Postoperative Sleep Disturbance and Functional Recovery Following Hip Fracture Surgery: A 13‐Month Prospective Longitudinal Cohort Study

ABSTRACT Background Sleep disturbances are common after hip fracture surgery and may adversely affect postoperative recovery. However, the longitudinal course of sleep quality and its relationship with functional recovery remain poorly understood. This study aimed to characterize postoperative changes in sleep quality and examine their association with functional outcomes in patients with hip fractures. Methods A total of 323 patients with hip fractures and no preoperative sleep disturbance were enrolled and followed for up to 13 months. Sleep quality and functional status were assessed at each postoperative visit with the Pittsburgh Sleep Quality Index (PSQI) and Barthel Index (BI). Temporal changes were characterized, and the association between PSQI and BI was evaluated using correlation analysis and adjusted linear mixed‐effects models. A clinically specified Firth penalized logistic regression model was also performed to identify risk factors for postoperative sleep disturbance (PSD). Results A total of 1521 follow‐up observations were included. The proportion of patients with PSD decreased from 88.2% at 1 month to 27.8% at 13 months after surgery, with the highest burden observed at 1 month (mean PSQI score, 7.72). Over time, PSQI scores decreased significantly ( β = −0.151, 95% CI −0.188 to −0.114, p < 0.001), whereas BI scores increased ( β = 3.041, 95% CI 2.701 to 3.380, p < 0.001). Higher PSQI scores were associated with lower BI scores ( r = −0.182, p < 0.001), and this association remained significant after multivariable adjustment ( β = −0.014, 95% CI −0.023 to −0.005, p = 0.002). Postoperative PSQI scores improved significantly across all subgroups over time (all p < 0.001). A significant follow‐up month × sex interaction ( β = 0.11, p = 0.013) indicated slower sleep recovery in males than females, whereas fracture type and surgical procedure did not significantly modify the longitudinal PSQI trajectories. A clinically prespecified Firth penalized logistic regression model showed that a higher preoperative ACCI score was independently associated with PSD at 1 month (adjusted OR = 1.226, 95% CI 1.005–1.495, p = 0.044). Conclusions PSD was common after hip fracture surgery and gradually improved over time, although recovery was slower in male patients. PSD was independently associated with worse functional status during follow‐up, although residual psychological confounding cannot be excluded, while a higher preoperative comorbidity burden may increase the risk of PSD. Sleep disturbance is a readily screenable and potentially modifiable factor, supporting routine sleep screening during postoperative rehabilitation and further evaluation of targeted sleep interventions.