Introduction
The global incidence of hip fracture is estimated at 16.75 million annually, and it is expected to increase further1,2. Hip fractures significantly impact mobility, leading to increased mortality, refracture rates, and the need for care3,4. Previous studies have shown that approximately 20–30% of patients who can walk pre-fracture fail to recover their pre-fracture walking status at 2 weeks or 6 months5,6. The inability to walk at hospital discharge is an independent predictor of 1-year mortality4. Therefore, enhancing postoperative mobility is crucial for reducing mortality and mitigating long-term care and health-care expenses.
The guidelines for hip fracture management have recommended that patients undergo physiotherapy assessment and, unless medically or surgically contraindicated, mobilization on the day after surgery because early mobilization is associated with survival and recovery for patients after hip fracture7,8,9. The National Hip Fracture Database defines mobilization as the ability to sit or stand out of bed with or without help7,10. Previous studies have shown that early mobilization leads to reduced mortality, reduced hospital stay, and increased ambulatory recovery7,11,12,13.
However, research on the impact of the early initiation of gait practice on gait recovery after surgery remains limited. A pivotal randomized controlled trial by Oldmeadow et al. demonstrated that patients who commenced ambulation practice within 48 h postoperatively exhibited superior gait reacquisition at one week and were more likely to be discharged to home compared to those who initiated ambulation practice later14. However, their study was limited by its single-center design, small sample size (n = 60), and focus on short-term outcomes, which may limit its generalizability to broader populations and clinical settings. Furthermore, the impact of early ambulation on longer-term outcomes, such as gait recovery at discharge, remains unclear.
To address these gaps, this multicenter cohort study aimed to evaluate the effects of early postoperative ambulation on gait recovery at one week postoperatively and at discharge. By utilizing a larger and more diverse sample, adjusting for a comprehensive set of confounding variables (such as cognitive impairment and preoperative mobility), and extending the assessment to discharge, this study provides a more robust and generalizable understanding of the benefits of early ambulation. We hypothesized that initiating ambulation within two days postoperatively would be associated with superior gait recovery at one week postoperatively and at discharge compared to late ambulation.
Methods
Study design and participants
This multicenter cohort study utilized the Nagano hip fracture database in Japan. This database contains information on the characteristics and rehabilitation outcomes of patients admitted with hip fracture, and between December 1, 2019, and July 31, 2023, data of 1613 surgically treated consecutive patients from 17 hospitals were submitted. These data were collected through a cloud electronic data capture system from each hospital. Patients who were ambulatory, aged 65 years or older before the fracture, had a femoral neck or trochanteric fracture, and admitted to an acute care hospital were included in the study. Patients with postoperative weight-bearing restrictions were excluded from the study. A flowchart of the patient selection process is presented in Fig. 1.
Patients’ recruitment and flow diagram.
Ethical considerations
This study was approved by the Ethics Committee of the School of Medicine, Shinshu University on November 12, 2019 (protocol number: 4541). All procedures complied with the ethical standards outlined in the Declaration of Helsinki. Due to the observational nature of the study and its reliance on routinely assessed medical evaluation items, the Ethics Committee of the School of Medicine, Shinshu University, waived the need to obtain informed consent. Instead, an opt-out approach was employed, with study details disclosed publicly at each participating site, allowing participants or their representatives to decline participation if desired. Detailed information on the study, including objectives, inclusion and exclusion criteria, and primary outcomes, was registered and made publicly available in the University Hospital Information Network (UMIN-CTR, unique identifier: UMIN000054114).
Measurement
The database contains the following information: patient background factors such as age, sex, body mass index, medical history (respiratory, cardiovascular, and neurological diseases), cognitive function before injury, mobility before injury, place of residence before admission, residence after discharge, and medical factors such as fracture type, surgical procedure, complications (deep vein thrombosis, peroneal nerve palsy, infection, and falls), waiting days for surgery, waiting days for rehabilitation from admission, and rehabilitation factors (postoperative days to the start of ambulation and rehabilitation intervention time). Mobility before injury was assessed based on the use of assistive devices and indoor movement. However, outdoor walking ability and walking distance were not specifically evaluated. Infections included pneumonia and urinary tract infections, and those diagnosed by a doctor were defined as having an infection. Cognitive function was assessed using the Degree of Daily Life Independence Score for People with Dementia (DDLIS-PD)15,16. This assessment has seven scales and is widely used to evaluate dementia in Japan. In this study, cognitive impairment was defined as DDLIS-PD grade II, with independence by support with some hindrances, or higher grades. FIM17 at 1 day after surgery, 1 week postoperatively, and at discharge and walking status at discharge (e.g., walking without aids, with one-point cane, with walker and other aids) were assessed as treatment outcomes.
Dependent variables
The main outcome was independent walking regardless of the use of aids, defined as mobility (walking) with FIM grade ≥ 5, following the approach outlined by Fu et al.17. A FIM score of ≥ 5 for walking indicates that the patient is able to walk with supervision or independently18. Walking ability was assessed 1 week postoperatively and at discharge. The secondary treatment outcome was walking recovery, which was defined as restoration of pre-fracture walking status and FIM grade ≥ 5. The other treatment outcomes were FIM score at discharge, walking status at discharge (ambulating without or with one-point cane, walker, or wheelchair), and days from operation to discharge.
Independent variables
The independent variable was early postoperative ambulation following hip fracture surgery. Patients were divided into two groups according to the interval between surgery and their first postoperative walk: the EA group (initiation of ambulation on postoperative day 1 or 2) and LA group (initiation of ambulation on postoperative day 3 or later). This classification was based on prior reports8,14 and an analysis of the relationship between ambulation initiation timing and independent walking at discharge in the present study. To establish the optimal cutoff point, receiver operating characteristic (ROC) analysis was performed, and the Youden index was used to determine the threshold, resulting in the classification of EA and LA groups. The ROC curve illustrating this analysis is presented in Supplementary Fig. S1. The robustness of this cutoff was further evaluated through stratified analysis (see Statistical Analyses for details). The date of ambulation initiation was defined as the day when a patient first engaged in ambulation, irrespective of the level of assistance required or the use of walking aids.
We adjusted for confounding variables identified both from previous studies and statistically significant differences in demographic and clinical characteristics between the two groups (Table 1), including age19,20,21,22,23; walking status before injury21,22,23,24,25; cognitive impairment12,20,21,22,24,25,26; preoperative medical history18,19,21,23 of respiratory, circulatory, and neurological diseases; fracture type19,21,23,24; and days from admission to surgery6,24, in the analysis of factors influencing postoperative walking independence.
Statistical analyses
Continuous variables are presented as medians (25th and 75th percentiles) and categorical variables as numbers and percentages. Baseline characteristics, rehabilitation factors, and treatment outcomes were compared between the EA and LA groups using Pearson’s chi-squared test, Fisher’s exact test, and the Mann–Whitney U test.
Non-parametric methods, including the Mann–Whitney U test, were applied for continuous variables due to deviations from normality based on Shapiro–Wilk test results. Fisher’s exact test was used for categorical variables when expected frequencies were less than five.
To determine the optimal cutoff for early ambulation, an ROC analysis was performed, and the threshold was determined using the Youden index27. Based on this cutoff, patients were classified into the EA (initiation of ambulation on postoperative day 1 or 2) and LA groups (initiation of ambulation on postoperative day 3 or later). To assess the robustness of this cutoff, a stratified analysis was conducted, in which the primary analysis was repeated across different patient subgroups.
We performed univariate logistic regression analysis to assess the association between early ambulation and independent walking at discharge without adjusting for confounding factors. Then, a multivariate logistic regression analysis (Model 1) and a generalized linear mixed model (Model 2) were conducted to ascertain whether early postoperative ambulation initiation affected independent walking at 1 week postoperatively and at discharge. In both analyses, the dependent variable was independent walking at 1 week postoperatively and at discharge, whereas the independent variable was early postoperative ambulation. All confounding factors were reduced to a single composite characteristic by applying a propensity score, which reflected the likelihood of study participants being assigned to the either the EA or LA group28. The propensity score was included as an adjustment factor in Model 1 and treated as a fixed effect in Model 2, with the hospital code incorporated as a random effect to account for inter-hospital variability. Missing values were excluded because they represented < 5% of the total28.
Furthermore, we conducted a stratified subgroup analysis using multivariate logistic regression analysis (Model 1 only) to examine the association between early ambulation and independent walking at discharge across different patient subgroups. The subgroup analysis aimed to determine whether the effect of early ambulation on independent walking at discharge varied based on specific patient characteristics, such as age, pre-fracture walking status, cognitive impairment, fracture type, and comorbidities.
Moreover, we conducted two additional subgroup analyses. First, a multivariate logistic regression analysis was performed in a subgroup restricted to patients who were able to engage in gait training during hospitalization to assess the association between early ambulation and walking independence at discharge. Second, a multivariate logistic regression model was applied to evaluate the adjusted odds ratios for walking independence at discharge by stratified postoperative ambulation timing (postoperative days 1–2 [reference], 3–4, and 5–6).
As a sensitivity analysis to assess the influence of missing data on postoperative walking timing and recovery, we performed logistic regression analysis with multiple imputations of missing sites. To address missing data in the dataset, we performed multiple imputation using the “mice” package in R, version 4.3.2 (https://www.r-project.org), creating 20 complete datasets using methods such as predictive mean matching for continuous variables and logistic regression for categorical variables. Subsequently, logistic regression models were applied to each imputed dataset, and the results across imputed datasets were integrated using Rubin’s rules29.
We further examined the potential impact of unknown confounders on the relationship between early postoperative ambulation and independent walking by calculating the E-values (https://www.evalue-calculator.com/)30,31. Statistical analyses were performed with R, version 4.3.2 (https://www.r-project.org). Statistical significance was set at P < 0.05.
Results
A total of 882 patients with a median age (25th, 75th percentiles) of 87 (81, 91) years from 10 acute hospitals were included, with 292 (33.1%) and 590 (66.9%) patients in the early ambulation (EA) and late-ambulation (LA) groups, respectively. The patients’ demographic data are presented in Table 1. The median ages (25th, 75th percentiles) of the EA and LA groups were 84 (79, 89) and 88 (83, 92) years, respectively; 78.8% and 46.6% walked without aids before the injury, and the percentages of cognitive impairment were 30.5% and 53.6%, respectively (Table 1).
The number of patients walking independently at 1-week postoperatively and at discharge was 156 (17.7%) and 292 (33.1%), respectively. The proportion of patients walking independently at 1-week postoperatively and at discharge decreased with increasing duration to postoperative first ambulation (Fig. 2). The EA group had a significantly higher percentage of independent walking patients, defined as walking Functional Independence Measure (FIM) grade ≥ 5 at 1-week postoperatively and at discharge, than the LA group (P < 0.0001; Table 2). The EA group had significantly better treatment outcomes than the LA group, including walking status (P < 0.0001) and FIM (P < 0.0001; Table 2) at discharge. The EA group showed a significant reduction in the median duration of hospital stay (P = 0.04) compared with the LA group. Furthermore, a higher proportion of patients in the EA group were discharged to home (EA: 19%; LA: 7.7%; P = 0.003).
Time to first ambulation after surgery and walking independently. Association between time to first ambulation after surgery and independent walking at 1 week postoperatively (left) and at discharge (right). The X-axis represents the number of days from surgery to the first ambulation in rehabilitation, while the Y-axis represents the proportion of patients walking independently (pink) and not walking independently (green). Earlier ambulation was associated with a higher proportion of independent walking, whereas delayed ambulation showed a decreasing trend in walking independence at both time points.
In the univariate logistic regression analysis, EA was significantly associated with independent walking both at 1 week postoperatively (odds ratio [OR] = 5.611, 95% confidence interval [CI] 3.879–8.114) and at discharge (OR = 5.314, 95% CI 3.918–7.207). Based on a multivariate logistic regression analysis (Model 1), EA (≤ 2 days from surgery to first ambulation) was associated with independent walking at 1 week postoperatively and at discharge after adjusting for confounders, with ORs of 3.27 (95% CI 2.17–4.94) and 3.33 (95% CI 2.38–4.69), respectively (Table 3). Furthermore, EA was associated with the recovery to pre-injury walking status at discharge after adjusting for confounders, with an OR of 3.05 (95% CI 1.59–5.93) (Table 4). These associations remained robust in the generalized linear mixed model (Model 2), which accounted for inter-hospital variability (Tables 3 and 4). Similarly, in the subgroup analysis, EA was significantly associated with independent walking at discharge across different patient subgroups (Supplementary Table S1 and Fig. S2). Additionally, a subgroup analysis restricted to patients who were able to initiate gait training during hospitalization was conducted. Consistent with the primary analysis, early ambulation remained significantly associated with walking independence at discharge in this restricted cohort (Supplementary Table S3). Further, a stratified analysis based on the timing of postoperative ambulation (postoperative days 1–2 [reference], 3–4, and 5–6) revealed that later initiation of ambulation was significantly associated with a lower odds of walking independence at discharge. The adjusted odds ratios were 0.42 (95% CI 0.28–0.63) for ambulation on days 3–4 and 0.26 (95% CI 0.15–0.51) for days 5–6 (Supplementary Table S4). A sensitivity analysis showed similar results in models incorporating multiple imputations of missing values for the completed data models. Subsequently, to evaluate the influence of potential unmeasured confounders on the relationship between early postoperative ambulation and independent walking at 1 week postoperatively and at discharge, we calculated the E-values, which were 3.02 and 3.05, respectively.
Furthermore, we examined the robustness of the early ambulation cutoff (≤ 2 days) through stratified analyses. In most subgroups, the optimal cutoff value for early ambulation remained at ≤ 2 days; however, in patients with cognitive impairment, the optimal cutoff changed to ≤ 3 days (Supplementary Table S2).
Discussion
This multicenter cohort study investigated the effect of early postoperative ambulation on gait recovery at the initial postoperative week and at discharge after hip fracture surgery in older patients, and the results showed that the initiation of ambulation within 2 postoperative days after hip fracture surgery was associated with independent walking at 1 week postoperatively and at discharge. Furthermore, early postoperative ambulation may influence the recovery of walking to pre-injury status. Other effects of early initiation of ambulation were reduced length of hospital stay, discharge home, and improved ability to perform activities of daily living.
The findings of this study align with the results of previous research14,32, which demonstrated that early ambulation improves functional outcomes. However, unlike prior studies, this was a multicenter cohort study that accounted for inter-hospital variability using a generalized linear mixed model. This approach confirmed the robustness of early ambulation as a predictor of postoperative gait recovery across diverse hospital settings. Furthermore, while most previous studies focused on short-term outcomes, our study demonstrated a longer-term effect on gait recovery, as observed at discharge. These findings suggest that early ambulation is broadly beneficial, regardless of variations in hospital rehabilitation protocols and policies.
In addition to accounting for inter-hospital variability, our study carefully adjusted for key confounding factors identified in previous research, including age19,20,21,22,23, pre-fracture mobility21,22,23,24,25, cognitive function12,20,21,22,24,25,26 comorbidities17,20,22, fracture type19,21,23,24, and days from admission to surgery6,24. Furthermore, stratified subgroup analyses confirmed the robustness of our findings, demonstrating that the significant association between early ambulation and independent walking at discharge was consistent across different patient characteristics.
Although the association between early postoperative ambulation initiation and gait reacquisition in this study is unclear, it is possible that early ambulation initiation prevented disuse and increased the amount of ambulation practice prior to discharge compared with late ambulation initiation, which may have resulted in the reacquisition of independent walking. Shimizu et al. and Marsault et al. reported that higher physical activity during hospitalization for hip fracture patients was associated with higher ability to perform activities of daily living at discharge33,34. However, these studies assessed physical activity for only a short period of 3 days to approximately 1 week33,34; therefore, future studies are needed to investigate the relationships between postoperative physical activity and independent walking.
Regarding the optimal cutoff for early ambulation, this study set the threshold at ≤ 2 days based on prior studies8,14 and the ROC analysis results. Stratified analyses confirmed that this cutoff was generally appropriate across most patient subgroups. However, in patients with cognitive impairment, the optimal cutoff was ≤ 3 days, suggesting that the impact of early ambulation may differ in this population. This finding implies that while early ambulation is beneficial, individualized rehabilitation strategies may be necessary for patients with cognitive impairment.
There are several unmeasured confounding factors in this study35, such as pain36, delirium37, nutritional status6,25,38, and inflammation6,32, which affect early postoperative ambulation initiation and gait reacquisition according to previous studies. Postoperative hip fracture-related pain was associated with a trochanteric fracture36; thus, we can potentially take into account the impact of pain on gait reacquisition by adjusting for the confounding factor of a trochanteric fracture. With regard to delirium, a previous study reported a strong association between delirium and cognitive decline39. The present study adjusted for cognitive decline as a confounding factor, which is believed to consider the influence of delirium. In the present study, to evaluate the influence of potential unmeasured confounders, we calculated the E-values, and the resulting E-values, 3.02 and 3.05, exceeded the odds ratio of unobserved confounders independently affecting the association between early ambulation and postoperative ambulation, suggesting a minimal effect of unknown or unmeasured confounders on this association.
This study has some limitations. First, the reasons for not being able to start walking early postoperatively were unclear. Oldmeadow et al. reported that early postoperative ambulation initiation failed when patients were medically unstable14. Other studies have found that reasons for delaying mobilization include pain, fatigue, and habitual cognitive status36. Additionally, while we accounted for inter-hospital variability using a generalized linear mixed model, we did not directly assess the specific policies of individual hospitals, decision-making processes of physicians regarding early ambulation initiation, or mechanisms of fracture (low-energy, high-energy, or pathological fractures). These institutional, clinical, and patient-related factors may influence the timing of ambulation and rehabilitation progress and should be explored in future research. Furthermore, while our stratified analysis confirmed the general validity of the 2-day cutoff, more research is needed to clarify why the optimal cutoff differed in patients with cognitive impairment. Second, the effect of early postoperative ambulation initiation on long-term gait reacquisition 6 months or 1 year later is unknown. Finally, in this study, walking independence was defined as walking FIM ≥ 5, which is the ability to walk independently with or without aids for at least 15 m, which indicates the least ability of independent indoor walking21, similar to the findings of Fu et al. Some walking status assessments were used in the previous studies, such as FIM17, the Iowa Level of Assistance scale14,32, the Cumulated Ambulation Score12, and walking status according to use of walking aids6,21. In the future, more applied assessments of walking ability will need to use outcomes other than FIM.
The strengths of this study include the large population with a multicenter cohort, the prospective design, and the systematic data collection, which eliminate single-center bias and may increase the generalizability of the results. Importantly, this study is the first to demonstrate that the effect of early ambulation remains significant after adjusting for inter-hospital variability, and that the optimal cutoff may differ in patients with cognitive impairment. These insights may contribute to tailored rehabilitation strategies, emphasizing that ambulation should begin as early as feasible while considering patient-specific factors.
In conclusion, the study results indicate that initiating ambulation within 2 postoperative days after hip fracture surgery affects independent walking at postoperative week 1 and at discharge. Furthermore, early postoperative ambulation may influence the recovery of walking to pre-injury status. However, the reasons for delayed ambulation postoperatively were unclear. Future investigation into the reasons for delayed ambulation could lead to a multidisciplinary team approach to encourage early ambulation within 2 days of surgery.
Data availability
The data that support the findings of this study are available from the corresponding author, Keisuke Nakamura, upon reasonable request.
References
Takusari, E. et al. Trends in hip fracture incidence in Japan: Estimates based on nationwide hip fracture surveys from 1992 to 2017. JBMR Plus 5, e10428. https://doi.org/10.1002/jbm4.10428 (2021).
Feng, J. N. et al. Global burden of hip fracture: The global burden of disease study. Osteoporos. Int. 35, 41–52 (2024).
Van Haecke, A. et al. Incidence and risk factors for bilateral proximal femoral fractures. Orthop. Traumatol. Surg. Res. 108, 102887. https://doi.org/10.1016/j.otsr.2021.102887 (2022).
Dubljanin-Raspopović, E. et al. Use of early indicators in rehabilitation process to predict 1-year mortality in elderly hip fracture patients. Hip Int. 22, 661–667 (2012).
Fukui, N., Watanabe, Y., Nakano, T., Sawaguchi, T. & Matsushita, T. Predictors for ambulatory ability and the change in ADL after hip fracture in patients with different levels of mobility before injury: A 1-year prospective cohort study. J. Orthop. Trauma. 26, 163–171 (2012).
Mashimo, S. et al. The impact of early mobility on functional recovery after hip fracture surgery. Disabil. Rehabil. 45, 4388–4393 (2023).
Goubar, A. et al. The 30-day survival and recovery after hip fracture by timing of mobilization and dementia: a UK database study. Bone Joint J. 103–B, 1317–1324 (2021).
National Clinical Guideline Centre (UK). The management of hip fracture in adults. UK: Royal College of Physicians; p. 2023 (Updated). (2011). https://www.nice.org.uk/
Sheehan, K. J. et al. Discharge after hip fracture surgery in relation to mobilisation timing by patient characteristics: Linked secondary analysis of the UK National hip fracture database. BMC Geriatr. 21, 694. https://doi.org/10.1186/s12877-021-02624-w (2021).
Royal College of Physicians. Falls and fragility fracture audit programme. National Hip Fracture Database (NHFD) Annual Report. In Royal College of Physicians (2019). https://www.nhfd.co.uk/
Sheehan, K. J. et al. Discharge after hip fracture surgery by mobilisation timing: Secondary analysis of the UK National hip fracture database. Age Ageing 50, 415–422 (2021).
Yamamoto, N. et al. Cumulated ambulation score as predictor of postoperative mobility in patients with proximal femur fractures. Arch. Orthop. Trauma. Surg. 143, 1931–1937 (2023).
Ferris, H., Brent, L. & Coughlan, T. Early mobilisation reduces the risk of in-hospital mortality following hip fracture. Eur. Geriatr. Med. 11, 527–533 (2020).
Oldmeadow, L. B. et al. No rest for the wounded: Early ambulation after hip surgery accelerates recovery. ANZ J. Surg. 76, 607–611 (2006).
Obayashi, K. & Masuyama, S. Pilot and feasibility study on elderly support services using communicative robots and monitoring sensors integrated with cloud robotics. Clin. Ther. 42, 364–371e4 (2020).
Nakanishi, K. et al. New score including daily life independence levels with dementia is associated with the onset of deep vein thrombosis in frail older adults. Geriatr. Gerontol. Int. 20, 414–421 (2020).
Fu, G. et al. Rapid preoperative predicting tools for 1-year mortality and walking ability of Asian elderly femoral neck fracture patients who planned for hip arthroplasty. J. Orthop. Surg. Res. 16, 455. https://doi.org/10.1186/s13018-021-02605-0 (2021).
Hamilton, B. B., Laughlin, J. A., Fiedler, R. C. & Granger, C. V. Interrater reliability of the 7-level functional independence measure (FIM). Scand. J. Rehabil Med. 26, 115–119 (1994).
Bellelli, G. et al. A prognostic model predicting recovery of walking independence of elderly patients after hip-fracture surgery. An experiment in a rehabilitation unit in Northern Italy. Osteoporos. Int. 23, 2189–2200 (2012).
Doherty, W. J. et al. Prediction of postoperative outcomes following hip fracture surgery: independent validation and recalibration of the Nottingham hip fracture score. J. Am. Med. Dir. Assoc. 22, 663–669e2 (2021).
Oba, T., Makita, H., Inaba, Y., Yamana, H. & Saito, T. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop. Traumatol. Surg. Res. 104, 1189–1192 (2018).
Tam, T. L., Tsang, K. K. & Lee, K. B. Development of a prognostic model to predict post-operative mobility of patients with fragility hip fractures: A retrospective cohort study. Int. J. Orthop. Trauma. Nurs. 38, 100770. https://doi.org/10.1016/j.ijotn.2020.100770 (2020).
Pajulammi, H. M., Pihlajamäki, H. K., Luukkaala, T. H. & Nuotio, M. S. Pre- and perioperative predictors of changes in mobility and living arrangements after hip fracture: A population-based study. Arch. Gerontol. Geriatr. 61, 182–189 (2015).
González de Villaumbrosia, C. et al. Predictive model of gait recovery at one month after hip fracture from a National cohort of 25,607 patients: the hip fracture prognosis (HF-Prognosis) tool. IJERPH. https://doi.org/10.3390/ijerph18073809 (2021).
Tomita, Y. et al. Clinical prediction model for postoperative ambulatory ability outcomes in patients with trochanteric fractures. Injury 52, 1826–1832 (2021).
Kamimura, T., Kobayashi, Y., Tamaki, S. & Koinuma, M. Impact of prefracture cognitive impairment and postoperative delirium on recovery after hip fracture surgery. J. Am. Med. Dir. Assoc. 25, 104961. https://doi.org/10.1016/j.jamda.2024.01.030 (2024).
Perkins, N. J. & Schisterman, E. F. The inconsistency of optimal cutpoints obtained using two criteria based on the receiver operating characteristic curve. Am. J. Epidemiol. 163, 670–675 (2006).
Biondi-Zoccai, G. et al. Are propensity scores really superior to standard multivariable analysis? Contemp. Clin. Trials 32, 731–740 (2011).
Marshall, A., Altman, D. G., Holder, R. L. & Royston, P. Combining estimates of interest in prognostic modelling studies after multiple imputation: current practice and guidelines. BMC Med. Res. Methodol. 9, 57. https://doi.org/10.1186/1471-2288-9-57 (2009).
VanderWeele, T. J. & Ding, P. Sensitivity analysis in observational research: Introducing the E-value. Ann. Intern. Med. 167, 268–274 (2017).
Mathur, M. B., Ding, P., Riddell, C. A. & VanderWeele, T. J. Web site and R package for computing E-values. Epidemiology 29, e45–e47 (2018).
Morri, M. et al. Which factors are independent predictors of early recovery of mobility in the older adults’ population after hip fracture? A cohort prognostic study. Arch. Orthop. Trauma. Surg. 138, 35–41 (2018).
Shimizu, T., Kanai, C. & Asakawa, Y. Relationship between independence in activities of daily living at discharge and physical activity at admission of older postoperative hip fracture rehabilitation inpatients: A retrospective case-control study. Physiother. Res. Int 29, e2070. https://doi.org/10.1002/pri.2070 (2024).
Marsault, L. et al. Objectively measured physical activity and its association with functional independence, quality of life and in-hospital course of recovery in elderly patients with proximal femur fractures: A prospective cohort study. Rehabil. Res.Pract. 2020, 5907652. https://doi.org/10.1155/2020/5907652 (2020).
Xu, B. Y., Yan, S., Low, L. L., Vasanwala, F. F. & Low, S. G. Predictors of poor functional outcomes and mortality in patients with hip fracture: A systematic review. BMC Musculoskelet. Disord. 20, 568. https://doi.org/10.1186/s12891-019-2950-0 (2019).
Münter, K. H., Clemmesen, C. G., Foss, N. B., Palm, H. & Kristensen, M. T. Fatigue and pain limit independent mobility and physiotherapy after hip fracture surgery. Disabil. Rehabil. 40, 1808–1816 (2018).
Marcantonio, E. R., Flacker, J. M., Michaels, M. & Resnick, N. M. Delirium is independently associated with poor functional recovery after hip fracture. J. Am. Geriatr. Soc. 48, 618–624 (2000).
Lieberman, D., Friger, M. & Lieberman, D. Inpatient rehabilitation outcome after hip fracture surgery in elderly patients: A prospective cohort study of 946 patients. Arch. Phys. Med. Rehabil. 87, 167–171 (2006).
Yang, Y. et al. Risk factors for postoperative delirium following hip fracture repair in elderly patients: A systematic review and meta-analysis. Aging Clin. Exp. Res. 29, 115–126 (2017).
Acknowledgements
We thank all the rehabilitation staff and cooperating hospitals for their contributions to data collection.
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Nakamura, K., Kurobe, Y., Sue, K. et al. Impact of early postoperative ambulation on gait recovery after hip fracture surgery: a multicenter cohort study. Sci Rep 15, 12893 (2025). https://doi.org/10.1038/s41598-025-97632-w
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-025-97632-w