| Sign In to gain access to subscriptions and/or personal tools. |
DOI: 10.1016/J.JHSB.2004.02.009
Risk Factors in Carpal Tunnel SyndromeFrom the Department of Trauma and Orthopaedics, Derbyshire Royal Infirmary, Derby, UK and the Clinical Epidemiology and Respiratory Medicine, Nottingham City Hospital, Nottingham, UK Correspondence: Mr John M. Geoghegan, 8 Cumbria Grange, Gamston, Nottingham, NG2 6LZ, UK. Tel.: +44-1332347141; fax: +44-1332254950; E-mail: john.geoghegan{at}ntlworld.com
We have undertaken a large case–control study using the UK General Practice Research Database to quantify the relative contributions of the common risk factors for carpal tunnel syndrome (CTS) in the community. Cases were patients with a diagnosis of CTS and, for each, four controls were individually matched by age, sex and general practice. Our dataset included 3,391 cases, of which 2,444 (72%) were women, with a mean age at diagnosis of 46 (range 16–96) years. Multivariate analysis showed that the risk factors associated with CTS were previous wrist fracture (OR = 2.29), rheumatoid arthritis (OR = 2.23), osteoarthritis of the wrist and carpus (OR = 1.89), obesity (OR = 2.06), diabetes (OR = 1.51), and the use of insulin (OR = 1.52), sulphonylureas (OR = 1.45), metformin (OR = 1.20) and thyroxine (OR = 1.36). Smoking, hormone replacement therapy, the combined oral contraceptive pill and oral corticosteroids were not associated with CTS. The results were similar when cases were restricted to those who had undergone carpal tunnel decompression.
Key Words: carpal tunnel syndrome • case control study • epidemiology • UK General Practice Research Database
Carpal Tunnel syndrome (CTS) is the most common compression neuropathy presenting in the UK, but relatively little is known about the contribution of the common risk factors in the community. Previous epidemiological studies have shown that a number of conditions are associated with CTS. These conditions can be classified as constitutional, hormonal and musculoskeletal factors. Constitutional risk factors include obesity (Ferry et al., 2000; Nathan et al., 2002; Nordstrom et al., 1997; Werner et al., 1994) and smoking (Nathan et al., 2002, 1996; Vessey et al., 1990). Hormonal risk factors include diabetes (Chammas et al., 1995; Ferry et al., 2000; Renard et al., 1994; Solomon et al., 1999; Stevens et al., 1992), hypothyroid-ism (Palumbo et al., 2000; Solomon et al., 1999), the use of the combined oral contraceptive pill (COCP) (Ferry et al., 2000; Sabour and Fadel, 1970), the use of hormonal replacement therapy (HRT) (Ferry et al., 2000; Solomon et al., 1999; Vessey et al., 1990), and corticosteroid use in the absence of inflammatory arthritis (Solomon et al., 1999). Musculoskeletal risk factors include rheumatoid arthritis (de Krom et al., 1990; Solomon et al., 1999; Stevens et al., 1992), osteoarthritis (Cooney and Choa, 1977; Ferry et al., 2000; Florack et al., 1992; Melone et al., 1987) and previous wrist fracture (Altissimi et al., 1986; Chapman et al., 1982; Solomon et al., 1999; Stewart et al., 1985). Previous studies have generally assessed small numbers of patients in specialist clinics (Altissimi et al., 1986; Chammas et al., 1995; Chapman et al., 1982; Dieck and Kelsey, 1985; Florack et al., 1992; Melone et al., 1987; Palumbo et al., 2000; Porter et al., 2002; Renard et al., 1994; Sabour and Fadel, 1970; Stewart et al., 1985; In the light of these limitations we have undertaken a large case–control study using the West Midlands General Practice Research Database, to quantify the relative contribution of these common risk factors for CTS in the community.
The UK General Practice Research Database (GPRD) represents the largest source of continuous data on illness and prescribing habits in general practice in the UK (Walley and Mantgani, 1997). The data represents the computerized recordings of routine general practice. The GPRD contains prescribing and diagnostic information collected as part of routine patient care for over 3 million patients, approximately 5% of the UK population, registered at 525 practices in the UK. For this study, we used the West Midland section of the GPRD, which represents approximately 10% of the total dataset, or approximately 383,000 patients.
Patients
Statistical analysis
Ethical approval
Our dataset included 3,391 cases of CTS and 13,564 matched controls. There were 2,444 (72%) women and 947 (28%) men. The mean age at diagnosis was 46 (range 16–96) years (standard deviation = 16). Sixty-six per cent of cases were aged between 30 and 59 years at diagnosis. The proportion of cases between the ages of 30 and 59 years was the same for women and the total case group, but this proportion was skewed towards an older age in the men (Fig 1). The mean annual general consultation rates were a strong predictor for having a diagnosis of CTS (Table 1). The greater the number of consultations that a patient had with their general practitioner, the greater the risk of CTS being diagnosed. Data were available to calculate the BMI in 61% of cases and 52% of controls. An increased BMI was found to be associated with CTS, such that a BMI of 25.1 or more gave an OR = 1.63 (95% CI: 1.45–1.84) (Table 1). The greater the BMI, the greater the risk of CTS (Table 1). The association between BMI and CTS was reduced, but there was no significant change, when adjusted for consultation rates (Table 1). There was evidence of a statistically significant interaction between BMI and age group, such that an increased BMI showed a stronger association with CTS in younger ages groups (BMI >30 + age < 30: OR = 4.81 (95% CI: 2.85–8.11)), and this trend decreased with increasing age (BMI >30 + age 70–80: OR = 1.64 (95% CI: 0.99–2.72)) (Fig 2). Smoking data was available on 56% of cases and 52% of controls. Smoking was not found to be associated with CTS (Table 1). A diagnosis of diabetes was significantly associated with CTS (Table 1). The association was reduced, but still present after adjusting for consultation rates (OR = 1.51, 95% CI: 1.24–1.84) (Table 1). In further multivariate analyses adjusting for BMI reduced the association with diabetes, but a positive association with CTS remained (OR = 1.99 95% CI: 1.22–3.11). The use of insulin and oral hypoglycaemics grouped as metformin and sulphonylureas were also associated with CTS. This association was reduced when adjusted for consultation rates. There was also evidence that thyroxine use was associated with CTS even after adjusting for consultation rates. We found no evidence of an association with the use of the oral contraceptive pill, the use of hormone replacement therapy or corticosteroid use for conditions other than inflammatory arthritis, after adjusting for consultation rates and CTS (Table 1). Rheumatoid arthritis, previous wrist fracture and osteoarthritis were all significantly associated with CTS (Table 1). Consultation rates slightly reduced the association for musculoskeletal conditions studied, however all factors still remained strongly associated with CTS: rheumatoid arthritis, OR = 2.23 (95% CI: 1.57–3.17): previous wrist fracture, OR = 2.29 (95% CI: 1.67–3.12): osteoarthritis, OR = 1.89 (95% CI: 1.65–2.17). Multi-variant analyses showed a positive interaction between wrist fracture and age revealing a bimodal pattern (Fig 3), such that fracture was important in the elderly with a small increase in the young. Rheumatoid arthritis and male gender showed the strongest association with CTS in this study: OR = 6.8 (95% CI: 3.11–14.85). Rheumatoid arthritis and female gender were also associated with CTS, OR = 2.74 (95% CI: 1.85–4.05).
The analysis was repeated limiting it to those patients with CTS who had undergone carpal tunnel decompression (CTD), adjusting for mean annual consultation rates (Table 2). There were 883 CTD (26% of all the CTS patients) with 72% being in women. The mean age of diagnosis was 54 years. The results in this group were very similar to those of the 3,391 CTS cases for each risk factor analysed (Table 2). However, there was a stronger association with a BMI >30.1 (moderately obese) and previous wrist fracture.
We have investigated a number of risk factors commonly believed to be important in the aetiology of CTS in the largest case–control study of CTS reported to date. The musculoskeletal conditions of rheumatoid arthritis, wrist fracture and osteoarthritis, and an increased BMI were all strong risk factors for CTS. Diabetes and hypothyroid treatment were also associated with CTS. The effects of these risk factors were independent of consulting behaviour. The combined oral contraceptive, hormone replacement therapy, steroid and smoking were not associated with CTS. The data in our study is derived from computerized general practice records used for routine clinical care. This has the advantage of reflecting real life experience rather than the research setting, and being a prospective continuous data collection which avoids recall bias. One disadvantage however, is that the general practitioner may have made the diagnosis of CTS without referral to hospital and without supporting evidence from neurophysiological investigations. Bias may also have occurred if our cases included patients without CTS but with other problems that were related to the various risk factors studied. General practitioners may have introduced bias if they were more likely to have diagnosed CTS because of prior knowledge of an earlier condition, but this would seem unlikely due to the poor understanding of the aetiology of CTS. Specific diagnostic criteria were not supplied for any of the illnesses reported by the participating doctors. However, data from practices is routinely validated by internal checks, and there are also specific audits of data supplied by individual practices. Only data meeting the minimum standards are added to the research database (Walley and Mantgani, 1997). It is likely that the diagnosis of CTS has good specificity but low sensitivity, which is important in a case–control study such as this. We repeated our analysis limiting it to those CTS cases that had undergone carpal tunnel decompression, in order to increase the specificity of the CTS diagnosis, and to check the validity of the diagnosis of CTS in the population studied. The results were very similar to the overall CTS group, illustrating that the results in the general population studied are robust. In the carpal tunnel decompression subgroup, it was noted that there was a stronger association with a BMI >30.1 (moderately obese), and previous wrist fracture. In our study the mean age at diagnosis of CTS was 46 years, the incidence peaked between 40 and 49 years, and our case sex mix was 3:1 women to men. This compares with previous studies (Ferry et al., 2000; McDiarmid et al., 2000; Nordstrom et al., 1998). Current smoking was not associated with CTS in our study, which is in agreement with two previous studies (Dieck and Kelsey, 1985; Ferry et al., 2000), but disagrees with others (Nathan et al., 2002, 1996; Vessey et al., 1990). We have shown that the greater the BMI, the greater the risk of CTS, and that obesity is a particularly important risk factor in people less than 30 years of age. Previous studies have assessed the association of obesity with CTS. Ferry et al. (2000) found an increased risk of CTS with obesity (adjusted odds ratio of 1.68), but gave no definition of obesity. Werner et al. (1994) showed that individuals classified as obese (BMI >29) were 2.5 times more likely than slender individuals (BMI <20) to be diagnosed with CTS. Nordstrom et al. (1997) were able to demonstrate that a unit increase in BMI increased the risk of CTS by 6%. Given the emerging epidemic of obesity, it will become an increasingly important risk factor for CTS, the age of presentation of CTS cases will decrease, and there will be a total increase in CTS case numbers. The prevalence of soft tissue hand lesions including CTS is higher in diabetic populations (Chammas et al.,1995; Ferry et al.,2000; Renard et al., 1994; Solomon et al.,1999; Stevens et al.,1992), particularly in type I diabetes mellitus (insulin dependant) (Renard et al.,1994), than in controls. Our study shows that diabetes is associated with a moderately increased risk of CTS and this effect was only slightly decreased when adjusted for BMI (OR = 1.86 (95% CI: 1.53–2.27)). Significantly, our study shows little difference between insulin and non-insulin dependant diabetes and their association with CTS. Diabetes causes a peripheral neuropathy through abnormal glycosylation of protein end products, but it is unlikely that this is the primary cause of CTS in diabetics (Solomon et al., 1999). The combined oral contraceptive pill (COCP) (Sabour and Fadel, 1970; Solomon et al., 1999) and hormone replacement therapy (HRT) (Ferry et al., 2000; Sabour and Fadel, 1970; Solomon et al., 1999) have been associated with CTS in previous studies. The hypothesis is that the exogenous oestrogens exert their effect through fluid retention, causing pressure on the median nerve. Our study shows no evidence of an association with the use of the COCP and CTS, which is consistent with the findings of Ferry et al. (2000). We also found no evidence of an association with HRT after adjusting for consultation rates and note that our study included the largest number of CTS cases on HRT. Corticosteroid use in the absence in inflammatory arthritis has been associated with a greater likelihood of undergoing carpal tunnel release (Solomon et al., 1999). After adjusting for consultation rates we found no evidence of such an association with CTS (OR = 1.09). Hypothyroidism has been associated with CTS (Ferry et al., 2000; Palumbo et al., 2000; Solomon et al., 1999), with symptoms persisting in the majority of patients after the hypothyroid state has been corrected (Palumbo et al., 2000). It is most likely that there is a mild compression within the carpal tunnel of a physiologically altered peripheral nerve resulting in the symptoms of CTS. We found the use of thyroxine was weakly associated with CTS, whereas previous small studies found a stronger association (Ferry et al., 2000; Solomon et al., 1999). It is broadly agreed that previous wrist fracture is associated with CTS (Altissimi et al.,1986; Chapman et al.,1982; Stewart et al.,1985), with a prevalence of 10% to 12.5% after distal radial fracture (Altissimi et al., 1986; Chapman et al.,1982; Stewart et al.,1985). Only two epidemiological studies with small numbers of cases have studied "arm fractures" (Ferry et al., 2000) and previous wrist fractures (de Krom et al., 1990) and found an association with CTS. Our study shows a strong association between previous wrist fracture and CTS with a positive interaction between wrist fracture and age with a bimodal pattern of distribution. Thus fracture is important in the elderly with a small peak in the young, which may be explained by the frequency of high-energy trauma in the younger age group (30–40 years), and increased falls and osteoporosis in the elderly (60 years +). We found rheumatoid arthritis to be the second highest risk factor for CTS in our study, second only to previous wrist fracture. Osteoarthritis (OA) has also been associated with CTS, whether this is OA of the spine or more generalized OA (Ferry et al., 2000). More specifically, trapeziometacarpal osteoarthritis or basal joint osteoarthritis has been positively associated with CTS (Florack et al., 1992; Melone et al., 1987). We found that osteoarthritis of the wrist and carpus was the fourth greatest risk factor for CTS in our study. Our study represents the largest community-based prospective case–control study of carpal tunnel syndrome, to date. The musculoskeletal conditions of previous wrist fracture, rheumatoid arthritis, and osteoarthritis of the wrist and carpus, an increase in BMI, diabetes and hypothyroidism are important risk factors for CTS. There are concerns of a developing epidemic of obesity amongst our younger generations, and thus obesity will probably become an increasingly important risk factor for CTS, and the age of presentation of CTS will decrease. This increased co-morbidity in the community may have far reaching economic and public health issues.
Although none of the authors has received or will receive benefits for personal orprofessional use from a commercial party related directly or indirectly to the subject of this article, benefits have been or will be received but will be directed solely to a research fund, foundation, educational institution, or other non-profit organization with which one or more of the authors are associated. Richard Hubbard is a Wellcome Trust Advanced Fellow. Received for publication September 9, 2003. Accepted for publication February 2, 2004.
  CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
 |
Sign In to gain access to subscriptions and/or personal tools.
TOP
Abstract
INTRODUCTION
