ORIGINAL ARTICLE

Acupuncture in Patients With Carpal Tunnel Syndrome

A Randomized Controlled Trial
Chun-Pai Yang, MD,*w z Ching-Liang Hsieh, MD, PhD,w y Nai-Hwei Wang, MD,Jz
Tsai-Chung Li, PhD,z# Kai-Lin Hwang, MSc, ** Shin-Chieh Yu, MD,* and
Ming-Hong Chang, MD w w zz

(P = 0.012). Acupuncture was well tolerated with minimal adverse

Objectives: To investigate the efficacy of acupuncture compared effects.
with steroid treatment in patients with mild-to-moderate carpal
tunnel syndrome (CTS) as measured by objective changes in nerve Conclusions: Short-term acupuncture treatment is as effective as
conduction studies (NCS) and subjective symptoms assessment in a short-term low-dose prednisolone for mild-to-moderate CTS. For
randomized, controlled study. those who do have an intolerance or contraindication for oral
steroid or for those who do not opt for early surgery, acupuncture
Methods: A total of 77 consecutive and prospective CTS patients treatment provides an alternative choice.
confirmed by NCS were enrolled in the study. Those who had fixed
sensory complaint over the median nerve and thenar muscle Key Words: acupuncture, carpal tunnel syndrome, CTS, steroid,
atrophy were excluded. The CTS patients were randomly divided global symptom score (GSS)
into 2 treatment arms: (1) 2 weeks of prednisolone 20 mg daily (Clin J Pain 2009;25:327–333)
followed by 2 weeks of prednisolone 10 mg daily (n = 39), and (2)
acupuncture administered in 8 sessions over 4 weeks (n = 38). A
validated standard questionnaire as a subjective measurement was
used to rate the 5 major symptoms (pain, numbness, paresthesia,
weakness/clumsiness, and nocturnal awakening) on a scale from 0
(no symptoms) to 10 (very severe). The total score in each of the 5 C arpal tunnel syndrome (CTS), which results from the
compression of the median nerve at the wrist, can be
caused by many different factors. Any condition that reduces
categories was termed the global symptom score (GSS). Patients
completed standard questionnaires at baseline and 2 and 4 weeks the dimensions of the tunnel or increases the volume of its
later. The changes in GSS were analyzed to evaluate the statistical content will predispose individuals to CTS, and many
significance. NCS were performed at baseline and repeated at the medical associations have been reported (ie, diabetes mellitus,
end of the study to assess improvement. All main analyses used renal failure, thyroid disease, rheumatoid arthritis), but most
intent-to-treat. cases are idiopathic.1–3 The typical symptoms of CTS include
sensory impairments, such as numbness or pain in the wrist,
Results: A total of 77 patients who fulfilled the criteria for mild-to- hand and fingers, which often occur during sleep and awaken
moderate CTS were recruited in the study. There were 38 in the CTS patients occasionally. Shaking or rubbing the hands
acupuncture group and 39 in the steroid group. The evaluation of usually relieves the symptoms. The motor symptoms of CTS
GSS showed that there was a high percentage of improvement in include weakness of the thenar muscle, and loss of hand
both groups at weeks 2 and 4 (P<0.01), though statistical dexterity and function. Both objective and subjective
significance was not demonstrated between the 2 groups symptoms can occur unilaterally or bilaterally. The best
(P = 0.15). Of the 5 main symptoms scores (pain, numbness, way to confirm the diagnosis is to carry out a median nerve
paresthesia, weakness/clumsiness, nocturnal awakening), only 1, conduction study (NCS) across the transverse carpal
nocturnal awakening, showed a significant decrease in acupuncture ligament. A characteristic of the condition is a focal
compared with the steroid group at week 4 (P = 0.03). Patients conduction slowing in NCS across the wrist segment.3–5
with acupuncture treatment had a significant decrease in distal Many conservative treatments are commonly used in
motor latency compared with the steroid group at week 4 mild and moderate CTS. For these patients, short-term
nonsurgical management may be desirable and may reduce
the number of patients undergoing surgical intervention.
Received for publication May 25, 2008; revised September 14, 2008; Among the conservative treatments, there is strong
accepted September 20, 2008. evidence that local corticosteroid injections, and to a lesser
From the Departments of *Neurology; JOrthopedics, Kuang Tien
General Hospital; wGraduate Institute of Acupuncture Science;
extent oral corticosteroids, provide short-term relief for
zGraduate Institute of Chinese Medical Science, College of Chinese CTS sufferers.6,7 In addition, splints are effective, especially
Medicine; #Biostatistics Center, China Medical University; if used full time6,7; however, many CTS patients report that
yChinese Medicine, China Medical University Hospital; **Depart- splinting restricts hand activity and hinders their ability to
ment of Public Health, Chung Shan Medical University; zHuang-
Kuang University; wwSection of Neurology, Taichung Veterans
work or perform daily activities.8 Local steroid injections
General Hospital, Taichung; and zzDepartment of Neurology, into the carpal tunnel may result in initial relief, but
National Yang-Ming University, Taipei, Taiwan. relapses are frequent, and mechanical or chemical nerve
Supported by KTGH grant. injury can occur.7,8 Oral steroids are better than nonsteroid
Reprints: Ming-Hong Chang, MD, Section of Neurology, Veterans
General Hospital, No 160, Chung-Kang Road, Section 3,
anti-inflammatory drugs and diuretics, but they can
Taichung, Taiwan, 40705 (e-mail: cmh50@ms10.hinet.net). produce side effects, which preclude their routine use for
Copyright r 2009 by Lippincott Williams & Wilkins CTS.7 Acupuncture is a complementary medical technique

Clin J Pain  Volume 25, Number 4, May 2009 www.clinicalpain.com | 327

Yang et al Clin J Pain  Volume 25, Number 4, May 2009

used for the treatment of painful disorders. However, at (Daling), PC-6 (Neiguan)] on the affected side in their 8
present, there is no conclusive evidence of the efficacy of sessions without modification for the specific symptoms of
acupuncture in treatment of CTS.6,8 In an attempt to the patients. We placed patients in the supine position to
investigate whether acupuncture is as effective and safe as make them more comfortable. Sterile disposable steel
steroid in the treatment of mild-to-moderate CTS, we needles (gauge and size: 0.25  40 mm) were used without
conducted a prospective, randomized clinical study under electrical stimulation or moxibustion. At each point, the
conditions similar to routine care. skin was wiped with alcohol and needles were inserted
perpendicularly at PC-6 to a depth of 1.0 to 1.5 inch and at
PC-7 they were inserted from 0.5 to 1.0 inch according to
PATIENTS AND METHODS the thickness of the patient’s wrist. The needles were
The study protocol was approved by the institutional manipulated by twirling with lifting-thrusting methods to
review board of our hospital. produce a characteristic sensation known as De Qi (an
awareness of numbness, soreness, swelling, heaviness, or
Patients radiating feeling from the point of needling deemed to
The patients, aged from 18 to 85, enrolled in this study indicate proper needle position and effective needling) and
had clinical symptoms and signs of CTS. CTS was were then left in place for 30 minutes. For patients with
diagnosed clinically based on the presence of at least one bilateral CTS, both wrists were needled and data were
of the following primary symptoms: (1) numbness, tingling reported separately.
pain, or paresthesia in the median nerve distribution; (2) However, we included only the more-affected hand
precipitation of these symptoms by repetitive hand activ- with a higher GSS in each individual for data analysis. As
ities, which could be relieved by resting, rubbing, and only 1 hand with a higher GSS score from each individual
shaking the hand; and (3) nocturnal awakening by such was used for analysis, the number of participants was equal
sensory symptoms. The diagnosis was often supported by a to the number of affected arms enrolled in the analysis set.
positive Tinel sign. All patients with clinically diagnosed All treatments were performed at the same facility by 1
CTS demonstrated median neuropathy at the wrist, acupuncturist. Additionally, the acupuncturist was asked to
confirmed by the presence of 1 or more of the following have the least possible communication with patients to
standard electrophysiologic criteria: (1) prolonged distal minimize bias. Complete details of the intervention are
motor latency (DML) to the abductor pollicis brevis (APB) presented in Table 1 in conformance to the standards for
(abnormal Z4.7 ms, stimulation over the wrist, 8 cm reporting interventions in controlled trial of acupuncture.13
proximal to the active electrode); (2) prolonged antidromic
distal sensory latency (DSL) to the second digit (abnormal Measures
Z3.1 ms; stimulation over the wrist, 14 cm proximal to the
active electrode); and (3) prolonged antidromic wrist-palm Electrophysiologic Assessment
sensory nerve conduction velocity (W-P SNCV) at a The median and ulnar nerves were studied with no
distance of 8 cm (W-P SNCV, abnormal <45 m/s).9–12 If abnormality in the ulnar nerves. Motor and sensory NCS
the patients fulfilled the criteria and gave written informed were performed using standard techniques of supramaximal
consent before randomization, they were enrolled in the percutaneous stimulation and surface electrode recording.
study. Possible side effects were fully explained. At their DML and DSL, motor nerve conduction velocity, com-
first visit, we assessed their medical and neurologic history, pound muscle action potential (CMAP), sensory nerve
gave them detailed physical and neurologic examinations, action potential (SNAP) amplitudes, and W-P SNCV were
biochemical and endocrine screenings (ie, fasting blood measured using the methods described by Delisa et al.9 The
sugar, thyroid stimulating hormone, free T4), NCS and
needle electromyography. Before treatment, the patients
were followed-up for 1 month. If improvement occurred
during observational periods, patients were excluded from TABLE 1. Standards for Reporting Interventions in Controlled
this study. After enrollment, the patients were randomized Trials of Acupuncture (STRICTA)
into 2 treatment arms: (1) a group receiving 2 weeks of Acupuncture A fixed and classic acupuncture points
20 mg prednisolone daily followed by 10 mg daily for rationale (PC-7 (Daling), PC-6 (Neiguan) in their
another 2 weeks; and (2) a group receiving acupuncture in 8 8 sessions
sessions over 4 weeks. The randomization was carried out Unilateral or bilateral points
according to computer-generated randomly allocated treat- Needling details
Depth of insertion: standard to each point
ment codes and data were kept by a person not involved in
according to classic acupuncture point
the care or evaluation of the patients or in the data analysis. Responses elicited: de qi sensation
All patients received complete global symptom score (GSS) Manual: twirling with lifting-thrusting
measurements at baseline, 2, and 4 weeks and NCS at method stimulation
baseline and 4 weeks later performed by the same blinded Needles retained for 30 min
evaluator throughout the entire study period. All patients Needle type: C&G, gauge and size:
were scheduled so as to avoid any overlap during which 0.25  40 mm
they could share clinical information and experiences with Treatment regimen Twice per week for 4 wk
each other. Cointervention None: no herbs, moxibustion, cupping,
rehabilitation advice regarding dietary
Acupuncture Treatment or lifestyle modifications
Practitioner License-certificated
Acupuncture consisted of 8 sessions of 30-minute background
duration, administrated over 4 weeks (2 sessions/wk). Each Control intervention Nil
patient had fixed and classic acupuncture points [PC-7

328 | www.clinicalpain.com r 2009 Lippincott Williams & Wilkins

Clin J Pain  Volume 25, Number 4, May 2009 Acupuncture in Patients With Carpal Tunnel Syndrome

electromyographic recording (Viking IV; Nicolet WI, improvement, we repeated NCS at the end of the
Madison, WI) of motor conduction studies were made assessment for those patients who completed the study.
with the filter band pass at 2 to 10 Hz, a sweep speed of But for the patients lost to follow-up and those who
2 ms/cm, and the amplifier gain adjusted for full reviewing received surgery, we decided not to repeat the NCS.
of the CMAP. For measurement of SNAP, the instrument Additional treatments (such as splinting and local injec-
settings were: filters, 20 Hz to 10 kHz; sweep, 2 ms/cm; gain, tions) or alterations in daily activities were not permitted
10 to 20 mV/cm. during the study.
Patients were excluded if any of the following were
present: (1) symptoms occurring less than 3 months before Safety Assessments
the study or symptoms improving during the 1-month Patients reported all serious adverse events with side
initial observation period (to exclude patients who might effects of both oral steroids treatment at weeks 2 and 4 and
have spontaneous resolution of symptoms); (2) severe CTS acupuncture treatment in each session. We recorded
that had progressed to visible muscle atrophy; (3) in our adverse side effects such as nausea, epigastric pain, tarry
study, mild CTS referred to patients with decreased stools, leg edema, cushingoid appearance, blood pressure,
conduction velocity over the palm-wrist segment and blood sugar along with ecchymosis, local paresthesia, or
delayed DSL, with normal median SNAP amplitude and bleeding to treat analysis for all enrolled patients.
CMAP amplitude of the APB. Moderate CTS referred to
patients with abnormally delayed DML and DSL with
either decreased median SNAP amplitude or decreased Statistical Analysis
CMAP amplitude of the APB muscle. Thus, CTS patients A last-observation-carried-forward approach was used
with the presence of either fibrillation potentials or to input missing data with the intent-to-treat analysis
reinnervation on needle EMG in the APB were excluded principle. Independent 2-sample t test was performed to
(to ensure the inclusion of only mildly or moderately compare the efficacy of the objective changes in nerve
affected individuals); (4) clinical or electrophysiologic conduction and subjective symptoms assessment between
evidence of accompanying conditions that could mimic the 2 groups for the baseline, 2-week and 4-week evalua-
CTS or interfere with its evaluation, such as cervical tions. Repeated measures analysis of variance with Bon-
radiculopathy, proximal median neuropathy, or significant ferroni adjustment for multiple testing was used to compare
polyneuropathy; (5) evidence of obvious underlying causes the changes in subjective symptoms assessment between
of CTS such as diabetes mellitus, rheumatoid arthritis, week 2 or 4 data and baseline data within each treatment
hypothyroidism (acromegaly), pregnancy, alcohol abuse or group. Paired t test was performed for objective changes in
drug usage (steroids or drugs acting through the central nerve conduction between week 4 data and baseline data
nervous system), use of vibrating machinery, and suspected within each treatment group. For 5 main symptoms score
malignancy or inflammation or autoimmune disease were of GSS and 6 measures of NCS, Bonferroni adjustment was
documented as underlying causes for CTS; (6) recent peptic made to control for type I error. All hypothesis testing were
ulcer or history of steroid intolerance; (7) prior unpleasant 2-tailed and level of significance was set at 0.05. All
experience with acupuncture or a bleeding diathesis; or (8) statistical analyses were performed using SPSS Version 15.0
cognitive impairment interfering with the patient’s ability to for Windows (SPSS Inc, Chicago, IL).
follow instructions and describe symptoms.
RESULTS
Clinical Assessments Enrollment of Patients and Baseline
Clinical assessments included the symptomatic ques- Characteristics
tionnaire modified from that used by Herskovitz et al14 and A total of 77 patients who fulfilled the inclusion and
by us in our previous study.10,11 We rated symptoms from 0 exclusion criteria agreed to participate in our study and
(no symptoms) to 10 (very severe symptoms) in each of 3 were randomly allocated to either the steroid or acupunc-
categories: pain, numbness, and paresthesia. Nocturnal ture treatment group. The baseline characteristics of the 2
awakening was scored by times awakened in 1 week: never, groups were similar in the intention-to-treat population
0; once or twice, 2; 3 or 4 times, 4; 5 to 7 times, 6; 8 to 10 (Table 2). Of the 77 patients, 3 patients in the acupuncture
times, 8; more than 10 times, 10. Weakness was scored group dropped out due to inability to take time off work,
according to the severity of the weakness: none, 0; mild, 2; and 4 patients in the steroid group did not finish the study
moderate, 3; severe, 4; very severe, 5; and assessed for due to intolerance of side effects of epigastric pain with
clumsiness by difficulty in manipulating small objects: none, nausea. No patients received surgery before the end of the
0; mild, 2; moderate, 3; severe, 4; very severe, 5. The total of
the scores of the 5 main symptoms was the GSS. Each
patient was directly questioned, and each score was based
on the patient’s subjective answers. Therefore, the maxi- TABLE 2. Summary of Baseline Characteristics of Study Patients
mum score was 50 (most severe symptoms) and the Acupuncture Group Steroid Group
minimum score was 0 (absence of symptoms). Furthermore, Number of patients 38 39
to ensure consistency, the evaluating physician was the Age (y) 49.3 (8.9) 49.9 (10.3)
same person on each occasion for each patient. Follow-up Sex (female/male) 38/6 39/8
assessments identical to the baseline procedure were Duration (mo) 7.6 (3.8) 7.7 (3.2)
performed at 2 and 4 weeks later. Baseline GSS 16.0 (8.7) 14.3 (7.5)
At the end of the study, neurologic examinations were
Values are number or mean (standard deviation, SD).
repeated, along with the same biochemical and endocrine GSS indicates global symptom score.
examinations as at baseline. To obtain objective evidence of

r 2009 Lippincott Williams & Wilkins www.clinicalpain.com | 329

Yang et al Clin J Pain  Volume 25, Number 4, May 2009

Fulfill inclusion and exclusion criteria

N=90

Excluded because
patients were not
interested or difficult
to find time to Total number of patients
cooperate N=13 randomized N=77

Randomized to Randomized to steroid

acupuncture group group N=39
N= 38

4 dropped out due to

3 dropped out due abdominal pain &
to take time off nausea
work

Completed 4 weeks study Completed 4 weeks study

N=35 N= 35

FIGURE 1. Flow chart of process and disposition of patients.

study. The dropout rate was low for both the steroid and significant difference between the 2 groups before treat-
acupuncture groups. We substituted baseline values for the ment. At the end of the study, there was a high percentage
missing data of the 7 patients who did not complete the of improvement in both the acupuncture and steroid groups
study (thus, setting differences compared with baseline to at weeks 2 and 4 (all P<0.01 for both groups), though
zero). Figure 1 illustrates patient enrollment and random statistical significance was not achieved between the 2
allocation of patients to study groups. There was no groups (P = 0.15) (Fig. 2A). Of the 5 parameter scores
difference in age, sex, or duration of symptoms between (pain, numbness, paresthesia, weakness/clumsiness, noctur-
treatment groups. nal awakening), only 1, nocturnal awakening showed a
significant decrease between the 2 groups. Patients with
Outcome of Treatment acupuncture treatment had significantly better improve-
Table 3 shows the changes in GSS for the 77 patients ment in nocturnal awakening compared with the steroid
who were available for the efficacy analysis. There was no group at week 4 (P = 0.03) (Fig. 2B).

TABLE 3. Cumulative Data of Global Symptom Score (GSS) Changes

GSS Score Numbness Pain Paresthesia Weakness Nocturnal Awakening
Baseline
Acupuncture 16.1 (8.8) 6.9 (3.1) 0.8 (2.2) 1.5 (2.8) 3.3 (3.5) 3.5 (3.8)
Steroid 14.3 (7.5) 7.1 (2.0) 0.6 (1.1) 1.0 (2.6) 2.6 (3.2) 3.0 (3.7)
Change* % Changew Change* Change* Change* Change* Change*
Week 2
Acupuncture  8.6 (6.3)  51.6 (22.4)  3.7 (2.5)  0.6 (1.9)  0.9 (1.9)  2.0 (2.5)  2.7 (3.5)
Steroid  7.3 (5.7)  51.1 (24.7)  3.3 (1.6)  0.3 (0.7)  0.7 (1.9)  1.5 (2.5)  1.9 (3.0)
Week 4
Acupuncture  11.7 (7.6)  70.0 (24.6)  4.9 (2.8)  0.8 (2.2)  1.4 (2.7)  3.0 (3.3)  3.5 (3.8)z
Steroid  9.3 (6.7)  64.7 (27.6)  4.0 (2.0)  0.3 (0.8)  0.8 (2.2)  1.9 (2.8)  1.5 (1.9)
*Week 2 or 4—baseline.
wWeek 2 or 4—baseline/baseline.
zP<0.05 after Bonferroni adjustment.
Values are mean (standard deviation, SD).

330 | www.clinicalpain.com r 2009 Lippincott Williams & Wilkins

Clin J Pain  Volume 25, Number 4, May 2009 Acupuncture in Patients With Carpal Tunnel Syndrome

A by 5% of the patients. Most adverse effects were related to

30
the local insertion of the needles, such as local pain after
Acupuncture session, ecchymosis, and local paresthesia during session.
25 Steroids Acupuncture was well tolerated by patients and no one
discontinued prematurely because of needle-related side
Global Symptom Score

20 effects. In the steroid treatment group, the most frequently

noted adverse effects were nausea and epigastralgia. Side
effects from steroid were reported by 18% of the patients.
15
Four patients dropped out due to intolerance of severe
epigastralgia with nausea.
10
∗∗

5 ∗∗
∗∗ DISCUSSION
∗∗ The present study is one of the most rigorous trials of
the efficacy of acupuncture treatment versus proven
0
0 1 2 3 4 5
standard drugs on CTS available. Its strength includes
interventions based on expert consensus by qualified and
Time (Weeks) experienced medical acupuncturists, assessment of the
B credibility of interventions, and outcome measurements as
recommended in guidelines for trials on CTS. The results of
Change from Baseline of Nocturnal Awakening

2 the current study showed that there was a high percentage

of improvement in both groups at week 4 with subjective
measurement of GSS, though statistical significance was
0
not achieved between the 2 groups. Furthermore, patients
with acupuncture treatment had significantly better im-
-2
provement in the main symptoms score of nocturnal
awakening compared with the steroid group at week 4. In
the assessment with objective measurement of NCS,
∗
-4 patients with acupuncture treatment had significantly better
improvement in DML compared with the steroid group at
week 4. It can be concluded that acupuncture treatment had
-6 Acupuncture at least equal, and in some cases, superior efficacy when
Steroids compared with steroid treatment not only in objective
changes in nerve conduction but also in subjective
-8 symptoms assessment. However, the disadvantage of
0 1 2 3 4 5 acupuncture is that it is time-consuming.
Time (Weeks) Several large surveys have also provided evidence that
acupuncture is a relatively safe treatment.15–18 Acupuncture
FIGURE 2. A, Change of total global symptom score for
treatments were well tolerated by our patients. Indeed, most
acupuncture and steroid groups over time. A significant
difference from baseline for weeks 2 and 4 were observed by patients found participation in the study to be pleasant and
repeated measures analysis of variance for both groups rewarding. Needle-related side effects like bruising and
(P<0.01);**P< 0.01 (B) Change of nocturnal awakening for soreness were more common in the acupuncture group than
acupuncture and steroid groups over time. A significant in the steroid group, but these were mild and did not affect
difference between acupuncture and steroid groups at week 4 treatment. No patient withdrew due to adverse effects.
was observed (P<0.05) by independent 2-sample t test. However, in the steroid group, 4 patients dropped out due
*P< 0.05. to intolerance of severe epigastralgia with nausea. Some
might ask why patients with acupuncture treatment had
Table 4 illustrates the outcome and severity of NCS significant improvement not only in objective changes in
findings including DML, CMAP amplitude of APB muscle, NCS but also in subjective symptoms assessment. Acu-
motor nerve conduction velocity, DSL, W-P SNCV, and puncture treatment is an invasive manual procedure; thus,
SNAP amplitudes of median nerves before and after separating the specific effects from nonspecific effects is
treatment in both groups. There was no significant extremely difficult.19 Various neurophysiologic and psycho-
difference between the 2 groups before treatment. After physiologic mechanisms underlying the analgesic effective-
treatment, there was a significant decrease in DML and ness of acupuncture have been hypothesized.19 However,
DSL, and a significant increase in W-P SNCV and SNAP even though acupuncture therapy has been used exten-
amplitudes within each treatment group (P<0.05) for both sively, its mechanisms of action in CTS are not precisely
steroid and acupuncture groups. In addition, there was known, in part because the pathophysiology of CTS itself is
significantly increased CMAP amplitude of the APB muscle not well understood. CTS etiology is thought to involve
in the steroid group (P<0.05). Patients with acupuncture compression of the distal median nerve due to an elevated
treatment had significantly better improvement in DML interstitial fluid pressure in the carpal tunnel. Ischemic
compared with steroid group at week 4 (P = 0.012) (Fig. 3). injury and mechanical deformity of the median nerve
produced by elevated pressure within the carpal tunnel
Adverse Side Effects leads to anoxic capillary damage, which in turn leads to
No serious adverse effects were noted. In the increased membrane permeability, exudative edema, and
acupuncture treatment group, side effects were reported subsequent fibrosis.14,20–22 Steroids are effective at reducing

r 2009 Lippincott Williams & Wilkins www.clinicalpain.com | 331

Yang et al Clin J Pain  Volume 25, Number 4, May 2009

TABLE 4. Improvement in Electrodiagnostic Measurements in Patients With Carpal Tunnel Syndrome who had Symptom Relief
Acupuncture Group Steroid Group
Electrodiagnostic Variable, With Normal Result Baseline After Treatment Baseline After Treatment Pw
DML (ms), <4.7 5.6 (0.9) 4.0 (0.7)* 5.6 (1.3) 4.7 (1.0)* 0.012
CMAP (mv), >6.5 7.2 (2.9) 7.2 (2.7) 7.2 (2.8) 7.6 (2.8)* NS
MNCV (mv), >50 53.1 (4.5) 53.7 (3.8) 51.9 (4.1) 52.4 (3.6) NS
DSL (ms), <3.1 3.7 (1.0) 3.3 (0.7)* 3.4 (0.8) 3.0 (0.6)* NS
W-P SNCV (m/s), >45 40.0 (8.6) 43.9 (8.0)* 43.3 (9.5) 48.6 (6.2)* NS
SNAP (mv), >15 15.4 (9.0) 18.4 (9.8)* 17.4 (9.3) 20.8 (9.9)* NS
*P<0.05 compared with baseline within group by paired t test with Bonferroni adjustment.
wThe change from baseline was compared between groups with independent t test.
Values are mean (standard deviation, SD).
CMAP indicates compound muscle action potential; DML, distal motor latency; DSL, distal sensory latency; MNCV, motor nerve conduction velocity;
NS, non-significant; SNAP, sensory nerve action potential; W-P SNCV, wrist-palm sensory nerve conduction velocity.

swelling because of their anti-inflammatory action. It is thus Furthermore, if both treatments are possibly effective, it is
reasonable to use oral steroids in the treatment of CTS and easy to explain and encourage patients to be recruited in
a short-term course of low-dose steroids can be of great current study. Recently, a Japanese study found that most
benefit in the treatment of mild-to-moderate people in Asian countries have knowledge about acupunc-
CTS.10,12,13,23,24 A recent study suggests that acupuncture ture and have received acupuncture treatment, and 60% of
may possess anti-inflammatory action via release of the patients could distinguish between sham and genuine
neuropeptides from nerve endings.25 There is also evidence needling.30 Our patients were also able to make this
that acupuncture processing in the brains of CTS patients distinction, so we did not choose sham acupuncture in
differs from that of healthy controls.26 It would be of great our study. Steroid treatment is one of the most common
interest to know what roles the peripheral and the central used drugs in clinical practice for treatment of mild-to-
mechanisms play in CTS patients after acupuncture moderate CTS. But in our society, most people are
treatment, although it is beyond the scope of this article. reluctant to take it. So, we set out to answer the clinically
In traditional Chinese medical literature, the acupuncture relevant question, ‘‘does acupuncture improve outcomes
point Neiguan has been shown to relieve insomnia.27 This among patients with mild-to-moderate CTS comparable to
may explain why patients who received acupuncture steroid treatment?’’ This is substantially different from the
treatment had significantly better improvement in noctur- question, ‘‘does acupuncture improve outcomes compar-
nal awakening compared with the steroid group at week 4. able to a sham procedure that appears to be similar to, but
The investigators are aware of and capable of using isn’t really, acupuncture?’’ Therefore, an active instead of
sham acupuncture28,29; however, the reason for our placebo control was used in this study, and the steroid
preference for an active drug rather than placebo was less treatment for CTS was chosen as a comparison.
ethical problem to adopt an active treatment arm for The natural history in CTS patients was not well
patients who looked for a treatment for their discomforts. characterized until a recent study by Padua et al.31 In their
study of 441 hands afflicted with idiopathic CTS, they
found that 21% of hands improved over 10 to 15 months of
1 follow-up without active intervention. Thus any therapeutic
Change from Baseline of motor distal latency

intervention should attempt to achieve a better than 21%

improvement rate.31 Genuine acupuncture is widely
0 accepted in Taiwan and oral steroid is considered as
an alternative conservative in previous studies.10,12,13,23,24
-1 Though there is no real placebo group in current study,
∗
however, a placebo effect or spontaneous resolution would
have been less likely to occur due to the patients’ more than
-2 21% improvement in GSS in both groups. In addition,
there was improvement in the objective measures, NCS, in
-3
patients after acupuncture and steroid treatment. Further-
more, in 1 previous study, nearly a quarter of the patients
had relief of symptoms within the first month of initial
-4 assessment.32,33 To decrease this confounding effect, any
patient whose symptoms occurred less than 3 months
-5 before the study or whose symptoms improved during the
N= 38 39 first observation period was excluded from current study.
Only 4 patients had marked relief of symptoms during the
Acupuncture Steroid observation period and they were excluded.
FIGURE 3. Change from baseline of motor distal latency (DML) Although we conclude that short-term acupuncture
between acupuncture and steroid groups by independent treatment is an effective and safe treatment for symptom-
sample t test. *P< 0.05. atic relief in CTS, some questions remain unanswered:

332 | www.clinicalpain.com r 2009 Lippincott Williams & Wilkins

Clin J Pain  Volume 25, Number 4, May 2009 Acupuncture in Patients With Carpal Tunnel Syndrome

1. Is acupuncture therapy effective for long-term symptom tional conduction techniques in electro diagnosis of carpal
relief of CTS? tunnel syndrome. Clin Neurophysiol. 2006;117:984–991.
2. Do symptoms recur once acupuncture is discontinued, 13. MacPherson H, White A, Cummings M, et al. Standards for
and is further acupuncture therapy effective in patients reporting interventions in controlled trials of acupuncture: the
STRICTA recommendations. J Altern Complement Med. 2002;
experiencing a recurrence? 8:85–89.
3. What is the mechanism of acupuncture on CTS? 14. Herskovitz S, Berger AR, Lipton RB. Low-dose, short-term
To answer these questions, we are currently conduct- oral prednisone in the treatment of carpal tunnel syndrome.
ing other studies. Future studies may also consider Neurology. 1995;45:1923–1925.
additional assessments using validated commonly used 15. White A, Hayhoe S, Hart A, et al. Survey of adverse events
disability scales such as the SF-36, Disability of Arm, following acupuncture (SAFA): a prospective study of 32,000
Shoulder and Hand questionnaire to make comparison of consultations. Acupunct Med. 2001;19:84–92.
the data to other published literature more relevant. 16. Melchart D, Weidenhammer W, Streng A, et al. prospective
investigation of adverse effects of acupuncture in 97733
patients. Arch Inter Med. 2004;164:104–105.
17. Yamashita H, Tsukayama H, Hori N, et al. Incidence of
adverse reactions associated with acupuncture. J Altern
CONCLUSIONS Complement Med. 2000;6:345–350.
Despite the limitations, this randomized, controlled 18. Witt C, Brinkhaus B, Jena S, et al. Acupuncture in patients with
study indicates that short-term acupuncture treatment is as osteoarthritis of the knee: a randomized trial. Lancet. 2005;366:
effective as short-term low-dose steroid for mild-to-moder- 136–143.
ate CTS. For those who do not tolerate oral steroid or for 19. Cabyoglu MT, Ergene N, Tan U. The mechanism of
those who do not opt for surgery, acupuncture treatment Acupuncture and clinical applications. Int J Neurosci. 2006;
provides an alternative choice. We now need to assess the 116:115–125.
20. Werner RA, Andary M. Carpal tunnel syndrome: pathophysio-
long-term effects of acupuncture on mild-to-moderate CTS
logy and clinical neurophysiology. Clin Neurophysiol. 2002;113:
in a large clinical trial. 1373–1381.
21. Keir PJ, Rempel DM. Pathomechanics of peripheral nerve
REFERENCES loading. Evidence in carpal tunnel syndrome. J Hand Ther.
2005;18:259–269.
1. Katz RT. Carpal tunnel syndrome: a practical review. Am Fam 22. Kiylioglu N, Akyol A, Guney E, et al. Sympathetic skin
Physican. 1994;49:1371–1379. response in idiopathic and diabetic carpal tunnel syndrome.
2. Bland JD. Treatment of carpal tunnel syndrome. Muscle Clin Neurol Neurosurg. 2005;108:1–7.
Nerve. 2007;36:167–171. 23. Hui AC, Wong SM, Leung CH, et al. A randomized controlled
3. Jablecki CK, Andary MT, So YT, et al. Literature review of trial of surgery vs. steroid injection for carpal tunnel syndrome.
the usefulness of nerve conduction studies and electromyo- Neurology. 2005;64:2074–2078.
graphy for the evaluation of patients with carpal tunnel 24. Hui AC, Wong SM, Wong KS, et al. Oral steroids in the
syndrome. Muscle Nerve. 1993;16:1392–1414. treatment of carpal tunnel syndrome. Ann Rheum Dis. 2001;
4. Practice Parameter for Electro diagnostic studies in carpal 60:813–814.
tunnel syndrome (summary statement). American Academy of 25. Zijlstra FJ, van den Berg-de Lange I, Huygen FJ, et al. Anti-
Neurology, American Association of Electro Diagnostic inflammatory actions of acupuncture. Mediators Inflamm. 2003;
Medicine, and American Academy of Physical Medicine and 12:59–69.
Rehabilitation. Neurology. 1993;43:2404–2405. 26. Napadow V, Kettner N, Liu J, et al. Hypothalamus and
5. Jablecki CK, Andary MT, Floeter MK, et al. Practice amygdala response to acupuncture stimuli in carpal tunnel
parameter: Electro Diagnostic studies in carpal tunnel syn- syndrome. Pain. 2007;130:254–266.
drome. Report of the American Association of Electro 27. Zhang Enquin (editor in chief), Chinese Acupuncture and
Diagnostic Medicine, American Academy of Neurology, and Moxibustion, 1990 Publishing House of Shanghai College of
the American Academy of Physical Medicine and Rehabilita- Traditional Chinese Medicine.
tion. Neurology. 2002;58:1589–1592. 28. Tam LS, Leung PC, Li TK, et al. Acupuncture in the treatment
6. O’Connor D, Marshall S, Massy-Westropp N. Non-surgical of rheumatoid arthritis: a double-blind controlled pilot study.
treatment (other than steroid injection) for carpal tunnel BMC Complement Altern Med. 2007;7:35.
syndrome. Cochrane Database Syst Rev. 2003;CD003219. 29. Park J, White A, Stevinson C, et al. Validating a new non-
7. Chang MH, Chiang HT, Lee SSJ, et al. Oral drug of choice in penetrating sham acupuncture device: two randomized con-
carpal tunnel syndrome. Neurology. 1998;51:390–393. trolled trials. Acupunct Med. 2002;20:168–174.
8. Gerritsen AA, de Krom MC, Struijs MA. Conservative treatment 30. Tsukayama H, Yamashita H, Kimura T, et al. Factors that
options for carpal tunnel syndrome: a systematic review of influence the applicability of sham needle in acupuncture
randomized controlled trials. J Neurol. 2002;249:272–280. trials: two randomized, single-blind, crossover trials with
9. Delisa JA, Mackenzie K, Baran EM. Manual of Nerve acupuncture-experienced subjects. Clin J Pain. 2006;22:
Conduction Velocity and Somatosensory Evoked Potentials. 346–349.
2nd ed. New York: Raven Press; 1987. 31. Padua L, Padua R, Aprile I, et al. Multiperspective follow-up
10. Chang MH, Liao KK, Chang SP, et al. Proximal slowing in of untreated carpal tunnel syndrome: a multicenter study.
carpal tunnel syndrome resulting from either conduction block Neurology. 2001;56:1459–1466.
or retrograde degeneration. J Neurol. 1993;240:287–290. 32. Destefano F, Nordtrom DL, Vierkant RA. Long-term
11. Chang MH, Ger LP, Hsieh PF, et al. A randomized clinical symptom outcomes of carpal tunnel syndrome and its
trial of oral steroids in the treatment of carpal tunnel syn- treatment. J Hand Surg. 1997;22:200–210.
drome: a long-term follow up. J Neurol Neurosurg Psychiatry. 33. Muller M, Tsui D, Schnurr R, et al. Effectiveness of hand
2002;73:710–714. therapy interventions in primary management of carpal
12. Chang MH, Liu LH, Lee YC, et al. Comparison of sensitivity tunnel syndrome: a systematic review. J Hand Ther. 2004;17:
of transcarpal median motor conduction velocity and conven- 210–228.

r 2009 Lippincott Williams & Wilkins www.clinicalpain.com | 333