Perioperative hyponatremia and the risk of balance-function deterioration during rehabilitation after cephalomedullary nailing for hip fracture in older adults: A retrospective cohort study.

Abstract

Balance recovery after hip fracture surgery is vital for mobility and fall prevention, yet modifiable perioperative predictors of poor rehabilitation gains remain unclear. Hyponatremia is common in older surgical patients and may impair neuromuscular function, but whether it represents a direct causal factor or an early perioperative risk marker for poor balance recovery is uncertain. We performed a retrospective cohort study of adults ≥65 years who underwent cephalomedullary nailing for hip fracture and completed inpatient rehabilitation (2019-2024). Perioperative hyponatremia was any serum sodium <135 mmol/L from 48 h pre-incision through postoperative day 3 or rehabilitation transfer. Hyponatremia burden was quantified as the area under the 135 mmol/L threshold (AUT135; mmol·h/L). The primary outcome was balance under-recovery, defined a priori as a Berg Balance Scale (BBS) increase <11.5 points from rehabilitation admission to discharge (hip fracture-specific MCID). Secondary outcomes included BBS recovery rate and Timed Up and Go (TUG) change; deterioration beyond MDC95 and in-rehabilitation falls were exploratory. Confounding was addressed using stabilized inverse probability of treatment weights and weighted regression with robust standard errors; sensitivity analyses additionally incorporated baseline rehabilitation function and management-related variables. Among 780 patients, 156 (20.0%) developed perioperative hyponatremia and 297 (38.1%) had balance under-recovery. Most hyponatremia was mild (121/156, 77.6%), and active sodium-directed management was usually conservative. Hyponatremia was associated with under-recovery (adjusted OR 1.62, 95% CI 1.17-2.25; q = 0.007), and greater AUT135 showed a graded association (aOR 1.17 per 50 mmol h/L, 95% CI 1.08-1.27; q = 0.001). Hyponatremia was also associated with slower BBS improvement (-1.3 points/week, 95% CI-2.1 to -0.5; q = 0.013) and smaller TUG gains (+1.4 s, 95% CI 0.5-2.3; q = 0.007). The primary association remained directionally similar after additional adjustment for admission BBS/TUG and for management-related variables. Perioperative hyponatremia and greater time-weighted sodium deficit were independently associated with poorer balance recovery and slower functional gains after hip fracture surgery and may be best interpreted as early risk-stratification signals of broader vulnerability. Prospective studies should test whether structured sodium monitoring and guideline-concordant management improve rehabilitation outcomes.