Clinical Biomechanics 30 (2015) 546–550
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Computer keyboarding biomechanics and acute changes in median nerve indicative of carpal tunnel syndrome Kevin K. Toosi a,b, Nathan S. Hogaboom b,c, Michelle L. Oyster b,d, Michael L. Boninger b,d,⁎ a
Department of Bioengineering, University of Pittsburgh, Pittsburgh, PA, USA Human Engineering Research Laboratories, VA Pittsburgh Healthcare System, Pittsburgh, PA, USA Department of Rehabilitation Science and Technology, University of Pittsburgh, Pittsburgh, PA, USA d Department of Physical Medicine and Rehabilitation, University of Pittsburgh, Pittsburgh, PA, USA b c
a r t i c l e
i n f o
Article history: Received 8 August 2014 Accepted 16 April 2015 Keywords: Carpal tunnel syndrome Computer keyboarding biomechanics Ultrasound Ergonomics
a b s t r a c t Background: Carpal tunnel syndrome is a common and costly peripheral neuropathy. Occupations requiring repetitive, forceful motions of the hand and wrist may play a role in the development of carpal tunnel syndrome. Computer keyboarding is one such task, and has been associated with upper-extremity musculoskeletal disorder development. The purpose of this study was to determine whether continuous keyboarding can cause acute changes in the median nerve and whether these changes correlate with wrist biomechanics during keyboarding. Methods: A convenience sample of 37 healthy individuals performed a 60-minute typing task. Ultrasound images were collected at baseline, after 30 and 60 min of typing, then after 30 min of rest. Kinematic data were collected during the typing task. Variables of interest were median nerve cross-sectional area, flattening ratio, and swelling ratio at the pisiform; subject characteristics (age, gender, BMI, wrist circumference, typing speed) and wrist joint angles. Findings: Cross-sectional area and swelling ratio increased after 30 and 60 min of typing, and then decreased to baseline after 30 min of rest. Peak ulnar deviation contributed to changes in cross-sectional area after 30 min of typing. Interpretation: Results from this study confirmed a typing task causes changes in the median nerve, and changes are influenced by level of ulnar deviation. Furthermore, changes in the median nerve are present until cessation of the activity. While it is unclear if these changes lead to long-term symptoms or nerve injury, their existence adds to the evidence of a possible link between carpal tunnel syndrome and keyboarding. Published by Elsevier Ltd.
1. Introduction Carpal tunnel syndrome (CTS) is the most widely reported nerve compression disorder (Werner and Andary, 2002) with symptoms, including numbness, tingling, pain, weakness, and fatigue in the affected hand, present in approximately 5% of the United States general population (Kwon et al., 2008), with as many as 90 per 100,000 requiring surgical intervention (Jerosch-Herold et al., 2014). According to Dartmouth–Hitchcock Medical Center, surgical treatment of carpal tunnel could cost more than $7000 per hand (http://www.dartmouthhitchcock.org). In addition, the income loss per CTS patient over a period of 6 years was estimated at $45,000–89,000 compared with controls (Foley et al., 2007). CTS symptoms can disturb the ability to perform work-related activities, potentially resulting in work disability (Turner et al., 2007).
⁎ Corresponding author at: Human Engineering Research Laboratories, 6425 Penn Avenue, Suite 400, Pittsburgh, PA 15206, USA. E-mail address: [email protected] (M.L. Boninger).
http://dx.doi.org/10.1016/j.clinbiomech.2015.04.008 0268-0033/Published by Elsevier Ltd.
The most commonly accepted theory describing pathogenesis of CTS is chronic compression of the median nerve within the carpal tunnel. Upon conducting a systemic review of studies of computer work and carpal tunnel syndrome, Thomsen et al. (2008) concluded that biomechanical factors, such as forceful exertions, repetition, and awkward postures, increase the risk of CTS by increasing carpal tunnel pressure, resulting in median nerve ischemia. Some investigators have suggested that the pathologic changes of the subsynovial connective tissue, including noninflammatory fibrosis and thickening, may also be a cause of carpal tunnel syndrome (Greening et al., 2001). Common methods for diagnosing CTS are typically limited to physical examinations or nerve conduction. Ultrasound is currently being explored as a cost-effective imaging study due to its non-invasiveness, precision, and accuracy (Beekman and Visser, 2003); a recent metaanalysis found ultrasound to have a high specificity (86.8%) and sensitivity (77.6%) in the diagnosis of CTS (Fowler et al., 2011). Unlike nerve conduction (Werner et al., 1997a,b), ultrasound has the potential to predict future CTS and identify risk factors. Quantitative ultrasound (QUS) techniques have been developed to quantify the immediate effects of an activity on nerve size and shape (Altinok et al., 2004;
K.K. Toosi et al. / Clinical Biomechanics 30 (2015) 546–550
Impink et al., 2010). Using this technique, direct relationships between nerve changes indicative of CTS and a given activity can be established. Epidemiological evidence suggests occupation plays a role in the development of CTS, with higher risk associated with jobs requiring prolonged repetitive and forceful motions, awkward or static postures, localized mechanical compression, and vibration (Palmer et al., 2007; Punnett, and and Wegman 2004; Roquelaure et al., 1997; Silverstein et al., 1987; van Rijn et al., 2009). While keyboarding is a highly repetitive task requiring small but forceful movements of the fingers, the evidence is conflicting as to whether there is a relationship between computer use and CTS (Gerr et al., 2006). Keyboard users exhibit a variety of preferred postures (Simoneau et al., 1999) that remain stable when typing (Baker et al., 2007) and are dependent on physical characteristics of the workstation (Marklin et al., 1999). Awkward typing postures have been identified as a risk factor for development of CTS symptoms; specifically, greater degrees of ulnar wrist deviation (Hunting et al., 1981) and migration from a neutral flexion/extension position (Simoneau and Marklin, 2001). In this study we aim to confirm our previous results, which indicated acute increase in size of the median nerve at the inlet of the carpal tunnel after keyboarding (Toosi et al., 2011). This paper involves a larger sample size and investigating changes that occur after a brief resting period, and also incorporate biomechanical variables and subject characteristics, which were chosen based on risk factors identified in the literature (Gelberman et al., 1981; Hunting et al., 1981; Silverstein et al., 1987). It was hypothesized that following 30 and 60 min of typing the median nerve will show increased cross-sectional area (CSA) and flattening ratio (FR) at the pisiform level, and an increased swelling ratio (SR) and that these same variables would then decrease to baseline after 30 min of rest. Increased CSA, FR and SR are associated with CTS (Beekman and Visser, 2003). It was also hypothesized that typing biomechanics (average peak ulnar deviation of the wrist, average peak wrist extension) and subject characteristics (BMI, gender, age, wrist circumference, typing speed) would predict median nerve CSA, SR, and FR after 30 and 60 min of typing. 2. Methods 2.1. Subjects A convenience sample of healthy individuals was recruited to participate in the study. Participants were included if they 1) were between 18 and 65 years old, 2) spoke English, 3) were self-report proficient typists (i.e., typing at least 40 words per minute), 4) used a computer keyboard at least 3 h a day, 4 days a week, and 5) typed using at least 8 fingers. Participants were excluded if they 1) had a history of wrist surgery or fracture, 2) self-reported or presented clinical condition that mimics CTS or interferes with its evaluation (e.g. proximal median neuropathy, cervical radiculopathy, or polyneuropathy), 3) selfreported history of CTS signs or symptoms, or 4) had a history of underlying disorders associated with CTS (e.g. diabetes mellitus, rheumatoid arthritis, acromegaly, or hypothyroidism). These criteria were chosen to eliminate potential confounders and reduce error when evaluating CTS symptoms. Participants were asked to refrain from intense physical activity such as sports, exercise, repetitive forceful arm tasks like yard work for 48 h, and vigorous typing, for 12 h prior to participation in this study since intense physical activity may affect the baseline measurements. The study was approved by the appropriate Institutional Review Board and subjects provided written informed consent prior to data collection.
547
demographic and hand and wrist pain questionnaires. Bi-lateral wrist circumference and length of finger segments were also measured. Participants were asked to complete two 30-minute typing tasks. Ultrasound images were collected using a Philips HD11 XE ultrasound machine with a 5–12 MHz 50 mm linear array transducer (Philips Medical Systems, Bothell, WA, USA) by a single investigator at four time points: before the first typing task began (baseline), immediately following the first and second typing task, and 30 min after cessation of the second typing task. Bilateral images were taken of the carpal tunnel at the distal radius and pisiform levels using a previously described technique (Impink et al., 2010), with primary emphasis on the median nerve. Keyboard used during the typing tasks was an L100 Dell keyboard (Dell, Inc., Round Rock, TX, USA), which was set in a flat position parallel to the table; set-up was held constant between participants. The keyboarding task was performed using an electronic keyboarding program, Typing Master Pro™ (Typing Master Finland, Inc., Helsinki, Finland), which presents a typing task for the keyboard user on the computer screen. All participants typed the same text at their normal rate and were instructed not to correct errors. Alternate input devices such as a mouse were not used. Kinematic data were collected using an Optotrack motion capture system (Northern Digital, Inc., Waterloo, Ontario, Canada). Hand, wrist, and finger movements were derived via tracking of active markers positioned on the dorsal surface of the dominant hand (Fig. 1). Sixty seconds of kinematic data were simultaneously collected at 5 and 25 min after typing began in each 30-minute session, using a sampling frequency of 60 Hz, providing total of four measurements. Data were not collected throughout the entire test as individual typing biomechanics remain relatively stable during typing tasks (Baker et al., 2007). 2.3. Data analysis 2.3.1. Ultrasound images An interactive image analysis program was developed to measure the median nerve at the levels of the distal radius and pisiform using the ultrasound images collected (Impink et al., 2010). Cross-sectional areas of the median nerve were determined by performing a boundary trace at each level, and were used to calculate SR (Eq. (1)), while major (or mediolateral) and minor (or anteroposterior) axes of the nerve were measured to calculate FR (Eq. (2)). CSA, SR and FR have all been found to be reliable measures related to CTS (Beekman and Visser, 2003). One investigator analyzed all images and was blinded to the subject and
2.2. Study procedure Participants were tested in the morning to limit the amount of activity performed on the day of testing. Participants completed
Fig. 1. Active markers affixed to the dorsal aspect of the dominant hand during the typing task.
548
K.K. Toosi et al. / Clinical Biomechanics 30 (2015) 546–550
time point. This technique has high reliability when a single investigator collects and analyzes all images (Impink et al., 2010). CSApisiform ¼ SR: CSAradius
ð1Þ
Lengthmajor ¼ FR: Lengthminor
ð2Þ
2.3.2. Keyboarding biomechanics Gross typing speed, adjusted typing speed, and accuracy were automatically collected throughout both 30-minute typing tasks using the electronic keyboarding program. Gross speed was defined as the speed at which the subject typed, and adjusted speed as gross speed minus any typing errors. The average gross typing speed (in words per minute) and typing accuracy (percentage of words typed correctly) were calculated for each 30-minute typing task, and averaged between the two typing tasks in the final analysis. Hand and wrist joint angles and ranges of motion (RoM) were calculated using a local coordinate system created with the active markers placed on the subject's dominant hand, wrist, and forearm. Kinematic variables of interest were peak and average wrist flexion/extension angles, peak and average ulnar/radial deviations, and RoM for the wrist and middle, index, and ring fingers. Joint angles were derived as described by Baker et al. (2007). Metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints were considered neutral (0°) when the metacarpal bones and phalanges formed a straight line, respectively; thus, ulnar deviation and wrist extension angles had negative values, and flexion and radial deviation were positive. The hand and forearm were modeled as rigid segments — their orientations were derived by tracking their local coordinate systems constructed on three hand and three forearm markers (Fig. 2). MCP and PIP joints were assumed to be joints with two degrees of freedom. MCP and PIP flexion/ extension and adduction/abduction angles were computed using the 3D angles between the corresponding metacarpals and proximal phalanges, or proximal and intermediate phalange vectors.
using standard summaries (e.g. means, standard deviations, percentiles, ranges) and graphical techniques (e.g. histograms, scatter plots). Normality was assessed using the Shapiro–Wilk test. A two-way factorial repeated-measures ANOVA was used to detect significant differences in ultrasound variables between dominant and non-dominant wrists at all four time points. Pair-wise multiple-comparison procedures were performed using a Bonferroni adjustment if the omnibus F-test was significant. Kinematic data were collected only on dominant arms; so images of the median nerve in dominant wrists were used in regression and correlation analyses. Subject characteristics of interest for correlation and regression were age, gender, BMI, typing speed, and wrist circumference; biomechanical parameters of interest included peak ulnar deviation and peak wrist extension. Wrist circumference moderately correlated with CSA (r = 0.428, P = 0.008), a measure that influences SR; thus, any correlation involving either CSA or SR included wrist circumference as a covariate to reduce confounding effects. Pearson's r was calculated to test correlations between baseline SR and CSA; subject characteristics of BMI, typing speed, and gender; and the biomechanical variable peak extension. Spearman's ρ was calculated for correlation analyses including baseline FR and peak ulnar deviation as these data were not normally distributed. Multiple linear regression models were developed to test whether wrist biomechanics or subject characteristics could predict median nerve measures after 30 and 60 min of typing. Dependent variables of interest were CSA, FR, and SR at 30- and 60-minute time points. Independent variables of interest were baseline values of median nerve measures, subject characteristics, and biomechanical variables. Models were then built to predict the value of the three median nerve measures at 30- and 60-minute time points as a function of their baseline values, biomechanics, and subject characteristic variables listed above. Wrist circumference was correlated with BMI and gender (r = 0.656 and r = − 0.603, respectively; both significant at the 0.01 level); thus, it was excluded in the regression analyses to reduce confounding effects and meet regression assumptions.
3. Results 2.4. Statistical analysis 3.1. Subjects Significance was set a priori to P b 0.05. Bonferonni corrections were applied for each median nerve measure used in the ANOVA, correlation, and regression models (α = 0.05/3 = 0.017). Statistical analyses were performed using SPSS version 21.0 (SPSS Inc., Chicago, IL, USA). All analyses were preceded by a detailed descriptive analysis of the data
Forty subjects were enrolled in the study, 37 of which were used in the final analysis. Data from one male and one female participant were excluded due to median nerve bifurcation; measurements can be influenced by the volume of blood in the bifurcating vessel and yield an inaccurate depiction of the nerve. Another female participant self-reported having symptoms of CTS and thus was excluded. Demographic variables are summarized in Table 1. The only baseline median nerve measure to correlate with subject characteristics was SR, which negatively correlated with age (r = −0.415, P = 0.012). No subject characteristics predicted changes in median nerve measures as a result of keyboarding.
Table 1 Demographic and subject characteristics; mm = millimeters; wpm = words per minute.
Fig. 2. Marker location on the dorsal aspect of the hand and forearm (circles) and the hand and forearm planes (lines and triangles) that were defined by their local coordinate systems.
Variable
Mean (SD) or N
Range or %
Age BMI Wrist circumference (mm) Gross speed (wpm) Accuracy Right handed Males
28.89 (6.6) 25.1 (4.6) 163.0 (13.8) 55.0 (10.9) 89.7 (4.8) 30 18
19–46 18.1–41.6 136–192 38–78 80–98 81.1% 48.6%
K.K. Toosi et al. / Clinical Biomechanics 30 (2015) 546–550
3.2. Ultrasound median nerve measures Baseline median nerve measures were highly correlated to measures at 30 and 60 minute time points (P b 0.01). CSA and SR significantly increased after 30 and 60 min of typing (P b 0.01). No differences existed in CSA or SR between 30- and 60-minute time points, nor between baseline and 90-minutes (P N 0.05). FR did not significantly change at any point during the task, and no differences were found in median nerve measures between dominant and non-dominant wrists (P N 0.05). Descriptive and ANOVA statistics for median nerve measures, including main effects and contrasts, are presented in Table 2. 3.3. Wrist and hand biomechanics Biomechanical variables (Table 3) did not correlate with baseline median nerve measures. When entered into linear-regression models, peak ulnar deviation significantly predicted CSA after 30 min of typing (β = −0.186, P = 0.017). Degrees of ulnar and radial deviation were negative and positive, respectively, which account for the negative beta value. No biomechanical variables significantly predicted other median nerve measures at any time point. 4. Discussion We previously demonstrated that a continuous keyboarding task caused acute changes in the ultrasonographic measures of the median nerve (Toosi et al., 2011); results of the current study were consistent with previous findings. The present study was able to quantify changes in median nerve size and shape in response to keyboarding and further relate those changes to hand/wrist biomechanics. It was hypothesized that changes in median nerve parameters previously linked to CTS (Beekman and Visser, 2003) would be observed after 30 and 60 min of keyboarding. Furthermore, keyboarding biomechanics and subject characteristics would influence these changes. In accordance with these hypotheses, CSA and SR increased after typing and returned to baseline after rest, and higher peak ulnar deviation angles predicted greater changes in median nerve measures after typing. CSA and SR have been highlighted as among the highest sensitivity and specificity when diagnosing CTS (Altinok et al., 2004). The changes in median nerve persisted over the two time points that we measured during the 60-minute typing task, and then returning to baseline when measured 30 min after stopping. We believe that the changes observed are likely a result of compression of the nerve during typing. Compression during long periods of activity, such as typing, can negatively affect the health of the median nerve and potentially
549
Table 3 Average and peak wrist angles (degrees) during typing, presented as means and standard deviations, with associated ranges. Wrist angles (°)
Mean (SD)
Range
Average flexion/extension Average ulnar/radial deviation Average peak flexion Average peak extension Average peak ulnar deviation Average peak radial deviation
−22.0 (11.2) −8.2 (11.6) −6.0 (11.3) −34.7 (13.1) −23.4 (9.4) 7.3 (6.3)
−47.3–4.6 −29.3–21.0 −28.59–17.4 −70.1 to −3.3 −38.7–51.24 0.4−21.6
contribute to the development of CTS (Barr and Barbe, 2002). It is possible that typing for long periods of time predisposes an individual to the development of CTS symptoms. Based on a return to baseline within 30 min it is possible that a resting period could be beneficial. There was no observable change in FR after 30 and 60 min of typing. Although the median nerve cross-sectional area at the pisiform level increased as a result of keyboarding, the two-dimensional nerve enlargement occurred in a proportional way. A potential reason for no change in FR is the location where the images were collected. Images in this study were limited to the pisiform level of the wrist, because our previous studies show less reliability at the hamate (Impink et al., 2010). Other studies have seen changes at the hamate and pisiform. Future investigations should include images at the hamate. In addition, it is possible that FR is less sensitive than CSA and SR for detecting shortterm changes, and probably could be used to investigate the effects of long-term tasks on the median nerve. Subjects with greater angles of ulnar deviation when typing experienced greater changes in their median nerves. Repetitive, forceful motions of the fingers – characteristic of typing – cause hypertrophy of flexor tendons within the carpal tunnel and raise carpal tunnel pressure (Werner and Andary, 2002). Performing the same motions while deviated from a neutral position amplifies the effect (Rempel et al. 1997; Werner et al., 1997a, 1997b). Most conventional keyboards are narrower than shoulder width of users. This naturally positions the wrist in ulnar deviation. A neutral wrist position can be achieved by using ergonomic technologies or workstation through alterations. In general, this study agrees with previous investigations that report keyboard users tend to type with extended and ulnarly-deviated wrists (Serina et al., 1999; Simoneau et al., 1999; Sommerich et al., 1996). Some of the biomechanics results did not match the stated hypotheses — no observable relationship existed between wrist flexion or extension and median nerve measures. Standardization of the workstation set-up likely altered natural typing biomechanics while standardization of the typing activity may have limited
Table 2 Cross-sectional area (CSA) in millimeters-squared (mm2), swelling ratio (SR), and flattening ratio (FR) in dominant (D) and non-dominant (N) hands at baseline, following 30 and 60 min of typing, and a 30-minute resting period after typing (i.e., 90-minute).
†
2
CSA (mm )
D N
SR‡
D N
FR
D N
†
Baseline mean [SD] (range)
30-minute⁎ mean [SD] (range)
60-minute⁎⁎ mean [SD] (range)
90-minute mean [SD] (range)
10.29 [1.79] (7.14–14.94) 10.47 [1.91] (6.57–14.46) 1.02 [0.14] (0.81–1.39) 1.05 [0.14] (0.86–1.41) 2.94 [0.76] (1.82–5.86) 2.77 [0.62] (1.64–4.28)
10.73 [2.12] (6.69–14.66) 10.91 [1.99] (7.17–15.34) 1.07 [0.17] (0.79–1.55) 1.12 [0.14] (0.86–1.48) 2.94 [0.65] (1.71–5.00) 2.84 [0.53] (2.03–6.26)
10.84 [1.81] (7.85–14.92) 10.69 [2.54] (7.31–15.22) 1.08 [0.16] (0.72–1.64) 1.10 [0.13] (0.87–1.47) 2.94 [0.67] (1.78–4.71) 2.80 [0.55] (2.03–4.41)
10.59 [1.77] (6.36–13.33) 10.29 [2.42] (7.50–14.19) 1.05 [0.16] (0.75–1.52) 1.06 [0.14] (0.85–1.37) 2.96 [0.74] (1.69–5.30) 2.88 [0.74] (1.83–4.99)
Significant main effect of typing on CSA (F(3,108) = 6.388, P = 0.001, η2p = .151). Significant main effect of typing on SR (F(3,108) = 7.970, P b 0.001, η2p = .181). ⁎ Significant increases from baseline in CSA (F(1,36) = 10.383, P = 0.003, η2p = .224) and SR (F(1,36) = 13.162, P = 0.001, η2p = .268) after 30 min of typing. ⁎⁎ Significant increases from baseline in CSA (F(1,36) = 15.293, P b 0.001, η2p = .298) and SR (F(1,36) = 14.252, P = 0.001, η2p = .284) after 60 min of typing. ‡
550
K.K. Toosi et al. / Clinical Biomechanics 30 (2015) 546–550
keystroke movement; ultimately, this would limit the amount of wrist flexion or extension. 4.1. Limitations The sample was limited to those without CTS symptoms. Our findings could change if symptomatic participants were included and could indicate that the median nerve changes were physiologic and may not lead to pathology. Future studies should include individuals with symptomatic CTS. The workstation setup was standardized to reduce variability, which could have elicited more uniform typing postures among participants. The use of a mouse was also eliminated to focus on effects of typing. It is likely that the use of pointing and typing devices in concert has a compound effect. Kinematic data were collected only on the dominant side while ultrasonographic measurements were collected on both. Typing is not a symmetrical task (Marklin et al., 1999) and thus by only collecting data on the dominant arm there is a risk that important information was missed. A control group, in which subjects held a similar wrist posture to typing, but avoided repetitive finger activity would have allow a stronger conclusion related to finger motion being responsible for the changes seen. 5. Conclusions This study serves to improve our understanding of median nerve changes after continuous keyboarding. CSA at the pisiform and SR as measured by ultrasound significantly increased after 30 and 60 min of keyboarding then returned to baseline after 30 min of rest. Greater changes were observed in those who typed with greater angles of ulnar deviation. While these changes may be part of a normal physiologic process, they also might represent part of the pathomechanics that leads to CTS. Further studies including individuals with CTS and following subjects longitudinally are needed. If acute median nerve changes as a result of activity are found to relate to risk of developing CTS, the methods used in this study could aid in the identification of risk to develop CTS and point to potential interventions. Acknowledgments This material is the result of work supported with resources and the use of facilities at the Human Engineering Research Laboratories, VA Pittsburgh Healthcare System, Pittsburgh, PA, USA. This study was supported by the U.S. Department of Veterans Affairs (B3142C) and the National Institutes of Health (T32HD049307). References Altinok, M.T., Baysal, O., Karakas, H.M., Firat, A.K., 2004. Sonographic evaluation of the carpal tunnel after provocative exercises. J. Ultrasound Med. 23, 1301–1306. Baker, N.A., Cham, R., Cidboy, E.H., Cook, J., Redfern, M.S., 2007. Kinematics of the fingers and hands during computer keyboard use. Clin. Biomech. 22, 34–43. Barr, A.E., Barbe, M.F., 2002. Pathophysiological tissue changes associated with repetitive movement: a review of the evidence. Phys. Ther. 82, 173–187. Beekman, R., Visser, L.H., 2003. Sonography in the diagnosis of carpal tunnel syndrome: a critical review of the literature. Muscle Nerve 27, 26–33.
Foley, M., Silverstein, B., Polissar, N., 2007. The economic burden of carpal tunnel syndrome: Long-term earnings of CTS claimants in Washington State. Am. J. Ind. Med. 50, 155–172. Fowler, J.R., Gaughan, J.P., Ilyas, A.M., 2011. The sensitivity and specificity of ultrasound for the diagnosis of carpal tunnel syndrome. Clin. Orthop. Relat. Res. 469, 1089–1094. Gelberman, R.H., Hergenroeder, P.T., Hargens, A.R., Lundborg, G.N., Akeson, W.H., 1981. The carpal tunnel syndrome: a study of carpal canal pressures. J. Bone Joint Surg. 63-A, 380–383. Gerr, F., Monteilh, C.P., Marcus, M.J., 2006. Keyboard use and musculoskeletal outcomes among computer users. Occup. Rehabil. 16, 265–277. Greening, J., Lynn, B., Leary, R., Warren, L., O'Higgins, P., Hall-Craggs, M., 2001. The use of ultrasound imaging to demonstrate reduced movement of the median nerve during wrist flexion in patients with non-specific arm pain. J. Hand Surg. (Br.) 26, 401–408. Hunting, W., Laubli, T., Grandjean, E., 1981. Postural and visual laods at VDT workplaces. I. Constrained postures. Ergonomics 24, 917–931. Impink, B.G., Gagnon, D., Collinger, J.L., Boninger, M.L., 2010. Repeatability of ultrasonographic median nerve measures. Muscle Nerve 41, 767–773. Jerosch-Herold, C., Shepstone, L., Wilson, E.C.F., Dyer, T., Blake, J., 2014. Clinical course, costs and predictive factors for response to treatment in carpal tunnel syndrome: the PALMS study protocol. BMC Musculoskelet. Disord. 15, 35. Kwon, B.C., Jung, K.I., Baek, G.H., 2008. Comparison of sonography and electrodiagnostic testing in the diagnosis of carpal tunnel syndrome. J. Hand. Surg. [Am.] 33, 65–71. Marklin, R.W., Simoneau, G.G., Monroe, J.F., 1999. Wrist and forearm posture from typing on split and vertically inclined computer keyboards. Hum. Factors 41, 559–569. Palmer, K.T., Harris, C., Coggon, D., 2007. Carpal tunnel syndrome and its relation to occupation: a systematic literature review. Occup. Med. 57, 57–66. Punnett, L., Wegman, D.H., 2004. Work-related musculoskeletal disorders: The epidemiologic evidence and the debate. J. Electromyogr. Kinesiol. 14, 13–23. Rempel, D., Keir, P.J., Smutz, W.P., Hargens, A., 1997. Effects of static fingertip loading on carpal tunnel pressure. J. Ortho. Res. 15, 422–426. Roquelaure, Y., Mechali, S., Dano, C., Fanello, S., Benetti, F., Bureau, D., et al., 1997. Occupational and personal risk factors for carpal tunnel syndrome in industrial workers. Scand. J. Work Environ. Health 23, 364–369. Serina, E.R., Tal, R., Rempel, D., 1999. Wrist and forearm postures and motions during typing. Ergonomics 42, 938–951. Silverstein, B.A., Fine, L.J., Armstrong, T.J., 1987. Occupational factors and carpal tunnel syndrome. Am. J. Ind. Med. 11, 343–358. Simoneau, G.G., Marklin, R.W., 2001. Effect of computer keyboard slope and height on wrist extension angle. Hum. Factors 43, 287–298. Simoneau, G., Marklin, R., Monroe, J., 1999. Wrist and forearm postures of users of conventional computer keyboards. Hum. Factors 41, 413–424. Sommerich, C.M., Marras, W.S., Parnianpour, M., 1996. A quantitative description of typing biomechanics. J. Occup. Rehabil. 6, 33–55. Thomsen, J.F., Gerr, F., Atroshi, I., 2008. Carpal tunnel syndrome and the use of computer mouse and keyboard: a systematic review. BMC Musculoskelet. Disord. 9, 134 (Accessed online: January 17, 2015. http://www.medscape.com/viewarticle/ 583079_3). Toosi, K.K., Impink, B.G., Baker, N.A., Boninger, M.L., 2011. Effects of computer keyboarding on ultrasonographic measures of the median nerve. Am. J. Ind. Med. 54, 826–833. Turner, J.A., Franklin, G., Fulton-Kehoe, D., Sheppard, L., Wickizer, T.M., Wu, R., et al., 2007. Early predictors of chronic work disability associated with carpal tunnel syndrome: a longitudinal workers' compensation cohort study. Am. J. Ind. Med. 50, 489–500. van Rijn, R.M., Huisstede, B.M., Koes, B.W., Burdorf, A., 2009. Associations between workrelated factors and the carpal tunnel syndrome–A systematic review. Scand. J. Work Environ. Health. 35, 19–36. Werner, R.A., Andary, M., 2002. Carpal tunnel syndrome: pathophysiology and clinical neurophysiology. J. Clin. Neurophysiol. 113, 1373–1381. Werner, R., Armstrong, T.J., Bir, C., Aylard, M.K., 1997a. Intracarpal canal pressures: the role of finger, hand, wrist, and forearm position. Clin. Biomech. 12, 44–51. Werner, R.A., Franzblau, A., Albers, J.W., Buchele, H., Armstrong, T.J., 1997b. Use of screening nerve conduction studies for predicting future carpal tunnel syndrome. Occup. Environ. Med. 54, 96–100. www.dartmouth-hitchcock.org (Accessed online: January 25, 2015).
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Computer keyboarding biomechanics and acute changes in median nerve indicative of carpal tunnel syndrome Kevin K. Toosi a,b, Nathan S. Hogaboom b,c, Michelle L. Oyster b,d, Michael L. Boninger b,d,⁎ a
Department of Bioengineering, University of Pittsburgh, Pittsburgh, PA, USA Human Engineering Research Laboratories, VA Pittsburgh Healthcare System, Pittsburgh, PA, USA Department of Rehabilitation Science and Technology, University of Pittsburgh, Pittsburgh, PA, USA d Department of Physical Medicine and Rehabilitation, University of Pittsburgh, Pittsburgh, PA, USA b c
a r t i c l e
i n f o
Article history: Received 8 August 2014 Accepted 16 April 2015 Keywords: Carpal tunnel syndrome Computer keyboarding biomechanics Ultrasound Ergonomics
a b s t r a c t Background: Carpal tunnel syndrome is a common and costly peripheral neuropathy. Occupations requiring repetitive, forceful motions of the hand and wrist may play a role in the development of carpal tunnel syndrome. Computer keyboarding is one such task, and has been associated with upper-extremity musculoskeletal disorder development. The purpose of this study was to determine whether continuous keyboarding can cause acute changes in the median nerve and whether these changes correlate with wrist biomechanics during keyboarding. Methods: A convenience sample of 37 healthy individuals performed a 60-minute typing task. Ultrasound images were collected at baseline, after 30 and 60 min of typing, then after 30 min of rest. Kinematic data were collected during the typing task. Variables of interest were median nerve cross-sectional area, flattening ratio, and swelling ratio at the pisiform; subject characteristics (age, gender, BMI, wrist circumference, typing speed) and wrist joint angles. Findings: Cross-sectional area and swelling ratio increased after 30 and 60 min of typing, and then decreased to baseline after 30 min of rest. Peak ulnar deviation contributed to changes in cross-sectional area after 30 min of typing. Interpretation: Results from this study confirmed a typing task causes changes in the median nerve, and changes are influenced by level of ulnar deviation. Furthermore, changes in the median nerve are present until cessation of the activity. While it is unclear if these changes lead to long-term symptoms or nerve injury, their existence adds to the evidence of a possible link between carpal tunnel syndrome and keyboarding. Published by Elsevier Ltd.
1. Introduction Carpal tunnel syndrome (CTS) is the most widely reported nerve compression disorder (Werner and Andary, 2002) with symptoms, including numbness, tingling, pain, weakness, and fatigue in the affected hand, present in approximately 5% of the United States general population (Kwon et al., 2008), with as many as 90 per 100,000 requiring surgical intervention (Jerosch-Herold et al., 2014). According to Dartmouth–Hitchcock Medical Center, surgical treatment of carpal tunnel could cost more than $7000 per hand (http://www.dartmouthhitchcock.org). In addition, the income loss per CTS patient over a period of 6 years was estimated at $45,000–89,000 compared with controls (Foley et al., 2007). CTS symptoms can disturb the ability to perform work-related activities, potentially resulting in work disability (Turner et al., 2007).
⁎ Corresponding author at: Human Engineering Research Laboratories, 6425 Penn Avenue, Suite 400, Pittsburgh, PA 15206, USA. E-mail address: [email protected] (M.L. Boninger).
http://dx.doi.org/10.1016/j.clinbiomech.2015.04.008 0268-0033/Published by Elsevier Ltd.
The most commonly accepted theory describing pathogenesis of CTS is chronic compression of the median nerve within the carpal tunnel. Upon conducting a systemic review of studies of computer work and carpal tunnel syndrome, Thomsen et al. (2008) concluded that biomechanical factors, such as forceful exertions, repetition, and awkward postures, increase the risk of CTS by increasing carpal tunnel pressure, resulting in median nerve ischemia. Some investigators have suggested that the pathologic changes of the subsynovial connective tissue, including noninflammatory fibrosis and thickening, may also be a cause of carpal tunnel syndrome (Greening et al., 2001). Common methods for diagnosing CTS are typically limited to physical examinations or nerve conduction. Ultrasound is currently being explored as a cost-effective imaging study due to its non-invasiveness, precision, and accuracy (Beekman and Visser, 2003); a recent metaanalysis found ultrasound to have a high specificity (86.8%) and sensitivity (77.6%) in the diagnosis of CTS (Fowler et al., 2011). Unlike nerve conduction (Werner et al., 1997a,b), ultrasound has the potential to predict future CTS and identify risk factors. Quantitative ultrasound (QUS) techniques have been developed to quantify the immediate effects of an activity on nerve size and shape (Altinok et al., 2004;
K.K. Toosi et al. / Clinical Biomechanics 30 (2015) 546–550
Impink et al., 2010). Using this technique, direct relationships between nerve changes indicative of CTS and a given activity can be established. Epidemiological evidence suggests occupation plays a role in the development of CTS, with higher risk associated with jobs requiring prolonged repetitive and forceful motions, awkward or static postures, localized mechanical compression, and vibration (Palmer et al., 2007; Punnett, and and Wegman 2004; Roquelaure et al., 1997; Silverstein et al., 1987; van Rijn et al., 2009). While keyboarding is a highly repetitive task requiring small but forceful movements of the fingers, the evidence is conflicting as to whether there is a relationship between computer use and CTS (Gerr et al., 2006). Keyboard users exhibit a variety of preferred postures (Simoneau et al., 1999) that remain stable when typing (Baker et al., 2007) and are dependent on physical characteristics of the workstation (Marklin et al., 1999). Awkward typing postures have been identified as a risk factor for development of CTS symptoms; specifically, greater degrees of ulnar wrist deviation (Hunting et al., 1981) and migration from a neutral flexion/extension position (Simoneau and Marklin, 2001). In this study we aim to confirm our previous results, which indicated acute increase in size of the median nerve at the inlet of the carpal tunnel after keyboarding (Toosi et al., 2011). This paper involves a larger sample size and investigating changes that occur after a brief resting period, and also incorporate biomechanical variables and subject characteristics, which were chosen based on risk factors identified in the literature (Gelberman et al., 1981; Hunting et al., 1981; Silverstein et al., 1987). It was hypothesized that following 30 and 60 min of typing the median nerve will show increased cross-sectional area (CSA) and flattening ratio (FR) at the pisiform level, and an increased swelling ratio (SR) and that these same variables would then decrease to baseline after 30 min of rest. Increased CSA, FR and SR are associated with CTS (Beekman and Visser, 2003). It was also hypothesized that typing biomechanics (average peak ulnar deviation of the wrist, average peak wrist extension) and subject characteristics (BMI, gender, age, wrist circumference, typing speed) would predict median nerve CSA, SR, and FR after 30 and 60 min of typing. 2. Methods 2.1. Subjects A convenience sample of healthy individuals was recruited to participate in the study. Participants were included if they 1) were between 18 and 65 years old, 2) spoke English, 3) were self-report proficient typists (i.e., typing at least 40 words per minute), 4) used a computer keyboard at least 3 h a day, 4 days a week, and 5) typed using at least 8 fingers. Participants were excluded if they 1) had a history of wrist surgery or fracture, 2) self-reported or presented clinical condition that mimics CTS or interferes with its evaluation (e.g. proximal median neuropathy, cervical radiculopathy, or polyneuropathy), 3) selfreported history of CTS signs or symptoms, or 4) had a history of underlying disorders associated with CTS (e.g. diabetes mellitus, rheumatoid arthritis, acromegaly, or hypothyroidism). These criteria were chosen to eliminate potential confounders and reduce error when evaluating CTS symptoms. Participants were asked to refrain from intense physical activity such as sports, exercise, repetitive forceful arm tasks like yard work for 48 h, and vigorous typing, for 12 h prior to participation in this study since intense physical activity may affect the baseline measurements. The study was approved by the appropriate Institutional Review Board and subjects provided written informed consent prior to data collection.
547
demographic and hand and wrist pain questionnaires. Bi-lateral wrist circumference and length of finger segments were also measured. Participants were asked to complete two 30-minute typing tasks. Ultrasound images were collected using a Philips HD11 XE ultrasound machine with a 5–12 MHz 50 mm linear array transducer (Philips Medical Systems, Bothell, WA, USA) by a single investigator at four time points: before the first typing task began (baseline), immediately following the first and second typing task, and 30 min after cessation of the second typing task. Bilateral images were taken of the carpal tunnel at the distal radius and pisiform levels using a previously described technique (Impink et al., 2010), with primary emphasis on the median nerve. Keyboard used during the typing tasks was an L100 Dell keyboard (Dell, Inc., Round Rock, TX, USA), which was set in a flat position parallel to the table; set-up was held constant between participants. The keyboarding task was performed using an electronic keyboarding program, Typing Master Pro™ (Typing Master Finland, Inc., Helsinki, Finland), which presents a typing task for the keyboard user on the computer screen. All participants typed the same text at their normal rate and were instructed not to correct errors. Alternate input devices such as a mouse were not used. Kinematic data were collected using an Optotrack motion capture system (Northern Digital, Inc., Waterloo, Ontario, Canada). Hand, wrist, and finger movements were derived via tracking of active markers positioned on the dorsal surface of the dominant hand (Fig. 1). Sixty seconds of kinematic data were simultaneously collected at 5 and 25 min after typing began in each 30-minute session, using a sampling frequency of 60 Hz, providing total of four measurements. Data were not collected throughout the entire test as individual typing biomechanics remain relatively stable during typing tasks (Baker et al., 2007). 2.3. Data analysis 2.3.1. Ultrasound images An interactive image analysis program was developed to measure the median nerve at the levels of the distal radius and pisiform using the ultrasound images collected (Impink et al., 2010). Cross-sectional areas of the median nerve were determined by performing a boundary trace at each level, and were used to calculate SR (Eq. (1)), while major (or mediolateral) and minor (or anteroposterior) axes of the nerve were measured to calculate FR (Eq. (2)). CSA, SR and FR have all been found to be reliable measures related to CTS (Beekman and Visser, 2003). One investigator analyzed all images and was blinded to the subject and
2.2. Study procedure Participants were tested in the morning to limit the amount of activity performed on the day of testing. Participants completed
Fig. 1. Active markers affixed to the dorsal aspect of the dominant hand during the typing task.
548
K.K. Toosi et al. / Clinical Biomechanics 30 (2015) 546–550
time point. This technique has high reliability when a single investigator collects and analyzes all images (Impink et al., 2010). CSApisiform ¼ SR: CSAradius
ð1Þ
Lengthmajor ¼ FR: Lengthminor
ð2Þ
2.3.2. Keyboarding biomechanics Gross typing speed, adjusted typing speed, and accuracy were automatically collected throughout both 30-minute typing tasks using the electronic keyboarding program. Gross speed was defined as the speed at which the subject typed, and adjusted speed as gross speed minus any typing errors. The average gross typing speed (in words per minute) and typing accuracy (percentage of words typed correctly) were calculated for each 30-minute typing task, and averaged between the two typing tasks in the final analysis. Hand and wrist joint angles and ranges of motion (RoM) were calculated using a local coordinate system created with the active markers placed on the subject's dominant hand, wrist, and forearm. Kinematic variables of interest were peak and average wrist flexion/extension angles, peak and average ulnar/radial deviations, and RoM for the wrist and middle, index, and ring fingers. Joint angles were derived as described by Baker et al. (2007). Metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints were considered neutral (0°) when the metacarpal bones and phalanges formed a straight line, respectively; thus, ulnar deviation and wrist extension angles had negative values, and flexion and radial deviation were positive. The hand and forearm were modeled as rigid segments — their orientations were derived by tracking their local coordinate systems constructed on three hand and three forearm markers (Fig. 2). MCP and PIP joints were assumed to be joints with two degrees of freedom. MCP and PIP flexion/ extension and adduction/abduction angles were computed using the 3D angles between the corresponding metacarpals and proximal phalanges, or proximal and intermediate phalange vectors.
using standard summaries (e.g. means, standard deviations, percentiles, ranges) and graphical techniques (e.g. histograms, scatter plots). Normality was assessed using the Shapiro–Wilk test. A two-way factorial repeated-measures ANOVA was used to detect significant differences in ultrasound variables between dominant and non-dominant wrists at all four time points. Pair-wise multiple-comparison procedures were performed using a Bonferroni adjustment if the omnibus F-test was significant. Kinematic data were collected only on dominant arms; so images of the median nerve in dominant wrists were used in regression and correlation analyses. Subject characteristics of interest for correlation and regression were age, gender, BMI, typing speed, and wrist circumference; biomechanical parameters of interest included peak ulnar deviation and peak wrist extension. Wrist circumference moderately correlated with CSA (r = 0.428, P = 0.008), a measure that influences SR; thus, any correlation involving either CSA or SR included wrist circumference as a covariate to reduce confounding effects. Pearson's r was calculated to test correlations between baseline SR and CSA; subject characteristics of BMI, typing speed, and gender; and the biomechanical variable peak extension. Spearman's ρ was calculated for correlation analyses including baseline FR and peak ulnar deviation as these data were not normally distributed. Multiple linear regression models were developed to test whether wrist biomechanics or subject characteristics could predict median nerve measures after 30 and 60 min of typing. Dependent variables of interest were CSA, FR, and SR at 30- and 60-minute time points. Independent variables of interest were baseline values of median nerve measures, subject characteristics, and biomechanical variables. Models were then built to predict the value of the three median nerve measures at 30- and 60-minute time points as a function of their baseline values, biomechanics, and subject characteristic variables listed above. Wrist circumference was correlated with BMI and gender (r = 0.656 and r = − 0.603, respectively; both significant at the 0.01 level); thus, it was excluded in the regression analyses to reduce confounding effects and meet regression assumptions.
3. Results 2.4. Statistical analysis 3.1. Subjects Significance was set a priori to P b 0.05. Bonferonni corrections were applied for each median nerve measure used in the ANOVA, correlation, and regression models (α = 0.05/3 = 0.017). Statistical analyses were performed using SPSS version 21.0 (SPSS Inc., Chicago, IL, USA). All analyses were preceded by a detailed descriptive analysis of the data
Forty subjects were enrolled in the study, 37 of which were used in the final analysis. Data from one male and one female participant were excluded due to median nerve bifurcation; measurements can be influenced by the volume of blood in the bifurcating vessel and yield an inaccurate depiction of the nerve. Another female participant self-reported having symptoms of CTS and thus was excluded. Demographic variables are summarized in Table 1. The only baseline median nerve measure to correlate with subject characteristics was SR, which negatively correlated with age (r = −0.415, P = 0.012). No subject characteristics predicted changes in median nerve measures as a result of keyboarding.
Table 1 Demographic and subject characteristics; mm = millimeters; wpm = words per minute.
Fig. 2. Marker location on the dorsal aspect of the hand and forearm (circles) and the hand and forearm planes (lines and triangles) that were defined by their local coordinate systems.
Variable
Mean (SD) or N
Range or %
Age BMI Wrist circumference (mm) Gross speed (wpm) Accuracy Right handed Males
28.89 (6.6) 25.1 (4.6) 163.0 (13.8) 55.0 (10.9) 89.7 (4.8) 30 18
19–46 18.1–41.6 136–192 38–78 80–98 81.1% 48.6%
K.K. Toosi et al. / Clinical Biomechanics 30 (2015) 546–550
3.2. Ultrasound median nerve measures Baseline median nerve measures were highly correlated to measures at 30 and 60 minute time points (P b 0.01). CSA and SR significantly increased after 30 and 60 min of typing (P b 0.01). No differences existed in CSA or SR between 30- and 60-minute time points, nor between baseline and 90-minutes (P N 0.05). FR did not significantly change at any point during the task, and no differences were found in median nerve measures between dominant and non-dominant wrists (P N 0.05). Descriptive and ANOVA statistics for median nerve measures, including main effects and contrasts, are presented in Table 2. 3.3. Wrist and hand biomechanics Biomechanical variables (Table 3) did not correlate with baseline median nerve measures. When entered into linear-regression models, peak ulnar deviation significantly predicted CSA after 30 min of typing (β = −0.186, P = 0.017). Degrees of ulnar and radial deviation were negative and positive, respectively, which account for the negative beta value. No biomechanical variables significantly predicted other median nerve measures at any time point. 4. Discussion We previously demonstrated that a continuous keyboarding task caused acute changes in the ultrasonographic measures of the median nerve (Toosi et al., 2011); results of the current study were consistent with previous findings. The present study was able to quantify changes in median nerve size and shape in response to keyboarding and further relate those changes to hand/wrist biomechanics. It was hypothesized that changes in median nerve parameters previously linked to CTS (Beekman and Visser, 2003) would be observed after 30 and 60 min of keyboarding. Furthermore, keyboarding biomechanics and subject characteristics would influence these changes. In accordance with these hypotheses, CSA and SR increased after typing and returned to baseline after rest, and higher peak ulnar deviation angles predicted greater changes in median nerve measures after typing. CSA and SR have been highlighted as among the highest sensitivity and specificity when diagnosing CTS (Altinok et al., 2004). The changes in median nerve persisted over the two time points that we measured during the 60-minute typing task, and then returning to baseline when measured 30 min after stopping. We believe that the changes observed are likely a result of compression of the nerve during typing. Compression during long periods of activity, such as typing, can negatively affect the health of the median nerve and potentially
549
Table 3 Average and peak wrist angles (degrees) during typing, presented as means and standard deviations, with associated ranges. Wrist angles (°)
Mean (SD)
Range
Average flexion/extension Average ulnar/radial deviation Average peak flexion Average peak extension Average peak ulnar deviation Average peak radial deviation
−22.0 (11.2) −8.2 (11.6) −6.0 (11.3) −34.7 (13.1) −23.4 (9.4) 7.3 (6.3)
−47.3–4.6 −29.3–21.0 −28.59–17.4 −70.1 to −3.3 −38.7–51.24 0.4−21.6
contribute to the development of CTS (Barr and Barbe, 2002). It is possible that typing for long periods of time predisposes an individual to the development of CTS symptoms. Based on a return to baseline within 30 min it is possible that a resting period could be beneficial. There was no observable change in FR after 30 and 60 min of typing. Although the median nerve cross-sectional area at the pisiform level increased as a result of keyboarding, the two-dimensional nerve enlargement occurred in a proportional way. A potential reason for no change in FR is the location where the images were collected. Images in this study were limited to the pisiform level of the wrist, because our previous studies show less reliability at the hamate (Impink et al., 2010). Other studies have seen changes at the hamate and pisiform. Future investigations should include images at the hamate. In addition, it is possible that FR is less sensitive than CSA and SR for detecting shortterm changes, and probably could be used to investigate the effects of long-term tasks on the median nerve. Subjects with greater angles of ulnar deviation when typing experienced greater changes in their median nerves. Repetitive, forceful motions of the fingers – characteristic of typing – cause hypertrophy of flexor tendons within the carpal tunnel and raise carpal tunnel pressure (Werner and Andary, 2002). Performing the same motions while deviated from a neutral position amplifies the effect (Rempel et al. 1997; Werner et al., 1997a, 1997b). Most conventional keyboards are narrower than shoulder width of users. This naturally positions the wrist in ulnar deviation. A neutral wrist position can be achieved by using ergonomic technologies or workstation through alterations. In general, this study agrees with previous investigations that report keyboard users tend to type with extended and ulnarly-deviated wrists (Serina et al., 1999; Simoneau et al., 1999; Sommerich et al., 1996). Some of the biomechanics results did not match the stated hypotheses — no observable relationship existed between wrist flexion or extension and median nerve measures. Standardization of the workstation set-up likely altered natural typing biomechanics while standardization of the typing activity may have limited
Table 2 Cross-sectional area (CSA) in millimeters-squared (mm2), swelling ratio (SR), and flattening ratio (FR) in dominant (D) and non-dominant (N) hands at baseline, following 30 and 60 min of typing, and a 30-minute resting period after typing (i.e., 90-minute).
†
2
CSA (mm )
D N
SR‡
D N
FR
D N
†
Baseline mean [SD] (range)
30-minute⁎ mean [SD] (range)
60-minute⁎⁎ mean [SD] (range)
90-minute mean [SD] (range)
10.29 [1.79] (7.14–14.94) 10.47 [1.91] (6.57–14.46) 1.02 [0.14] (0.81–1.39) 1.05 [0.14] (0.86–1.41) 2.94 [0.76] (1.82–5.86) 2.77 [0.62] (1.64–4.28)
10.73 [2.12] (6.69–14.66) 10.91 [1.99] (7.17–15.34) 1.07 [0.17] (0.79–1.55) 1.12 [0.14] (0.86–1.48) 2.94 [0.65] (1.71–5.00) 2.84 [0.53] (2.03–6.26)
10.84 [1.81] (7.85–14.92) 10.69 [2.54] (7.31–15.22) 1.08 [0.16] (0.72–1.64) 1.10 [0.13] (0.87–1.47) 2.94 [0.67] (1.78–4.71) 2.80 [0.55] (2.03–4.41)
10.59 [1.77] (6.36–13.33) 10.29 [2.42] (7.50–14.19) 1.05 [0.16] (0.75–1.52) 1.06 [0.14] (0.85–1.37) 2.96 [0.74] (1.69–5.30) 2.88 [0.74] (1.83–4.99)
Significant main effect of typing on CSA (F(3,108) = 6.388, P = 0.001, η2p = .151). Significant main effect of typing on SR (F(3,108) = 7.970, P b 0.001, η2p = .181). ⁎ Significant increases from baseline in CSA (F(1,36) = 10.383, P = 0.003, η2p = .224) and SR (F(1,36) = 13.162, P = 0.001, η2p = .268) after 30 min of typing. ⁎⁎ Significant increases from baseline in CSA (F(1,36) = 15.293, P b 0.001, η2p = .298) and SR (F(1,36) = 14.252, P = 0.001, η2p = .284) after 60 min of typing. ‡
550
K.K. Toosi et al. / Clinical Biomechanics 30 (2015) 546–550
keystroke movement; ultimately, this would limit the amount of wrist flexion or extension. 4.1. Limitations The sample was limited to those without CTS symptoms. Our findings could change if symptomatic participants were included and could indicate that the median nerve changes were physiologic and may not lead to pathology. Future studies should include individuals with symptomatic CTS. The workstation setup was standardized to reduce variability, which could have elicited more uniform typing postures among participants. The use of a mouse was also eliminated to focus on effects of typing. It is likely that the use of pointing and typing devices in concert has a compound effect. Kinematic data were collected only on the dominant side while ultrasonographic measurements were collected on both. Typing is not a symmetrical task (Marklin et al., 1999) and thus by only collecting data on the dominant arm there is a risk that important information was missed. A control group, in which subjects held a similar wrist posture to typing, but avoided repetitive finger activity would have allow a stronger conclusion related to finger motion being responsible for the changes seen. 5. Conclusions This study serves to improve our understanding of median nerve changes after continuous keyboarding. CSA at the pisiform and SR as measured by ultrasound significantly increased after 30 and 60 min of keyboarding then returned to baseline after 30 min of rest. Greater changes were observed in those who typed with greater angles of ulnar deviation. While these changes may be part of a normal physiologic process, they also might represent part of the pathomechanics that leads to CTS. Further studies including individuals with CTS and following subjects longitudinally are needed. If acute median nerve changes as a result of activity are found to relate to risk of developing CTS, the methods used in this study could aid in the identification of risk to develop CTS and point to potential interventions. Acknowledgments This material is the result of work supported with resources and the use of facilities at the Human Engineering Research Laboratories, VA Pittsburgh Healthcare System, Pittsburgh, PA, USA. This study was supported by the U.S. Department of Veterans Affairs (B3142C) and the National Institutes of Health (T32HD049307). References Altinok, M.T., Baysal, O., Karakas, H.M., Firat, A.K., 2004. Sonographic evaluation of the carpal tunnel after provocative exercises. J. Ultrasound Med. 23, 1301–1306. Baker, N.A., Cham, R., Cidboy, E.H., Cook, J., Redfern, M.S., 2007. Kinematics of the fingers and hands during computer keyboard use. Clin. Biomech. 22, 34–43. Barr, A.E., Barbe, M.F., 2002. Pathophysiological tissue changes associated with repetitive movement: a review of the evidence. Phys. Ther. 82, 173–187. Beekman, R., Visser, L.H., 2003. Sonography in the diagnosis of carpal tunnel syndrome: a critical review of the literature. Muscle Nerve 27, 26–33.
Foley, M., Silverstein, B., Polissar, N., 2007. The economic burden of carpal tunnel syndrome: Long-term earnings of CTS claimants in Washington State. Am. J. Ind. Med. 50, 155–172. Fowler, J.R., Gaughan, J.P., Ilyas, A.M., 2011. The sensitivity and specificity of ultrasound for the diagnosis of carpal tunnel syndrome. Clin. Orthop. Relat. Res. 469, 1089–1094. Gelberman, R.H., Hergenroeder, P.T., Hargens, A.R., Lundborg, G.N., Akeson, W.H., 1981. The carpal tunnel syndrome: a study of carpal canal pressures. J. Bone Joint Surg. 63-A, 380–383. Gerr, F., Monteilh, C.P., Marcus, M.J., 2006. Keyboard use and musculoskeletal outcomes among computer users. Occup. Rehabil. 16, 265–277. Greening, J., Lynn, B., Leary, R., Warren, L., O'Higgins, P., Hall-Craggs, M., 2001. The use of ultrasound imaging to demonstrate reduced movement of the median nerve during wrist flexion in patients with non-specific arm pain. J. Hand Surg. (Br.) 26, 401–408. Hunting, W., Laubli, T., Grandjean, E., 1981. Postural and visual laods at VDT workplaces. I. Constrained postures. Ergonomics 24, 917–931. Impink, B.G., Gagnon, D., Collinger, J.L., Boninger, M.L., 2010. Repeatability of ultrasonographic median nerve measures. Muscle Nerve 41, 767–773. Jerosch-Herold, C., Shepstone, L., Wilson, E.C.F., Dyer, T., Blake, J., 2014. Clinical course, costs and predictive factors for response to treatment in carpal tunnel syndrome: the PALMS study protocol. BMC Musculoskelet. Disord. 15, 35. Kwon, B.C., Jung, K.I., Baek, G.H., 2008. Comparison of sonography and electrodiagnostic testing in the diagnosis of carpal tunnel syndrome. J. Hand. Surg. [Am.] 33, 65–71. Marklin, R.W., Simoneau, G.G., Monroe, J.F., 1999. Wrist and forearm posture from typing on split and vertically inclined computer keyboards. Hum. Factors 41, 559–569. Palmer, K.T., Harris, C., Coggon, D., 2007. Carpal tunnel syndrome and its relation to occupation: a systematic literature review. Occup. Med. 57, 57–66. Punnett, L., Wegman, D.H., 2004. Work-related musculoskeletal disorders: The epidemiologic evidence and the debate. J. Electromyogr. Kinesiol. 14, 13–23. Rempel, D., Keir, P.J., Smutz, W.P., Hargens, A., 1997. Effects of static fingertip loading on carpal tunnel pressure. J. Ortho. Res. 15, 422–426. Roquelaure, Y., Mechali, S., Dano, C., Fanello, S., Benetti, F., Bureau, D., et al., 1997. Occupational and personal risk factors for carpal tunnel syndrome in industrial workers. Scand. J. Work Environ. Health 23, 364–369. Serina, E.R., Tal, R., Rempel, D., 1999. Wrist and forearm postures and motions during typing. Ergonomics 42, 938–951. Silverstein, B.A., Fine, L.J., Armstrong, T.J., 1987. Occupational factors and carpal tunnel syndrome. Am. J. Ind. Med. 11, 343–358. Simoneau, G.G., Marklin, R.W., 2001. Effect of computer keyboard slope and height on wrist extension angle. Hum. Factors 43, 287–298. Simoneau, G., Marklin, R., Monroe, J., 1999. Wrist and forearm postures of users of conventional computer keyboards. Hum. Factors 41, 413–424. Sommerich, C.M., Marras, W.S., Parnianpour, M., 1996. A quantitative description of typing biomechanics. J. Occup. Rehabil. 6, 33–55. Thomsen, J.F., Gerr, F., Atroshi, I., 2008. Carpal tunnel syndrome and the use of computer mouse and keyboard: a systematic review. BMC Musculoskelet. Disord. 9, 134 (Accessed online: January 17, 2015. http://www.medscape.com/viewarticle/ 583079_3). Toosi, K.K., Impink, B.G., Baker, N.A., Boninger, M.L., 2011. Effects of computer keyboarding on ultrasonographic measures of the median nerve. Am. J. Ind. Med. 54, 826–833. Turner, J.A., Franklin, G., Fulton-Kehoe, D., Sheppard, L., Wickizer, T.M., Wu, R., et al., 2007. Early predictors of chronic work disability associated with carpal tunnel syndrome: a longitudinal workers' compensation cohort study. Am. J. Ind. Med. 50, 489–500. van Rijn, R.M., Huisstede, B.M., Koes, B.W., Burdorf, A., 2009. Associations between workrelated factors and the carpal tunnel syndrome–A systematic review. Scand. J. Work Environ. Health. 35, 19–36. Werner, R.A., Andary, M., 2002. Carpal tunnel syndrome: pathophysiology and clinical neurophysiology. J. Clin. Neurophysiol. 113, 1373–1381. Werner, R., Armstrong, T.J., Bir, C., Aylard, M.K., 1997a. Intracarpal canal pressures: the role of finger, hand, wrist, and forearm position. Clin. Biomech. 12, 44–51. Werner, R.A., Franzblau, A., Albers, J.W., Buchele, H., Armstrong, T.J., 1997b. Use of screening nerve conduction studies for predicting future carpal tunnel syndrome. Occup. Environ. Med. 54, 96–100. www.dartmouth-hitchcock.org (Accessed online: January 25, 2015).
