Exploring predictors of welfare dependency 1, 3, and 5 years after mental health-related absence in danish municipalities between 2010 and 2012 using flexible machine learning modelling.

BACKGROUND: Using XGBoost (XGB), this study demonstrates how flexible machine learning modelling can complement traditional statistical modelling (multinomial logistic regression) as a sensitivity analysis and predictive modelling tool in occupational health research. DESIGN: The study predicts welfare dependency for a cohort at 1, 3, and 5 years of follow-up using XGB and multinomial logistic regression (MLR). The models' predictive ability is evaluated using tenfold cross-validation (internal validation) and geographical validation (semi-external validation). In addition, we calculate and graphically assess Shapley additive explanation (SHAP) values from the XGB model to examine deviation from linearity assumptions, including interactions. The study population consists of all 20-54 years old on long-term sickness absence leave due to self-reported common mental disorders (CMD) between April 26, 2010, and September 2012 in 21 (of 98) Danish municipalities that participated in the Danish Return to Work program. The total sample of 19.664 observations is split geospatially into a development set (n = 9.756) and a test set (n = 9.908). RESULTS: There were no practical differences in the XGB and MLR models' predictive ability. Industry, job skills, citizenship, unemployment insurance, gender, and period had limited importance in predicting welfare dependency in both models. On the other hand, welfare dependency history and reason for sickness absence were strong predictors. Graphical SHAP-analysis of the XGB model did not indicate substantial deviations from linearity assumptions implied by the multinomial regression model. CONCLUSION: Flexible machine learning models like XGB can supplement traditional statistical methods like multinomial logistic regression in occupational health research by providing a benchmark for predictive performance and traditional statistical models' ability to capture important associations for a given set of predictors as well as potential violations of linearity. TRIAL REGISTRATION: ISRCTN43004323.

Does occupational forward bending of the back increase long-term sickness absence risk? A 4-year prospective register-based study using device-measured compositional data analysis.

OBJECTIVE: Forward bending of the back is common in many jobs and a risk factor for sickness absence. However, this knowledge is based on self-reported forward bending that is generally imprecise. Thus, we aimed to investigate the dose-response relation between device-measured forward bending at work and prospective register-based risk of long-term sickness absence (LTSA). METHODS: At baseline, 944 workers (93% from blue-collar jobs) wore accelerometers on their upper back and thigh over 1-6 workdays to measure worktime with forward bending (>30˚ and >60˚) and body positions. The first event of LTSA (≥6 consecutive weeks) over a 4-year follow-up were retrieved from a national register. Compositional Cox proportional hazard analyses were used to model the association between worktime with forward bending of the back in an upright body position and LTSA adjusted for age, sex, body mass index (BMI), occupational lifting/carrying, type of work, and, in an additional step, for leisure time physical activity (PA) on workdays. RESULTS: During a mean worktime of 457 minutes/day, the workers on average spent 40 and 10 minutes on forward bending >30˚ and >60˚ in the upright position, respectively. Five more minutes forward bending >30˚ and >60˚ at work were associated with a 4% [95% confidence interval (CI) 1.01-1.07] and 8% (95% CI 1.01-1.16) higher LTSA risk, respectively. Adjustment for leisure-time PA did not influence the results. CONCLUSION: We found a dose-response association between device-measured forward bending of the back and prospective LTSA risk. This knowledge can be integrated into available feasible methods to measure forward bending of the back for improved workplace risk assessment and prevention.