Workforce Characterization and Biopsychosocial Predictors of Musculoskeletal Symptoms in Chilean Salmon Processing: A Cross-Sectional Sparse Learning Study

Work-related musculoskeletal symptoms are common in physically demanding food-processing environments, yet salmon-processing workers in southern Chile remain under-characterized. This cross-sectional field study characterized 100 full-time workers from a salmon-processing facility in Puerto Montt, Chile, and identified candidate predictors of self-reported regional pain. Participants completed surveys on demographics, work history, lifestyle, psychosocial, and sleep-related factors; anthropometrics and grip strength were also measured. Binary outcomes captured 12-month self-reported pain in the low back, shoulder, and upper extremity. Pain prevalence was high across regions: low back 52%, shoulder 44%, and upper extremity 50%. Sparse learning methods, including LASSO and the Sparse Wrapper AlGorithm, were applied within logistic regression models to reduce a broad biopsychosocial feature set. Sleep-related variables recurred across pain outcomes, while anthropometric and physical-capacity measures appeared more region-specific. Findings support data-driven, field-oriented screening strategies for applied ergonomics surveillance.