The time variation pattern of mechanical exposure and the reporting of low back pain

Creating mechanical exposure profiles for multi-component jobs for use in epidemiological studies of work-related musculoskeletal disorders is challenging but becoming more common. Once time-varying mechanical exposure profiles are created, some form of exposure index must be derived. Available techniques typically provide information about the relative amount of time exposed to different levels and durations of stressors. Little or no information concerning the sequence or history of exposure delivery is maintained. The current paper examines an approach to the processing of time varying exposures for epidemiological studies. Shift-long exposure data of 68 production operators (34 cases; 34 controls) from a large epidemiological study of low back pain reporting were used to create individual shift-long internal mechanical exposure profiles. These were processed through a first-order system with varying time constants to generate a range of modelled responses from which two exposure indices were then created. The utility of these mechanical exposure indices was evaluated using low back pain reporting as the health outcome from the same epidemiological study. The modelled response using the first-order system appeared to provide additional exposure information, perhaps time-history sensitive exposure information that was not captured by the unprocessed exposure profiles. Risk estimates from the peak value exposure index derived from an internal exposure processed with a time constant of 2?s were almost doubled compared with the risk estimates of the unprocessed peak value, and the final value, taken from the internal exposure processed with a time constant of 5?000?s, generated higher risk estimates over the cumulative/average exposures. This study suggests a potential analysis approach for epidemiological studies of musculoskeletal disorders where long-duration mechanical exposure data is available.