OCCUPATIONAL APPLICATIONSExposure Variation Analysis (EVA) is a method for describing variation in exposure, capturing not only the intensity of exposure (e.g., muscle activity amplitudes or posture levels) but also how long exposures of different intensities occur in uninterrupted sequences. The present review of studies applying EVA found that this method has been used across a range of occupational sectors, from computer-based office work to physically demanding jobs, such as assembly and caregiving, and for a range of exposures, including electromyography and postures. EVA offers a way to identify patterns of exposure and recovery that may contribute to fatigue, discomfort, or disorders, or conversely promote health. The flexibility of EVA makes it suitable even for use with wearable sensor data. By adopting EVA, ergonomics professionals can better assess risk and inform the design of tasks that reduce sustained exposures and promote more balanced activity patterns.
Keywords: Muscle activity; ergonomics; physical behavior; posture; temporal patterns.
Background: Time-varying exposures, such as muscle activity and postural angles, are key risk factors for work-related musculoskeletal disorders when accumulated for long, uninterrupted periods. Traditional exposure metrics, however, often ignore the temporal structure of exposure. Exposure Variation Analysis (EVA) captures both the levels and sequence duration of exposures, thus offering a two-dimensional documentation of exposure variation. Purpose: We aimed to summarize and critically appraise the use of EVA, examine methodological variability between studies, and identify opportunities for future development. Methods: We performed a comprehensive review of studies published between 1991 and 2024 that explicitly utilized EVA. Searches were performed using citation tracking and database queries. Key information was extracted on exposure types, level and sequence duration intervals, EVA derivatives, and statistical analyses of EVA. Results: A total of 94 studies were included in this review, mainly in occupational and public health sciences and sports. EVA has most frequently been used for analyzing electromyographic and postural data, but physical behaviors are a growing application. We found a wide variation in how studies defined categories for level and sequence duration intervals. Several studies simplified or post-processed the EVA matrix to extract summary measures, such as marginal distributions or variability indices. However, only one study accounted for the compositional structure of EVA data. Most statistical analyses were conducted on summary measures, not the full EVA matrix. Conclusions: EVA has been applied in a range of research areas, mainly to document exposures rather than analyzing association with health outcomes. The large variety of level and sequence duration intervals highlight the need for standardization and further development. Future work should focus on defining reference interval sets for different exposure types, developing EVA derivatives, analyzing log-term effects on health of different EVA structures, and developing appropriate statistical models reflecting these analysis needs.