Geriatric cognitive frailty and short-term prognosis following hip fracture surgery

Abstract

Background

To evaluate the prevalence of cognitive frailty in geriatric patients with hip fractures and to assess its association with short-term postoperative complications following surgical intervention.

Methods

This single-centre observational study enrolled geriatric patients (aged ≥ 65 years) undergoing surgical intervention for hip fracture. Patients were stratified into cognitive frailty and non-cognitive frailty groups based on assessments using the FRAIL scale and the Montreal Cognitive Assessment (MoCA). Clinical outcomes, including duration of hospitalisation and postoperative complications (e.g., cardiovascular, pulmonary, cerebrovascular, gastrointestinal events, and delirium), were compared between the cohorts. Multivariable logistic regression analysis was employed to identify independent risk factors for postoperative complications.

Results

A total of 159 patients (mean age: 81.35 ± 8.44 years) were enrolled, of whom 28.3% were classified as having cognitive frailty. Compared to the non-frail group, the cognitive frailty group exhibited a significantly longer hospital stay (15.6 ± 3.3 days vs. 13.9 ± 2.7 days, P < 0.05) and a higher incidence of postoperative complications (24.4% vs. 8.8%, P = 0.011). Multivariate analysis identified cognitive frailty (OR = 2.73, 95% CI: 1.03–7.25) and the Charlson Comorbidity Index (CCI) (OR = 1.31, 95% CI: 1.01–1.70) as independent risk factors for postoperative complications.

Conclusion

Cognitive frailty is highly prevalent among elderly patients with hip fractures and represents an independent predictor of short-term postoperative complications. Early identification of cognitive frailty could enhance risk stratification and perioperative management for this vulnerable population.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-025-06412-8.

Introduction

The ageing global population has resulted in a marked rise in age-related conditions, with hip fractures constituting a significant public health challenge. Elderly patients sustaining hip fractures face elevated risk of postoperative complications and protracted recovery periods owing to diminished physiological reserves and frequent comorbidities. Although surgical techniques have advanced, this population continues to experience unacceptably high rates of short-term mortality and complications [1, 2].

Frailty, characterised by a decline in physiological reserves across multiple systems, is a well-established predictor of poor postoperative outcomes [3–6]. Cognitive impairment, defined as deterioration in one or more cognitive domains (memory, attention, executive function, language, or visuospatial abilities) in older adults, is most commonly assessed using the Montreal Cognitive Assessment (MoCA), with score < 26 indicating mild cognitive impairment. The IANA/IAGG 2013 consensus defines cognitive frailty as the co-occurrence of physical frailty and mild cognitive impairment in individuals without dementia [7, 8]. Notably, this condition may be reversible in its early stages, underscoring the clinical importance of timely identification and intervention [9, 10]. Given that cognitive frailty substantially elevates the risk of adverse outcomes, including prolonged hospitalisation, postoperative complications, and mortality, early recognition and targeted management may significantly improve patient prognosis [9–11]. Although it is well-established that cognitive impairment predicts adverse outcomes after hip fracture, few studies have specifically focused on cognitive frailty, defined as the coexistence of physical frailty and mild cognitive impairment without dementia, and its distinct predictive value for postoperative complications. This subtle distinction may improve patient risk stratification. Despite the IANA/IAGG consensus definition, operational criteria for cognitive frailty vary substantially across studies. For example, Zhang et al.’s scoping review illustrated the field’s methodological heterogeneity and lack of consensus. While most studies used MoCA < 26 and FRAIL ≥ 3 to define cases, they applied different cut-off nuances variations and intervention combinations, demonstrating inconsistencies in diagnostic thresholds and intervention effects across trials [12]. In this study, cognitive frailty was defined as the simultaneous presence of: (1) physical frailty (FRAIL scale ≥ 3 points) and (2) cognitive impairment (MoCA score < 26 points), excluding dementia, based on validated criteria [7, 12–14]. Therefore, this study specifically focuses on elderly hip fracture patients without dementia to investigate the prevalence of cognitive frailty and its impact on short-term postoperative complications. The results provide evidence for optimising risk stratification and perioperative management in this vulnerable population.

Methods

Trial design and oversight

This single-centre clinical observational study was conducted with approval from the Medical Ethics Committee of Haidian Hospital (Beijing, China). All participants provided written informed consent, indicating their full understanding of the study procedures and associated risk.

Patients

The study enrolled patients admitted with hip fractures admitted to the Orthopedic Department of Beijing Haidian Hospital between January 2019 and July 2024. Inclusion criteria comprised: (1) age ≥ 65 years; and (2) surgical intervention for hip fracture within 48 h of admission. Exclusion criteria included: (1) polytrauma with concomitant fractures or organ injuries; (2) pathological fractures; (3) refusal of surgical treatment; (4) diagnosis of Alzheimer’s disease, dementia or other psychiatric conditions; and (5) inability to complete simple questionnaires; General clinical data were collected from enrolled patients, including age, gender, height, weight, underlying diseases and other relevant information.

Diagnosis and definitions

Enrolled patients underwent assessment with collection of general clinical data. Participants were stratified into two groups based on cognitive frailty status: cognitive frailty group and non-cognitive frailty group. All patients (≥ 65 years) completed standardised physical and cognitive assessment within 48 h of admission. Physical frailty was evaluated using the FRAIL scale, comprising five components (fatigue, resistance, ambulation, illness, and weight loss; 1 point per item). Using total scores, patients were classified as: robust (0 points), pre-frail (1–2 points), or frail (≥ 3 points) [13]. For this study’s purposes, consistent with previous validated studies, frail was defined as FRAIL score ≥ 3 and non‐frail was FRAIL score < 3. Cognitive function was assessed using the Montreal Cognitive Assessment (MoCA), a 30-item screening instrument. Per prior validation studies, we used a cutoff score < 26 to define mild cognitive impairment [12, 14]. Cognitive frailty was operationally defined as co-occurrence of mild cognitive impairment (MoCA < 26) and physical frailty (FRAIL ≥ 3) in dementia-free patients, aligning with IANA/IAGG 2013 consensus criteria [7] and recent clinical studies [11–14]. Patients with a clinical dementia prior to hospitalisation, as documented in medical records or through previous neurological assessments (e.g., Clinical Dementia Rating ≥ 1), were systematically excluded to isolate the prognostic impact of mild cognitive impairment. All assessments were conducted by trained evaluators following standardised procedures. Prior to assessment, evaluators established rapport through friendly conversation and verified the patient’s physical comfort status. For patients aged 80 years or older, evaluators adapted their approach by speaking more slowly, increasing vocal volume, and when necessary, supplementing verbal instructions with gestures or providing magnifying glasses. During assessment, evaluators took care not to rush participants. When incorrect answers were given, evaluators provided neutral responses (e.g., “Thank you, let’s proceed to the next item”) to minimise participant frustration. Rest periods were introduced whenever significant fatigue or pain was observed, ensuring evaluation accuracy. Demographic data and standard laboratory values were systematically recorded. Comorbidities were assessed using the Charlson Comorbidity Index (CCI), adapted to evaluate common geriatric conditions with appropriate weighting based on each condition’s potential prognostic impact [15].

Short-term postoperative complications and length of hospital stay served as the primary outcome measures. Postoperative medical records of both groups were systematically compared, with particular focus on hospitalisation duration and complication rates. Short-term postoperative complications were defined as any adverse events occurring during the index hospitalisation, including cardiovascular events (acute myocardial infarction, acute heart failure, arrhythmias requiring medication), pulmonary complications (pulmonary embolism, new pulmonary infections, acute respiratory failure), cerebrovascular events (new ischemic or haemorrhagic stroke), gastrointestinal complications (e.g., GI bleeding, perforation, ileus), and neurologic events (delirium). All-cause in‐hospital mortality was also recorded. All complications were graded according to the Clavien-Dindo classification system, with Grade III or higher considered major complications [16, 17].

Statistical methods

First, the normality of quantitative data was evaluated. Then, normally distributed quantitative data (age, body mass index, hospital stay) were expressed as mean ± standard deviation, while categorical data (gender, hypertension, diabetes and others) were expressed as number (percentage). Next, independent samples t-tests were used to compare the cognitive frailty and non-cognitive frailty groups for quantitative data, and chi-square tests were used for categorical data. For the regression analysis, univariable logistic regression was first performed for each candidate predictor of major postoperative complications. Variables with P < 0.20 in these univariable models were then entered into a multivariable logistic regression to identify independent risk factors. Finally, results are presented as β coefficients, odds ratios (OR = exp[β]), 95% confidence intervals (CI), standard errors (SE), and P-values, with two-tailed P < 0.05 considered statistically significant. All data were analysed using SPSS 29.0 software.

Results

General characteristics of patients

A total of 257 patients were recruited between January 2019 and July 2024. After applying exclusion criteria, including multiple injuries and other fractures or organ injury (n = 15), pathological fractures due to tumours (n = 13), refusal to undergo surgery (n = 21), Alzheimer’s disease, dementia, or other mental illnesses (n = 32), and inability to cooperate with simple questionnaires (n = 17), 159 patients aged 65–98 years (mean age: 81.35 ± 8.44 years) were enrolled in the study. The cohort comprised 69 males (43.4%) and 90 females (56.6%). The FRAIL scale scores ranged from 0 to 5 (mean: 2.50 ± 1.39), with 51/159 (32.1%) of patients classified as physically frail (FRAIL ≥ 3). MoCA scores ranged from 16 to 28 (mean: 22.21 ± 3.92), and 82/159 (51.6%) scored < 26, indicating mild cognitive impairment. Based on cognitive frailty assessments at admission, patients were stratified into a cognitive frailty group (n = 45) and a non-cognitive frailty group (n = 114) (Fig. 1). Of the 159 patients, 45 (28.3%) were classified as cognitively frail. The cognitive frailty group exhibited significantly higher FRAIL scores (4.27 ± 0.72 vs. 1.81 ± 0.89, P < 0.001) and lower MoCA scores (18.47 ± 1.91 vs. 23.69 ± 3.51, P < 0.001) compared to the non-cognitive frailty group. Additionally, the cognitively frail group was older (84.07 ± 7.94 vs. 80.27 ± 8.42 years, P = 0.010), had a lower BMI (23.84 ± 2.76 vs. 24.92 ± 2.24, P = 0.012), and higher CCI scores (5.84 ± 1.81 vs. 7.30 ± 1.06, P = 0.001). The hospital stays of patients in the cognitive frailty group and the non-cognitive frailty group was recorded, and the average hospital stay was compared between the two groups. The average hospital stay in the cognitive frailty group was significantly longer than in the non-cognitive frailty group (15.6 ± 3.3 vs. 13.9 ± 2.7, P = 0.001, Fig. 2). A comparison of baseline characteristics between the groups is presented in Table 1.

Fig. 1.

Flowchart of participants

Fig. 2.

Comparison of hospital stay between cognitive frailty and non-cognitive frailty groups

Table 1.

Baseline characteristics of patients

	Cognitively frail (n = 45)	Non-cognitively frail (n = 114)	p value
Age (y)	84.07 ± 7.94	80.27 ± 8.42	0.010
Gender, male (%)	20 (44.4%)	49 (43.0%)	0.832
BMI (kg/m2)	23.84 ± 2.76	24.92 ± 2.24	0.012
Smoking, n (%)	14 (31.1%)	30 (26.3%)	0.561
Comorbidities
Hypertension, n (%)	22 (48.9%)	34 (29.8%)	0.075
Diabetes, n (%)	11 (24.4%)	17 (14.9%)	0.264
Respiratory disease, n (%)	6 (13.3%)	16 (14.0%)	0.813
Heart disease, n (%)	8 (17.8%)	11 (9.6%)	0.778
CCI	5.84 ± 1.81	7.30 ± 1.06	0.001
FRAIL	4.27 ± 0.72	1.81 ± 0.89	< 0.001
MoCA	18.47 ± 1.91	23.69 ± 3.51	< 0.001
Hospital stay (d)	15.62 ± 3.27	13.89 ± 2.71	0.001

BMI Body Mass Index, CCI Charlson Comorbidity Index, MoCA Montreal Cognitive Assessment

Postoperative complications between the two patient groups

Complications occurred in 11 of 45 cognitive frail patients compared with 10 of 114 non-cognitive frail patients. Notably, three cognitive frail patients and two non-cognitive frail patients experienced multiple concurrent complications. All three deaths occurred in the cognitive frailty group, as detailed in Table 2.

Table 2.

Complications comparing the two groups

	Cognitively frail (n = 45)	Non-cognitively frail (n = 114)	P-value
Major postoperative complications, n (%)	11 (24.4%)	10 (8.8%)	0.011
Cardiovascular complications, n (%)	2 (4.4%)	1 (0.9%)	0.193
Pulmonary complications, n (%)	6 (13.3%)	4 (3.5%)	0.031
Cerebrovascular complications, n (%)	1 (2.2%)	1 (0.9%)	0.487
Digestive system complications, n (%)	1 (2.2%)	5 (4.5%)	1.000
Delirium, n (%)	4 (8.9%)	1 (0.9%)	0.023
Death, n (%)	3(6.7%)	0(0%)	0.022

Risk factors associated with internal complications following hip fracture surgery

In the univariable analysis, demographic and relevant clinical indicators were included as independent variables in a binary logistic regression model, with major postoperative complications as the dependent variable. Supplementary Table S1 presents the univariable logistic regression results for all candidate predictors of major postoperative complications. Variables meeting the screening criterion, including cognitive frailty, Charlson Comorbidity Index, and age, showed potential association with major postoperative complications (P < 0.2). These variables (P < 0.2) were subsequently entered into a multivariable logistic regression model. A Logistic regression analysis was conducted with postoperative complications during hospitalization after hip fracture surgery as the dependent variable. After adjusting for potential confounding factors such as age, cognitive frailty (OR = 2.726, 95% CI: 1.025–7.251, P = 0.044) and CCI (OR = 1.309, 95% CI: 1.006–1.704, P = 0.045) were identified as independent risk factors for postoperative complications (Table 3).

Table 3.

Multivariate analysis of major post-operative complication

Variables	β	Standard errors	OR	95% CI	P-value
Cognitive Frailty	1.003	0.499	2.726	1.025–7.251	0.044
CCI	0.270	0.134	1.309	1.006–1.704	0.045
Age	0.033	0.030	1.034	0.975–1.096	0.269

OR odds ratio, CI confidence interval, CCI Charlson Comorbidity Index

Discussion

In this prospective observational cohort study of elderly hip fracture inpatients, we found that 28.3% of patients met criteria for cognitive frailty, defined as concurrent physical frailty (FRAIL ≥ 3) and mild cognitive impairment (MoCA < 26) in the absence of dementia. Compared with their non-cognitively frail counterparts, cognitively frail patients demonstrated significantly higher rates of short-term postoperative complications and in-hospital mortality, along with prolonged hospital stay. Logistic regression analysis identified both cognitive frailty and Charlson Comorbidity Index as independent predictors of major postoperative complications.

Tosteson et al. used a Cox proportional hazards model in 1,944 hip fracture patients and reported an adjusted hazard ratio of 6.28 for mortality within six months post-fracture, which declined to 1.04 beyond six months, indicating that excess risk is largely confined to the early postoperative period [18]. These findings demonstrate that the mortality risk is highest during the immediate postoperative period, supporting our emphasis on short-term outcomes as crucial indicators for improving prognosis. In a two-year prospective study of 1,944 hip fracture patients aged 66 years or older, Söderqvist et al. observed an in-hospital mortality rate of 4%, increasing to 16% at four months and exceeding 33% at one year [19]. Our cohort showed a 6.7% in-hospital mortality rate among cognitively frail patients, compared with no deaths in the non-cognitively frail group. While this suggests a possible association between cognitive frailty and early postoperative mortality, these results should be interpreted cautiously due to our study’s limited sample size and advanced patient age.

With the continuous advancement of medical technology, the prognosis of hip fracture surgery is no longer determined solely by surgical factors, but rather by multiple factors. Among these, patients’ preoperative underlying conditions are significant contributors. This study found that for hip fractures patients, assessing cognitive frailty is crucial for prognosis evaluation. Currently, research in this area remains limited. This study further analysed factors influencing postoperative complications during hospitalisation by performing logistic regression analysis on differences between cognitive frailty and non-cognitive frailty groups. The results showed that cognitive frailty and Charlson Comorbidity Index were independent risk factors for postoperative complications in elderly hip fractures patients. A meta-analysis on preoperative prediction of mortality after hip fracture surgery found that patients’ preoperative mobility was an independent risk factor [20]. Consistent with these findings, our study performed preoperative cognitive frailty assessment to observe different prognostic outcomes. Cognitive frailty evaluation encompasses patients’ activity tolerance, underlying disease status, and cognitive function. Cognitive frailty represents a comprehensive manifestation of multi-system imbalance, with its essence being the disruption or dysregulation of homeostasis at both physiological and molecular levels. Emerging evidence suggests this phenomenon may be mediated through oxidative stress and chronic inflammatory pathways [21–23]. The diagnosis construct of cognitive frailty necessitates dual assessments of cognitive function and physical frailty status. Current frailty evaluation tools show considerable heterogeneity, encompassing the FRAIL scale, Frailty Index, and Fred phenotypic criteria. A recent scoping review in maintenance haemodialysis patients identified six disparate diagnostic approaches, ranging from Fried phenotypic criteria with MoCA to Clinical Dementia Rating incorporation, revealing strikingly variable prevalence estimates (4.6%−56.4%), that highlight ongoing standardisation challenges [12]. Based on previous research findings, five simple and practical frailty evaluation criteria were proposed, which are particularly suitable for bedridden elderly patients. Consequently, this study employed this frailty assessment method to evaluate hospitalised patients with hip fractures. The selection of MoCA to define cognitive impairment aligns with prior validation studies [9, 24]. Although the IAGG/IANA consensus defines the cognitive component of cognitive frailty using a Clinical Dementia Rating (CDR) score, a MoCA score < 26 was applied in this study. This choice was guided by MoCA’s widespread use, high sensitivity for early cognitive changes, and practicality in acute orthopaedic settings, while the consensus requirement of excluding dementia was still followed. It is acknowledged that differences in cognitive assessment tools may contribute to variability in prevalence estimates, and that future research directly comparing MoCA- and CDR-based definitions in surgical populations would be helpful in harmonising diagnostic approaches. Our findings indicate that cognitive frail patients experienced a high incidence of postoperative complications. Furthermore, a review of the literature suggests that frailty has been associated with osteoporosis and low bone mass, serving as an independent risk factor for reduced bone mass in patients [25]. However, few studies have explored the relationship between cognitive frailty and hip fractures, highlighting the need for further research in this field.

This study has limitation as a single-centre prospective observational cohort. The sample size was determined by the enrolment capacity of a single hospital during the study period, which may affect the precision and generalisability of our findings. Future multicentre collaboration would enable larger cohorts and stronger validation of these results. Although low bone mass and osteoporosis are well-established contributors to frailty, we did not collect bone density data in this cohort, representing a limitation. Additionally, the use of MoCA (rather than CDR) to operationalise the cognitive component may represent a methodological difference from the IAGG/IANA consensus definition, and thus may affect comparability with studies applying their criteria. This prospective cohort study is among the first to apply standardised FRAIL and MoCA assessments to elderly hip fracture inpatients, allowing precise identification of cognitive frailty prior to surgery. All evaluations were performed by trained evaluators following standardised procedures, with surgeries conducted within 48 h of admission to minimise timing bias. By integrating both physical and cognitive measures, we examined their combined impact on short-term postoperative outcomes, providing novel insights for risk stratification and perioperative management.

We present novel prospective evidence indicating that preoperative cognitive frailty predicts short-term postoperative complications in elderly hip fracture patients. The underlying biological mechanisms, however, remain unclear. Further large-scale, multicentre comparative studies and systematic reviews are required to standardise definitions of cognitive frailty and confirm its prognostic value. Implementing routine preoperative screening for cognitive frailty could improve risk stratification and facilitate tailored perioperative management.

Conclusion

Cognitive frailty is prevalent among elderly patients undergoing hip fracture surgery and is independently associated with an increased risk of postoperative complications, prolonged hospitalisation, and in-hospital mortality. Early identification of cognitive frailty could enhance risk stratification and optimise perioperative management for this vulnerable population.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

MMSE

Mini-Mental State Examination

OR

odds ratio

CI

confidence interval

CCI

Charlson Comorbidity Index

MoCA

Montreal Cognitive Assessment

Author contributions

Study concept and design: BL and JC; data acquisition: CH and BL; data analysis and interpretation: BL, YWW, YS; manuscript drafting: BL; critical revision of the manuscript: BL. All authors read and approved the final manuscript.

Funding

None.

Data availability

The datasets supporting the findings of this study are available from Beijing Haidian Hospital, but restrictions apply to their availability. These data were obtained under license for the current study and are not publicly accessible. However, data can be made available from the authors upon reasonable request and subject to permission from Beijing Haidian Hospital.

Declarations

Ethics approval and consent to participate

This study had approval from the Medical Ethics Committee of Beijing Haidian Hospital, Signed informed consent was obtained from all subjects. The current study followed the Declaration of Helsinki. All methods were performed in accordance with the relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.