Association between sleep duration, depression and gynecological cancer in the United States: a national health and nutrition examination survey analysis2007–2018

Objective This study investigates the relationship between gynecological cancer and sleep duration, depression, and evaluates machine learning models for predicting gynecological cancer risk using data from 30,213 female participants in the NHANES (2007–2018), with 405 diagnosed cases. Methods Depression was assessed via PHQ-9, and sleep duration was categorized as short (< 7 hours), normal (7–9 hours), and long (> 9 hours). Multivariable logistic regression and six machine learning algorithms (AdaBoost, RF, Boost Tree, ANN, XGBoost, SVM) were employed to analyze associations and predict cancer occurrence. Results The average age of cancer patients was 53.0 years, higher than the non-cancer group’s 46.6 years. Significant differences between groups were found in race, education, PIR, smoking, alcohol use, BMI, waist circumference, hypertension, diabetes, and depression. Depression increased gynecological cancer risk (OR = 2.28, 95%CI: 1.80–2.91), while sleep duration showed no significant association. The Boost Tree model had the best performance (AUC = 0.70, sensitivity = 4%, specificity = 99%). Conclusion Depression significantly raises gynecological cancer risk. The Boost Tree model effectively predicts gynecological cancer, with potential for clinical early screening.