Developing and validating prediction models for low back pain and neck pain in office workers: a cross-sectional study

Abstract

Office workers frequently report low back pain (LBP) and neck pain (NP) related to sedentary work. While several prognostic risk factors of muscle performances have been identified, accurate predictive models are essential. This study aimed to develop and validate prediction models of LBP and NP among office workers. A cross-sectional study was conducted with 170 office workers aged 20–59 years, all with over one year of experience and daily computer use exceeding six hours. Participants underwent muscles performance tests for LBP and NP, including assessments of trunk muscles stability and endurance, lumbar lordotic curve, a multi-joint strength test and the neck strength and endurance tests. Univariate and multivariate logistic regression were used to develop a predicting models, calculating odds ratio (OR) with 95% confidence interval (CI). Model performance was validated using Receiver Operating Characteristic (ROC) curves, area under the curve (AUC), and validity indexes. The prevalence rates were 59.4% for LBP and 72.4% for NP. Higher body fat percentage was associated with an increased risk of LBP (OR:1.06, 95%CI:1.01–1.12), while a greater lumbar lordotic curve was associated with reduced risk (OR:0.96, 95%CI:0.92–0.99). For NP, greater deep neck flexor endurance was significantly associated with a lower risk (OR:0.98, 95%CI:0.97–0.99). The prediction models demonstrated acceptable discrimination, with an AUC of 0.65 (95%CI: 0.56–0.73) for LBP and 0.64 (95%CI: 0.54–0.73) for NP. The developed predictive models show potential for integration into clinical decision-making to guide the development of workplace safety and intervention programs aimed at preventing LBP and NP in office workers.