Abstract

Abstract Background/Objectives: Sleep quality is a critical determinant of recovery after total joint arthroplasty (TJA), yet multidimensional trajectories of sleep health remain poorly defined. This study aimed to describe the trajectory of sleep health and sleep quality before and after hip or knee arthroplasty, and to examine the influence of sex and surgical site. Methods: A single-centre, prospective cohort study was conducted including 316 patients scheduled for primary hip or knee arthroplasty. Sleep was assessed two months before surgery and at one and six months postoperatively, using the Pittsburgh Sleep Quality Index (PSQI) and the RU-SATED scale. Statistical analyses used linear mixedeffects models with random intercepts for participants. Effect sizes (Cohen’s d) were derived from the residual variance of the mixed models. Results: At baseline, 40% of patients had poor sleep quality. Estimated marginal mean PSQI scores were 5.30 (95% CI: 4.87 to 5.73) preoperatively, 5.71 (5.27 to 6.15) at one month, and 4.19 (3.73 to 4.65) at six months, representing a reduction of 1.11 points (95% CI: 0.69 to 1.54; Cohen’s d = 0.49, 95% CI: 0.30 to 0.68; p < 0.001) from baseline to six months. The proportion of patients with poor sleep quality (PSQI > 5) fell from 40% to 28%, a 12 percentage-point absolute reduction reflecting individual-level transitions across the validated clinical threshold. Exploratory component analyses showed improvements in subjective sleep quality, sleep duration, sleep efficiency, and sleep disturbance. RU-SATED scores increased from 9.72 (95% CI: 9.46 to 9.98) to 10.45 (10.19 to 10.72) at six months, an improvement of 0.73 points (95% CI: 0.47 to 1.00; Cohen’s d = 0.44, 95% CI: 0.27 to 0.60; p < 0.001). Women undergoing knee arthroplasty had significantly worse sleep scores than men at baseline and one month postoperatively, though both sexes followed parallel recovery trajectories (timeby- sex interaction p > 0.30), with the absolute sex difference narrowing by six months. In hip arthroplasty, no significant sex differences or time-by-sex interactions were observed after adjusting for the age imbalance between sexes. Conclusions: Arthroplasty was associated with significant improvements in multidimensional sleep health by six months, though the first postoperative month represents a period of stagnation or slight decline. Women undergoing knee arthroplasty consistently reported worse sleep than men, although the recovery trajectory was parallel between sexes. These findings highlight the potential value of integrating sleep assessment into perioperative care, particularly for women scheduled for knee arthroplasty, though whether targeted sleep interventions improve clinical outcomes remains to be established through prospective intervention studies.