Subjects
This cohort-based study was approved by our institutional review board, and informed consent was obtained from all participants. This study was conducted as a substudy of the Korean Longitudinal Study on Health and Aging (KLoSHA).
The KLoSHA was a cohort study of health, aging, and common geriatric diseases in the elderly Korean population. The data, collected from September 2005 to August 2006 from residents of Seongnam City, South Korea, were also used later for this cross-sectional study. Candidates were drawn randomly from a list of all individuals who were more than sixty-five years of age; a computer-generated list of resident registration numbers was used for this purpose. A total of 1118 residents were invited to participate in this study by letter and telephone, and 696 agreed (a response rate of 62.3%). Demographic data were archived, and posteroanterior hand radiographs were made of all participants. All radiographic images were acquired digitally with use of a picture archiving and communication system (PACS) (IMPAX; Agfa HealthCare, Mortsel, Belgium), which was used to carry out assessments. A specially trained nurse surveyed the patients regarding symptoms of pain, numbness, or a tingling sensation, including the laterality of the symptoms and their locations in the hands. Nerve conduction studies were performed on 434 of the 696 participants; this group of 434 patients had been chosen randomly with the use of a computer-generated list. Upper-extremity nerve conduction studies were performed on the dominant side only by a rehabilitation physician. The participant’s skin temperature was maintained above 31.0°C13. Sensory nerve conduction studies were performed with use of antidromic methods, with stimulation at the wrist and recording with use of ring electrodes from the index finger (electrode-to-wrist distance, 14 cm). Motor nerve conduction studies were performed by recording with surface electrodes over the abductor pollicis brevis muscle following stimulation of the median nerve at the wrist 8 cm proximal to the recording site, and at the antecubital fossa.
Fifty-eight of the 434 individuals were excluded because of a history of trauma to the hand or wrist or of surgery in the upper extremity, and another eight participants with rheumatoid arthritis were excluded as well. We analyzed the data of 368 participants with hand radiographs and nerve conduction study data. The prevalence of basal joint arthritis of the thumb as seen on radiographs was determined, and prevalence of carpal tunnel syndrome and the correlation between basal joint arthritis and carpal tunnel syndrome were investigated on the dominant side. The mean age (and standard deviation) of the participants was 75 ± 7 years (range, sixty-six to ninety-four years); 192 (52.2%) were men and 176 were women. After data collection, the subjects were divided into four age groups (sixty-six to seventy years, seventy-one to seventy-five years, seventy-six to eighty years, and eighty-one years or older) according to a previous study in which the authors analyzed the prevalence of basal joint arthritis with a five-year spread14. Subgroup analyses were then performed by age and sex.
Diagnosis of Basal Joint Arthritis of the Thumb on Radiographs
Osteoarthritis of the basal joint was graded from 0 to 4, with use of the Kellgren and Lawrence criteria15, on the basis of hand radiographs. According to this scale, grade 0 indicates no osteoarthritis; grade 1, minimal osteophytes and possible cyst formation (doubtful osteoarthritis); grade 2, definite osteophytes and possible cysts (minimal osteoarthritis); grade 3, moderate osteophytes, narrowing of the joint space, subchondral sclerosis, and deformity of the bone ends (moderate osteoarthritis); and grade 4, large osteophytes, severe sclerosis, and narrowing of the joint space (severe osteoarthritis). Unlike the Eaton classification, the Kellgren and Lawrence system uses hand radiographs rather than dedicated thumb views16-21. In the present study, grades 2, 3, and 4 were considered to indicate the presence of basal joint osteoarthritis according to previously published reports regarding the prevalence of hand osteoarthritis16-19.
Hand radiographs were evaluated by two orthopaedic hand surgeons blinded to the study subjects’ information, including age, sex, symptoms, and nerve conduction study results. Intraobserver reliability was evaluated by asking an examiner to repeat all radiographic assessments after three weeks. Interobserver reliability was evaluated by having two examiners each perform all radiographic assessments independently. The intraobserver and interobserver reliabilities of the radiographic grading scale assessments were tested with use of Cohen kappa coefficients. The kappa coefficient was 0.897 (p = 0.028) for intraobserver reliability and 0.826 for interobserver reliability (p = 0.035). Given the high interobserver correlation, the Kellgren and Lawrence grades assigned by only one examiner were used for the data analyses.
The diagnosis of basal joint arthritis of the thumb was based on radiographs only; there was no specific history-taking for basal joint arthritis or tests such as a grinding test performed at the time of data collection.
Diagnosis of Carpal Tunnel Syndrome
A positive result for carpal tunnel syndrome on nerve conduction studies was defined as a median motor nerve distal onset latency of >4.0 ms or a median sensory nerve distal onset latency of >3.2 ms. Only participants with a positive result on nerve conduction studies and who reported experiencing pain, numbness, and/or a tingling sensation in the dominant hand were diagnosed as having carpal tunnel syndrome. Those with only ulnar-side symptoms and those who had a discrepancy between the symptoms and the results of the nerve conduction studies were deemed to have a negative finding for carpal tunnel syndrome. A prolongation of motor distal latency and a decrease of motor conduction velocity were both considered to correspond to the electrophysiological severity of carpal tunnel syndrome22-26.
Two participants had a previous diagnosis of carpal tunnel syndrome but did not have surgical treatment and thus were included in the study. No participant had surgery for carpal tunnel syndrome or basal joint arthritis. The history of treatments other than surgery was not available.
Statistical Analysis
A sample size of seventy-one participants was required to detect a Spearman correlation of greater than r = 0.3 between the radiographic grade of basal joint arthritis and the electrophysiological severity of carpal tunnel syndrome with a power of 80% (meaning that the chance that a true correlation would be overlooked is <20%) at a level of significance of p < 0.05.
Chi-square analysis was performed to compare the prevalence of carpal tunnel syndrome between those with and those without basal joint arthritis of the thumb. The Fisher exact test was used to analyze subgroups with sample sizes of less than five. Spearman rho analysis was performed to evaluate the relationship between the electrophysiological severity of carpal tunnel syndrome and the radiographic severity of basal joint arthritis of the thumb. Analysis was performed to determine the correlation of age with the prevalence of carpal tunnel syndrome or basal joint arthritis. P values of <0.05 were considered significant.
Source of Funding
The KLoSHA was supported by grants from the Seongnam city government in South Korea (Grant No. 800-20050211) and Pfizer Global Pharmaceuticals (Grant No. 06-05-039), which were used for laboratory fees and salaries. The funding sources did not play a role in the performance of the investigations for this particular substudy of KLoSHA, and none of the authors received financial support.
Prevalence of Basal Joint Arthritis of the Thumb
Radiographic evaluation for arthritis of the basal joint, according to the Kellgren and Lawrence criteria, demonstrated grade 0 in 30.2% (111) of the 368 participants, grade 1 in 56.8% (209), grade 2 in 12.0% (forty-four), grade 3 in 0.8% (three), and grade 4 in 0.3% (one). Overall, basal joint arthritis of the thumb (a Kellgren and Lawrence grade of ≥2) was diagnosed in 13.0% (forty-eight) of the 368 patients in the entire series, 15.1% (twenty-nine) of the 192 male subjects, and 10.8% (nineteen) of the 176 female subjects. There was no significant difference according to sex (p = 0.220). The correlation between age and the prevalence of basal joint arthritis was significant (p < 0.01): it was diagnosed in 7.5% (nine) of the 120 individuals between sixty-six and seventy years old, 9.6% (eleven) of the 114 individuals between seventy-one and seventy-five years old, 9.8% (five) of the fifty-one individuals between seventy-six and eighty years old, and 27.7% (twenty-three) of the eighty-three individuals who were eighty-one years old or older.
Prevalence of Carpal Tunnel Syndrome
The overall prevalence of carpal tunnel syndrome was 11.7% (forty-three of 368), and the prevalence was significantly higher in women (18%, thirty-one of 176) than in men (6%, twelve of 192) (p < 0.01). The prevalence of carpal tunnel syndrome was 12.5% (fifteen of 120) in the individuals between sixty-six and seventy years old, 7.9% (nine of 114) in the individuals between seventy-one and seventy-five years old, 15.7% (eight of fifty-one) in the individuals between seventy-six and eighty years old, and 13.3% (eleven of eighty-three) in the individuals who were eighty years old or more. No significant correlation was found between age and the prevalence of carpal tunnel syndrome (p = 0.697).
The prevalence of carpal tunnel syndrome was 16.7% (eight of forty-eight) in the group with basal joint arthritis (a Kellgren and Lawrence grade of ≥2) and 10.9% (thirty-five of 320) in the group without basal joint arthritis; these prevalences were not significantly different (p = 0.249). If basal joint arthritis was defined as a Kellgren and Lawrence grade of ≥1, the prevalence of carpal tunnel syndrome was 12.5% (thirty-two of 257) in the participants with arthritis and 9.9% (eleven of 111) in those without arthritis; again, these prevalences were not significantly different (p = 0.485).
Table I presents the prevalences of carpal tunnel syndrome according to sex and age in the groups with basal joint arthritis (a Kellgren and Lawrence grade of ≥2) and without basal joint arthritis. No significant differences were observed between these subgroups.
A total of 198 participants had positive nerve conduction studies, and forty-three of them had correlating symptoms and thus were considered to have true carpal tunnel syndrome. The remaining 155 participants with positive nerve conduction studies but no symptoms were considered not to have carpal tunnel syndrome. Thirty-six additional participants reported symptoms of carpal tunnel syndrome but had normal results on nerve conduction studies; these participants were also considered not to have carpal tunnel syndrome.
Correlation Between Electrophysiological Severity of Carpal Tunnel Syndrome and Radiographic Grade of Basal Joint Arthritis
In the entire group of study subjects, the mean motor distal latency (and standard deviation), sensory distal latency, and motor conduction velocity were 3.68 ± 0.63 ms, 3.84 ± 2.1 ms, and 56.5 ± 25.9 m/s, respectively; motor distal latency and motor conduction velocity were not correlated with the Kellgren and Lawrence grade of the basal joint arthritis (p = 0.154 for motor distal latency and 0.662 for motor conduction velocity).
In the group of 198 subjects with positive nerve conduction studies, the mean motor distal latency, sensory distal latency, and motor conduction velocity were 3.98 ± 0.68 ms, 4.72 ± 2.56 ms, and 54.23 ± 6.08 m/s, respectively; again, motor distal latency and motor conduction velocity were not correlated with the Kellgren and Lawrence grade (p = 0.247 for motor distal latency and p = 0.841 for motor conduction velocity).
In the group of forty-three subjects diagnosed as having true carpal tunnel syndrome, the mean motor distal latency, sensory distal latency, and motor conduction velocity were 4.13 ± 0.73 ms, 4.88 ± 2.45 ms, and 55.42 ± 8.14 m/s, respectively. Correlations with the Kellgren and Lawrence grade were also insignificant in these forty-three patients with carpal tunnel syndrome (p = 0.603 for motor distal latency and p = 0.998 for motor conduction velocity).
In addition, no significant correlation was found between motor distal latency or motor conduction velocity and the Kellgren and Lawrence grade in any of the age or sex subgroups (Table II).
This study indicates that carpal tunnel syndrome is not associated with basal joint arthritis of the thumb found on radiographs of elderly Koreans (more than sixty-five years old). Furthermore, no correlation was found between the electrophysiological severity of carpal tunnel syndrome and the radiographic severity of basal joint arthritis of the thumb. These findings suggest that the prevalence of carpal tunnel syndrome is similar between those with and those without radiographic evidence of basal joint arthritis of the thumb. However, this was a study of a general population and not of patients seeking medical care for basal joint arthritis; thus, our results do not necessarily contradict previous findings that carpal tunnel syndrome is common in patients who receive surgical treatment for basal joint arthritis8-12.
In the present study, we used the Kellgren and Lawrence criteria to assess the prevalence of thumb basal joint arthritis on hand radiographs, as has been done in previous studies evaluating hand osteoarthritis16-19, although we recognize that some investigators have used the Eaton classification instead5,12. On the basis of the Kellgren and Lawrence criteria, the radiographic prevalence of basal joint arthritis of the thumb was 13.0% in all study subjects (more than sixty-five years old), 15.1% in men, and 10.8% in women, rates that were comparable with the results of other Asian studies16,18,19 but lower than those reported for white populations evaluated with the same radiographic criteria17,20. Kellgren and Lawrence reported a prevalence of radiographically evident basal joint arthritis of 19.1% in men and 33.4% in women in a population of English individuals ranging in age from fifty-five to sixty-four years20. In a study of Finnish adults, the prevalence was >30% for women aged sixty-five to seventy-four years and >35% for women aged seventy-five years or more17. However, in the Hizen-Oshima Study, 10.2% of 551 Japanese women who were forty years old or more had radiographic evidence of basal joint arthritis18, and Zhang et al. reported a prevalence of 19.4% for men and 14.5% for women aged sixty years or older in a Chinese population19. These differences may be explained in part by genetic and environmental factors. Several investigators have hypothesized that genetic factors account for approximately 65% of cases of osteoarthritis seen on hand radiographs16,19,27,28. In addition to genetic factors, differences between the lifestyles of Asian and white individuals may contribute to observed disparities in the prevalence of osteoarthritis. For example, chopsticks require palmar thumb force, whereas forks and knives require lateral thumb force16,21,29. Palmar force results in mechanical stress to the interphalangeal joint of the thumb, while lateral force stresses the basal joint of the thumb16,29. This has led to speculation that chopstick use is linked to a low prevalence of basal joint arthritis of the thumb in Asian populations16,18,19. Furthermore, unlike what has been observed in white populations, Korean women have a lower prevalence of basal joint arthritis than Korean men, which concurs with what was found in the Chinese study by Zhang et al.19. In that study, it was speculated that the genetic predisposition for development of hand osteoarthritis is lower in Chinese people, and that the prevalence of osteoarthritis in these joints among Chinese people is more likely to be determined by use-related factors19.
In the present study, the prevalence of carpal tunnel syndrome in all study subjects was 11.7%. Ferry et al. estimated a prevalence of between 7% and 16% in a study from the U.K., which is consistent with our findings30. The subjects enrolled in that U.K. study represented a random sample obtained from a group of individuals between eighteen and seventy-five years of age who were registered with a general practice serving a suburban population in Manchester, England. In addition, in the study by Ferry et al. carpal tunnel syndrome was diagnosed with a hand pain questionnaire and nerve conduction test results with use of electrophysiological criteria similar to ours, not by clinical examination.
In our study, 198 of 368 participants had positive nerve conduction studies and only forty-three of them had correlating symptoms. We speculate that the high frequency of positive nerve conduction studies in asymptomatic participants may have resulted from the inherent characteristics of the nerve conduction study criteria—i.e., the absolute sensory or motor distal latency cutoff values used to diagnose carpal tunnel syndrome. Ferry et al. estimated the proportion of the population with delayed median nerve conduction who would be expected to have hand symptoms30. Of those determined to have delayed conduction on the basis of a combined sensory distal latency cutoff of >3.7 ms or motor distal latency cutoff of >4.5 ms, thresholds that were higher than ours, 67% had “unlikely” or no symptoms. In contrast, Atroshi et al. used a technique that included measurements of median nerve distal sensory latency (long finger-wrist) and wrist-palm sensory conduction velocity as well as ulnar nerve distal sensory latency (small finger-wrist)31. In their study, the criterion used for the diagnosis of median neuropathy was the median-ulnar sensory latency difference in each individual, with 0.8 ms or longer considered abnormal. In their study, the false-positive rate for nerve conduction studies was 18%31.
Florack et al. reported that 39% (ninety-five) of 246 patients who underwent basal joint arthroplasty of the thumb had carpal tunnel syndrome9, a rate that is considerably higher than the prevalence of carpal tunnel syndrome in general populations. They noted that the prevalence in the literature was similar to theirs9. However, almost all individuals with both conditions were patients who underwent surgical treatment, and Florack et al. did not have a control group. These authors postulated that the association between carpal tunnel syndrome and thumb basal joint arthritis might result from alterations in the osseous anatomy of the carpal canal caused by degenerative arthritis or a midcarpal collapse pattern, and they also suggested that inflammation in the thumb basal joint might cause median nerve compression. However, they acknowledged that inflammatory changes were not identified in prior histological studies of the flexor tenosynovium in patients with carpal tunnel syndrome. We believe that the absence of a correlation between the electrophysiological severity of carpal tunnel syndrome and the radiographic grade of basal joint arthritis indicates that these two conditions are not associated in the general population. We suspect that those with both conditions are more likely to seek medical attention as a result of greater symptom severity, and that this led to the higher prevalence of carpal tunnel syndrome among those undergoing surgery for basal joint arthritis in previous studies9,30-32. In addition, there can be an underlying difference between patients who seek medical treatment for pain and dysfunction of the basal joint and those with asymptomatic basal joint arthritis.
Several limitations of the present study should be noted. First, the study was of elderly Koreans; thus, our findings may not apply to other ethnicities. For example, although the prevalence of electrophysiologically confirmed carpal tunnel syndrome in the present study was similar to that reported in Western studies, radiographically evident basal joint arthritis of the thumb was less common in the present study than has been reported in studies of white individuals17,20. Therefore, there can be disparate etiologies of basal joint arthritis, which may limit the generalizability of our findings. Second, while subjective symptoms and electrophysiological tests were used for the diagnosis of carpal tunnel syndrome, clinical examinations such as hand diagrams and provocation tests such as the Tinel and Phalen tests, which may have been useful to improve diagnostic accuracy, were not performed. Furthermore, false-positive and negative rates of electrophysiological tests may have affected the observed prevalence of carpal tunnel syndrome. Third, although the sample was large enough for analysis of the correlation between electrophysiological severity of carpal tunnel syndrome and the severity of basal joint arthritis seen on radiographs, chi-square analysis for comparison of the prevalence of carpal tunnel syndrome between the groups with and without basal joint arthritis was underpowered. If the prevalence of carpal tunnel syndrome in the group without basal joint arthritis was 10%, and we wanted to set 15% (a 5% increase and an effect size of 0.5) as a meaningful difference in carpal tunnel syndrome prevalence in the group with basal joint arthritis group, with a power of 80%, then an adequate sample size would be 686, with 343 in each group. Fourth, nerve conduction studies were performed on the dominant side only; thus, the prevalence of carpal tunnel syndrome and the correlation between basal joint arthritis and carpal tunnel syndrome were determined for dominant hands. This should have underestimated the prevalence of carpal tunnel syndrome in the general population, although the correlation between basal joint arthritis and carpal tunnel syndrome would not be affected. Fifth, we used motor distal latency and motor conduction velocity to assess the electrophysiological severity of carpal tunnel syndrome, as motor distal latency and motor conduction velocity were shown to be correlated with carpal tunnel syndrome severity in some studies22-26. However, the authors of several studies have concluded that symptoms are not correlated with the electrophysiological grades of carpal tunnel syndrome26,33,34. Finally, there may not be a direct correlation between the radiographic grade of basal joint arthritis and symptom severity14,35.
In summary, the prevalence of carpal tunnel syndrome was similar in elderly Koreans (more than sixty-five years old) with and without basal joint arthritis of the thumb. Furthermore, no correlation was found between the electrophysiological severity of carpal tunnel syndrome and the severity of basal joint arthritis of the thumb. These results suggest that there is not a statistical association between carpal tunnel syndrome and basal joint arthritis of the thumb in the general population, although limitations in our study methodology may preclude a comprehensive analysis. Additional studies are warranted to explore the association between carpal tunnel syndrome and basal joint arthritis of the thumb in different ethnic populations or in patients seeking medical care.