A prediction model for sleep disorders in shift workers of a chemical fiber enterprise

Objective: To construct a prediction model for sleep disorders in shift workers of a chemical fiber enterprise, so as to provide the basis for early identification and prevention of sleep disorders in shift workers. Methods: Shift workers were sampled from a chemical fiber enterprise in Tongxiang City, Zhejiang Province using a cluster sampling method from August 2022 to July 2024. Demographic information, length of service and average weekly working hours were collected through questionnaire surveys. Depressive symptoms, anxiety symptoms and sleep disorders were evaluated using the Pittsburgh Sleep Quality Index, Patient Health Questionnaire and Generalized Anxiety Disorder Questionnaire, respectively. The shift workers were randomly divided into a training set and a validation set at a ratio of 7∶3. Predictive factors were selected using a multivariable logistic regression model based on the training set, and a nomograph model for prediction of sleep disorders in shift workers was established. The predictive values of the model were evaluated using the receiver operating characteristic (ROC) curve and calibration curve based on the training set and validation set. Results: Totally 673 shift workers were included, with a median age of 32 (interquartile range, 12) years. There were 493 males, accounting for 73.25%. There were 471 (69.99%) workers in the training set and 202 (30.01%) workers in the validation set. There were 274 workers with sleep disorders, accounting for 40.71%. The equation for the prediction model was ln[p/(1-p)]=-8.391+1.906×average weekly working hours+1.822×depressive symptoms+1.667×anxiety symptoms. The area under the ROC curve was 0.769 (95%CI: 0.661-0.835) for the training set and 0.655 (95%CI: 0.593-0.737) for the validation set, and Hosmer-Lemeshow test showed a good fitting effect (both P>0.05). Conclusion: The nomograph model constructed by average weekly working hours, depressive symptoms and anxiety symptoms can be used to predict the risk of sleep disorders in shift workers of a chemical fiber enterprise.