Carpal tunnel syndrome (CTS) is the most common nerve entrapment syndrome.1 2 Its prevalence in the general populations ranges between 1% and 5%.1^–3 Women are at greater risk for developing CTS than men.2 4 5 The age distribution of CTS is bimodal with a peak between ages 50 and 59 years and a second peak in those aged 70 years or older.4^–6

Epidemiological studies have shown the association of CTS with forceful repetitive work and vibration.7^–10 Most of the previous studies on the risk factors of CTS have been conducted among small and selected occupational populations. Few population-based studies have assessed the role of physical work load factors in CTS.11 12 Moreover, most of the previous studies have investigated the effects of forceful manual tasks combined with repetition on the risk of CTS. There is relatively little information regarding the separate and joint effects of physical work load factors on CTS. It is unclear whether, in the absence of forceful activities, exposure to work tasks demanding only repetitive movements of the hand or wrist, or work tasks requiring only the use of vibrating tools can cause CTS.

In this study we investigated the relations between exposures to a single or a combination of physical work load factors and the prevalence of CTS in a large representative sample of the general population.

MATERIALS AND METHODS

Population

In this national health examination survey, the Health 2000 Survey, the target population consisted of men and women aged 30 years or over residing in Finland between the fall of 2000 and the spring of 2001. The main emphasis of the Health 2000 Survey was to obtain up-to-date information in Finland on cardiovascular, respiratory, musculoskeletal and mental diseases and disability, their determinants and treatments. To obtain a representative sample of the whole Finnish population, a two-stage stratified cluster sampling design was used and sample stratified according to the five university hospital regions, each containing roughly one million inhabitants.13 From each university hospital region 16 healthcare districts were sampled as clusters (altogether, n = 80). The study plan was accepted at the local ethics committee and study subjects gave their written informed consent.

The original sample consisted of 8028 subjects aged 30 years or over. Of them 51 were deceased before interview, 6986 (87.6%) were interviewed and 6354 (79.7%) participated in the health examination. Subjects with missing information on CTS (n = 100) were excluded, leaving 6254 (78.4%) subjects eligible for the analysis.

Information was gathered by means of interview and clinical health examination at five field clinics. A home interview was used to gather information on background and sociodemographic characteristics, living habits, work and work ability. At the comprehensive health examination, specially trained nurses carried out an interview on physical activity, psychosocial factors, and musculoskeletal complaints, and physicians performed a standardised physical examination including the status of the upper extremities.

Outcome

The diagnosis of possible or probable CTS was based on (1) pain or paraesthesia or decreased sensitivity present in the thumb or index or middle finger on the Katz hand diagram, and (2) either positive Tinel’s test, combined wrist flexion and carpal compression test, decreased sensation in the median nerve distribution, or weakness of thumb abduction or wasting of the thenar eminence.14 A probable case of CTS was defined as classic or probable Katz hand diagram, that is, symptoms in two of the three radial fingers, and positive findings in at least two of the four clinical tests. Information was also gathered regarding surgery due to CTS.

Quality assessment

Two pilot studies were carried out 7 and 3 months before the field work started, in order to test and improve the methods. All staff members attended a 3-week training course. Quality assurance and quality control measures included training, written instructions, observation, video recording with feedback on examination technique, and repeated and parallel measurements. To study the repeatability of the physical examination of CTS, a sub-sample of 94 subjects underwent the standard clinical examination by two field physicians. The agreement between the two examiners was good for a positive clinical sign of CTS (kappa coefficient 0.68).

Determinants

Occupational physical factors

Physical work load factors were assessed by the home interview. All former jobs practised longer than 1 year were listed with their duration. The presence (frequency or duration per day) of the following physical exposures in the latest and each of the five most long-lasting former jobs were scrutinised. We aggregated the five most long-lasting former jobs into one group for the analysis.

• Working with hands above the shoulder level for at least 1 h.

• Manual handling of loads (manually lifting, carrying, pushing or pulling items) heavier than 5 kg at least two times per minute at the minimum of 2 h daily.

• Manual handling of loads heavier than 20 kg for at least 10 times.

• Working with a vibrating tool for at least 2 h.

• Work demanding high handgrip forces, for example, squeezing, twisting, holding burdens or tools corresponding with a grip force of 3 kg for at least 1 h.

• Repetitive movements of the hands or wrists (eg, packing and sorting out) for at least 2 h.

Other determinants

The home interview elicited information on age, gender, education and smoking history. As a part of the health examination, body weight and height were measured and body mass index was calculated. The subjects were defined as (1) current smokers, if they smoked cigarettes, cigars or a pipe at the time of interview; (2) former smokers if they had smoked at least for 1 year in the past and were not current smokers; (3) occasional smokers; and (4) never smokers.

We obtained information on rheumatoid arthritis, diabetes and hypothyroidism through clinical examination. Leisure time physical exercise causing mild breathlessness or sweating, was assessed by a single global question and classified into three groups: ⩽1, 2–3 and ⩾4 times per week.

The presence and severity of depressive symptoms was assessed by the modified version of the Beck Depression Inventory (BDI-II). The subjects with score 0–9 were classified as free from depression, those with score 10–17 as having mild to moderate depression, and subjects with score 18 or higher as having moderate to severe depression. The presence of somatisation was assessed with a 13-item somatisation part of Symptom Check List-90.15 We excluded three questions on pain and scored the remaining 10 items on a five-point Likert scale. The total score for each subject ranged from 0 to 40, with higher scores reflecting higher levels of somatisation.

Work-related psychosocial factors were measured only for those who had held a job during the preceding 12 months (61% of the population used in the main analyses). Job satisfaction and the threat of being bullied or mentally abused were assessed by a single global question with five levels. Social atmosphere of workplace was assessed by four questions and the score (1–5) for each of these questions was added up, and workplaces with score ⩽13 were classified as good to fair and those with score ⩾14 as poor. Social support and job strain were measured with the Karasek Job Content Questionnaire.16

Statistical methods

Statistical significance (two-tailed p<0.05) was assessed by a chi-squared test for categorical variables and by a two-sample t test for continuous variables. Survey logistic regression models were run to study the associations between physical work load factors and CTS. We used population weighting in estimating the prevalences, odds ratios, and confidence intervals to correct the age, gender, residential district, and language distributions of the study sample to correspond to those of the Finnish general population. Age (continuous), gender, body mass index (continuous), smoking (four groups), rheumatoid arthritis, diabetes, hypothyroidism, somatisation (continuous), and depressive symptoms (three groups), were used as covariates in the multivariable logistic regression models. Moreover, we investigated whether work-related psychosocial factors confounded the relationships between physical exposures and CTS. This additional analysis was performed among a subgroup of subjects who had held a job during the preceding 12 months (n = 3814). Handgrip with high forces was correlated with manual handling of loads heavier than 5 kg (Pearson’s correlation coefficient (r) = 0.49, p<0.001), manual handling of loads heavier than 20 kg (r = 0.50, p<0.001), work with hand above shoulder level (r = 0.43, p<0.001), repetitive movements of the hand or wrist (r = 0.38, p<0.001), and work with vibrating tools (r = 0.38, p<0.001). Therefore, separate and joint effects of physical work load factors on CTS were investigated by stratification and by logistic regression analysis. Multiple imputation method was used to predict the values of missing information on smoking (13.8% missing) or somatisation (7% missing). We used STATA, version 8.2, software for the analyses.

RESULTS

The subjects of the study group (n = 6254) were younger, had a higher level of education and were less often current smokers than those with no information on CTS (n = 184–1723). They had less often diabetes, rheumatoid arthritis and depressive symptoms and were less frequently exposed to physical work load factors than the subjects with no information on CTS. No differences were found regarding gender, body mass index and hypothyroidism.

Among the study group, women were on average 2.2 years older than men (table 1). Men more often were current smokers and held more frequently a job and had a higher body mass index than women. On the other hand, women were physically more active in their leisure time and reported more commonly somatic and depressive symptoms than men. There was no difference between genders regarding education.

Men more often than women reported work tasks demanding working with hands above the shoulder level, handling of loads heavier than 5 or 20 kg, handgrip with high forces and working with vibrating tools (table 1). The level of exposure to repetitive movements of the hand or wrist did not differ between genders.

The prevalence of possible or probable CTS was 2.1% in men and 5.3% in women (table 1). Probable CTS was also more common in women than in men. Of men, 0.7% and, of women, 1.8% had been operated because of CTS.

Table 2 shows age- and gender-adjusted odds ratios for the associations between CTS and exposure to physical load factors in the latest or five most long-lasting former jobs. Exposure to physical load factors only in the former jobs was associated with operated CTS only. Work tasks demanding working with hands above the shoulder level, repetitive movements of the hand or wrist, and handgrip with high forces in the former jobs were related to a higher rate of surgery due to CTS.

The prevalence of possible or probable CTS increased with increasing duration of exposure to work tasks requiring handgrip with high forces or vibrating tools (table 2). The prevalence of probable CTS increased with a longer exposure to work tasks demanding manual handling of loads heavier than 20 kg, repetitive movements of the hands, or vibrating tools. The risk of surgery due to CTS increased with increasing duration of exposure to manual handling of loads.

In analyses of both genders combined, work tasks requiring manual handling of loads heavier than 5 kg, handgrip with high forces, repetitive movements of the hand or wrist, or vibrating tools were significantly related to an increased prevalence of possible or probable CTS after adjustment for age, gender, body mass index, smoking and somatisation (table 3). With all work-related physical factors in the model and controlled for the above mentioned covariates, only vibration and handgrip with high forces showed an association with CTS. Further adjustment for leisure time physical activity, rheumatoid arthritis, diabetes, hypothyroidism, and depressive symptoms did not change the main results (odds ratio 1.9, 95% CI 1.2 to 2.9 for using vibrating tools and odds ratio 1.7, 95% CI 1.2 to 2.5 for handgrip with high forces).

Moreover, in an additional analysis in a sub-sample of actively working subjects, further adjustment for work-related psychosocial factors; job satisfaction, job strain and social support; threat of being bullied or mentally abused; and social climate at workplace did not affect the relations between physical work load factors and CTS (data not shown).

In the gender-specific analyses, CTS was statistically significantly associated with work tasks requiring vibrating tools in men and with work tasks demanding handgrip with high forces or repetitive movements of the hand or wrist in women (table 3). The relation of CTS with work tasks requiring handgrip with high forces in men (p = 0.057) and with vibration in women (p = 0.069) did not reach statistical significance.

In age-specific analyses, work requiring manual handling of loads was related to a higher prevalence of CTS in people aged 45 years or older and not in younger subjects (table 4). On the other hand, tasks demanding handgrip with high forces or vibrating tools were associated with an increased prevalence of CTS only among the working age population and not in those who had retired. Repetitive movements of the hand or wrist showed a bimodal relationship with CTS, the risk being higher in younger and older people and not in middle-aged subjects (45–64 years).

In stratified analyses controlled for other covariates, the association between handgrip with high forces and CTS was independent of working with hands above the shoulder level and handling of loads heavier than 5 or 20 kg (table 5). There were joint effects between handgrip with high forces and work with vibrating tools, between forceful activities (handgrip with high forces or handling of loads) and repetitive movements of the hand, and between repetitive movements of the wrist and work with vibrating tools. Work tasks demanding only vibrating tools or only repetitive movements of the hand or wrist were not related to CTS. Odds ratio for joint effect between repetitive movements of the hand and work with vibrating tools reduced to 2.0 (95% CI 1.1 to 3.6) after further adjustment for forceful manual work.

DISCUSSION

Our findings from a large representative sample of the general population showed that work tasks demanding handgrip with high forces or the use of vibrating tools are related to an increased prevalence of CTS. We found joint effects on CTS risk of exposure to handgrip with high forces and the use of vibrating tools, of handgrip with high forces and repetitive movements of the hand or wrist, and of using vibrating tools and repetitive movements of the hand or wrist. Regarding most exposures, only exposure in the most recent job was associated with CTS, suggesting that termination of exposure decreases the risk of CTS.

In line with other studies2 4 5 we found a higher prevalence of CTS in women than in men. There were no major gender differences in the associations between most physical load factors and CTS. The slight differences were mainly due to low power of our study especially for men. Moreover, this study had low power to detect the association between vibration and CTS in women because only 3.3% of women were exposed to vibration in their latest job. It seems that repetitive movements of the hand or wrist have a stronger effect on CTS in women than in men. The effects of vibration and forceful gripping are difficult to disentangle, since holding a vibrating tool always involves exertion of gripping force.

Consistent with previous population-based studies4 5 our population has shown a bimodal age distribution for CTS.6 Moreover, for manual handling of loads and repetitive movements of the hand the risk of CTS seemed to be elevated even after retirement. The mechanisms underlying these associations are not fully understood. Forceful or repetitive use of the hand or wrist could lead to increased interstitial fluid pressure within the carpal tunnel.17 18 Increased intracarpal canal pressure causes compression of the carpal tunnel contents, which may lead to poor blood circulation in the flexor synovial cells and median nerve. The ischaemia over a long period of time causes synovial thickening, and intraneural fibrosis and demyelination of the median nerve.19 Long-term forceful or repetitive hand activities may have long-lasting effects and may induce irreversible damages to the flexor synovial cells and median nerve.

Previous epidemiological studies have shown associations between CTS and forceful work, repetitive activities and vibration.7–10 20 In line with our findings a positive interaction between high force and highly repetitive work tasks for the risk of CTS has also been reported.21 We found that highly forceful work tasks alone may increase the risk of CTS but not highly repetitive work alone or using a vibrating tool alone.

To our knowledge, this is the largest population-based study on CTS using clinical examination with standardised criteria. The examining physicians were specially trained using feedback from video recordings. Repeatability tested between two examining physicians showed good agreement for a positive clinical sign of CTS.6 Other advantages of the current study are high participation rate and population level information on physical exposures obtained with face-to-face interview. The classification of physical exposures was based on current knowledge of the potentially harmful quantitative and qualitative features of physical work. We had information on both the latest and earlier job exposures. The large sample size enabled us to study the effects of single and combined exposures.

A limitation of this study is the cross-sectional nature, restricting the temporal relation to be confirmed between the exposure and the outcome. Outcome can affect the self-assessment of the exposures: for example, subjects with a significant health condition may recall their exposures more accurately than healthy subjects, due to increased attentiveness. Symptomatic subjects may also over-report their exposures.22 This differential exposure misclassification can lead to overestimation of risk estimates. We tried to reduce the likelihood of such an information bias by avoiding subjective wording in the exposure definition (eg, lifting “heavy” loads). Instead, objective dimensions were provided (eg, weight in kilograms). Also, the exposure assessment at the home interview took place some weeks before the clinical examination and the diagnosis of CTS. On the other hand, CTS is associated with major disability, and people with CTS may have changed jobs or retired, resulting in those with CTS having low exposure at current job. This could lead to underestimation of risk estimates.

Other limitation is the reliance on clinically confirmed CTS, in default of nerve conduction studies to confirm the diagnosis of CTS. Considering nerve conduction studies as a golden standard test, a combination of a classic or probable hand diagram and either a positive Tinel’s or Phalen’s test can correctly diagnose 79% of CTS cases.23 The probability even increases to 92% with a combination of classic or probable Katz hand diagram and positive Tinel’s and Phalen’s test. Our prevalence estimate (3.8%) was also close to those obtained in a Swedish population study based on symptoms and physical examination (3.8%), symptoms and electrodiagnostic measurement (4.9%), or symptoms, physical examination and electrodiagnostic measurement (2.7%).2 Moreover, in this latter study consistent associations emerged between physical work load factors and the three different CTS outcomes.

We adjusted the obtained odds ratios for most known risk factors of CTS; sociodemographic status, lifestyle factors, medical conditions, and psychological factors. The observed associations between physical work load factors and CTS are less likely due to confounding by these factors.

In summary, work tasks demanding handgrip with high forces or the use of vibrating tools are associated with CTS. The association is stronger if these work tasks are accompanied by repetitive movements of the hand or wrist.