological indices between left and right sides. SNCV, DSL from the palm to the wrist, MNCV, and DML for the median nerve on the unaffected side in stroke patients were significantly prolonged compared with the hemiparetic side and the controls. Also, in stroke patients, SNAP, SNAP from the palm to the wrist, and median nerve CMAP were significantly reduced on the unaffected side versus amplitudes on the hemiparetic side and in controls.
SNCV, DML, and CMAP were slower or reduced in the disused hand group than in the functioning hand group, although the differences were not statistically significant. In both groups, however, these indices were significantly slower or reduced relative to those in controls. SNAP and SNAP from the palm to the wrist were significantly reduced in the disused hand group compared with controls; no significant changes in SNAP or SNAP from the palm to the wrist were observed between the functioning hand and control groups. DSL from the palm to the wrist was significantly slower in the disused hand group than in both the functioning hand and control groups.
The percentage of cases 2 SD below or beyond the control mean was higher in the disused hand group than in the functioning hand group for all electrophysiological indices except for MNCV and CMAP. The highest occurrence of abnormal values was 62%, for the distal latency from the palm to the wrist in the disused hand group, followed by 41% for the DML in the disused hand group (Table 4⇓).
Gender and Sidedness
CTS generally occurs more frequently in women than in men, and in both genders the dominant hand is more likely to be affected than the nondominant hand.13 However, no significant gender-related differences and no side-to-side differences in patients were seen for any electrophysiological parameter (data not shown).
Mobility With and Without Assist Devices
No significant differences in the 7 electrophysiological indices were seen between patients using and not using assist devices for ambulation, except for DSL from the palm to the wrist, which showed a significant delay in the device-using group compared with the no-device group (2.184±1.033 versus 1.934±0.341, P=0.0365). No significant differences in the occurrence of neurological symptoms and signs indicative of CTS were seen between device and nondevice groups (data not shown).
In our electrophysiological study we demonstrated that after stroke subclinical CTS occurs on the “unaffected” side relative to the hemiparetic side or to control subjects. All 7 electrophysiological indices were significantly prolonged or reduced for the median nerve on the unaffected side relative to paretic side. Because no significant difference was seen in the SNCV of the ulnar nerve between unaffected and hemiparetic sides, the observed electrophysiological abnormalities of the median nerve on the unaffected side cannot be explained by postulating a subclinical compression neuropathy, which is fairly common among the elderly.14 Neurological examination disclosed a relatively high prevalence of mildly decreased sensation in the median nerve distribution as well as Tinel’s sign, and these were seen more frequently in the disused hand group than in the functioning hand group. Indeed, palm-to-wrist DSL was significantly more prolonged in the disused hand group than in the functioning hand group. In addition, SNAP and SNAP from the palm to the wrist were significantly reduced in the disused hand relative to those in controls. Decreased sensation in the median nerve distribution and atrophy of the thenar muscles on the unaffected side in patients after a stroke may have been overlooked, because these findings were mild in the present study.
The prevalence of CTS in the general population has been reported to be 125 to 220 per 100 000.8 Repetitive use of the hand or wrist is a common cause of CTS.4 5 6 7 8 The mechanism involves increased pressure within the carpal canal due to nonspecific flexor tenosynovitis.8 Stroke patients with disuse of a hand may overuse their unaffected hand, particularly when assistive devices are used for mobility. In this study we expected that patients using assist devices would have more overuse of the hand than those not using a device, but 6 of 7 electrophysiological parameters did not differ significantly between device users and nonusers. However, palm-to-wrist DSL was more abnormal in the disused hand group than in the functioning hand group; a higher percentage of patients with a disused hand deviated >2 SD from control means for the indices than when patients had a functioning hand. Therefore, the determinant of CTS on the unaffected side is not use of an assist device but functional severity of palsy of the opposite hand.
Sensory fibers constitute 94% of median nerve fibers15 and are the first to be affected by compression in the carpal tunnel. Not only sensory parameters such as SNCV, SNAP, or DSL from the palm to the wrist but also motor parameters showed involvement, as evidenced by prolonged DML or reduced CMAP in both functioning and disused hand groups. This implies that continuing repetitive use of the unaffected hand results in involvement of both sensory and motor fibers of the median nerve.
Significant deficits in the unaffected hand of hemiparetic stroke patients compared with normal subjects have been reported with regard to grip strength, flexion of fingers and thumb, pinch strength, kinesthesia for the thumb, gross manual dexterity, fine manual dexterity, and motor coordination.1 2 3 We believe that CTS represents 1 cause of these symptoms, because “centric” sensory deficits such as s kinesthesia may be seen in CTS.16
To prevent CTS on the intact side in patients after stroke, wrist splints17 or administration of nonsteroidal anti-inflammatory drugs16 or possibly vitamin B618 may prove useful if such treatment is begun soon after stroke. If definite symptomatic CTS occurs in patients such as the 3 who were the index cases for the present study, surgical therapy19 should be considered preferable to local steroid injections20 or short-term treatment with oral prednisone21 because such conservative therapy cannot prevent progression of thenar muscle atrophy.
- Received September 29, 1998.
- Revision received November 16, 1998.
- Accepted November 24, 1998.
- Copyright © 1999 by American Heart Association
Colbebatch JG, Gandevia S. The distribution of muscular weakness in upper motor neuron lesions affecting the arm. Brain. 1989;112:749–763.
Smutok MA, Grafman G, Salazar AM, Sweeney JK, Jonas BS, DiRocco PJ. Effects of unilateral brain damage on contralateral side and ipsilateral upper extremity function in hemiplegia. Phys Ther. 1989;69:195–203.
Desrosiers J, Bourbonnalis D, Bravo G, Roy PM, Guay M. Performance of the “unaffected” upper extremity of elderly stroke patients. Stroke. 1996;27:1564–1570.
Cannon L, Bernnacki E, Walter S. Personal and occupational factors associated with carpal tunnel syndrome. J Occup Med. 1981;23;255–258.
Dawson MW. Entrapment neuropathies of the upper extremities. N Engl J Med. 1993;2013–2018.
Scandinavian Stroke Study Group. Multicenter trial of hemodilution in ischemic stroke: background and study protocol. Stroke. 1985;16:885–890.
Buchthal F, Rosenfalk A. Evoked action potentials and conduction velocity in human sensory nerves. Brain Res. 1966;3:1–22.
Report of the Quality Standards Subcommittee of the American Academy of Neurology. Practice parameter for carpal tunnel syndrome. Neurology. 1993;43:2406–2409. Summary statement.
Chang MH, Chiang HT, Lee SSJ, Ger LP, Lo YK. Oral drug in carpal tunnel syndrome. Neurology. 1998;51:390–393.