Multivariate, longitudinal analysis of the impact of changes in office work environments on surface electromyography measures

Abstract

Purpose: To detect impacts of changes in work environment and worker-equipment interface variables upon surface electromyography (EMG) measures using multivariate, longitudinal analysis.

Methods: For 33 office workers, yearly measurements (1999-2001) were taken during normal work. Independent variables were related to work environment (expert-observed equipment dimensions, work organization on questionnaire) and interface (expert-observed postures, self-reported workstation-equipment relative fit i.e. inside or outside guidelines-informed location, and 30 min video-based task analysis). Internal mechanical exposure (EMG) was recorded bilaterally from extensor carpi radialis brevis (ECRB) and upper trapezius sites, each side, also for 30 min. Dependent variables were amplitude probability distribution functions (APDF 50 and 90%) and gaptime for entire record EMG (over all tasks) and task-specific EMG (for four separate tasks). Multivariate mixed models used independent variables to predict EMG measures (4 muscle sites × (1 entire record + 4 task specific) = 20 models total).

Results: Among EMG measures, 9/16 means and 2/16 variances were significantly different across years (p < 0.1). Environment and interface variables explained part of the variation in EMG measures in 13/20 models. The most consistent predictors included: (1) increased monitor distance predicted reduced APDFs and increased gaptimes; (2) wrist extension <20° predicted decreases in left ECRB APDFs; (3) keyboard location within guidelines predicted improvements in all right ECRB EMG measures during keyboarding; and (4) longer task duration predicted higher APDFs and lower gaptimes.

Conclusion: Longitudinal analysis with multivariate models can detect the impacts of changes in environment and interface exposures on EMG measures among office workers.