Sleep Disorders in Patients With PCOS After Ketogenic Diet Intervention: An Analysis of Influencing Factors and...

To explore the factors influencing sleep disorders in patients with polycystic ovary syndrome (PCOS) after ketogenic diet intervention and establish a predictive model.

Data of 220 PCOS patients undergoing ketogenic diet intervention at Beijing Shijitan Hospital, Capital Medical University, from January 2021 to December 2023 were retrospectively collected. Patients were randomly divided into the modelling group (132 patients) and the validation group (88 patients) in a 3:2 ratio. The modelling group was further divided into the sleep disorder group (56 patients) and the non-sleep disorder group (76 patients). Univariate and binary logistic regression analyses were conducted to determine the influencing factors of sleep disorders after administering a ketogenic diet intervention in patients with PCOS. The predictive model was constructed using SPSS, and analyses of receiver operating characteristic (ROC) curves, calibration curves and decision curve analysis (DCA) were performed in the R language to evaluate the clinical practicality of the model.

In the modelling group, anxiety (odds ratio [OR] = 1.768, 95% confidence interval [CI]: 1.284-2.434, p < 0.001), depression (OR = 1.494, 95% CI: 1.193-1.872, p < 0.001), blood sugar status (impaired fasting glucose/tolerance) (OR = 5.278, 95% CI: 1.533-18.177, p = 0.008) and low-density lipoprotein cholesterol (LDL-c; OR = 1.619, 95% CI: 1.201-2.181, p = 0.002) were significant factors affecting sleep disorders. The prediction model incorporates these factors (X₁-X₄), and the model expression is Logit(P) = β_Constant + (β₁X₁) + (β₂X₂) + (β₃X₃) + (β₄X₄). The calibration curve showed good agreement between predicted risks and actual risks. ROC analysis showed that the area under the curve was 0.9328 (95% CI: 0.892-0.954) for the modelling group and 0.8431 (95% CI: 0.777-0.899) for the validation group, both indicating that the model has high accuracy. DCA curves showed that the model has significant positive net benefits and good clinical utility.

Anxiety, depression, blood sugar status and LDL-c are key factors influencing sleep disorders in patients with PCOS after ketogenic diet intervention. A prediction model featuring high accuracy and remarkable clinical utility was successfully established.