• Computer use
• musculoskeletal disorder
• epidemiology
• exposure assessment
• back disorders
• occupational health practice
• ergonomics
• neurobehavioural effects
• neuropsychology
• longitudinal studies
• cross sectional studies
• video display units
• vibration, solvents
• mercury
• lead
• wruld
• repetitive strain injury
• musculoskeletal

The results of a large prospective study of computer users show little association between input device use and musculoskeletal outcomes. Ordinarily, this study would simply be one more addition to the already large body of literature examining associations between computer use and musculoskeletal outcomes and one more citation included in the ‘negative association’ column of review papers. This is not an ordinary study, however. Unique to this investigation was its use of newly developed memory resident computer registration software that allows for the automated collection of computer and mouse activity over months or years. Using data collected with this software, IJmker et al1 observed no association between computer and mouse use and severe distal upper extremity or neck–shoulder symptoms. Even more provocative was the simultaneous observation by IJmker et al1 (see page 502) that self-reported daily duration of computer use was significantly associated with distal upper extremity and neck–shoulder symptoms. With over one billion computers in use worldwide, the health effects of computer work continue to be relevant to occupational health practice.

To our knowledge, nearly all previous large studies examining associations between computer use and musculoskeletal outcomes used self-reported metrics of computer use (most commonly hours per day or hours per week) as no other method was feasible for large epidemiological studies. Prior to the publication of the report by IJmker et al,1 only one study was available in the literature in which computer use was ascertained with computer registration software.2 Although associations between computer activity and musculoskeletal symptoms were observed in that study, the small sample size (N=27) and other methodological limitations make it difficult to extend the results to other populations. Given their inclusion of new objective metrics of computer use duration, do the results of IJmker et al1 supersede the …