- Email: [email protected]
Waist circumference and waist-to-hip ratio in carpal tunnel syndrome: A case–control study
Journal of the Neurological Sciences 338 (2014) 207–213
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Waist circumference and waist-to-hip ratio in carpal tunnel syndrome: A case–control study Mauro Mondelli a,⁎, Alessandro Aretini a, Federica Ginanneschi b, Giuseppe Greco c, Stefano Mattioli d a
EMG Service, Local Health Unit 7, Siena, Italy Dpt. Neurological, Neurosurgical and Behavioural Sciences, University of Siena, Italy EMG Service, Local Health Unit 7, “Nottola” Hospital, Montepulciano, Siena, Italy d Dpt. Medical and Surgical Sciences, University of Bologna, Italy b c
a r t i c l e
i n f o
Article history: Received 10 November 2013 Received in revised form 5 January 2014 Accepted 8 January 2014 Available online 15 January 2014 Keywords: Body mass index Carpal tunnel syndrome Hand ratio Nerve entrapment syndrome Obesity Waist circumference Waist-to-hip ratio Wrist ratio
a b s t r a c t Background: The association between carpal tunnel syndrome (CTS) and high body mass index (BMI) and some hand measures is well known. No study has been specifically focused on waist circumference (WC) and waist-tohip-ratio (WHR). The aim of this prospective case–control study is to evaluate the association between CTS and WC, WHR and other body and hand anthropometric measures. Methods: We consecutively enrolled one “idiopathic” CTS case for two controls in 3 outpatient electromyography labs. The main anthropometric measures were BMI, WC, WHR, wrist ratio (WR) and hand ratio (HR). We performed univariate and multivariate analyses. Results: Female cases and controls were 250 and 474 and male cases and controls were 120 and 273, respectively. At univariate analysis there were differences in many anthropometric measures between cases and controls. At multivariate logistic regression analyses high BMI, WC and WHR and abnormal HR and WR were independent risk factors for CTS. Crossing two categories between BMI, WC and WHR, the overweight subjects, especially females, were at risk only if they had very high WC or high WHR. The risk increased if they were obese. Conclusions: High WC/WHR doubles the risk of CTS, the risk further increased if overweight/obese subjects have also very high WC or high WHR. The obese subjects were always at risk regardless of WC and WHR values. Metabolic causes of this association with CTS were hypothesised. BMI is not the only and most powerful body predictor of “idiopathic” CTS, but also WHR and WC should be considered. These measures may not be interchangeable and it may be desirable to consider the utility of their joint use. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Carpal tunnel syndrome (CTS) is the most frequent focal peripheral neuropathy. Many risk factors for CTS were identified as female gender, age, diabetes, rheumatoid arthritis, thyroid dysfunction, renal failure, pregnancy, hand and wrist trauma, use of oral contraception, smoking, and occupations that involved forced and repetitive exertions of hand and wrist, high handgrip forces and using vibrating tools [1–7]. Many studies have focused on hand/wrist and body anthropometric characteristics. There is almost unanimous agreement that a high body mass index (BMI), a squared wrist and a short hand predispose to CTS [4,5,8–13]. Recently the World Health Organization (WHO) suggested that waist circumference (WC) and waist-to-hip-ratio (WHR) are better predictors than BMI in some diseases [14]. In literature there are no studies on WC and WHR expressly designed to demonstrate whether high WC and high WHR may be independent risk factors for CTS.
⁎ Corresponding author at: Servizio di EMG, ASL 7, Via Pian d'Ovile, 9, 53100 Siena, Italy. Tel.: +39 0577 535904; fax: +39 0577 535971. E-mail address: [email protected] (M. Mondelli). 0022-510X/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jns.2014.01.012
The purpose of our prospective case–control study is to evaluate the association between CTS and WC, WHR and other anthropometric measures. 2. Methods 2.1. Enrolment and definition of cases and controls We decided in advance to recruit one case with CTS for two controls. We prospectively enrolled cases among all consecutive patients, regardless of age, gender and occupation, who were admitted to three outpatient electromyography labs to perform an electrodiagnostic testing (EDX) for the first time because of CTS symptoms. We prospectively enrolled convenience controls among all the other consecutive patients, regardless of age, gender and occupation, who were admitted to the same three outpatient electromyography labs to perform EDX for the first time because of upper limb complaints, other than CTS. All patients who underwent hand and wrist surgery, with polyneuropathy, amyotrophic lateral sclerosis, diabetes, rheumatic or thyroid diseases, renal failure, gout, history of alcoholism, malignancy in the previous 5 years, hand, wrist and arm trauma with or without fracture, onset of symptoms during
208
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
pregnancy or lactation, and previous intake of medication considered toxic to the peripheral nervous system were excluded from cases and controls. The collection of cases lasted six months (from 15 January to 15 July 2011) and that of controls further eight months (up to 15 March 2012), the time necessary to obtain a ratio of 1 case to 2 controls. CTS diagnosis was made on the basis of the clinical findings and delay of distal conduction velocity of the median nerve according to a consensus conference on criteria for the classification of CTS in epidemiological studies [15]. For clinical diagnosis, the inclusion criteria were those recommended by the American Academy of Neurology summarized here as: dull, aching discomfort in the hand, forearm, or upper arm; paraesthesia, weakness or clumsiness of the hand provoked or worsened by sleep, sustained hand or arm position, repetitive actions of the hand or wrist and mitigated by changing posture or by shaking the hand [16]. Sensory deficit in the median innervated region of the hand and motor deficit or hypotrophy of median innervated thenar muscles could be present, but they were non-necessary to fulfil the diagnosis in the presence of typical symptoms. Local ethics committee approved the study and all patients gave informed consent.
2.2. Electrophysiological methods To confirm the clinical diagnosis of CTS we performed EDX according to our protocol inspired by the American Association of Neuromuscular & Electrodiagnostic Medicine (AANEM) and based upon the recommendations of Werner et al. [17–19]. The EDX protocol included “standard” and “optional” tests. “Standard” EDX included motor conduction velocity in the elbow– wrist segment of the median nerve and below the elbow–wrist segment of the ulnar nerve, recording from the abductor pollicis brevis and abductor digiti minimi muscles, respectively. Distal motor latency (DML) was calculated at a fixed distance of 7 cm from the point of stimulation at the wrist to the muscle from which compound muscle action potential was recorded. Sensory conduction velocity (SCV) was orthodromically measured in the third and fourth finger–wrist (M4) tracts for the median nerve and in the fourth finger–wrist tract for the ulnar nerve (U4). Differences between U4–M4 SCV and median–ulnar DML were also calculated. The amplitudes of sensory nerve and compound muscle action potentials were measured but not used for diagnosis of CTS. If at least one absolute parameter of the median nerve plus one differential value at “standard” EDX were abnormal, the diagnosis of CTS was electrophysiologically confirmed. When none or only one value of “standard” protocol was abnormal, the following “optional” tests were performed: SCV in the first finger–wrist tracts for the median (M1) and radial (R1) nerves, differences between R1–M1 SCVs, and differences between the latencies of the median and ulnar nerves in 8 cm palm-to-wrist segment and between the second lumbrical–second interosseous muscles' DML. The patients were definitively included in the cases when at least one absolute plus one differential value or two differential values were abnormal at “standard” or “optional” EDX. Skin temperature of the hand was maintained above 32 °C with an infrared lamp and measured with a digital thermometer. Neurographic values that differed by at least 2 SDs from the mean of the normative data of each lab were considered abnormal. These abnormal values varied little between the three labs. In the lab that identified the most cases, DML of the median nerve in adults was considered significantly delayed if it was more than 4.4 ms. The M1, M3 and M4 SCVs were considered significantly slowed if they were below 40.8, 45.3 and 42.7 m/s, respectively. We considered significantly abnormal the differences N9.5 and N10.2 m/s between U4–M4 and R1–M1 SCVs, respectively, N0.44 ms in median–ulnar 8 cm palm-to-wrist latency, N1.56 ms in median–ulnar DML recording from abductor pollicis brevis and abductor digiti minimi muscles and N0.7 ms in median–ulnar DML recording from the second lumbrical–interosseous muscles.
The EDX was performed in the controls according to the clinical suspicion. However “standard” EDX (to confirm the absence of CTS) was mandatory normal to enrol the patient in the control group. In addition the patients with clinical diagnosis of CTS and normal distal conduction velocity of the median nerve and patients with asymptomatic distal delay of the median nerve were excluded from the cases and controls. All three electromyographers were experienced, received the same neurophysiological training, and used the same standardised methods for clinical and electrophysiological diagnosis of CTS.
2.3. Anthropometric measurements We measured the external hand and wrist dimensions (in mm) in cases and controls using a standard sliding calliper (accurate to 0.1 mm) from the palm side. The fingers were fully extended on a flat and hard support surface. The measurements were: 1) wrist width: maximum transverse distance between the borders at the level of the distal flexor wrist crease; 2) wrist depth: anterior–posterior depth at the level of the distal flexor wrist crease; 3) hand length: distance of the volar surface between the distal flexor crease of the wrist to the tip of the third finger; and 4) palm width: maximum distance of the volar surface between the second and fifth metacarpal heads. Based on these measures two ratios were calculated: wrist ratio (WR): wrist depth/wrist width and hand-ratio (HR): hand length/ palm width. The difference in the measurements between dominant and nondominant hands was randomly tested in 20 cases and 40 controls. There were no significant interside differences in all hand measures using the non-parametric sign test. In the patients with bilateral symptoms we measured the hand with worst symptoms or if there was no difference between sides, the dominant hand. Therefore we analysed the data at patient level and not at hand level because including a patient with bilateral CTS as two cases may be a source of statistical bias and the results may be overstated if the correlation between the two hands is not taken into account [20]. We measured waist and hip circumferences (HC), in cm, according to WHO recommendation. The subject is standing upright during the measurement, with arms relaxed at the side, feet evenly spread apart and body weight evenly distributed. The WC was measured at the end of several consecutive natural breaths, at a level parallel to the floor, midpoint between the top of the iliac crest and the lower margin of the last palpable rib in the midaxillary line. The HC was calculated at a level parallel to the floor, at the largest circumference of the buttocks. Both measurements were made with a stretch-resistant tape that is wrapped snugly around the subject, but not to the point that the tape is constricting. WHR was calculated dividing WC by HC [14]. Height and weight were also measured and BMI calculated (kg/m2).
2.4. Reliability of measurements Four examiners (three electromyographers and one neurophysiological technician) performed all anthropometric measures and then clinical and electrophysiological examinations. The four examiners underwent a common training to standardise the measurement techniques. The interexaminer agreement of all body and hand measures was tested in a single blind measurement session with 17 volunteers of various body sizes (12 women and 5 men, mean age 46.6 ± 10.2 years [range 30–64], mean height 166.8 ± 7.8 cm [range 150–178], and mean weigh 74 ± 19.3 kg [range 49–120]). There were no significant differences in all anthropometric measures between the four operators (Friedman test) and the intrameasure and intermeasure errors of WC and HC were less than 2 cm according to WHO recommendations [14].
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
2.5. Statistical analysis Descriptive statistics was given as mean and SD. As the distribution of data was not normal (Kolmogorov–Smirnoff test with Lilliefors correction), the Mann–Whitney nonparametric test was employed to test the differences between cases and controls. Multivariate logistic regression analysis was also performed to estimate odds ratio (OR) and 95% confidence interval with the aim to assess the strength of association between CTS and independent categorical variables represented by BMI (in three categories: b25, ≥ 25 and b30, ≥ 30), WHR (binary, cut-off for females and males at 0.85 and 0.95, respectively) and WC (in three categories, cut-offs for females and males at 81–88 and 95–102 cm, respectively). The cut-offs were those recommended by WHO according to gender and Europids [14]. All models were also adjusted for age (continuous variable, years), enrolment centre and occupation (four categories: non-at-risk workers and students — taken as the reference category, at-risk workers, retired and unemployed subjects, housewives) [21]. The pensioners were included in the “at-risk” workers category if symptoms started before the retirement and they had had an at-risk occupation. We also utilised the hand anthropometric measures to adjust the multivariate models: WR (binary, cut-off at 0.7, normal values ≤0.7) and HR (binary, cut-off at 2.2, normal values N2.2). WR N 0.7 and HR ≤ 2.2 were proposed as hand/wrist anthropometric ratios at risk for CTS [8,22,23]. We included both categorical variables of hand and wrist ratios in the same model, because HR and WR were not correlated. Conversely, to avoid collinearity, we did not include in the same model the categorical body variables (BMI, WC and WHR), because of the strong correlation between them. In addition, with the aim to explore the relevance of different categories of body measures, we created three complex categorical variables, crossing the categories of BMI and WC, BMI and WHR, and WC and WHR, respectively. The reference categories of these variables were “BMI b 25 and normal WC”, “BMI b 25 and normal WHR” and “normal WC and normal WHR”. All statistics were run on STATA 12.0 (College Park, TX, USA) software packages. An alpha-error of 0.05 was accepted.
209
Table 1 Descriptive data of demographic and anthropometric variables and differences between female (1a) and male (1b) cases and controls. 1a Variable
Age (years) Weight (kg) Height (cm) Waist circumference (cm) Hip circumference (cm) Wrist width (mm) Wrist depth (mm) Hand length (mm) Palm width (mm) BMI (kg/m2) Waist-to-hip ratio (WHR) Wrist ratio (WR) Hand ratio (HR)
Cases (n. 250)
Controls (n. 474)
Differences cases vs. controls
Mean ± SD
Mean ± SD
Za
P
54.4 66.2 158.4 89.5 104 53.5 38.2 171.4 77.2 26.4 0.86 0.75 2.23
51.8 63.0 159.9 84.4 101.0 53.3 37.5 174.0 76.5 24.7 0.83 0.7 2.28
−1.84 −4.11 −2.81 −6.0 −4.39 −0.23 −2.96 −3.68 −1.75 −5.85 −5.0 −3.85 −5.2
0.066 b0.0001 0.005 b0.0001 b0.0001 0.82 0.003 b0.0001 0.08 b0.0001 b0.0001 b0.0001 b0.0001
± ± ± ± ± ± ± ± ± ± ± ± ±
15 11.4 6.6 11.5 9.5 3.8 3.1 8.8 4.5 4.5 0.07 0.04 0.13
± ± ± ± ± ± ± ± ± ± ± ± ±
16.6 11.5 7.3 11.5 10.4 3.8 3.52 9.0 4.4 4.4 0.07 0.04 0.12
1b Variable
Age (years) Weight (kg) Height (cm) Waist circumference (cm) Hip circumference (cm) Wrist width (mm) Wrist depth (mm) Hand length (mm) Palm width (mm) BMI (kg/m2) Waist-to-hip ratio (WHR) Wrist ratio (WR) Hand ratio (HR) a
Cases (n. 120)
Controls (n. 273)
Differences cases vs. controls
Mean ± SD
Mean ± SD
Za
P
57.9 82.5 170.7 100.1 104.4 60.5 43.2 185.5 87.8 28.3 0.96 0.73 2.12
50.3 78.9 173.2 94.9 102.7 59.3 41.5 190.3 85.2 26.3 0.93 0.70 2.24
−4.27 −2.64 −3.55 −4.38 −1.49 −2.65 −5.05 −4.33 −5.3 −4.95 −5.41 −3.35 −8.0
b0.0001 0.008 b0.0001 b0.0001 0.14 0.008 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 0.001 b0.0001
± ± ± ± ± ± ± ± ± ± ± ± ±
16.3 14.4 7.4 11.5 8.5 3.7 3.2 9.3 4.9 4.0 0.06 0.04 0.13
± ± ± ± ± ± ± ± ± ± ± ± ±
16.3 12.8 7.2 11.0 9.1 4 3.2 9.7 5.1 3.7 0.08 0.04 0.13
Z value of Mann–Whitney U test.
3. Results We enrolled 1117 patients, 370 cases and 747 controls. All were Caucasian. Female cases and controls were 250 and 474 and male cases and controls were 120 and 273, respectively. The cases (no. 370) with bilateral or unilateral CTS were 233 (63%) and 137 (37%), respectively. The dominant hand was the worst or the only affected hand in 312 cases (84.3%). The controls (no. 747) included: 536 patients (71.8%) with clinical examination and EDX negative for neuromuscular diseases, 102 (13.7%) had upper limb radiculopathy or whiplash injury, 45 (6%) ulnar nerve mononeuropathy, 36 (4.8%) other mononeuropathy or brachial plexopathy and 28 (3.7%) myopathy including patients with asymptomatic or paucisymptomatic hyperCKaemia. We excluded from cases and controls 42 patients with abnormal EDX without CTS symptoms or signs, 35 patients with CTS symptoms and normal EDX and 6 patients with CTS symptoms and only one abnormal “standard” or “optional” EDX. Table 1 shows age and anthropometric findings, grouped for gender, and differences between case and control groups. There were significant differences for almost all variables of male and female groups. Table 2 reports the results of univariate and multivariate analyses grouped for gender, adjusted for age, enrolment centre, occupation, HR and WR. Both analyses demonstrated that overweight females and males, or females and males with high WC or high WHR, were at risk for CTS. The risk further increased in females and males with BMI ≥ 30 (obese) or with very high WC. The results of the analyses regarding more complex categorical variables (BMI and WC, BMI and WHR, WC and WHR) are reported in
Tables 3–5. The overweight (BMI between 25 and 29.9) subjects, especially females, were at risk only if they had very high WC or high WHR, the risk further increased if they were obese. The obesity associated with very high WC or with high WHR tripled the risk of CTS. On the contrary, females and males with normal BMI and high WC and WHR were not at risk (see Tables 3 and 4). Table 5 shows the analyses crossing the categories of WC and WHR. These two variables resulted to be strongly interrelated. Table 6 shows the results of HR and WR, in univariate and multivariate analyses. At both analyses the two anthropometric hand variables were independent risk factors for CTS in both genders. Abnormal HR and WR at least doubled the risk of CTS. 4. Discussion In our outpatient-based population with CTS, we investigated the anthropometric risk factors including WC and WHR, never examined previously in detail in literature. The original finding of our study is that high WC or WHR are independent risk factors for CTS, they doubled the risk of CTS in both genders. In addition we demonstrated that there is a relation between BMI and WC or WHR when they are considered together as risk factors. The results of multivariate regression analyses, adjusted for age, wrist and hand dimension, occupation and enrolment centre, showed that high BMI, if associated with high value of WC/WHR, was a major risk factor than high BMI alone. Obese people (BMI ≥ 30) are anyway at high risk of CTS regardless of WC or WHR value, on the contrary
210
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
Table 2 Univariate and multivariate analyses of body measures as risk factors for CTS among females (2a) and males (2b). 2a Females = no. 724
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
98 111
282 148
41
44
1.0 2.16 (1.53–3.04) 2.68 (1.64–4.39)
1.0 1.80 (1.23–2.65) 2.22 (1.28–3.84)
Waist circumference (cm) ≤80 (normal) 60 81–88 (high) 61
203 123
N88 (very high)
129
148
1.0 1.68 (1.10–2.56) 2.95 (2.00–4.33)
1.0 1.82 (1.14–2.93) 2.56 (1.67–3.92)
Waist-to-hip ratio (WHR) ≤0.85 (normal) 115 N0.85 (high) 135
304 170
1.0 2.10 (1.53–2.88)
1.0 2.08 (1.45–2.99)
Males = no. 393
Univariate
Multivariatea
Cases (no. = 120)
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
1.0 2.24 (1.30–3.88) 3.89 (1.95–7.78)
1.0 1.67 (0.90–3.09) 3.14 (1.46–6.73)
1.0 1.36 (0.78–2.36) 2.92 (1.72–4.98)
1.0 0.98 (0.52–1.85) 2.13 (1.16–3.89)
1.0 2.22 (1.40–3.51)
1.0 1.87 (1.10–3.12)
BMI (kg/m2) b25 25–29 ≥30
2b
BMI (kg/m2) b25 25–29
23 66
104 133
≥30
31
36
Waist circumference (cm) ≤94 (normal) 40 95–102 (high) 30
138 76
N102 (very high)
50
59
Waist-to-hip ratio (WHR) ≤0.95 (normal) 68 N0.95 (high) 52
203 70
a Multivariate unconditional logistic regression models were executed including BMI or WC or WHR as dependent variable; OR was adjusted for age, enrolment centre, occupational status, WR, and HR.
overweight subjects (BMI 25–29.9) are at risk only if WC is very high or WHR is high. Hence, WC and WHR may individuate subjects at risk, independently from the simple value of BMI. Many papers demonstrated association between high BMI or obesity and CTS [2–5,9–12,24–27]. Only a few studies denied this association [28–31]. The association between high BMI and CTS was commonly explained with the increased fatty tissue within the carpal tunnel or to increased hydrostatic pressure in the carpal tunnel in overweight and obese subjects exerting a compressive effect on the median nerve. But Werner et al. questioned this hypothesis and supposed a localized metabolic mechanism causing endoneurial oedema and intrafascicular swelling of the median nerve responsible of SCV delay of the median nerve in obese subjects [32]. There are no direct findings in the literature on WC/WHR as risk factors for CTS. Some recent literature data could suggest that WC/WHR might have a role as risk factor, but these studies were aimed at other purposes: to investigate the association of the atherosclerosis with CTS and the role of glucose metabolism abnormalities in “idiopathic” CTS [33,34]. BMI is a good general indicator of body fat content and WC reflects abdominal adiposity. High WC is an indicator of metabolic dysfunction
and is related to “metabolic syndrome”. Balci and Utku demonstrated that the “metabolic syndrome” is three times more frequent in patients with CTS [35]. Nakamichi and Tachibana showed an association between low-density lipoproteins and the presence of CTS. In addition cross sectional area, measured with ultrasonography, and DML and SCV of the median nerve were correlated with serum low-density lipoproteins [36]. An epidemiological research suggested the association between CTS and cardiovascular risk factors [33]. Plastino et al. showed strong association between CTS and glucose metabolism anomalies. Insulin resistance was documented in 80% of “idiopathic” CTS patients, of whom 45% had impaired glucose tolerance and WC and BMI were increased in CTS groups in respect to controls [34]. It is likely that high WHR and high WC associated with high BMI may be a risk factor for CTS via metabolic changes. Finally, our study confirms numerous literature reports showing that two well-known ratios of the hand and wrist are independent risk factors for CTS. A short and wide hand and squared wrist increased risk for CTS or association with median nerve abnormality at the carpal tunnel [8,10–12,22,23,31,37–42]. Our results showed that WR and HR were not influenced by any body anthropometric and demographic factors and abnormal values (WR N 0.7 and HR ≤ 2.2) at least doubled the risk of CTS regardless of BMI and WC/WHR values. HR ≤2.2 almost quadrupled the risk for CTS in males.
4.1. Strength and limits of the study We measured external dimensions of the hand and wrist with a calliper as surrogate of the dimensions of carpal canal. For screening purposes a simple estimation of external wrist dimension can be sufficient because hand shape matched the dimension of carpal tunnel [13]. The WC and HC were obtained according to standardised method suggested by WHO [14]. In addition the interobserver variability of all anthropometric measurements between the four examiners was negligible and our measures reliable. Because female gender is a well-known risk factor for CTS and females have different anthropometric characteristics than males, the statistics was performed grouping by gender. At multivariate regression analysis ORs were calculated adjusting the models for age and occupation of subjects and for centre of enrolment considered as confounding factors. We did not perform a detailed study of specific ergonomic exposure of the occupation, but we took into consideration only the type of occupation at the onset of the symptoms as surrogate of occupational risk factor for CTS according to a checklist of the job-titles considered at risk [21]. We did not stratify obese subjects into three common subgroups (obesity class 1: 30–34.9, severe obesity class 2: 35–39.9 and extreme obesity class 3: ≥ 40) [14], because the number of cases/controls in some subgroups was too small to obtain reliable statistical results. Therefore, for all analyses, obesity was considered for BMI ≥ 30. The enrolled CTS patients had clinical symptoms and two electrophysiological abnormalities of which at least one comparative test or short segment conduction, because the abnormality of only one test could be random. Two or more abnormal EDX tests greatly reduced the probability of including false positive cases of CTS [18]. Use of comparison latencies in addition to absolute latency/conduction velocity can improve sensitivity and specificity and helps to control for other confounding variables such as temperature, age, gender, height, hand size and other patient-specific variability [43]. Therefore our inclusion criteria were restrictive (coexistence of symptoms and electrophysiological abnormalities). Some cases with mild CTS symptoms and without two abnormal EDX tests were not included in this study, also subjects with asymptomatic delay of the median nerve were excluded from cases and controls, but the overall number was negligible (less than 10%).
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
211
Table 3 Univariate and multivariate analyses regarding body mass index (BMI) and waist circumference (WC) crossed categories as risk factors for CTS among females (3a) and males (3b). 3a Females = no. 724
BMI b25 and normal WC BMI b25 and high WC BMI b 25 with very high WC Overweight with normal WC Overweight with high WC Overweight with very high WC Obesity with normal WC Obesity with high WC Obesity with very high WC
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
57 28 13 3 33 75 – – 41
180 75 27 23 48 77 – – 44
1.0 1.17 (0.69–1.99) 1.52 (0.73–3.15) 0.41 (0.11–1.43) 2.17 (1.26–3.73) 3.07 (1.95–4.83) – – 2.94 (1.72–5.02)
– 1.42 (0.79–2.54) 1.42 (0.61–3.27) 0.36 (0.09–1.40) 2.06 (1.12–3.80) 2.56 (1.57–4.18) – – 2.57 (1.43–4.64)
Univariate
Multivariatea
3b Males = no. 393
BMI b25 and normal WC BMI b25 with high WC BMI b 25 with very high WC Overweight with normal WC Overweight with high WC Overweight with very high WC Obesity with normal WC Obesity with high WC Obesity with very high WC a
Cases (no. = 120)
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
19 4 0 21 24 21 0 2 29
92 8 4 45 63 25 1 5 30
1.0 2.42 (0.65–9.00) – 2.26 (1.09–4.68) 1.84 (0.93–3.67) 4.07 (1.83–9.06) – 1.94 (0.35–10.86) 4.68 (2.20–9.98)
– 1.95 (0.47–8.15) – 2.11 (0.94–4.73) 1.27 (0.58–2.79) 2.39 (1.00–5.70) – 1.44 (0.20–10.25) 3.75 (1.64–8.58)
Multivariate unconditional logistic regression model, OR was adjusted for age, enrolment centre, occupational status, WR, and HR.
Convenience controls were enrolled among all the consecutive patients with upper limb complaints, other than CTS, instead of healthy subjects. Although we performed “standard” EDX to confirm the absence of CTS in the controls, a possible enrolment bias could be due to the existence of risk factors (for upper limb disorders other than CTS) similar to risk factors for CTS. On the other hand, some of the major risk factors for CTS, as well as wrist dimensions and hand/wrist movement repetition, are quite peculiar to this disease, and other risk factors, such as obesity, are yet studied in the case of other upper limb disorders, as in the ulnar neuropathy at the elbow [44]. Hence, this possible bias could have determined, at most, an underestimation of some our results.
Another enrolment bias may also be the site of patient recruitment. Our study is based on population from EMG labs. Even if our EMG labs are primary care outpatient services and examine only unselect patients, our cohort may be different than that captured through general practice and may be not representative of the general population. Using patient controls may have underestimated the association between obesity and CTS. A previous case–control study demonstrated that, using two different control groups, the occurrence of overweight/obesity for hospital-based controls was higher than that for population-based controls. The number of cases and controls of this study was small enough (38 men cases and 152 controls) [45]. Our
Table 4 Univariate and multivariate analyses regarding body mass index (BMI) and waist-to-hip ratio (WHR) crossed categories as risk factors for CTS among females (4a) and males (4b). 4a Females = no. 724
BMI b25 with normal WHR BMI b 25 with high WHR Overweight with normal WHR Overweight with high WHR Obesity with normal WHR Obesity with high WHR
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
68 30 38 73 9 32
225 57 73 75 6 38
1.0 1.74 (1.03–2.94) 1.72 (1.07–2.78) 3.22 (2.08–4.99) 4.96 (1.67–14.73) 2.79 (1.60–4.85)
– 2.11 (1.17–3.82) 1.44 (0.84–2.47) 2.91 (1.80–4.70) 5.83 (1.83–18.63) 2.28 (1.23–4.25)
Univariate
Multivariatea
4b Males = no. 393
BMI b25 with normal WHR BMI b 25 with high WHR Overweight with normal WHR Overweight with high WHR Obesity with normal WHR Obesity with high WHR a
Cases (no. = 120)
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
18 5 40 26 10 21
93 11 100 33 10 26
1.0 2.35 (0.72–7.69) 2.07 (1.10–3.89) 4.07 (1.91–8.69) 5.17 (1.78–14.96) 4.17 (1.87–9.33)
– 1.74 (0.47–6.47) 1.51 (0.74–3.06) 2.75 (1.20–2.27) 4.27 (1.37–13.28) 3.17 (1.29–7.77)
Multivariate unconditional logistic regression model, OR adjusted for age, enrolment centre, occupational status, WR, and HR.
212
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
Table 5 Univariate and multivariate analyses regarding waist circumference (WC) and waist-to-hip ratio (WHR) crossed categories as risk factors for CTS among females (5a) and males (5b). 5a Females = no. 724
Normal WC with normal WHR Normal WC with high WHR High WC with normal WHR High WC with high WHR Very high WC with normal WHR Very high WC with high WHR
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
57 3 36 25 22 107
190 13 86 37 28 120
1.0 0.77 (0.21–2.80) 1.40 (0.85–2.28) 2.25 (1.24–4.09) 2.62 (1.38–4.98) 2.97 (1.98–4.47)
– 1.59 (0.41–6.10) 1.50 (0.87–2.60) 2.86 (1.47–5.56) 2.35 (1.18–4.71) 2.72 (1.72–4.30)
Univariate
Multivariatea
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
133 5 56 20 14 45
1.0 5.64 (1.62–19.62) 1.37 (0.72–2.61) 2.22 (0.96–5.13) 4.61 (1.97–10.75) 3.05 (1.66–5.58)
– 3.83 (1.04–14.06) 0.99 (0.48–2.07) 1.54 (0.59–4.01) 2.95 (1.17–7.43) 2.33 (1.17–4.65)
5b Males = no. 393 Cases (no. = 120) Normal WC with normal WHR Normal WC with high WHR High WC with normal WHR High WC with high WHR Very high WC with normal WHR Very high WC with high WHR a
33 7 19 11 16 34
Multivariate unconditional logistic regression model, OR adjusted for age, enrolment centre, occupational status, WR, and HR.
study did not use hospital-based controls but outpatients referring to EMG labs.
Disclosures The authors state that there are no conflicts of interest and that they have not received any financial support.
4.2. Conclusions Abdominal adiposity measures (WC and WHR) are correlated with BMI, but the degree of association could vary, suggesting that these measures may provide different information and thus may not be interchangeable and it is advisable to obtain BMI and WC together and consider the utility of joint use of the two indicators.
Table 6 Univariate and multivariate analyses regarding hand ratio (HR) and wrist ratio (WR) as risk factors for CTS among females (6a) and males (6b). 6a Females = no. 724
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
Hand ratio (HR) N2.2 (not at risk) ≤2.2 (at risk)
138 112
343 131
1.0 2.13 (1.53–2.94)
1.0 2.68 (1.85–3.88)
Wrist ratio (WR) ≤0.7 (not at risk) N0.7 (at risk)
85 165
239 235
1.0 1.97 (1.43–2.73)
1.0 2.68 (1.85–3.87)
Males = no. 393
Univariate
Multivariatea
Cases (no. = 120)
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
6b
Hand ratio (HR) N2.2 (not at risk) ≤2.2 (at risk)
31 89
160 113
1.0 4.07 (2.47–6.69)
1.0 3.69 (2.18–6.25)
Wrist ratio (WR) ≤0.7 (not at risk) N0.7 (at risk)
46 74
148 125
1.0 1.90 (1.22–2.97)
1.0 2.32 (1.36–3.95)
a Multivariate unconditional logistic regression model, OR adjusted for age, enrolment centre, occupational status, and BMI.
References [1] Stevens JC, Beard CM, O'Fallon WM, Kurland LT. Conditions associated with carpal tunnel syndrome. Mayo Clin Proc 1992;67:541–8. [2] Nordstrom DL, Vierkant RA, DeStefano F, Layde PM. Risk factors for carpal tunnel syndrome in a general population. Occup Environ Med 1997;54:734–40. [3] Karpitskaya Y, Novak CB, Mackinnon SE. Prevalence of smoking, obesity, diabetes mellitus, and thyroid disease in patients with carpal tunnel syndrome. Ann Plast Surg 1997;48:269–73. [4] Becker J, Nora DB, Gomes I, Stringari FF, Seitensus R, Panosso JS, et al. An evaluation of gender, obesity, age and diabetes mellitus as risk factors for carpal tunnel syndrome. Clin Neurophysiol 2002;113:1429–34. [5] Geoghegan JM, Clark DI, Bainbridge LC, Smith C, Hubbard R. Risk factors in carpal tunnel syndrome. J Hand Surg Br 2002;29:315–20. [6] Barcenilla A, March LM, Chen JS, Sambrook PN. Carpal tunnel syndrome and its relationship to occupation: a meta-analysis. Rheumatology (Oxford) 2012;51:250–61. [7] Pourmemari MH, Viikari-Juntura E, Shiri R. Smoking and carpal tunnel syndrome: a meta-analysis. Muscle Nerve Jun 12 2013. http://dx.doi.org/10.1002/mus.23922. [8] Johnson EW, Gatens T, Poindexter D, Bowers D. Wrist dimensions: correlation with median sensory latencies. Arch Phys Med Rehabil 1983;64:556–7. [9] Werner RA, Albers JW, Franzblau A, Armstrong TJ. The relationship between body mass index and the diagnosis of carpal tunnel syndrome. Muscle Nerve 1994;17:632–6. [10] Kouyoumdjian JA, Zanetta DM, Morita MP. Evaluation of age, body mass index, and wrist index as risk factors for carpal tunnel syndrome severity. Muscle Nerve 2002;25:93–7. [11] Boz C, Ozmenoglu M, Altunayoglu V, Velioglu S, Alioglu Z. Individual risk factors for carpal tunnel syndrome: an evaluation of body mass index, wrist index and hand anthropometric measurements. Clin Neurol Neurosurg 2004;106:294–9. [12] Moghtaderi A, Izadi S, Sharafadinzadeh N. An evaluation of gender, body mass index, wrist circumference and wrist ratio as independent risk factors for carpal tunnel syndrome. Acta Neurol Scand 2005;112:375–9. [13] Chiotis K, Dimisianos N, Rigopoulou A, Chrysanthopoulou A, Chroni E. Role of anthropometric characteristics in idiopathic carpal tunnel syndrome. Arch Phys Med Rehabil 2013;94:737–44. [14] World Health Organization. Waist circumference and waist–hip ratio, report of a WHO expert consultation. http://whqlibdoc.who.int/publications/2011/9789241501491_ eng.pdf; December 8–11 2008 . [Retrieved March 21, 2012]. [15] Rempel D, Evanoff B, Amadio PC, de Krom M, Franklin G, Franzblau A, et al. Consensus criteria for the classification of carpal tunnel syndrome in epidemiologic studies. Am J Public Health 2013;88:1447–51. [16] Quality Standards Subcommittee of the American Academy of Neurology. Practice parameter for carpal tunnel syndrome (summary statement). Neurology 1993;43:2406–9. [17] American Association of Electrodiagnostic Medicine, American Academy of Neurology, American Academy of Physical Medicine and Rehabilitation. Practice parameter
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
[18] [19]
[20] [21]
[22] [23]
[24]
[25]
[26] [27] [28] [29]
[30]
for electrodiagnostic studies in carpal tunnel syndrome: summary statement. Muscle Nerve 2002;25:918–22. Werner RA. Evaluation of work-related carpal tunnel syndrome. J Occup Rehabil 2006;16:207–22. Mondelli M, Baldasseroni A, Aretini A, Ginanneschi F, Padua L. Prevalent involvement of thenar motor fibres in vineyard workers with carpal tunnel syndrome. Clin Neurophysiol 2010;121:1251–5. Padua L, Pasqualetti P, Rosenbaum R. One patient, two carpal tunnels: statistical and clinical analysis — by hand or by patient? Clin Neurophysiol 2005;116:241–3. Colombini D, Occhipinti E, Cairoli S, Battevi N, Menoni O, Ricci MG, et al. Musculoskeletal conditions of the upper and lower limbs as an occupational disease: what kind and under what conditions. Consensus document of a national working group (ISPESL). Med Lav 2003;94:312–29. Chroni E, Paschalis C, Arvaniti C, Zotou K, Nikolakopoulou A, Papapetropoulos T. Carpal tunnel syndrome and hand configuration. Muscle Nerve 2001;24:1607–11. Kamolz LP, Beck H, Haslik W, Högler R, Rab M, Schrögendorfer KF, et al. Carpal tunnel syndrome: a question of hand and wrist configurations? J Hand Surg Br 2004;29:321–4. Stallings SP, Kasdan ML, Soergel TM, Corwin HM. A case–control study of obesity as a risk factor for carpal tunnel syndrome in a population of 600 patients presenting for independent medical examination. J Hand Surg Am 1997;22:211–5. Nathan PA, Keniston RC, Myers LD, Meadows KD. Obesity as a risk factor for slowing of sensory conduction of the median nerve in industry. A cross-sectional and longitudinal study involving 429 workers. J Occup Med 1992;34:379–83. Bland JD. The relationship of obesity, age, and carpal tunnel syndrome: more complex than was thought? Muscle Nerve 2005;32:527–32. Seror P, Seror R. Prevalence of obesity and obesity as a risk factor in patients with severe median nerve lesion at the wrist. Joint Bone Spine 2013;80:632–7. Cannon LJ, Bernacki EJ, Walter SD. Personal and occupational factors associated with carpal tunnel syndrome. J Occup Med 2013;23:255–8. Gorsche RG, Wiley JP, Renger RF, Brant RF, Gemer TY, Sasyniuk TM. Prevalence and incidence of carpal tunnel syndrome in a meat packing plant. Occup Environ Med 1999;56:417–22. Hakim AJ, Cherkas L, El Zayat S, MacGregor AJ, Spector TD. The genetic contribution to carpal tunnel syndrome in women: a twin study. Arthritis Rheum 2002;47:275–9.
213
[31] Farmer JE, Davis TR. Carpal tunnel syndrome: a case–control study evaluating its relationship with body mass index and hand and wrist measurements. J Hand Surg Eur Vol 2008;33:445–8. [32] Werner RA, Jacobson JA, Jamadar DA. Influence of body mass index on median nerve function, carpal canal pressure, and cross-sectional area of the median nerve. Muscle Nerve 2004;30:481–5. [33] Shiri R, Heliövaara M, Moilanen L, Viikari J, Liira H, Viikari-Juntura E. Associations of cardiovascular risk factors, carotid intima–media thickness and manifest atherosclerotic vascular disease with carpal tunnel syndrome. BMC Musculoskelet Disord Apr 26 2011;12:80. [34] Plastino M, Fava A, Carmela C, De Bartolo M, Ermio C, Cristiano D, et al. Insulin resistance increases risk of carpal tunnel syndrome: a case–control study. J Peripher Nerv Syst 2011;16:186–90. [35] Balci K, Utku U. Carpal tunnel syndrome and metabolic syndrome. Acta Neurol Scand 2007;116:113–7. [36] Nakamichi K, Tachibana S. Hypercholesterolemia as a risk factor for idiopathic carpal tunnel syndrome. Muscle Nerve 2005;32:364–7. [37] Dekel S, Coates R. Primary carpal stenosis as a cause of ‘idiopathic’ carpal-tunnel syndrome. Lancet 1979;2(8150):1024. [38] Bleecker ML, Bohlman M, Moreland R, Tipton A. Carpal tunnel syndrome: role of carpal canal size. Neurology 1985;35:1599–604. [39] Gordon C, Johnson EW, Gatens PF, Ashton JJ. Wrist ratio correlation with carpal tunnel syndrome in industry. Am J Phys Med Rehabil 1988;67:270–2. [40] Radecki P. A gender specific wrist ratio and the likelihood of a median nerve abnormality at the carpal tunnel. Am J Phys Med Rehabil 1994;73:157–62. [41] Nakamichi K, Tachibana S. Small hand as a risk factor for idiopathic carpal tunnel syndrome. Muscle Nerve 1994;18:664–6. [42] Lim PG, Tan S, Ahmad TS. The role of wrist anthropometric measurement in idiopathic carpal tunnel syndrome. J Hand Surg Eur Vol 2008;33:645–7. [43] Werner RA, Andary M. Electrodiagnostic evaluation of carpal tunnel syndrome. Muscle Nerve 2008;44:597–607. [44] Frost P, Johnsen B, Fuglsang-Frederiksen A, Svendsen SW. Lifestyle risk factors for ulnar neuropathy and ulnar neuropathy-like symptoms. Muscle Nerve 2013;48:507–15. [45] Wieslander G, Norbäck D, Göthe CJ, Juhlin L. Carpal tunnel syndrome (CTS) and exposure to vibration, repetitive wrist movements, and heavy manual work: a casereferent study. Occup Environ Med 1989;46:43–7.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Waist circumference and waist-to-hip ratio in carpal tunnel syndrome: A case–control study Mauro Mondelli a,⁎, Alessandro Aretini a, Federica Ginanneschi b, Giuseppe Greco c, Stefano Mattioli d a
EMG Service, Local Health Unit 7, Siena, Italy Dpt. Neurological, Neurosurgical and Behavioural Sciences, University of Siena, Italy EMG Service, Local Health Unit 7, “Nottola” Hospital, Montepulciano, Siena, Italy d Dpt. Medical and Surgical Sciences, University of Bologna, Italy b c
a r t i c l e
i n f o
Article history: Received 10 November 2013 Received in revised form 5 January 2014 Accepted 8 January 2014 Available online 15 January 2014 Keywords: Body mass index Carpal tunnel syndrome Hand ratio Nerve entrapment syndrome Obesity Waist circumference Waist-to-hip ratio Wrist ratio
a b s t r a c t Background: The association between carpal tunnel syndrome (CTS) and high body mass index (BMI) and some hand measures is well known. No study has been specifically focused on waist circumference (WC) and waist-tohip-ratio (WHR). The aim of this prospective case–control study is to evaluate the association between CTS and WC, WHR and other body and hand anthropometric measures. Methods: We consecutively enrolled one “idiopathic” CTS case for two controls in 3 outpatient electromyography labs. The main anthropometric measures were BMI, WC, WHR, wrist ratio (WR) and hand ratio (HR). We performed univariate and multivariate analyses. Results: Female cases and controls were 250 and 474 and male cases and controls were 120 and 273, respectively. At univariate analysis there were differences in many anthropometric measures between cases and controls. At multivariate logistic regression analyses high BMI, WC and WHR and abnormal HR and WR were independent risk factors for CTS. Crossing two categories between BMI, WC and WHR, the overweight subjects, especially females, were at risk only if they had very high WC or high WHR. The risk increased if they were obese. Conclusions: High WC/WHR doubles the risk of CTS, the risk further increased if overweight/obese subjects have also very high WC or high WHR. The obese subjects were always at risk regardless of WC and WHR values. Metabolic causes of this association with CTS were hypothesised. BMI is not the only and most powerful body predictor of “idiopathic” CTS, but also WHR and WC should be considered. These measures may not be interchangeable and it may be desirable to consider the utility of their joint use. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Carpal tunnel syndrome (CTS) is the most frequent focal peripheral neuropathy. Many risk factors for CTS were identified as female gender, age, diabetes, rheumatoid arthritis, thyroid dysfunction, renal failure, pregnancy, hand and wrist trauma, use of oral contraception, smoking, and occupations that involved forced and repetitive exertions of hand and wrist, high handgrip forces and using vibrating tools [1–7]. Many studies have focused on hand/wrist and body anthropometric characteristics. There is almost unanimous agreement that a high body mass index (BMI), a squared wrist and a short hand predispose to CTS [4,5,8–13]. Recently the World Health Organization (WHO) suggested that waist circumference (WC) and waist-to-hip-ratio (WHR) are better predictors than BMI in some diseases [14]. In literature there are no studies on WC and WHR expressly designed to demonstrate whether high WC and high WHR may be independent risk factors for CTS.
⁎ Corresponding author at: Servizio di EMG, ASL 7, Via Pian d'Ovile, 9, 53100 Siena, Italy. Tel.: +39 0577 535904; fax: +39 0577 535971. E-mail address: [email protected] (M. Mondelli). 0022-510X/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jns.2014.01.012
The purpose of our prospective case–control study is to evaluate the association between CTS and WC, WHR and other anthropometric measures. 2. Methods 2.1. Enrolment and definition of cases and controls We decided in advance to recruit one case with CTS for two controls. We prospectively enrolled cases among all consecutive patients, regardless of age, gender and occupation, who were admitted to three outpatient electromyography labs to perform an electrodiagnostic testing (EDX) for the first time because of CTS symptoms. We prospectively enrolled convenience controls among all the other consecutive patients, regardless of age, gender and occupation, who were admitted to the same three outpatient electromyography labs to perform EDX for the first time because of upper limb complaints, other than CTS. All patients who underwent hand and wrist surgery, with polyneuropathy, amyotrophic lateral sclerosis, diabetes, rheumatic or thyroid diseases, renal failure, gout, history of alcoholism, malignancy in the previous 5 years, hand, wrist and arm trauma with or without fracture, onset of symptoms during
208
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
pregnancy or lactation, and previous intake of medication considered toxic to the peripheral nervous system were excluded from cases and controls. The collection of cases lasted six months (from 15 January to 15 July 2011) and that of controls further eight months (up to 15 March 2012), the time necessary to obtain a ratio of 1 case to 2 controls. CTS diagnosis was made on the basis of the clinical findings and delay of distal conduction velocity of the median nerve according to a consensus conference on criteria for the classification of CTS in epidemiological studies [15]. For clinical diagnosis, the inclusion criteria were those recommended by the American Academy of Neurology summarized here as: dull, aching discomfort in the hand, forearm, or upper arm; paraesthesia, weakness or clumsiness of the hand provoked or worsened by sleep, sustained hand or arm position, repetitive actions of the hand or wrist and mitigated by changing posture or by shaking the hand [16]. Sensory deficit in the median innervated region of the hand and motor deficit or hypotrophy of median innervated thenar muscles could be present, but they were non-necessary to fulfil the diagnosis in the presence of typical symptoms. Local ethics committee approved the study and all patients gave informed consent.
2.2. Electrophysiological methods To confirm the clinical diagnosis of CTS we performed EDX according to our protocol inspired by the American Association of Neuromuscular & Electrodiagnostic Medicine (AANEM) and based upon the recommendations of Werner et al. [17–19]. The EDX protocol included “standard” and “optional” tests. “Standard” EDX included motor conduction velocity in the elbow– wrist segment of the median nerve and below the elbow–wrist segment of the ulnar nerve, recording from the abductor pollicis brevis and abductor digiti minimi muscles, respectively. Distal motor latency (DML) was calculated at a fixed distance of 7 cm from the point of stimulation at the wrist to the muscle from which compound muscle action potential was recorded. Sensory conduction velocity (SCV) was orthodromically measured in the third and fourth finger–wrist (M4) tracts for the median nerve and in the fourth finger–wrist tract for the ulnar nerve (U4). Differences between U4–M4 SCV and median–ulnar DML were also calculated. The amplitudes of sensory nerve and compound muscle action potentials were measured but not used for diagnosis of CTS. If at least one absolute parameter of the median nerve plus one differential value at “standard” EDX were abnormal, the diagnosis of CTS was electrophysiologically confirmed. When none or only one value of “standard” protocol was abnormal, the following “optional” tests were performed: SCV in the first finger–wrist tracts for the median (M1) and radial (R1) nerves, differences between R1–M1 SCVs, and differences between the latencies of the median and ulnar nerves in 8 cm palm-to-wrist segment and between the second lumbrical–second interosseous muscles' DML. The patients were definitively included in the cases when at least one absolute plus one differential value or two differential values were abnormal at “standard” or “optional” EDX. Skin temperature of the hand was maintained above 32 °C with an infrared lamp and measured with a digital thermometer. Neurographic values that differed by at least 2 SDs from the mean of the normative data of each lab were considered abnormal. These abnormal values varied little between the three labs. In the lab that identified the most cases, DML of the median nerve in adults was considered significantly delayed if it was more than 4.4 ms. The M1, M3 and M4 SCVs were considered significantly slowed if they were below 40.8, 45.3 and 42.7 m/s, respectively. We considered significantly abnormal the differences N9.5 and N10.2 m/s between U4–M4 and R1–M1 SCVs, respectively, N0.44 ms in median–ulnar 8 cm palm-to-wrist latency, N1.56 ms in median–ulnar DML recording from abductor pollicis brevis and abductor digiti minimi muscles and N0.7 ms in median–ulnar DML recording from the second lumbrical–interosseous muscles.
The EDX was performed in the controls according to the clinical suspicion. However “standard” EDX (to confirm the absence of CTS) was mandatory normal to enrol the patient in the control group. In addition the patients with clinical diagnosis of CTS and normal distal conduction velocity of the median nerve and patients with asymptomatic distal delay of the median nerve were excluded from the cases and controls. All three electromyographers were experienced, received the same neurophysiological training, and used the same standardised methods for clinical and electrophysiological diagnosis of CTS.
2.3. Anthropometric measurements We measured the external hand and wrist dimensions (in mm) in cases and controls using a standard sliding calliper (accurate to 0.1 mm) from the palm side. The fingers were fully extended on a flat and hard support surface. The measurements were: 1) wrist width: maximum transverse distance between the borders at the level of the distal flexor wrist crease; 2) wrist depth: anterior–posterior depth at the level of the distal flexor wrist crease; 3) hand length: distance of the volar surface between the distal flexor crease of the wrist to the tip of the third finger; and 4) palm width: maximum distance of the volar surface between the second and fifth metacarpal heads. Based on these measures two ratios were calculated: wrist ratio (WR): wrist depth/wrist width and hand-ratio (HR): hand length/ palm width. The difference in the measurements between dominant and nondominant hands was randomly tested in 20 cases and 40 controls. There were no significant interside differences in all hand measures using the non-parametric sign test. In the patients with bilateral symptoms we measured the hand with worst symptoms or if there was no difference between sides, the dominant hand. Therefore we analysed the data at patient level and not at hand level because including a patient with bilateral CTS as two cases may be a source of statistical bias and the results may be overstated if the correlation between the two hands is not taken into account [20]. We measured waist and hip circumferences (HC), in cm, according to WHO recommendation. The subject is standing upright during the measurement, with arms relaxed at the side, feet evenly spread apart and body weight evenly distributed. The WC was measured at the end of several consecutive natural breaths, at a level parallel to the floor, midpoint between the top of the iliac crest and the lower margin of the last palpable rib in the midaxillary line. The HC was calculated at a level parallel to the floor, at the largest circumference of the buttocks. Both measurements were made with a stretch-resistant tape that is wrapped snugly around the subject, but not to the point that the tape is constricting. WHR was calculated dividing WC by HC [14]. Height and weight were also measured and BMI calculated (kg/m2).
2.4. Reliability of measurements Four examiners (three electromyographers and one neurophysiological technician) performed all anthropometric measures and then clinical and electrophysiological examinations. The four examiners underwent a common training to standardise the measurement techniques. The interexaminer agreement of all body and hand measures was tested in a single blind measurement session with 17 volunteers of various body sizes (12 women and 5 men, mean age 46.6 ± 10.2 years [range 30–64], mean height 166.8 ± 7.8 cm [range 150–178], and mean weigh 74 ± 19.3 kg [range 49–120]). There were no significant differences in all anthropometric measures between the four operators (Friedman test) and the intrameasure and intermeasure errors of WC and HC were less than 2 cm according to WHO recommendations [14].
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
2.5. Statistical analysis Descriptive statistics was given as mean and SD. As the distribution of data was not normal (Kolmogorov–Smirnoff test with Lilliefors correction), the Mann–Whitney nonparametric test was employed to test the differences between cases and controls. Multivariate logistic regression analysis was also performed to estimate odds ratio (OR) and 95% confidence interval with the aim to assess the strength of association between CTS and independent categorical variables represented by BMI (in three categories: b25, ≥ 25 and b30, ≥ 30), WHR (binary, cut-off for females and males at 0.85 and 0.95, respectively) and WC (in three categories, cut-offs for females and males at 81–88 and 95–102 cm, respectively). The cut-offs were those recommended by WHO according to gender and Europids [14]. All models were also adjusted for age (continuous variable, years), enrolment centre and occupation (four categories: non-at-risk workers and students — taken as the reference category, at-risk workers, retired and unemployed subjects, housewives) [21]. The pensioners were included in the “at-risk” workers category if symptoms started before the retirement and they had had an at-risk occupation. We also utilised the hand anthropometric measures to adjust the multivariate models: WR (binary, cut-off at 0.7, normal values ≤0.7) and HR (binary, cut-off at 2.2, normal values N2.2). WR N 0.7 and HR ≤ 2.2 were proposed as hand/wrist anthropometric ratios at risk for CTS [8,22,23]. We included both categorical variables of hand and wrist ratios in the same model, because HR and WR were not correlated. Conversely, to avoid collinearity, we did not include in the same model the categorical body variables (BMI, WC and WHR), because of the strong correlation between them. In addition, with the aim to explore the relevance of different categories of body measures, we created three complex categorical variables, crossing the categories of BMI and WC, BMI and WHR, and WC and WHR, respectively. The reference categories of these variables were “BMI b 25 and normal WC”, “BMI b 25 and normal WHR” and “normal WC and normal WHR”. All statistics were run on STATA 12.0 (College Park, TX, USA) software packages. An alpha-error of 0.05 was accepted.
209
Table 1 Descriptive data of demographic and anthropometric variables and differences between female (1a) and male (1b) cases and controls. 1a Variable
Age (years) Weight (kg) Height (cm) Waist circumference (cm) Hip circumference (cm) Wrist width (mm) Wrist depth (mm) Hand length (mm) Palm width (mm) BMI (kg/m2) Waist-to-hip ratio (WHR) Wrist ratio (WR) Hand ratio (HR)
Cases (n. 250)
Controls (n. 474)
Differences cases vs. controls
Mean ± SD
Mean ± SD
Za
P
54.4 66.2 158.4 89.5 104 53.5 38.2 171.4 77.2 26.4 0.86 0.75 2.23
51.8 63.0 159.9 84.4 101.0 53.3 37.5 174.0 76.5 24.7 0.83 0.7 2.28
−1.84 −4.11 −2.81 −6.0 −4.39 −0.23 −2.96 −3.68 −1.75 −5.85 −5.0 −3.85 −5.2
0.066 b0.0001 0.005 b0.0001 b0.0001 0.82 0.003 b0.0001 0.08 b0.0001 b0.0001 b0.0001 b0.0001
± ± ± ± ± ± ± ± ± ± ± ± ±
15 11.4 6.6 11.5 9.5 3.8 3.1 8.8 4.5 4.5 0.07 0.04 0.13
± ± ± ± ± ± ± ± ± ± ± ± ±
16.6 11.5 7.3 11.5 10.4 3.8 3.52 9.0 4.4 4.4 0.07 0.04 0.12
1b Variable
Age (years) Weight (kg) Height (cm) Waist circumference (cm) Hip circumference (cm) Wrist width (mm) Wrist depth (mm) Hand length (mm) Palm width (mm) BMI (kg/m2) Waist-to-hip ratio (WHR) Wrist ratio (WR) Hand ratio (HR) a
Cases (n. 120)
Controls (n. 273)
Differences cases vs. controls
Mean ± SD
Mean ± SD
Za
P
57.9 82.5 170.7 100.1 104.4 60.5 43.2 185.5 87.8 28.3 0.96 0.73 2.12
50.3 78.9 173.2 94.9 102.7 59.3 41.5 190.3 85.2 26.3 0.93 0.70 2.24
−4.27 −2.64 −3.55 −4.38 −1.49 −2.65 −5.05 −4.33 −5.3 −4.95 −5.41 −3.35 −8.0
b0.0001 0.008 b0.0001 b0.0001 0.14 0.008 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 0.001 b0.0001
± ± ± ± ± ± ± ± ± ± ± ± ±
16.3 14.4 7.4 11.5 8.5 3.7 3.2 9.3 4.9 4.0 0.06 0.04 0.13
± ± ± ± ± ± ± ± ± ± ± ± ±
16.3 12.8 7.2 11.0 9.1 4 3.2 9.7 5.1 3.7 0.08 0.04 0.13
Z value of Mann–Whitney U test.
3. Results We enrolled 1117 patients, 370 cases and 747 controls. All were Caucasian. Female cases and controls were 250 and 474 and male cases and controls were 120 and 273, respectively. The cases (no. 370) with bilateral or unilateral CTS were 233 (63%) and 137 (37%), respectively. The dominant hand was the worst or the only affected hand in 312 cases (84.3%). The controls (no. 747) included: 536 patients (71.8%) with clinical examination and EDX negative for neuromuscular diseases, 102 (13.7%) had upper limb radiculopathy or whiplash injury, 45 (6%) ulnar nerve mononeuropathy, 36 (4.8%) other mononeuropathy or brachial plexopathy and 28 (3.7%) myopathy including patients with asymptomatic or paucisymptomatic hyperCKaemia. We excluded from cases and controls 42 patients with abnormal EDX without CTS symptoms or signs, 35 patients with CTS symptoms and normal EDX and 6 patients with CTS symptoms and only one abnormal “standard” or “optional” EDX. Table 1 shows age and anthropometric findings, grouped for gender, and differences between case and control groups. There were significant differences for almost all variables of male and female groups. Table 2 reports the results of univariate and multivariate analyses grouped for gender, adjusted for age, enrolment centre, occupation, HR and WR. Both analyses demonstrated that overweight females and males, or females and males with high WC or high WHR, were at risk for CTS. The risk further increased in females and males with BMI ≥ 30 (obese) or with very high WC. The results of the analyses regarding more complex categorical variables (BMI and WC, BMI and WHR, WC and WHR) are reported in
Tables 3–5. The overweight (BMI between 25 and 29.9) subjects, especially females, were at risk only if they had very high WC or high WHR, the risk further increased if they were obese. The obesity associated with very high WC or with high WHR tripled the risk of CTS. On the contrary, females and males with normal BMI and high WC and WHR were not at risk (see Tables 3 and 4). Table 5 shows the analyses crossing the categories of WC and WHR. These two variables resulted to be strongly interrelated. Table 6 shows the results of HR and WR, in univariate and multivariate analyses. At both analyses the two anthropometric hand variables were independent risk factors for CTS in both genders. Abnormal HR and WR at least doubled the risk of CTS. 4. Discussion In our outpatient-based population with CTS, we investigated the anthropometric risk factors including WC and WHR, never examined previously in detail in literature. The original finding of our study is that high WC or WHR are independent risk factors for CTS, they doubled the risk of CTS in both genders. In addition we demonstrated that there is a relation between BMI and WC or WHR when they are considered together as risk factors. The results of multivariate regression analyses, adjusted for age, wrist and hand dimension, occupation and enrolment centre, showed that high BMI, if associated with high value of WC/WHR, was a major risk factor than high BMI alone. Obese people (BMI ≥ 30) are anyway at high risk of CTS regardless of WC or WHR value, on the contrary
210
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
Table 2 Univariate and multivariate analyses of body measures as risk factors for CTS among females (2a) and males (2b). 2a Females = no. 724
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
98 111
282 148
41
44
1.0 2.16 (1.53–3.04) 2.68 (1.64–4.39)
1.0 1.80 (1.23–2.65) 2.22 (1.28–3.84)
Waist circumference (cm) ≤80 (normal) 60 81–88 (high) 61
203 123
N88 (very high)
129
148
1.0 1.68 (1.10–2.56) 2.95 (2.00–4.33)
1.0 1.82 (1.14–2.93) 2.56 (1.67–3.92)
Waist-to-hip ratio (WHR) ≤0.85 (normal) 115 N0.85 (high) 135
304 170
1.0 2.10 (1.53–2.88)
1.0 2.08 (1.45–2.99)
Males = no. 393
Univariate
Multivariatea
Cases (no. = 120)
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
1.0 2.24 (1.30–3.88) 3.89 (1.95–7.78)
1.0 1.67 (0.90–3.09) 3.14 (1.46–6.73)
1.0 1.36 (0.78–2.36) 2.92 (1.72–4.98)
1.0 0.98 (0.52–1.85) 2.13 (1.16–3.89)
1.0 2.22 (1.40–3.51)
1.0 1.87 (1.10–3.12)
BMI (kg/m2) b25 25–29 ≥30
2b
BMI (kg/m2) b25 25–29
23 66
104 133
≥30
31
36
Waist circumference (cm) ≤94 (normal) 40 95–102 (high) 30
138 76
N102 (very high)
50
59
Waist-to-hip ratio (WHR) ≤0.95 (normal) 68 N0.95 (high) 52
203 70
a Multivariate unconditional logistic regression models were executed including BMI or WC or WHR as dependent variable; OR was adjusted for age, enrolment centre, occupational status, WR, and HR.
overweight subjects (BMI 25–29.9) are at risk only if WC is very high or WHR is high. Hence, WC and WHR may individuate subjects at risk, independently from the simple value of BMI. Many papers demonstrated association between high BMI or obesity and CTS [2–5,9–12,24–27]. Only a few studies denied this association [28–31]. The association between high BMI and CTS was commonly explained with the increased fatty tissue within the carpal tunnel or to increased hydrostatic pressure in the carpal tunnel in overweight and obese subjects exerting a compressive effect on the median nerve. But Werner et al. questioned this hypothesis and supposed a localized metabolic mechanism causing endoneurial oedema and intrafascicular swelling of the median nerve responsible of SCV delay of the median nerve in obese subjects [32]. There are no direct findings in the literature on WC/WHR as risk factors for CTS. Some recent literature data could suggest that WC/WHR might have a role as risk factor, but these studies were aimed at other purposes: to investigate the association of the atherosclerosis with CTS and the role of glucose metabolism abnormalities in “idiopathic” CTS [33,34]. BMI is a good general indicator of body fat content and WC reflects abdominal adiposity. High WC is an indicator of metabolic dysfunction
and is related to “metabolic syndrome”. Balci and Utku demonstrated that the “metabolic syndrome” is three times more frequent in patients with CTS [35]. Nakamichi and Tachibana showed an association between low-density lipoproteins and the presence of CTS. In addition cross sectional area, measured with ultrasonography, and DML and SCV of the median nerve were correlated with serum low-density lipoproteins [36]. An epidemiological research suggested the association between CTS and cardiovascular risk factors [33]. Plastino et al. showed strong association between CTS and glucose metabolism anomalies. Insulin resistance was documented in 80% of “idiopathic” CTS patients, of whom 45% had impaired glucose tolerance and WC and BMI were increased in CTS groups in respect to controls [34]. It is likely that high WHR and high WC associated with high BMI may be a risk factor for CTS via metabolic changes. Finally, our study confirms numerous literature reports showing that two well-known ratios of the hand and wrist are independent risk factors for CTS. A short and wide hand and squared wrist increased risk for CTS or association with median nerve abnormality at the carpal tunnel [8,10–12,22,23,31,37–42]. Our results showed that WR and HR were not influenced by any body anthropometric and demographic factors and abnormal values (WR N 0.7 and HR ≤ 2.2) at least doubled the risk of CTS regardless of BMI and WC/WHR values. HR ≤2.2 almost quadrupled the risk for CTS in males.
4.1. Strength and limits of the study We measured external dimensions of the hand and wrist with a calliper as surrogate of the dimensions of carpal canal. For screening purposes a simple estimation of external wrist dimension can be sufficient because hand shape matched the dimension of carpal tunnel [13]. The WC and HC were obtained according to standardised method suggested by WHO [14]. In addition the interobserver variability of all anthropometric measurements between the four examiners was negligible and our measures reliable. Because female gender is a well-known risk factor for CTS and females have different anthropometric characteristics than males, the statistics was performed grouping by gender. At multivariate regression analysis ORs were calculated adjusting the models for age and occupation of subjects and for centre of enrolment considered as confounding factors. We did not perform a detailed study of specific ergonomic exposure of the occupation, but we took into consideration only the type of occupation at the onset of the symptoms as surrogate of occupational risk factor for CTS according to a checklist of the job-titles considered at risk [21]. We did not stratify obese subjects into three common subgroups (obesity class 1: 30–34.9, severe obesity class 2: 35–39.9 and extreme obesity class 3: ≥ 40) [14], because the number of cases/controls in some subgroups was too small to obtain reliable statistical results. Therefore, for all analyses, obesity was considered for BMI ≥ 30. The enrolled CTS patients had clinical symptoms and two electrophysiological abnormalities of which at least one comparative test or short segment conduction, because the abnormality of only one test could be random. Two or more abnormal EDX tests greatly reduced the probability of including false positive cases of CTS [18]. Use of comparison latencies in addition to absolute latency/conduction velocity can improve sensitivity and specificity and helps to control for other confounding variables such as temperature, age, gender, height, hand size and other patient-specific variability [43]. Therefore our inclusion criteria were restrictive (coexistence of symptoms and electrophysiological abnormalities). Some cases with mild CTS symptoms and without two abnormal EDX tests were not included in this study, also subjects with asymptomatic delay of the median nerve were excluded from cases and controls, but the overall number was negligible (less than 10%).
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
211
Table 3 Univariate and multivariate analyses regarding body mass index (BMI) and waist circumference (WC) crossed categories as risk factors for CTS among females (3a) and males (3b). 3a Females = no. 724
BMI b25 and normal WC BMI b25 and high WC BMI b 25 with very high WC Overweight with normal WC Overweight with high WC Overweight with very high WC Obesity with normal WC Obesity with high WC Obesity with very high WC
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
57 28 13 3 33 75 – – 41
180 75 27 23 48 77 – – 44
1.0 1.17 (0.69–1.99) 1.52 (0.73–3.15) 0.41 (0.11–1.43) 2.17 (1.26–3.73) 3.07 (1.95–4.83) – – 2.94 (1.72–5.02)
– 1.42 (0.79–2.54) 1.42 (0.61–3.27) 0.36 (0.09–1.40) 2.06 (1.12–3.80) 2.56 (1.57–4.18) – – 2.57 (1.43–4.64)
Univariate
Multivariatea
3b Males = no. 393
BMI b25 and normal WC BMI b25 with high WC BMI b 25 with very high WC Overweight with normal WC Overweight with high WC Overweight with very high WC Obesity with normal WC Obesity with high WC Obesity with very high WC a
Cases (no. = 120)
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
19 4 0 21 24 21 0 2 29
92 8 4 45 63 25 1 5 30
1.0 2.42 (0.65–9.00) – 2.26 (1.09–4.68) 1.84 (0.93–3.67) 4.07 (1.83–9.06) – 1.94 (0.35–10.86) 4.68 (2.20–9.98)
– 1.95 (0.47–8.15) – 2.11 (0.94–4.73) 1.27 (0.58–2.79) 2.39 (1.00–5.70) – 1.44 (0.20–10.25) 3.75 (1.64–8.58)
Multivariate unconditional logistic regression model, OR was adjusted for age, enrolment centre, occupational status, WR, and HR.
Convenience controls were enrolled among all the consecutive patients with upper limb complaints, other than CTS, instead of healthy subjects. Although we performed “standard” EDX to confirm the absence of CTS in the controls, a possible enrolment bias could be due to the existence of risk factors (for upper limb disorders other than CTS) similar to risk factors for CTS. On the other hand, some of the major risk factors for CTS, as well as wrist dimensions and hand/wrist movement repetition, are quite peculiar to this disease, and other risk factors, such as obesity, are yet studied in the case of other upper limb disorders, as in the ulnar neuropathy at the elbow [44]. Hence, this possible bias could have determined, at most, an underestimation of some our results.
Another enrolment bias may also be the site of patient recruitment. Our study is based on population from EMG labs. Even if our EMG labs are primary care outpatient services and examine only unselect patients, our cohort may be different than that captured through general practice and may be not representative of the general population. Using patient controls may have underestimated the association between obesity and CTS. A previous case–control study demonstrated that, using two different control groups, the occurrence of overweight/obesity for hospital-based controls was higher than that for population-based controls. The number of cases and controls of this study was small enough (38 men cases and 152 controls) [45]. Our
Table 4 Univariate and multivariate analyses regarding body mass index (BMI) and waist-to-hip ratio (WHR) crossed categories as risk factors for CTS among females (4a) and males (4b). 4a Females = no. 724
BMI b25 with normal WHR BMI b 25 with high WHR Overweight with normal WHR Overweight with high WHR Obesity with normal WHR Obesity with high WHR
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
68 30 38 73 9 32
225 57 73 75 6 38
1.0 1.74 (1.03–2.94) 1.72 (1.07–2.78) 3.22 (2.08–4.99) 4.96 (1.67–14.73) 2.79 (1.60–4.85)
– 2.11 (1.17–3.82) 1.44 (0.84–2.47) 2.91 (1.80–4.70) 5.83 (1.83–18.63) 2.28 (1.23–4.25)
Univariate
Multivariatea
4b Males = no. 393
BMI b25 with normal WHR BMI b 25 with high WHR Overweight with normal WHR Overweight with high WHR Obesity with normal WHR Obesity with high WHR a
Cases (no. = 120)
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
18 5 40 26 10 21
93 11 100 33 10 26
1.0 2.35 (0.72–7.69) 2.07 (1.10–3.89) 4.07 (1.91–8.69) 5.17 (1.78–14.96) 4.17 (1.87–9.33)
– 1.74 (0.47–6.47) 1.51 (0.74–3.06) 2.75 (1.20–2.27) 4.27 (1.37–13.28) 3.17 (1.29–7.77)
Multivariate unconditional logistic regression model, OR adjusted for age, enrolment centre, occupational status, WR, and HR.
212
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
Table 5 Univariate and multivariate analyses regarding waist circumference (WC) and waist-to-hip ratio (WHR) crossed categories as risk factors for CTS among females (5a) and males (5b). 5a Females = no. 724
Normal WC with normal WHR Normal WC with high WHR High WC with normal WHR High WC with high WHR Very high WC with normal WHR Very high WC with high WHR
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
57 3 36 25 22 107
190 13 86 37 28 120
1.0 0.77 (0.21–2.80) 1.40 (0.85–2.28) 2.25 (1.24–4.09) 2.62 (1.38–4.98) 2.97 (1.98–4.47)
– 1.59 (0.41–6.10) 1.50 (0.87–2.60) 2.86 (1.47–5.56) 2.35 (1.18–4.71) 2.72 (1.72–4.30)
Univariate
Multivariatea
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
133 5 56 20 14 45
1.0 5.64 (1.62–19.62) 1.37 (0.72–2.61) 2.22 (0.96–5.13) 4.61 (1.97–10.75) 3.05 (1.66–5.58)
– 3.83 (1.04–14.06) 0.99 (0.48–2.07) 1.54 (0.59–4.01) 2.95 (1.17–7.43) 2.33 (1.17–4.65)
5b Males = no. 393 Cases (no. = 120) Normal WC with normal WHR Normal WC with high WHR High WC with normal WHR High WC with high WHR Very high WC with normal WHR Very high WC with high WHR a
33 7 19 11 16 34
Multivariate unconditional logistic regression model, OR adjusted for age, enrolment centre, occupational status, WR, and HR.
study did not use hospital-based controls but outpatients referring to EMG labs.
Disclosures The authors state that there are no conflicts of interest and that they have not received any financial support.
4.2. Conclusions Abdominal adiposity measures (WC and WHR) are correlated with BMI, but the degree of association could vary, suggesting that these measures may provide different information and thus may not be interchangeable and it is advisable to obtain BMI and WC together and consider the utility of joint use of the two indicators.
Table 6 Univariate and multivariate analyses regarding hand ratio (HR) and wrist ratio (WR) as risk factors for CTS among females (6a) and males (6b). 6a Females = no. 724
Univariate
Multivariatea
Cases (no. = 250)
Controls (no. = 474)
OR (95% CI)
OR (95% CI)
Hand ratio (HR) N2.2 (not at risk) ≤2.2 (at risk)
138 112
343 131
1.0 2.13 (1.53–2.94)
1.0 2.68 (1.85–3.88)
Wrist ratio (WR) ≤0.7 (not at risk) N0.7 (at risk)
85 165
239 235
1.0 1.97 (1.43–2.73)
1.0 2.68 (1.85–3.87)
Males = no. 393
Univariate
Multivariatea
Cases (no. = 120)
Controls (no. = 273)
OR (95% CI)
OR (95% CI)
6b
Hand ratio (HR) N2.2 (not at risk) ≤2.2 (at risk)
31 89
160 113
1.0 4.07 (2.47–6.69)
1.0 3.69 (2.18–6.25)
Wrist ratio (WR) ≤0.7 (not at risk) N0.7 (at risk)
46 74
148 125
1.0 1.90 (1.22–2.97)
1.0 2.32 (1.36–3.95)
a Multivariate unconditional logistic regression model, OR adjusted for age, enrolment centre, occupational status, and BMI.
References [1] Stevens JC, Beard CM, O'Fallon WM, Kurland LT. Conditions associated with carpal tunnel syndrome. Mayo Clin Proc 1992;67:541–8. [2] Nordstrom DL, Vierkant RA, DeStefano F, Layde PM. Risk factors for carpal tunnel syndrome in a general population. Occup Environ Med 1997;54:734–40. [3] Karpitskaya Y, Novak CB, Mackinnon SE. Prevalence of smoking, obesity, diabetes mellitus, and thyroid disease in patients with carpal tunnel syndrome. Ann Plast Surg 1997;48:269–73. [4] Becker J, Nora DB, Gomes I, Stringari FF, Seitensus R, Panosso JS, et al. An evaluation of gender, obesity, age and diabetes mellitus as risk factors for carpal tunnel syndrome. Clin Neurophysiol 2002;113:1429–34. [5] Geoghegan JM, Clark DI, Bainbridge LC, Smith C, Hubbard R. Risk factors in carpal tunnel syndrome. J Hand Surg Br 2002;29:315–20. [6] Barcenilla A, March LM, Chen JS, Sambrook PN. Carpal tunnel syndrome and its relationship to occupation: a meta-analysis. Rheumatology (Oxford) 2012;51:250–61. [7] Pourmemari MH, Viikari-Juntura E, Shiri R. Smoking and carpal tunnel syndrome: a meta-analysis. Muscle Nerve Jun 12 2013. http://dx.doi.org/10.1002/mus.23922. [8] Johnson EW, Gatens T, Poindexter D, Bowers D. Wrist dimensions: correlation with median sensory latencies. Arch Phys Med Rehabil 1983;64:556–7. [9] Werner RA, Albers JW, Franzblau A, Armstrong TJ. The relationship between body mass index and the diagnosis of carpal tunnel syndrome. Muscle Nerve 1994;17:632–6. [10] Kouyoumdjian JA, Zanetta DM, Morita MP. Evaluation of age, body mass index, and wrist index as risk factors for carpal tunnel syndrome severity. Muscle Nerve 2002;25:93–7. [11] Boz C, Ozmenoglu M, Altunayoglu V, Velioglu S, Alioglu Z. Individual risk factors for carpal tunnel syndrome: an evaluation of body mass index, wrist index and hand anthropometric measurements. Clin Neurol Neurosurg 2004;106:294–9. [12] Moghtaderi A, Izadi S, Sharafadinzadeh N. An evaluation of gender, body mass index, wrist circumference and wrist ratio as independent risk factors for carpal tunnel syndrome. Acta Neurol Scand 2005;112:375–9. [13] Chiotis K, Dimisianos N, Rigopoulou A, Chrysanthopoulou A, Chroni E. Role of anthropometric characteristics in idiopathic carpal tunnel syndrome. Arch Phys Med Rehabil 2013;94:737–44. [14] World Health Organization. Waist circumference and waist–hip ratio, report of a WHO expert consultation. http://whqlibdoc.who.int/publications/2011/9789241501491_ eng.pdf; December 8–11 2008 . [Retrieved March 21, 2012]. [15] Rempel D, Evanoff B, Amadio PC, de Krom M, Franklin G, Franzblau A, et al. Consensus criteria for the classification of carpal tunnel syndrome in epidemiologic studies. Am J Public Health 2013;88:1447–51. [16] Quality Standards Subcommittee of the American Academy of Neurology. Practice parameter for carpal tunnel syndrome (summary statement). Neurology 1993;43:2406–9. [17] American Association of Electrodiagnostic Medicine, American Academy of Neurology, American Academy of Physical Medicine and Rehabilitation. Practice parameter
M. Mondelli et al. / Journal of the Neurological Sciences 338 (2014) 207–213
[18] [19]
[20] [21]
[22] [23]
[24]
[25]
[26] [27] [28] [29]
[30]
for electrodiagnostic studies in carpal tunnel syndrome: summary statement. Muscle Nerve 2002;25:918–22. Werner RA. Evaluation of work-related carpal tunnel syndrome. J Occup Rehabil 2006;16:207–22. Mondelli M, Baldasseroni A, Aretini A, Ginanneschi F, Padua L. Prevalent involvement of thenar motor fibres in vineyard workers with carpal tunnel syndrome. Clin Neurophysiol 2010;121:1251–5. Padua L, Pasqualetti P, Rosenbaum R. One patient, two carpal tunnels: statistical and clinical analysis — by hand or by patient? Clin Neurophysiol 2005;116:241–3. Colombini D, Occhipinti E, Cairoli S, Battevi N, Menoni O, Ricci MG, et al. Musculoskeletal conditions of the upper and lower limbs as an occupational disease: what kind and under what conditions. Consensus document of a national working group (ISPESL). Med Lav 2003;94:312–29. Chroni E, Paschalis C, Arvaniti C, Zotou K, Nikolakopoulou A, Papapetropoulos T. Carpal tunnel syndrome and hand configuration. Muscle Nerve 2001;24:1607–11. Kamolz LP, Beck H, Haslik W, Högler R, Rab M, Schrögendorfer KF, et al. Carpal tunnel syndrome: a question of hand and wrist configurations? J Hand Surg Br 2004;29:321–4. Stallings SP, Kasdan ML, Soergel TM, Corwin HM. A case–control study of obesity as a risk factor for carpal tunnel syndrome in a population of 600 patients presenting for independent medical examination. J Hand Surg Am 1997;22:211–5. Nathan PA, Keniston RC, Myers LD, Meadows KD. Obesity as a risk factor for slowing of sensory conduction of the median nerve in industry. A cross-sectional and longitudinal study involving 429 workers. J Occup Med 1992;34:379–83. Bland JD. The relationship of obesity, age, and carpal tunnel syndrome: more complex than was thought? Muscle Nerve 2005;32:527–32. Seror P, Seror R. Prevalence of obesity and obesity as a risk factor in patients with severe median nerve lesion at the wrist. Joint Bone Spine 2013;80:632–7. Cannon LJ, Bernacki EJ, Walter SD. Personal and occupational factors associated with carpal tunnel syndrome. J Occup Med 2013;23:255–8. Gorsche RG, Wiley JP, Renger RF, Brant RF, Gemer TY, Sasyniuk TM. Prevalence and incidence of carpal tunnel syndrome in a meat packing plant. Occup Environ Med 1999;56:417–22. Hakim AJ, Cherkas L, El Zayat S, MacGregor AJ, Spector TD. The genetic contribution to carpal tunnel syndrome in women: a twin study. Arthritis Rheum 2002;47:275–9.
213
[31] Farmer JE, Davis TR. Carpal tunnel syndrome: a case–control study evaluating its relationship with body mass index and hand and wrist measurements. J Hand Surg Eur Vol 2008;33:445–8. [32] Werner RA, Jacobson JA, Jamadar DA. Influence of body mass index on median nerve function, carpal canal pressure, and cross-sectional area of the median nerve. Muscle Nerve 2004;30:481–5. [33] Shiri R, Heliövaara M, Moilanen L, Viikari J, Liira H, Viikari-Juntura E. Associations of cardiovascular risk factors, carotid intima–media thickness and manifest atherosclerotic vascular disease with carpal tunnel syndrome. BMC Musculoskelet Disord Apr 26 2011;12:80. [34] Plastino M, Fava A, Carmela C, De Bartolo M, Ermio C, Cristiano D, et al. Insulin resistance increases risk of carpal tunnel syndrome: a case–control study. J Peripher Nerv Syst 2011;16:186–90. [35] Balci K, Utku U. Carpal tunnel syndrome and metabolic syndrome. Acta Neurol Scand 2007;116:113–7. [36] Nakamichi K, Tachibana S. Hypercholesterolemia as a risk factor for idiopathic carpal tunnel syndrome. Muscle Nerve 2005;32:364–7. [37] Dekel S, Coates R. Primary carpal stenosis as a cause of ‘idiopathic’ carpal-tunnel syndrome. Lancet 1979;2(8150):1024. [38] Bleecker ML, Bohlman M, Moreland R, Tipton A. Carpal tunnel syndrome: role of carpal canal size. Neurology 1985;35:1599–604. [39] Gordon C, Johnson EW, Gatens PF, Ashton JJ. Wrist ratio correlation with carpal tunnel syndrome in industry. Am J Phys Med Rehabil 1988;67:270–2. [40] Radecki P. A gender specific wrist ratio and the likelihood of a median nerve abnormality at the carpal tunnel. Am J Phys Med Rehabil 1994;73:157–62. [41] Nakamichi K, Tachibana S. Small hand as a risk factor for idiopathic carpal tunnel syndrome. Muscle Nerve 1994;18:664–6. [42] Lim PG, Tan S, Ahmad TS. The role of wrist anthropometric measurement in idiopathic carpal tunnel syndrome. J Hand Surg Eur Vol 2008;33:645–7. [43] Werner RA, Andary M. Electrodiagnostic evaluation of carpal tunnel syndrome. Muscle Nerve 2008;44:597–607. [44] Frost P, Johnsen B, Fuglsang-Frederiksen A, Svendsen SW. Lifestyle risk factors for ulnar neuropathy and ulnar neuropathy-like symptoms. Muscle Nerve 2013;48:507–15. [45] Wieslander G, Norbäck D, Göthe CJ, Juhlin L. Carpal tunnel syndrome (CTS) and exposure to vibration, repetitive wrist movements, and heavy manual work: a casereferent study. Occup Environ Med 1989;46:43–7.