Precision of measurements of physical workload during standardised manual handling. Part I: surface electromyography of m. trapezius, m. infraspinatus and the forearm extensors.

Though surface electromyography (EMG) has been widely used in studies of occupational exposure, its precision in terms of the variance between-days and between-subjects has seldom been evaluated. This study aimed at such an evaluation. Six women performed three different work tasks: 'materials picking', 'light assembly', and 'heavy assembly', repeated on 3 different days. EMG was recorded from m. trapezius, m. infraspinatus and the forearm extensors. Normalisation was made to a maximal (MVE), and a submaximal (RVE), reference contraction. Variance components between days (within subjects) and between subjects were derived for the 10th, 50th and 90th percentiles, as well as for muscular rest parameters. For the task 'heavy assembly', the coefficient of variation between days (CV(BD)) was 8% for m. trapezius (right side, 50th percentile, MVE normalised values). Larger variabilities were found for m. infraspinatus (CV(BD) 15%), and the forearm extensors (CV(BD) 33%). Between-subjects variability (CV(BS)) was greater, 16% for m. trapezius and 57% for m. infraspinatus, 29% for the forearm extensors. RVE normalisation resulted in larger CV(BD), while reducing CV(BS). The between-days and between-subjects variability may be used to optimise sampling strategy, and to assess the bias in epidemiological studies. The bias caused by measurement procedures per se is acceptable.