Computer mouse use predicts acute pain but not prolonged or chronic pain in the neck and shoulder

Transcription

1 Computer mouse use predits aute pain but not prolonged or hroni pain in the nek and shoulder J H Andersen, 1 M Harhoff, 2 S Grimstrup, 2 I Vilstrup, 1 C F Lassen, 3 L P A Brandt, 4 A I Kryger, 3,5 E Overgaard, 1 K D Hansen, 6 S Mikkelsen 3 1 Department of Oupational Mediine, Herning Hospital, 7400 Herning, Denmark; 2 Department of Biostatistis, University of Copenhagen, Copenhagen, Denmark; 3 Department of Oupational Mediine, Copenhagen University Hospital, Glostrup, Denmark; 4 Department of Oupational and Environmental Mediine, The University Hospital of Odense, Odense, Denmark; 5 Department of Oupational and Environmental Mediine, Copenhagen University Hospital, Bisbebjerg, Denmark; 6 Division of Biostatistis, University of California at Berkeley, CA, USA Correspondene to: Johan Hviid Andersen, Department of Oupational Mediine, Herning Hospital, 7400 Herning, Denmark; hejha@ringamt.dk Aepted 19 July 2007 Published Online First 6 August 2007 ABSTRACT Bakground: Computer use may have an adverse effet on musuloskeletal outomes. This study assessed the risk of nek and shoulder pain assoiated with objetively reorded professional omputer use. Methods: A omputer programme was used to ollet data on mouse and keyboard usage and weekly reports of nek and shoulder pain among 2146 tehnial assistants. Questionnaires were also ompleted at baseline and at 12 months. The three outome measures were: (1) aute pain (measured as weekly pain); (2) prolonged pain (no or minor pain in the nek and shoulder region over four onseutive weeks followed by three onseutive weeks with a high pain sore); and (3) hroni pain (reported pain or disomfort lasting more than 30 days and quite a lot of trouble during the past 12 months). Results: Risk for aute nek pain and shoulder pain inreased linearly by 4% and 10%, respetively, for eah quartile inrease in weekly mouse usage time. Mouse and keyboard usage time did not predit the onset of prolonged or hroni pain in the nek or shoulder. Women had higher risks for nek and shoulder pain. Number of keystrokes and mouse liks, length of the average ativity period, and miro-pauses did not influene reports of aute or prolonged pain. A few psyhosoial fators predited the risk of prolonged pain. Conlusions: Mostomputerworkershavenoorminor nek and shoulder pain, few experiene prolonged pain, and even fewer, hroni nek and shoulder pain. Moreover, there seems to be no relationship between omputer use and prolonged and hroni nek and shoulder pain. Musuloskeletal omplaints have been regarded as an adverse effet of omputer use, and numerous studies have been onduted, most of them pointing to an assoiation between nek and upper extremity pain and duration of omputer use. 1 6 Most of these assoiations have been based on ross-setional omparisons of self-reported omputer, keyboard or mouse use and self-reported pain omplaints. 7 A minority of studies whih have used linial outome measures have either found too few ases with traditional medial diagnoses to perform meaningful analyses or have used diagnostis that were ontroversial. 3 6 Reently, methodologies for assessing work ativity during omputer use have been evaluated, and ativitybased reordings of omputer use seem to be reliable and in agreement with measures obtained by observational tehniques A reent review alled for future studies to improve our understanding of safe levels of omputer use by measuring the duration of omputer use in a more objetive way, differentiating between mouse and keyboard use. 12 In a substudy of the NUDATA (Nek and Upper extremity Disorders Among Tehnial Assistants) study we found that even though self-reports of mouse and keyboard use were orrelated with objetive reordings, partiipants overestimated their mouse use two to three times, and their keyboard usage up to six times when ompared with the reordings. 13 The aim of this study was to examine if objetive measures of mouse and keyboard usage predit the onset of aute, prolonged and hroni pain in the nek and right shoulder. METHODS Subjets From a ohort of 6943 omputer users, whih has been desribed in detail elsewhere, a subgroup of 2146 partiipants installed the software program WorkPaeReorder (WPR), whih was developed by Nihe Software Ltd, New Zealand. Reordings of keyboard and mouse usage were obtained for up to 52 weeks from 1585 women (mean age 42.0 years, standard deviation (SD) 8.0) and 561 men with a mean age of 42.0 years (SD 9.0). The omputer workers were mainly tehnial assistants (75%). The remainder were mahine tehniians, seretaries, building assistants and other tehniians. Most of their mouse usage was when working with omputer-assisted design (CAD) tools. Measurements of keyboard and mouse usage From personal omputers (PCs) running Mirosoft Windows it is possible to obtain information on all ativations of the keyboard and mouse, irrespetive of the appliation being used. For example, keystroke ount, number of single and double mouse liks, depression of keys and mouse buttons, and software appliations used an be logged. The amount of time using the keyboard and the mouse was inferred from these ativations, and total keying time was defined as all keying ativity ourring without non-ativity intervals lasting 5 s or more, and total mouse usage was defined similarly. The assumptions behind the intervals were that (1) if two onseutive events on the keyboard are 5 s apart, it is likely that the person stopped using their fingers or holding them over the keyboard, and (2) if mouse usage events (inluding slight movements) are 5 s apart the person probably removed their hand from the mouse. The measurements for this study were aggregated into weekly parameters for usage times. Table 1 shows the overall mean exposures during the reording period among women and men. 126 Oup Environ Med 2008;65: doi: /oem

2 Parameters whih were orrelated with keyboard or mouse time above 0.5 were exluded from the analyses. These inlude total numbers of mouse liks, single mouse liks, keystrokes and keystrokes on the numeri pad. Quartile distribution of mouse usage was 0, 2.1, 5.2, 9.0 and 46 h per week. Quartiles for keyboard use were 0, 0.4, 0.9, 1.7 and 22 h per week. Reording of omputer ativity among ohort members only was seured by a personal log-on to the WPR programme. Outome measures Eah Friday partiipants were prompted to answer an on-sreen omplaint questionnaire as a part of the WPR monitoring proedure. If the partiipant was not using the omputer on Friday, the software prompted next time they did use the omputer until Tuesday in the following week for omplaints in the previous week. The questionnaire asks the partiipants to rate their degree of pain within the last 7 days on a sale from 0 to 7 with the verbal anhors: no pain, very mild pain, mild pain, mild to moderate pain, moderate pain, moderate to severe pain, severe pain and very severe pain. Levels 5, 6 and 7 were ollapsed in the analyses giving six levels of pain on an ordinal sale from 0. For eah region a different image was displayed highlighting the body area of interest. The weekly levels of nek and right shoulder pain onstitute the outome measures: (1) aute nek pain level and (2) aute shoulder pain level. If there were missing responses to the weekly pain assessments, imputed values were alulated if there were values in the week before and in the week after the missing value. If the week before and the week after were the same, the missing value was given that value. In the ase where there was one ategory differene between the week before and the week after the missing value, the missing value was randomly seleted from the week before or the week after a missing value. Differenes of more than two ategories were not imputed and not inluded in the analysis. In an attempt to apture partiipant s transition from no or minor pain to a more pronouned pain intensity, the omputer programme identified inident prolonged pain ases defined as a partiipant averaging 2.5 or less in the nek and shoulder region over four onseutive weeks followed by three onseutive weeks with a pain sore average of 4 or more in that region. The two inident pain outomes onstitute the third and fourth outome measures: (3) inident prolonged nek pain and (4) inident prolonged shoulder pain. At 1-year follow-up, partiipants ompleted a postal questionnaire, and on the basis of answers to questions relating to nek and shoulder omplaints, two measures of hroni pain were onstruted: partiipants who reported pain or disomfort lasting more than 30 days and had experiened at least quite a lot of trouble during the past 12 months but had been free of pain above 3 in that region at baseline. Outome 5 is thus hroni nek pain, and outome 6 hroni shoulder pain. From personal date logs we restrited the analysis to reordings prior to ompleting the 1-year questionnaire. Bakground variables From a baseline questionnaire, a series of ovariates regarding physial, psyhosoial and individual harateristis was inluded in the analyses. These harateristis have been desribed in detail elsewhere In short, these variables were individual fators (gender, age, type A behaviour, negative affetivity, private soial support, medial diseases with potential to influene nek and upper limb pain status, and aidents involving injury of the nek or shoulder), psyhosoial work environment fators (job demands, job ontrol, soial support, and time pressure at work) and ergonomi fators (abnormal position of mouse or keyboard, lak of arm/wrist support, height of sreen, and adjustable work desk and hair). Seniority was assessed by the length of time partiipants had used a omputer at work to the same extent as urrently, divided into less than 3 years, 4 7 years, 8 10 years and more than 10 years. Statistial analyses The relationship between nek pain and right shoulder pain, and the independent variables was analysed in three different ways. Firstly, aute nek and shoulder pain were analysed by autoregressive ordinal regression analyses with the reorded exposure variables from the week preeding the reports of pain. Earlier pain reporting was inluded in the models as the mean over the four preeding weeks based on the orrelation data struture of pain reporting. Restrited ubi splines were used to examine the relationships. 17 The basi point in the models was an attempt to isolate the diret influene of mouse and keyboard usage on the pain level, adjusted for orrelations in individuals and feedbak effets. The aim of these analyses is to estimate an intra-individual effet of the exposure on the following week s pain level. The analyses were performed with R. We further used a GEE approah in a simple ordinal regression model to ompare the results obtained from the autoregressive models. The results were very similar, and we report the results from the autoregressive analysis in this paper beause we find them most suitable with this kind of data orrelation struture. Fixed ovariates from baseline were inluded in all models. The seond set of analyses involved analysing inident prolonged pain by Cox regression. Covariates from baseline on individual, psyhosoial and ergonomi risk fators were inluded in all the models, and the interquartile ranges of weekly mouse and keyboard use preeding the ourrene of prolonged pain were tested. None of the interation terms Table 1 Exposure variables Distributional harateristis of the reorded exposure variables from WorkPaeReorder among men and women Female (n = ) Male (n = ) Mean (SD) Min Max Mean (SD) Min Max Computer use in hours per week 9.2 (6.2) (6.5) Mouse usage in hours per week 6.1 (4.9) (4.8) Keyboard usage in hours per week 1.3 (1.3) (1.5) 0 21 Mouse liks (per 25 liks per minute) 1.0 (0.3) (0.3) Keystrokes (per 100 per minute) 1.5 (0.2) (0.2) Average period of ative mouse use (per 10 min) 0.3 (0.2) (0.2) Average period of ative mouse breaks (per minute) 0.7 (0.3) (0.3) Average period of ative keyboard use (per 2 min) 0.1 (0.1) (0.1) Average period of ative keyboard breaks (per minute) 0.6 (0.5) (0.4) Oup Environ Med 2008;65: doi: /oem

3 between any of the reorded exposure variables ontributed signifiantly to any of models. Analyses were onduted using the Stata statistial pakage (v 9.0; Stata, College Station, TX). Thirdly, for the purpose of analysis of hroni nek and shoulder pain, we used a logisti regression model to analyse the rosssetional assoiations between registered umulative mouse and keyboard use in the year preeding the follow-up questionnaire. In this model, mouse and keyboard usage was summed aross the 52 weeks to give a umulative number of hours working with the mouse and keyboard within the year preeding the report of hroni nek and shoulder pain in the 1-year follow-up questionnaire. The quartiles for hours per year with a mouse were 0, 160, 273, 413 and 1590 h per year. For a keyboard the quartiles were 0, 30, 50, 80, and 550 h per year. Due to a shortage of hroni pain ases, we only inluded mouse and keyboard usage time, age, gender and seniority in these models. RESULTS The 2146 partiipants were reruited from an original ohort of 6943 tehniians. There were several reasons why some subjets were not inluded in this part of the study: lak of an employer s approval for installing the software programme, partiipants using more than one omputer at work simultaneously, working with platforms other than Windows 95/98/ NT4, and of ourse the partiipants refraining from installing the programme for other reasons. Among the partiipants who installed WorkPaeReorder, 11.1% graded their nek pain within the last 7 days as 4 or more on the pain sale from 0 to 7, and 8.0% graded their shoulder pain to be 4 or more. The same perentages for the whole ohort were 10.7% and 7.7%. Self-reported mouse use was a little higher in the WPR group with a mean weekly usage of 13.4 vs 11.4 h per week in the full ohort. Keyboard time was 8.8 h per week in both groups. More women installed the programme: 73.8% vs 62.6% in the baseline ohort. WPR data files were returned by 2146 partiipants, and weekly observations were obtained. Overall, 84% (n = 1795) of respondents returned data files for 52 weeks, and 95% of the data files inluded at least 30 weeks. A total of weekly questionnaires were ompleted (85%). Aute pain Table 2 shows the perentage of partiipants reporting pain in the nek and the right shoulder during the reording period. Most partiipants did not experiene nek and shoulder pain. However, 6.3% of males and 9.1% of females experiened severe pain in the nek in any 1 week during the study period, and 9.8% males and 10.2% females reported severe shoulder pain. Figure 1 illustrates a linear assoiation between mouse usage time and aute nek and shoulder pain with signifiant differenes from a zero effet. The effet on nek pain orresponds to a 0.6% inrease in risk for a 1 h inrease in weekly usage, and for shoulder pain the risk inrease per hour is 2%. In table 3 the risk inrease is shown for an inrease in the interquartile range of mouse usage with an odds ratio (OR) for aute nek pain of 1.04 (95% CI 1.00 to 1.09), and for aute shoulder pain an OR of 1.10 (95% CI 1.05 to 1.16). Figure 1 also illustrates a non-effet of keyboard usage on both aute nek and shoulder pain. From table 3 it an also be seen that none of the variables for mouse and keyboard speed, miro-pauses or length of ativity periods had any main effets on aute pain. Women had an inreased risk for aute nek pain (OR 1.21, 95% CI 1.16 to 1.27) and for aute shoulder pain (OR 1.22, 95% CI 1.16 to 1.28). Prolonged and hroni pain Table 4 shows the inidene of prolonged and hroni nek and shoulder pain for men and women. Onset of prolonged nek pain was seen in 302 partiipants and prolonged shoulder pain ourred in 227; the inidene rates were for prolonged nek pain and for prolonged shoulder pain. Few partiipants fulfilled our riteria for inident hroni pain at 1- year follow-up. A total of 1.75% of males and 2.13% of females reported new hroni nek pain during 1 year, and the orresponding perentages for hroni shoulder pain were 1.89% of males and 2.5% of females. Table 3 reveals that assoiation between mouse usage time and aute pain does not hold for prolonged pain in the nek or shoulder. None of the reorded omputer use exposure variables had an effet on prolonged pain in the nek and shoulder. Females showed an inreased risk of around 80% for prolonged pain. Among the potential onfounders inluded, only a few were signifiant besides gender. Overall dissatisfation with the physial lay-out of the work station was assoiated with prolonged nek pain (OR 1.76, 95% CI 1.22 to 2.55). Low soial support in the work plae was assoiated with nek pain (OR 1.56, 95% CI 1.21 to 2.03). Prolonged shoulder pain was assoiated with the ombination of high job demands and low job ontrol (OR 2.30, 95% CI 1.24 to 4.28). None of the other fixed variables from baseline on ergonomi and individual risk fators were assoiated signifiantly with prolonged nek and shoulder pain. Table 5 shows assoiations between umulative mouse and keyboard usage time and the risk for hroni nek and shoulder pain. Time with mouse or keyboard did not predit hroni pain at 1-year follow-up. Neither age nor seniority was signifiantly assoiated with hroni shoulder pain, but a trend was seen for more nek pain with greater seniority (p = 0.04 for trend). DISCUSSION This is the first prospetive study of nek and shoulder pain among omputer users whih benefits from ontinuously reorded objetive omputer usage, weekly measures of pain and the opportunity to identify inident ases with prolonged pain from a definition inluded in the omputer programme. In this study we found that most partiipants reported no or minor pain problems in the nek and shoulder. The prevalene of weekly severe pain was low, the inidene of prolonged nek pain was 0.31%, and the inidene of prolonged shoulder pain was 0.23%. Analyses of risk fators related to omputer work showed that for every quartile inrease in weekly mouse usage, the risk for aute nek pain inreased by 4% and the risk for aute shoulder pain by 10%. Analyses of time with mouse and keyboard showed no threshold level for the time variables. Exlusion of the 5% of subjets with zero values for mouse and keyboard usage some time in the 52-week period of reording did not hange the linear relationship or the perentages of inrease for every quartile of inrease in mouse or keyboard usage. None of the other reorded variables suh as typing/ liking speed, average ativity period lengths or miro-pauses, influened the results. Prolonged and hroni nek and shoulder pain were not related to omputer use, neither to mouse use nor keyboard use. Our speifi definition of inident prolonged pain is open to disussion. The partiipants should fulfil the option of four onseutive weeks with no or minor pain followed by 3 weeks with more severe pain. There are probably other ways to advane from minor pain to severe pain. We tested if any ombination of long-lasting severe pain for three, four or six onseutive weeks was assoiated with our exposure variables in 128 Oup Environ Med 2008;65: doi: /oem

4 Figure 1 Adjusted log odds with 95% onfidene intervals for aute nek and shoulder pain in relation to mouse (left) and keyboard use (right) in the preeding week. The vertial arrows show the knots used for modelling with restrited ubi splines in the autoregressive ordinal regression models. The urves show the linear assoiation between mouse and keyboard usage with all other variables at a fixed level. Table 2 Overall perentages and within perentages of aute nek and shoulder pain among 2146 omputer users with weekly reordings for 52 weeks Aute nek pain Aute shoulder pain Males (n = 561) Females (n = 1585) Males (n = 561) Females (n = 1585) (n = ) (n = 1737) (n = ) (n = 5775) (n = ) (n = 1533) (n = ) No pain Very mild pain Mild pain Moderate pain Moderate to severe pain Severe pain (n = 5240) Oup Environ Med 2008;65: doi: /oem

5 Table 3 Reorded variables Risk estimates for aute* and prolonged pain{ in the nek and right shoulder in relation to reorded exposure variables Nek pain Shoulder pain Aute pain Prolonged pain Aute pain Prolonged pain OR adj * (95% CI) HR adj { (95% CI) OR adj * (95% CI) HR adj { (95% CI) Mouse variables Usage time (interquartile range) 1.04 (1.00 to 1.09) 1.01 (0.97 to 1.06) 1.10 (1.05 to 1.16) 1.02 (0.96 to 1.08) Speed - mouse liks per 25 liks/min 0.99 (0.97 to 1.02) 0.84 (0.63 to 1.12) 1.02 (0.99 to 1.05) 1.11 (0.79 to 1.56) Average ativity periods per 10 min 1.01 (0.99 to 1.02) 1.02 (0.92 to 1.13) 0.99 (0.97 to 1.01) 0.98 (0.86 to 1.12) Miro-pauses per minute 0.97 (0.94 to 1.00) 0.96 (0.75 to 1.24) 1.02 (0.99 to 1.06) 0.85 (0.62 to 1.16) Keyboard variables Usage time (interquartile range) 1.01 (0.98 to 1.03) 1.08 (0.80 to 1.47) 1.01 (0.98 to 1.04) 0.87 (0.60 to 1.26) Speed per 100 keystrokes/min 0.99 (0.96 to 1.02) 0.85 (0.63 to 1.16) 0.95 (0.92 to 0.98) 0.72 (0.49 to 1.06) Average ativity periods per 2 min 1.00 (0.98 to 1.01) 1.06 (0.96 to 1.16) 0.99 (0.98 to 1.01) 0.95 (0.77 to 1.16) Miro-pauses per minute 1.01 (0.97 to 1.04) 0.95 (0.84 to 1.07) 0.95 (0.91 to 0.98) 1.08 (0.94 to 1.23) Female gender (male/female) 1.21 (1.16 to 1.27) 1.80 (1.28 to 2.52) 1.22 (1.16 to 1.28) 1.81 (1.19 to 2.73) Age (10-year inrements) 1.04 (1.02 to 1.07) 0.94 (0.80 to 1.11) 1.06 (1.03 to 1.09) 1.02 (0.84 to 1.25) *Aute pain with odds ratios (OR) obtained by autoregressive ordinal logisti regression analysis; {prolonged pain with hazard ratios (HR) obtained by Cox regression analysis with mutual adjustments for fixed ovariates. a different way from our definition of prolonged pain. There were no suh indiations (results not shown). An explanation for an effet on aute pain, but not on prolonged or hroni pain, ould be reverse ausation in that the level of pain ould influene the amount of mouse and keyboard usage in the subsequent week. We tested whether level of any musuloskeletal pain predited the following week s usage time using a random effets regression model and found no effets (p = 0.49 for men and p = 0.30 for women). Some of the fixed ovariates from the baseline assessment were risk fators for nek and shoulder pain. The most onsistent risk fator was female gender, but being unsatisfied with the physial lay-out of the workplae was also assoiated with nek pain even though none of the speifi ergonomi risk fators showed any ontribution to the models. Asking about satisfation with the physial lay-out of the workplae was meant to apture other ergonomi fators besides those atually asked about, but it ould also serve to indiate overall dissatisfation with the work environment, tasks or other fators. Low soial support at work was a risk fator for nek pain but not for shoulder pain, and high job strain was preditive of prolonged shoulder pain. Table 4 Inidene of prolonged (per week) and hroni (1-year) pain in the nek and right shoulder within 1 year of reordings Outome Subjets Cases Time at risk (weeks) Inidene rate per week /1-year inidene (%) Prolonged nek pain Males Females Total Chroni nek pain Males % Females % Total % Prolonged shoulder pain Males Female Total Chroni shoulder pain Males % Females % Total % An explanation for the low ourrene of hroni pain in this study ould be seletion out of the ohort beause of nek and shoulder pain. Our data do not support this. The levels of musuloskeletal pain were no different among the 6943 subjets at baseline, the 5658 who ompleted 1-year follow-up, the 1335 drop-outs from baseline to 1-year follow-up, and the 2146 who partiipated in this study. Furthermore, overall sikness absene for more than 2 weeks, obtained from the Danish entral register, did not reveal higher sikness absene among non-responders at baseline or at follow-up, whih strongly points against healthy worker seletion in this group of omputer workers. One interesting aspet of our study was that it was possible to ondut a large sale epidemiologial study using objetive measures of exposure. The partiipation rate with respet to installing and returning WPR data was rather low (31%), however, but partiipants who returned data did not differ very muh from those who did not with respet to questionnaire data on omputer work and musuloskeletal pain. The partiipants in this study reported slightly higher weekly mouse usage than the full ohort and there was an over-representation of women. So this study inluded more heavy female mouse users than the original ohort, but we do not think that the assoiations between omputer use and pain-related variables would be severely biased by this. Considering that the study overed more than 3000 work plaes with different software platforms and poliies for software installation, we find the Table 5 Risk fators* for hroni nek and shoulder pain at 1-year follow-up Chroni nek pain OR (95% CI) Chroni shoulder pain OR (95% CI) Mouse usage, quartiles of 0.77 (0.55 to 1.07) 1.11 (0.86 to 1.44) hours per year Keyboard usage, quartiles 1.05 (0.74 to 1.51) 0.91 (0.68 to 1.21) of hours per year Male/female 1.43 (0.62 to 3.30) 1.20 (0.55 to 2.58) Seniority (3 years (n = 529) 1.00 (ref ) 1.00 (ref ) 4 7 years (n = 780) 1.06 (0.36 to 3.07) 0.58 (0.24 to 1.41) 8 10 years (n = 442) 1.88 (0.65 to 5.44) 0.53 (0.18 to 1.54).10 years (n = 375) 2.53 (0.84 to 7.56) 1.08 (0.43 to 2.69) Age, 10-year inrements 0.85 (0.52 to 1.39) 0.89 (0.55 to 1.45) *OR, odds ratios with 95% onfidene intervals obtained from a logisti regression model. 130 Oup Environ Med 2008;65: doi: /oem

6 Main messages Most omputer workers have no or minor nek and shoulder pain omplaints, few experiene prolonged pain, and even fewer, hroni nek and shoulder pain. Objetively measured mouse and keyboard ativities were positively assoiated with aute nek and shoulder pain but were not assoiated with prolonged or hroni nek and shoulder pain. Earlier results from this ohort based on self-reports of mouse and keyboard use ould be biased. partiipation rate quite satisfatory. Muh higher partiipation rates in large epidemiologial studies using eletroni ativity reordings an probably be expeted in studies onentrated in a few large ompanies. In general, the ohort members were mouse users more than they were keyboard users. The mean mouse usage was around 6 h per week, but there was a large variation with 10% of the weeks without any mouse use and around 10% with more than 15 h per week. Total omputer use was 9.2 h per week and was strongly orrelated with mouse use with a orrelation oeffiient of Keyboard usage was at a low level, so the results for keyboard users should be autiously interpreted and may not hold for heavy keyboard users. The maximum keyboard use in our ohort was around 20 h per week. Partiipation in the study and answering questions about pain every week might indue heightened awareness of pain and aggravation of any minor sensations, but the reorded levels of omputer ativity were not known to the partiipants. In addition, the length of the monitoring time period makes it unlikely that the risk estimates for mouse and keyboard usage were flawed by information bias. Limitations of the study inlude the fixed ovariates, whih were self-assessed only at baseline, and we do not know the exat extent to whih ergonomi or psyhosoial work plae fators hanged during the reording period. The results from other analyses in the NUDATA study do not point to differenes in risk estimates for ergonomi risk fators when omparing those who made hanges in how they arried out their job and those who did not Findings from the NUDATA study on nek and shoulder pain in relation to self-reported usage times for mouse and keyboard have been previously reported. 6 This study onfirms the linear relationship between mouse usage time and aute pain, but it strongly questions the existene of any relationship between omputer use and prolonged and hroni pain. Conerning 1- year inident hroni pain, we found an effet of mouse use on shoulder pain when exposure was based on self-reports, but this assoiation disappears in this study where we have inluded objetive measurements of mouse use. This ontradition raises the possibility of severe information bias, probably assoiated with the setting and launhing of studies of omputer use and musuloskeletal pain. Even though the NUDATA study was launhed as a general study of work and health among tehniians, disussions among tehniians and writings in the media have entred on lay understandings of the mouse arm, and under these irumstanes the possibility for information bias is huge. In earlier studies, linial outomes are rare, and existing evidene for a detrimental effet on the nek and shoulder of omputer use is sparse. From the NUDATA study we an onlude that most omputer workers have no or minor Poliy impliations The ommon opinion that omputer use has an adverse health effet is questionable. Future researh on the relationship between omputer work and adverse health effets should utilise objetive assessments of mouse and keyboard use. pain omplaints, few experiene prolonged pain, and even fewer, hroni pain. Moreover, there seems to be no relationship between omputer use and our measures of prolonged and hroni nek and shoulder pain. A reent review 12 found stronger risk estimates for the hand arm region than for the nek shoulder region based on the few prospetive studies identified, and the possibility for hand arm disorders should be further studied. The results from the NUDATA study of distal arm pain will be reported in a separate paper. Aknowledgements: We thank the Danish Assoiation of Professional Tehniians for useful assistane, and we espeially thank the study partiipants for their time and effort. Funding: This study was supported by grant from the Danish Medial Researh Counil and grant from the Danish Ministry of Employment, National Work Environment Authority. Competing interests: None. REFERENCES 1. Punnett L, Bergqvist U. Visual display unit work and upper extremity musuloskeletal disorders: a review of epidemiologial findings. Arbete oh Halsa 16. Solna: Arbetslivsinstitutet, Blatter BM, Bongers PM. Duration of omputer use and mouse use in relation to musuloskeletal disorders of the nek and upper limb. Int J Ind Ergon 2002;30: Gerr F, Marus M, Ensor C, et al. A prospetive study of omputer users: I. Study design and inidene of musuloskeletal symptoms and disorders. Am J Ind Med 2002;41(4): Korhonen T, Ketola R, Toivonen R, et al. Work related and individual preditors for inident nek pain among offie employees working with video display units. Oup Environ Med 2003;60(7): Jensen C. Development of nek and hand-wrist symptoms in relation to duration of omputer use at work. Sand J Work Environ Health 2003;29(3): Brandt LPA, Andersen JH, Lassen CF, et al. Nek and shoulder symptoms and disorders in a Danish ohort of 6943 omputer workers. Sand J Work Environ Health 2004;30: Gerr F, Marus M, Monteilh C. Epidemiology of musuloskeletal disorders among omputer users: lesson learned from the role of posture and keyboard use. J Eletromyogr Kinesiol 2004;14(1): Homan MM, Armstrong TJ. Evaluation of three methodologies for assessing work ativity during omputer use. AIHA J 2003;64(1): Blangsted AK, Hansen K, Jensen C. Validation of a ommerial software pakage for quantifiation of omputer use. Int J Ind Ergon 2004;34: Heinrih J, Blatter BM, Bongers PM. A omparison of methods for the assessment of postural load and duration of omputer use. Oup Environ Med 2004;61(12): Douwes M, de Kraker H, Blatter BM. Validity of two methods to assess omputer use: self-report by questionnaire and omputer use software. Int J Ind Ergon 2007;37: Ljemker S, Huysmans M, Blatter BM, et al. Should offie workers spend fewer hours at their omputer? A systemati review of the literature. Oup Environ Med 2007;64(4): Mikkelsen S, Vilstrup I, Lassen CF, et al. Validity of questionnaire self-reports on omputer, mouse and keyboard usage during a 4 week period. Oup Environ Med 2007:64; Lassen CF, Mikkelsen S, Kryger A, et al. Elbow and wrist/hand symptoms among 6943 omputer operators. A 1-year follow-up study (the NUDATA study). Am J Ind Med 2004;46: Kryger AI, Andersen JH, Lassen CF, et al. Is intensive use of the arms an oupational hazard among omputer users? From the NUDATA study. Oup Environ Med 2003;60(11):e Andersen JH, Thomsen JF, Overgaard E, et al. Computer use and arpal tunnel syndrome: a 1-year follow-up study. JAMA 2003;289(22): Diggle PJ, Heagerty P, Liang KY, et al. Analysis of longitudinal data. Oxford: Oxford University Press, Oup Environ Med 2008;65: doi: /oem