Key points are not available for this paper at this time.
Background Sleep disorders represent a common and impactful complication following acute ischemic stroke (AIS). This study aimed to identify clinical risk factors and evaluate the predictive value of serum hypocretin (Hcrt), neurofilament heavy chain (NfH), and interleukin-1 beta (IL-1β) for post-stroke sleep disorders. Methods We conducted a retrospective observational study of 256 patients with AIS. Patients were classified into sleep disorder ( n = 161) and non-sleep disorder ( n = 95) groups based on their Pittsburgh Sleep Quality Index scores 7 days after stroke onset. Fasting serum levels of Hcrt, NfH, and IL-1β were measured upon admission. We utilized multivariate logistic regression and receiver operating characteristic (ROC) curves to evaluate predictive performance. The combined model was internally validated using 1,000 bootstrap resamples to assess optimism-corrected discriminative performance. Results Sleep disorders were present in 62.9% of patients. Nine independent risk factors were identified: age ≥ 65 years (OR = 2.059), snoring history (OR = 1.980), prior stroke (OR = 2.036), lower ADL scores (OR = 1.839), higher HAMD (OR = 1.726) and NIHSS scores (OR = 1.677), decreased serum Hcrt (OR = 1.863), elevated NfH (OR = 2.020), and elevated IL-1β (OR = 1.793; all p 0.05). Individual biomarker AUCs ranged from 0.742 to 0.781, whereas the combined three-biomarker model achieved a significantly superior AUC of 0.874 (sensitivity 88.82%, specificity 71.58%). Bootstrap internal validation yielded a mean optimism-corrected AUC of 0.861 (95% CI: 0.812–0.903), indicating robust model performance with minimal overfitting. Conclusion Clinical variables alongside altered levels of Hcrt, NfH, and IL-1β serve as independent predictors of post-stroke sleep disorders. The combined three-biomarker panel, reflecting neuroendocrine dysregulation, axonal injury, and systemic inflammation, demonstrates substantially superior predictive accuracy over individual biomarkers and offers a clinically practical tool for early identification of high-risk patients.
Wang et al. (Thu,) studied this question.