Abstract
[Purpose] We aimed to clarify the relationship between walking ability and cognitive function two weeks post-surgery in patients with hip fractures classified by age group. [Participants and Methods] In this single-center retrospective observational study, a total of 473 patients with hip fractures aged 75 years and older were included. The patients were classified into “old” (270 patients) and “super-old” (203 patients) groups. Multiple logistic regression analysis and receiver operating characteristic analysis were employed for data analysis. [Results] In the old group, the surgery type, Barthel index, Mini-Mental State Examination scores, and the three-day cumulative ambulation score were associated with the ability to walk independently at two weeks post-surgery. In the super-old group, Barthel index, Mini-Mental State Examination scores, and the three-day cumulative ambulation score were associated with the ability to walk independently at two weeks post-surgery. The Mini-Mental State Examination cutoff values (area under the curve) required for independent walking at two weeks post-surgery were ≥20 (0.80) for the old group and ≥24 (0.84) for the super-old group. [Conclusion] The findings suggest that maintaining cognitive function within the normal range may be important for super-old patients aged 90 years and older to be able to walk independently two weeks post-surgery.
Key words: Cognitive function, Hip fracture, Walking ability
INTRODUCTION
In Japan, the number of patients with hip fractures continues to rise as the population rapidly ages. Affected individuals range from older to much older patients. Most of the factors related to functional recovery in patients with hip fractures are baseline characteristics and biological and sociodemographic factors1). However, there is a lack of reliable and highly predictive clinical prediction models for older patients with hip fractures2).
Because baseline characteristics, complication rates, and mortality rates differ by age group3), establishing functional outcome indicators is needed for postoperative rehabilitation for different age groups. Age influences multiple outcomes in patients with hip fractures, and functional outcomes such as walking ability vary accordingly4). Aging affects both physical and cognitive functions.
Approximately 40% of patients with hip fractures experience cognitive impairment5). The degree of cognitive function is an important factor in determining the discharge destination6), functional outcomes7), and walking ability8), and strongly affects the progress of postoperative rehabilitation. Since cognitive decline is common in this population, cognitive assessments should not only classify patients as having normal or impaired cognitive function, but also incorporate cutoff values for each outcome indicator (e.g., discharge to home: Mini-Mental State Examination ≥17 points9); one-year mortality: Montreal cognitive assessment score <15 points10)). However, the relationship between short-term walking ability and cognitive function across different age groups has not been clarified. Understanding short-term walking ability in patients with hip fractures, categorized by age group, would provide valuable information for the development of postoperative rehabilitation plans.
The novelty of this study lies in classifying hip fracture patients into different age groups and verifying the relationship between early postoperative clinical assessment indicators and short-term walking ability. Assessment based on age stratification that reflects a super-aging society may contribute to improved rehabilitation medicine. In this study, we hypothesized that the relationship between short-term walking ability and cognitive function differs across the classified age groups. We aimed to clarify the relationship between walking ability and cognitive function two weeks after surgery in patients with hip fractures classified by age group.
PARTICIPANTS AND METHODS
The Institutional Review Board of Aizawa Hospital approved this study (approval number: 2025-018). Personal information was anonymized and protected. The requirement for obtaining written informed consent was waived because of the retrospective observational nature of the study. Using the opt-out method, information related to this study was posted on the hospital’s website.
This retrospective observational study was conducted at a single secondary emergency hospital (hereinafter referred to as “our hospital”). Patients included in the study were managed in the orthopedic or convalescent rehabilitation ward. Rehabilitation services are provided year-round at the hospital, and physical and occupational therapies begin on the day of admission or the following day.
The participants comprised 730 patients aged 75 years and older who underwent treatment for hip fracture at our hospital from January 2023 to August 2025. The exclusion criteria were as follows: inability to walk independently prior to hospitalization, high-energy trauma, basicervical femoral neck fracture, total hip replacement, missing cognitive function assessment, and death.
Data collected from medical records included age, sex, fracture type (femoral neck or trochanteric), surgery type (bipolar hip arthroplasty or osteosynthesis), days to surgery, postoperative hospitalization period, prior to injury activities of daily living, cognitive function, activity to sit or stand out of bed on the day after surgery, early postoperative mobility, and walking ability. Age stratification followed the criteria established by Japan Geriatrics Society11), classifying patients aged 75–89 years as old and those aged 90 years and older as super-old. Walking ability was assessed using the Functional Ambulation Categories12, 13). The Functional Ambulation Categories is a six-point scale based on the level of independence in ambulation (0, unable to walk; 1, moderate to significant assistance required; 2, mild assistance required; 3, supervision required; 4, independent on level ground; 5, independent regardless of location). In this study, to minimize bias due to differences in evaluation timing, walking ability at two weeks post-surgery was used as the main outcome. Walking ability two weeks post-surgery is a clinical indicator when considering discharge destinations after acute care management. Patients with a Functional Ambulation Categories score of 4 or higher were classified as the independent walking group, and those with a score of 3 or lower were classified as the non-independent walking group.
Cognitive function was assessed using the Mini-Mental State Examination administered upon admission. The Mini-Mental State Examination is a simple screening method for evaluating cognitive decline, with scores ranging from 0 to 30 points; lower scores indicate greater cognitive impairment.
Early postoperative mobility was assessed using the cumulative ambulation score14), and the total cumulative ambulation score from postoperative days one to three (three-day cumulative ambulation score) was calculated. The three-day cumulative ambulation score is a more accurate assessment than the single-day cumulative ambulation score15) and has been reported to be associated with short-term walking ability16).
Prior to injury activities of daily living were evaluated using the Barthel index. The ability to perform activities of daily living before injury is also a factor that affects walking ability17).
Post-hoc power analysis of the sample size (independent walking group/non-independent walking group old group 102 patients/168 patients, super-old group 26 patients/177 patients) was performed using G*Power18, 19). The effect size (d) was calculated as 1.24/1.40 based on the mean and standard deviation of the Mini-Mental State Examination scores (old group 24.6 ± 3.7/17.3 ± 7.4, super-old group 24.2 ± 3.7/15.9 ± 7.5). With an alpha error of 5%, the resulting power reached 1.00 for the old group and 0.99 for the super-old group.
For group comparisons across different age groups, the χ2 test or the Mann–Whitney U test was used. Multiple logistic regression analysis (forward selection using the likelihood ratio method) was performed to identify factors associated with walking ability two weeks after surgery in each age group. Walking ability served as the dependent variable. The independent variables included sex, fracture type, surgery type, days to surgery, Barthel index, Mini-Mental State Examination, activity to sit or stand out of bed on the day after surgery, and the three-day cumulative ambulation score. Receiver operating characteristic analysis was then used to determine the Mini-Mental State Examination, three-day cumulative ambulation score, and Barthel index scores for independent walking at two weeks post-surgery, with cut-off values calculated using Youden’s index. The results of the multiple logistic regression and receiver operating characteristic analyses were also used to perform sensitivity analysis. IBM SPSS Windows version 26 (IBM Japan, Ltd., Tokyo, Japan) was used, and the significance level was set at 5%.
RESULTS
Of the 730 participants, 257 were excluded due to inability to walk independently prior to hospitalization (208), high-energy trauma (19), basicervical femoral neck fracture (seven), total hip replacement (four), missing cognitive function assessment (seven), and death (12). The final analysis included 473 patients. The patients were classified into two groups, with 270 patients in the old group and 203 patients in the super-old group. By two weeks after surgery, 102 patients in the old group and 26 patients in the super-old group were able to walk independently (Fig. 1).
Fig. 1.
Flowchart of patient participation.
Table 1 shows the basic characteristics of the participants and the results of the group comparisons. The participants comprised 104 men and 369 women, with an average age of 87.8 ± 6.0 years.
Table 1. Participant characteristics and group comparisons.
| Item | Total (n=473) | Old patients (aged 75–89 years) (n=270) |
Super-old patients (aged 90 years or older)
(n=203) |
|||||||
| Independent walking group (n=102) | Non-independent walking group (n=168) | Effect size | Independent walking group (n=26) | Non-independent walking group (n=177) | Effect size | |||||
| Age (years)† | 87.8 ± 6.0 | 83.0 [79.3–86.0] | 85.0 [82.0–88.0] | ** | 0.22 | 91.0 [90.0–93.0] | 93.0 [91.0–95.0] | ** | 0.21 | |
| Sex (n, %)‡ | 0.06 | 0.15 | ||||||||
| Female | 369 (78.0) | 74 (27.4) | 131 (48.5) | 17 (8.4) | 147 (72.4) | |||||
| Male | 104 (22.0) | 28 (10.4) | 37 (13.7) | 9 (4.4) | 30 (14.8) | |||||
| Fracture type (n, %)‡ | ** | 0.24 | ** | 0.20 | ||||||
| Femoral neck | 257 (54.3) | 76 (28.1) | 85 (31.5) | 19 (9.4) | 77 (37.9) | |||||
| Trochanteric | 216 (45.7) | 26 (9.6) | 83 (30.7) | 7 (3.4) | 100 (49.3) | |||||
| Surgery type (n, %)‡ | ** | 0.21 | ** | 0.22 | ||||||
| Bipolar hip arthroplasty | 234 (49.5) | 68 (25.2) | 76 (28.1) | 19 (9.4) | 71 (35.0) | |||||
| Osteosynthesis | 239 (50.5) | 34 (12.6) | 92 (34.1) | 7 (3.4) | 106 (52.2) | |||||
| Days to surgery (days)† | 2.0 ± 3.5 | 1.0 [1.0–2.0] | 1.0 [1.0–2.0] | 0.03 | 1.0 [1.0–1.0] | 1.0 [1.0–2.0] | 0.07 | |||
| Postoperative hospitalization period (days)† | 26.1 ± 15.0 | 19.0 [16.3–23.0] | 24.0 [18.0–32.3] | ** | 0.30 | 19.5 [16.3–24.3] | 23.0 [18.0–33.0] | ** | 0.19 | |
| Barthel index (points)† | 84.9 ± 18.2 | 100.0 [95.0–100.0] | 90.0 [75.0–95.0] | ** | 0.47 | 100.0 [95.0–100.0] | 85.0 [70.0–90.0] | ** | 0.40 | |
| Mini-mental state examination (points)† | 18.8 ± 7.5 | 25.0 [22.0–28.0] | 17.5 [12.0–23.3] | ** | 0.50 | 25.0 [24.0–26.0] | 17.0 [11.0–22.0] | ** | 0.40 | |
| Activity to sit or stand out of bed on the day after surgery (n, %)‡ | ** | 0.24 | 0.12 | |||||||
| Yes | 410 (86.7) | 100 (37.0) | 137 (50.7) | 25 (12.3) | 148 (72.9) | |||||
| No | 63 (13.3) | 2 (0.7) | 31 (11.5) | 1 (0.5) | 29 (14.3) | |||||
| Three-day cumulated ambulation score (points)† | 7.1 ± 2.5 | 9.0 [8.0–9.0] | 7.0 [5.0–8.0] | ** | 0.50 | 9.0 [8.0–9.0] | 6.0 [5.0–8.0] | ** | 0.32 | |
Data are provided as mean ± SD, median [interquartile range], or number (percentage). † Mann–Whitney U test, ‡ χ2 test, **p<0.01.
The proportion of patients who were able to walk independently two weeks after surgery was compared between age groups. The patients in the super-old group showed a lower proportion of independent walkers (26 patients, 12.8%) than those in the old group (102 patients, 37.8%) (p<0.01, adjusted residual 6.1, φ=0.28).
Surgery type, the three-day cumulative ambulation score, Barthel index, and Mini-Mental State Examination scores were associated with independent walking ability two weeks post-surgery in the old group. The three-day cumulative ambulation score, Barthel index, and Mini-Mental State Examination scores were associated with independent walking ability in the super-old group (Table 2).
Table 2. Factors associated with walking ability at two weeks post-surgery.
| Item | Odds ratio | 95% confidence interval | ||
| Old patients †1 | Surgery type | ** | 3.08 | 1.53–6.21 |
| Barthel index | * | 1.04 | 1.01–1.08 | |
| Mini-mental state examination | ** | 1.23 | 1.14–1.33 | |
| Three-day cumulated ambulation score | ** | 1.63 | 1.32–2.01 | |
| Super-old patients †2 | Barthel index | ** | 1.14 | 1.04–1.25 |
| Mini-mental state examination | * | 1.17 | 1.02–1.33 | |
| Three-day cumulated ambulation score | ** | 1.84 | 1.26–2.69 |
†1Modelχ2 test: p<0.01, Hosmer–Lemeshow test: p=0.71, percentage of correct classifications: 85.6%, Nagelkerke R2: 0.57. †2Modelχ2 test: p<0.01, Hosmer–Lemeshow test: p=0.93, percentage of correct classifications: 91.1%, Nagelkerke R2: 0.53. *p<0.05, **p<0.01.
Figure 2 shows the cutoff values (area under the curve) for determining independent walking ability two weeks after surgery. The Mini-Mental State Examination cutoff values were ≥20 (0.80) for the old group and ≥24 (0.84) for the super-old group. The three-day cumulative ambulation score cutoff values were ≥9 (0.79) for the old group and ≥8 (0.77) for the super-old group. The Barthel index cutoff values were ≥98 (0.77) for the old group and ≥93 (0.85) for the super-old group).
Fig. 2.
Cutoff values for determining walking independence at two weeks post-surgery.
The results are presented as follows: Item (figure symbols)/cutoff value/area under the curve/95% confidence interval/sensitivity/specificity. The circular mark on the receiver operating characteristic curve represents the cutoff value. Old patients: Mini-Mental State Examination (a)/ ≥20/0.80/0.75-0.85/0.93/0.57, three-day cumulative ambulation score (b)/ ≥9/0.79/0.74-0.85/0.73/0.76, Barthel index (c)/ ≥98/0.77/0.71-0.82/0.68/0.76. Super-old patients: Mini-Mental State Examination (d)/ ≥24/0.84/0.77-0.91/0.81/0.85, three-day cumulative ambulation score (e)/ ≥8/0.77/0.68-0.85/0.81/0.64, Barthel index (f)/ ≥93/0.85/0.78-0.91/0.81/0.77.
DISCUSSION
The walking ability in patients with hip fractures classified by age group two weeks after surgery was examined. The proportion of patients who were able to walk independently at two weeks post-surgery was 37.8% in the old group and 12.8% in the super-old group. Factors associated with independent walking were similar between the two groups, although the cutoff values for the Mini-Mental State Examination, three-day cumulative ambulation score, and Barthel index differed. These results indicate that functional outcome indicators should be established separately for each age group.
The Mini-Mental State Examination cutoff values for determining independent walking ability were ≥20 points in the old group and ≥24 points in the super-old group. These values suggest that the super-old group may be important higher cognitive function than the old group to walk independently. It has been shown that the temporal relationship between cognitive function and physical ability is reversed across age groups (in the 75–85 age group, decline in walking ability is associated with decline in cognitive function, but in the 85–90 age group, decline in cognitive function is associated with decline in walking ability)20). In addition, some research21) has reported that physical function and cognitive function are strongly related in much older adults (90 years and older), and it is thought that the fact that cognitive function is maintained in the super-old group compared to the old group is associated with their independent walking ability. According to the classification of cognitive impairment severity, cognitive function in the old group fell within the mild (20–23 points)22) or moderate range (11–20 points)23). However, in the super-old group, cognitive function fell within the very mild (≥24 points)22) or mild range (21–25 points)23). Earlier studies demonstrated that cognitive function affects walking ability8) and that patients older than 90 years who regained walking ability had a lower incidence of dementia24). The present study adds to this evidence by suggesting that maintaining cognitive function within the normal range is associated with for short-term independent walking in super-old patients aged 90 years and older.
For super-old patients with MMSE scores below the cutoff value (≥24 points), increasing opportunities for walking in daily living activities is important, even in situations where the use of walking aids is essential. If the person forgets to use them, adjusting an environment where they could move safely without the aids would be necessary. Furthermore, we believe that the possibility of early discharge to the pre-injury living environment of the patient and home-based rehabilitation should be considered to improve walking functions.
The multiple logistic regression analysis identified the Mini-Mental State Examination, three-day cumulative ambulation score, and Barthel index as factors associated with independent walking ability two weeks post-surgery in both the old and super-old groups, confirming the robustness of the findings. By contrast, the cutoff values for the Mini-Mental State Examination, three-day cumulative ambulation score, and Barthel index differed between groups. The super-old group showed higher Mini-Mental State Examination scores than the old group, whereas the three-day cumulative ambulation score and Barthel index scores were lower in the super-old group than in the old group. These results suggest that the relationship between high and low values of each indicator might not be consistent when comparing cutoff values for the independent walking ability across age groups.
This study had several limitations. First, we used a single-center design and the interpretation of the results remains limited. The possibility of variability in rehabilitation and potential measurement bias could not be ruled out. However, we demonstrated that functional outcome indicators need to be set individually for each age group. We believe that conducting a multi-center collaborative study in the future would increase the likelihood of generalization. Second, although patients were classified into an old group aged 75–89 and a super-old group aged 90 and above, further subdivision and inclusion of patients under 75 years of age are necessary. Additional classifications based on age-related characteristics, such as body composition, may also strengthen the analysis. Third, this study excluded patients with an inability to walk independently prior to hospitalization, so interpretation of the results is limited to older people with high physical function. Therefore, generalizability might be limited and selection bias could be introduced. In addition, studies with larger samples are needed to clarify potential sex differences. Fourth, walking ability two weeks after surgery was used as the outcome measure, and long-term walking ability remained unexamined. Fifth, in the analyses presented in this study, we did not include confounding factors that could affect both cognitive function and walking ability, such as comorbidities, frailty, nutritional status, delirium, and postoperative complications.
In conclusion, the cutoff values for Mini-Mental State Examination, three-day cumulative ambulation score, and Barthel index used to determine independent walking ability two weeks post-surgery differed between the old and super-old groups. The findings also suggest that super-old patients aged 90 and above may be important cognitive function within the normal range to achieve independent walking ability two weeks post-surgery. Future research should investigate functional outcome measures classified by age group.
Funding
The authors received no financial support.
Conflict of interest
The authors declare no conflicts of interest.
Acknowledgments
The authors thank the physical and occupational therapists at Aizawa Hospital for their support in conducting the study.
REFERENCES
- 1.Araiza-Nava B, Méndez-Sánchez L, Clark P, et al. : Short- and long-term prognostic factors associated with functional recovery in elderly patients with hip fracture: a systematic review. Osteoporos Int, 2022, 33: 1429–1444. [DOI] [PubMed] [Google Scholar]
- 2.Kaizu Y, Tamura S, Saito H, et al. : Clinical prediction models for nonmortality outcomes in older adults with hip fractures: a systematic review. J Gerontol A Biol Sci Med Sci, 2023, 78: 2363–2370. [DOI] [PubMed] [Google Scholar]
- 3.Shigemoto K, Sawaguchi T, Higashikawa T, et al. : Differences in age-related characteristics among elderly patients with hip fractures. J Orthop Sci, 2023, 28: 1131–1135. [DOI] [PubMed] [Google Scholar]
- 4.Hagino T, Sato E, Tonotsuka H, et al. : Prediction of ambulation prognosis in the elderly after hip fracture. Int Orthop, 2006, 30: 315–319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Seitz DP, Adunuri N, Gill SS, et al. : Prevalence of dementia and cognitive impairment among older adults with hip fractures. J Am Med Dir Assoc, 2011, 12: 556–564. [DOI] [PubMed] [Google Scholar]
- 6.Lee SW, Elsakr C, Joung KM, et al. : Outcomes after inpatient rehabilitation for older adults with hip fractures: a cross-sectional study. Clin Rehabil, 2025, 39: 35–46. [DOI] [PubMed] [Google Scholar]
- 7.Sheehan KJ, Williamson L, Alexander J, et al. : Prognostic factors of functional outcome after hip fracture surgery: a systematic review. Age Ageing, 2018, 47: 661–670. [DOI] [PubMed] [Google Scholar]
- 8.González de Villaumbrosia C, Barba R, Ojeda-Thies C, et al. : Predictive factors of gait recovery after hip fracture: a scoping review. Age Ageing, 2025, 54: afaf057. [DOI] [PubMed] [Google Scholar]
- 9.Chow IH, Miller T, Pang MY: Predictive factors for home discharge after femoral fracture surgery: a prospective cohort study. Eur J Phys Rehabil Med, 2023, 59: 743–753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wong RM, Ng RW, Chau WW, et al. : Montreal cognitive assessment (MoCA) is highly correlated with 1-year mortality in hip fracture patients. Osteoporos Int, 2022, 33: 2185–2192. [DOI] [PubMed] [Google Scholar]
- 11.Ouchi Y, Rakugi H, Arai H, et al. Joint Committee of Japan Gerontological Society (JGLS) and Japan Geriatrics Society (JGS) on the definition and classification of the elderly: Redefining the elderly as aged 75 years and older: proposal from the Joint Committee of Japan Gerontological Society and the Japan Geriatrics Society. Geriatr Gerontol Int, 2017, 17: 1045–1047. [DOI] [PubMed] [Google Scholar]
- 12.Holden MK, Gill KM, Magliozzi MR, et al. : Clinical gait assessment in the neurologically impaired. Reliability and meaningfulness. Phys Ther, 1984, 64: 35–40. [DOI] [PubMed] [Google Scholar]
- 13.Mehrholz J, Wagner K, Rutte K, et al. : Predictive validity and responsiveness of the functional ambulation category in hemiparetic patients after stroke. Arch Phys Med Rehabil, 2007, 88: 1314–1319. [DOI] [PubMed] [Google Scholar]
- 14.Foss NB, Kristensen MT, Kehlet H: Prediction of postoperative morbidity, mortality and rehabilitation in hip fracture patients: the cumulated ambulation score. Clin Rehabil, 2006, 20: 701–708. [DOI] [PubMed] [Google Scholar]
- 15.Ferriero G, Kristensen MT, Invernizzi M, et al. : Psychometric properties of the cumulated ambulation score: a systematic review. Eur J Phys Rehabil Med, 2018, 54: 766–771. [DOI] [PubMed] [Google Scholar]
- 16.Yamamoto N, Tomita Y, Ichinose A, et al. : Cumulated ambulation score as predictor of postoperative mobility in patients with proximal femur fractures. Arch Orthop Trauma Surg, 2023, 143: 1931–1937. [DOI] [PubMed] [Google Scholar]
- 17.Endo T, Tsukuda Y, Oishi A, et al. : Predictors of walking ability at discharge post proximal femoral fracture in an acute care hospital. J Phys Ther Sci, 2025, 37: 341–347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Faul F, Erdfelder E, Lang AG, et al. : G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods, 2007, 39: 175–191. [DOI] [PubMed] [Google Scholar]
- 19.Faul F, Erdfelder E, Buchner A, et al. : Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods, 2009, 41: 1149–1160. [DOI] [PubMed] [Google Scholar]
- 20.Stijntjes M, Aartsen MJ, Taekema DG, et al. : Temporal relationship between cognitive and physical performance in middle-aged to oldest old people. J Gerontol A Biol Sci Med Sci, 2017, 72: 662–668. [DOI] [PubMed] [Google Scholar]
- 21.Legdeur N, Badissi M, Venkatraghavan V, et al. : The temporal relation of physical function with cognition and the influence of brain health in the oldest-old. Gerontology, 2025, 71: 13–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zec RF, Landreth ES, Vicari SK, et al. : Alzheimer disease assessment scale: useful for both early detection and staging of dementia of the Alzheimer type. Alzheimer Dis Assoc Disord, 1992, 6: 89–102. [PubMed] [Google Scholar]
- 23.Perneczky R, Wagenpfeil S, Komossa K, et al. : Mapping scores onto stages: mini-mental state examination and clinical dementia rating. Am J Geriatr Psychiatry, 2006, 14: 139–144. [DOI] [PubMed] [Google Scholar]
- 24.Takeuchi R, Mutsuzaki H, Shimizu Y, et al. : Factors affecting ambulatory ability in patients aged 90 years and older following proximal femoral fractures. J Rural Med, 2017, 12: 63–67. [DOI] [PMC free article] [PubMed] [Google Scholar]