Frailty and intraoperative hypotension in older adults undergoing hip fracture surgery: a retrospective cohort study

Abstract

Background

Frailty is common among older adults and is associated with adverse postoperative outcomes. Hip fractures are increasingly prevalent in aging populations and frequently require urgent surgical intervention. Intraoperative hypotension (IOH) is a common perioperative complication associated with serious postoperative morbidity. However, the relationship between frailty and IOH remains unclear. This study aimed to investigate the association between frailty, assessed using the 11-item modified frailty index (mFI-11), and IOH in older adults undergoing hip fracture surgery.

Methods

This retrospective cohort study included 522 patients aged ≥ 65 years who underwent hip fracture surgery at a tertiary care hospital in southern Thailand between January 2019 and June 2024. Patients were classified as frail (mFI-11 ≥ 0.27), prefrail (> 0 to < 0.27), or robust (0). IOH was defined as a mean arterial pressure (MAP) < 65 mmHg or a ≥ 20% reduction from baseline MAP in patients with hypertension. Multivariable logistic regression was used to evaluate the association between frailty and IOH. Adjusted quantile regression was performed to assess the association between frailty and cumulative duration of IOH.

Results

Among 522 patients, 124 (23.8%) were classified as frail, 341 (65.3%) as prefrail, and 57 (10.9%) as robust. IOH occurred in 298 patients (57.1%). Frailty was not independently associated with IOH. Compared with robust patients, the adjusted odds ratios for IOH were 0.78 (95% CI 0.42–1.43) for prefrail patients and 0.77 (95% CI 0.38–1.53) for frail patients. Independent predictors of IOH included intertrochanteric fracture, general anesthesia, preoperative antihypertensive medication administration on the day of surgery, longer operative duration, and colloid infusion. The multivariable model demonstrated modest discrimination (AUC 0.665, 95% CI 0.618–0.713). Although the incidence of IOH was similar across frailty groups, frail patients experienced longer cumulative duration of IOH. In adjusted analyses, frail patients had longer cumulative duration of IOH than prefrail patients (p = 0.045). Overall postoperative complication rates were similar among groups.

Conclusions

Frailty assessed using the mFI-11 was not independently associated with IOH occurrence in older adults undergoing hip fracture surgery. However, frail patients tended to experience a longer cumulative duration of IOH once it developed, suggesting a possible link between frailty and the persistence rather than the occurrence of intraoperative hemodynamic instability. These findings highlight the importance of vigilant intraoperative hemodynamic monitoring and individualized perioperative management in frail older adults.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12877-026-08028-y.

Introduction

Hip fracture is a common and serious condition among older adults and represents a major and growing public health concern. The incidence increases substantially with age, largely due to reduced bone mineral density and an increased risk of falls [1, 2]. Hip fractures are associated with functional decline, loss of independence, and increased mortality [3, 4]. Surgical treatment is therefore typically required to relieve pain and restore mobility.

Ageing is accompanied by a progressive decline in physiological reserve across multiple organ systems, contributing to the development of frailty. Frailty is characterized by increased vulnerability to stressors and impaired ability to maintain physiological homeostasis [5, 6]. It has been consistently associated with adverse postoperative outcomes, including complications and mortality, across a wide range of surgical populations. Several validated tools are available to assess frailty; however, no single instrument has been universally accepted [7, 8]. The modified frailty index (mFI), based on the accumulation-of-deficits model, is a widely used and validated tool that has demonstrated associations with adverse outcomes in orthopedic surgery [9, 10].

Intraoperative hypotension (IOH) is a common complication during anesthesia, including both general anesthesia and neuraxial techniques. IOH has been associated with serious postoperative complications such as acute kidney injury, myocardial infarction, stroke, postoperative delirium, prolonged hospital stay, and increased mortality [11, 12]. Identifying factors associated with IOH is therefore essential for improving perioperative risk stratification and management in older adults.

Although frailty is an established predictor of adverse perioperative outcomes, its association with IOH remains unclear. Frailty is characterized by reduced physiological reserve across multiple organ systems, including impaired autonomic regulation, decreased cardiovascular adaptability, and diminished resilience to physiological stressors. These alterations may impair compensatory responses to anesthetic-induced vasodilation, blood loss, and perioperative fluid shifts, potentially increasing susceptibility to IOH [5, 7]. Nevertheless, evidence regarding the association between frailty and IOH remains limited and inconsistent, particularly among older adults undergoing hip fracture surgery. Therefore, this study aimed to investigate the association between frailty, assessed using the modified frailty index, and IOH in older adults undergoing hip fracture surgery.

Methods

Study design and setting

This retrospective cohort study was conducted at Songklanagarind Hospital, a university-affiliated tertiary care center in southern Thailand. Electronic medical records from January 2019 to June 2024 were reviewed to investigate the association between frailty and IOH in older adults undergoing hip fracture surgery.

Ethics approval and consent to participate

This study was approved by the Human Research Ethics Committee of the Faculty of Medicine, Prince of Songkla University (REC.67-121-8-1). The requirement for informed consent was waived due to the retrospective nature of the study. All procedures were performed in accordance with the ethical standards of the institutional research committee and the Declaration of Helsinki and its subsequent amendments.

Participants

Patients aged ≥ 65 years with ASA physical status II–IV who sustained low-energy hip fractures and underwent surgery were eligible. Exclusion criteria were: (1) preoperative shock (blood pressure < 90/60 mmHg), (2) ongoing vasopressor infusion prior to surgery, (3) multiple traumatic injuries, and (4) missing data required for mFI-11 calculation.

Of 574 eligible patients, 52 were excluded due to incomplete data for calculating the 11-item modified frailty index or multiple trauma, leaving 522 patients for analysis (Fig. 1).

Fig. 1.

Patient flow diagram

Perioperative management

At our institution, spinal anesthesia was the preferred anesthetic technique for hip fracture surgery and was used unless contraindicated or when prolonged operative duration or specific clinical considerations favored general anesthesia. Intraoperative hemodynamic management was performed at the discretion of the attending anesthesiologist. Ephedrine and norepinephrine were the primary vasopressors used for the treatment of hypotension. Continuous vasopressor infusion was considered when repeated bolus administration was required or when hypotension was severe or persistent. Fluid administration was individualized according to clinical assessment, estimated blood loss, and hemodynamic status. Blood transfusion was administered at the discretion of the attending anesthesiologist, generally when hemoglobin concentrations fell below 8 g/dL and according to the patient’s clinical status.

Variables

Frailty

Frailty was assessed using the 11-item Modified Frailty Index (mFI-11) [9], a validated deficit-accumulation measure derived from the Canadian Study of Health and Aging Frailty Index and adapted using variables available in the American College of Surgeons National Surgical Quality Improvement Program database. The mFI-11 comprises 11 variables: non-independent functional status, diabetes mellitus, chronic obstructive pulmonary disease or recent pneumonia, congestive heart failure, myocardial infarction, history of coronary revascularization or angina, hypertension requiring medication, peripheral vascular disease, impaired sensorium, transient ischemic attack or cerebrovascular accident without deficit, and cerebrovascular accident with deficit. The mFI-11 score was calculated as the proportion of deficits present out of the 11 assessed variables, yielding a score ranging from 0 to 1. Patients were classified as robust (0 deficits; mFI-11 = 0), prefrail (1–2 deficits; mFI-11 > 0 and < 0.27), or frail (≥ 3 deficits; mFI-11 ≥ 0.27), consistent with previously published studies.

Intraoperative hypotension (IOH)

IOH was defined as either (1) a mean arterial pressure (MAP) < 65 mmHg or (2) a ≥ 20% reduction from baseline MAP in patients with a documented history of hypertension [13–15]. Vasopressor administration was recorded as an indicator of clinically significant hypotension but was not required for the definition of IOH.

MAP values were obtained directly from the electronic medical record. Baseline MAP was derived from routine blood pressure measurements recorded during the preoperative ward stay. For each patient, the highest and lowest MAP values recorded during this period were identified, and baseline MAP was defined as the average of these two values. This approach was intended to reduce the influence of transient blood pressure fluctuations that may occur immediately before anesthesia induction. Blood pressure was measured at 3–5-minute intervals using non-invasive monitoring, whereas invasive arterial pressure monitoring was used in selected high-risk patients to provide continuous real-time measurements.

Cumulative duration of IOH

Cumulative duration of IOH was defined as the total duration during which blood pressure met the predefined criteria for IOH. For patients monitored using non-invasive blood pressure measurements, the cumulative duration of IOH was determined from serial blood pressure recordings obtained at 3–5-minute intervals. In patients with invasive arterial pressure monitoring, continuous arterial pressure data were available. The cumulative duration of IOH episodes occurring during surgery was recorded in minutes.

Minimum alveolar concentration (MAC)

End-tidal volatile anesthetic concentrations were recorded and expressed as age-adjusted MAC values, as automatically calculated by the anesthesia monitoring system.

Postoperative outcomes

Postoperative complications during the same admission included stroke, major adverse cardiovascular events (MACE), and acute kidney injury (AKI).

Outcomes

The primary outcome was the association between frailty (mFI-11) and IOH. Secondary outcomes included postoperative complications and hospital length of stay.

Data collection

Data were extracted from electronic medical records, including demographic characteristics, comorbidities, anesthetic techniques, intraoperative hemodynamics, and postoperative outcomes. Chronic antihypertensive medication use was recorded as the number of regularly prescribed antihypertensive medications. In addition, administration of antihypertensive medications on the day of surgery was recorded separately. All data were anonymized prior to analysis.

Sample size

Sample size was calculated based on a pilot study (n = 62), in which the incidence of IOH was 71% in frail patients and 53% in non-frail patients. The expected non-frail-to-frail participant ratio was 2.6:1. Using a two-sided α of 0.05 and a power of 80%, the required sample size was estimated to be 313 patients (87 frail and 226 non-frail). After allowing for 10% incomplete data, the target sample size was 349 patients. The final study cohort (n = 522) exceeded this requirement.

Statistical analysis

Continuous variables were summarized as mean ± standard deviation (SD) or median (IQR) and compared using one-way analysis of variance (ANOVA) or the Kruskal–Wallis test, as appropriate. Categorical variables were compared using the chi-square test or Fisher’s exact test.

Multivariable logistic regression was performed to assess the association between frailty and IOH. Variables considered for multivariable analysis were selected based on either clinical relevance or univariable significance (p < 0.20). A stepwise logistic regression approach was then used to identify the final multivariable model. Frailty status, age, and sex were retained in the model regardless of statistical significance because of their clinical relevance. Continuous variables were assessed for linearity prior to inclusion in the regression model. Model discrimination was evaluated using the receiver operating characteristic (ROC) curve.

To evaluate the association between frailty and cumulative duration of IOH, quantile regression was performed to estimate adjusted median differences while accounting for potential confounders. Covariates included age, sex, fracture type, anesthetic technique, preoperative antihypertensive medication administration on the day of surgery, and operative time. All analyses were performed using RStudio, with p < 0.05 considered statistically significant.

Results

A total of 574 patients were screened for eligibility. After exclusion of 52 patients because of missing data required for mFI-11 calculation (n = 21) or multiple traumatic injuries (n = 31), 522 patients were included in the final analysis (Fig. 1). Among the 522 patients included in the analysis, 23.8% were classified as frail, 65.3% as prefrail, and 10.9% as robust. Baseline demographic characteristics, including age, sex, body mass index, fracture type, fracture side, and emergency surgery status, were comparable among frailty groups (all p > 0.05) (Table 1).

Table 1.

Baseline characteristics and intraoperative outcomes according to frailty status

	Frail n = 124 (23.8%)	Prefrail n = 341 (65.3%)	Robust n = 57 (10.9%)	Total n = 522	P value
Age (years), median (IQR)	82 (76,86)	83 (76,87)	84 (76,88)	83 (76,87)	0.40
Sex, n (%)					0.83
male	40 (32.3)	100 (29.3)	17 (29.8)	157 (30.1)
female	84 (67.7)	241 (70.7)	40 (70.2)	365 (69.9)
BMI (kg/m2), median (IQR)	23.0 (20.0,25.7)	21.9 (19.5,25.2)	21.9 (19.5,24.8)	22.2 (19.6,25.3)	0.13
Fracture type					0.53
Femoral neck	62 (50.0)	180 (52.8)	27 (47.4)	269 (51.5)
Intertrochanteric	56 (45.2)	145 (42.5)	30 (52.6)	231 (44.3)
Subtrochanteric	6 (4.8)	15 (4.4)	0 (0)	21 (4.0)
Femoral head	0 (0)	1 (0.3)	0 (0)	1 (0.2)
Fracture side
Right	63 (50.8)	163 (47.8)	29 (50.9)	255 (48.9)	0.80
Left	61 (49.2)	178 (52.2)	28 (49.1)	267 (51.1)
ASA classification, n (%)					< 0.001
II	17 (13.7)	121 (35.5)	27 (47.4)	165 (31.6)
III	106 (85.5)	218 (63.9)	30 (52.6)	354 (67.8)
IV	1 (0.8)	2 (0.6)	0 (0)	3 (0.6)
Emergency surgery, n (%)	45 (36.3)	135 (39.6)	29 (50.9)	209 (40.0)	0.17
Comorbidity, n (%)
Hypertension	112 (90.3)	227 (66.6)	0 (0)	339 (64.9)	< 0.001
Myocardial infarction	5 (4.0)	1 (0.3)	0 (0)	6 (1.2)	0.009
Heart failure	14 (11.3)	3 (0.9)	0 (0)	17 (3.3)	< 0.001
Diabetes mellitus	79 (63.7)	65 (19.1)	0 (0)	144 (27.6)	< 0.001
Chronic kidney disease	27 (21.8)	56 (16.4)	2 (3.5)	85 (16.3)	0.008
End-stage renal disease	14 (11.3)	8 (2.4)	0 (0)	22 (4.2)	< 0.001
Transient ischemic attack	2 (1.6)	3 (0.9)	0 (0)	5 (1.0)	0.78
Prior stroke	45 (36.3)	44 (12.9)	0 (0)	89 (17.1)	< 0.001
Chronic antihypertensive medication use, n (%)					< 0.001
None	11 (8.9)	104 (30.5)	52 (91.2)	167 (32.0)
1 drug	44 (35.5)	120 (35.2)	3 (5.3)	167 (32.0)
>1 drug	69 (55.6)	117 (34.3)	2 (3.5)	188 (36.0)
Anesthetic technique, n (%)					0.054
General anesthesia	38 (30.6)	70 (20.5)	10 (17.5)	118 (22.6)
Spinal anesthesia	86 (69.4)	271 (79.5)	47 (82.5)	404 (77.4)
IOH, n (%)	74 (59.7)	191 (56.0)	33 (57.9)	298 (57.1)	0.77
Duration of IOH, min, median (IQR)	27.5 (11.25,58.75)	20 (5.0,40.0)	10 (5.0,20.0)	20 (10,40)	0.0015
Arterial line monitoring, n (%)	35 (28.2)	47 (13.8)	5 (8.8)	87 (16.7)	< 0.001
Intraoperative vasopressor requirement, n (%)					0.81
No vasopressor	50 (40.3)	151 (44.3)	24 (42.1)	225 (43.1)
Bolus dose < 3 doses	27 (21.8)	87 (25.5)	14 (24.6)	128 (24.5)
Bolus dose ≥ 3 doses	30 (24.2)	68 (19.9)	13 (22.8)	111 (21.3)
Continuous infusion	17 (13.7)	35 (10.3)	6 (10.5)	58 (11.1)

Values are n (%) unless otherwise indicated

IQR interquartile range, BMI body mass index, IOH intraoperative hypotension

Frail patients had a significantly higher comorbidity burden. They were more likely to have higher ASA physical status and a greater prevalence of cardiovascular and metabolic conditions, including hypertension, myocardial infarction, diabetes mellitus, chronic kidney disease, heart failure, and prior stroke (all p < 0.01). Consistent with these findings, chronic antihypertensive medication use was more common among frail patients, whereas most robust patients were not receiving antihypertensive medications (p < 0.001). Frail patients were also more likely to undergo arterial line monitoring (28.2% vs. 13.8% vs. 8.8%, p < 0.001). However, anesthetic technique, intraoperative vasopressor requirements, and the incidence of IOH were comparable across frailty groups (all p > 0.05) (Table 1).

Although the incidence of IOH did not differ significantly among frailty groups (59.7% in frail, 56.0% in prefrail, and 57.9% in robust patients; p = 0.77), cumulative duration of IOH increased progressively with frailty severity. The median cumulative duration of IOH was 27.5 (IQR 11.25–58.75) minutes in frail patients, compared with 20 (IQR 5.0–40.0) minutes in prefrail patients and 10 (IQR 5.0–20.0) minutes in robust patients (p = 0.0015) (Table 1).

In univariable analyses, frailty status was not associated with the occurrence of IOH (p = 0.77) (Table 2). Patients who developed IOH were more likely to have intertrochanteric fractures (50.3% vs. 36.2%, p = 0.003), preoperative antihypertensive medication administration on the day of surgery (p = 0.013), undergo general anesthesia (27.2% vs. 16.5%, p = 0.009), and have longer operative durations (median 175 vs. 165 min, p = 0.03).

Table 2.

Factors associated with IOH

Factors	IOH, (n = 298)	No IOH, (n = 224)	Total, (n = 522)	P value
Age				0.896
65–74 years	55 (18.5)	45 (20.1)	100 (19.2)
75–84 years	129 (43.3)	95 (42.4)	224 (42.9)
≥ 85 years	114 (38.3)	84 (37.5)	198 (37.9)
Frailty group, n (%)				0.77
Frail	74 (24.8)	50 (22.3)	124 (23.8)
Prefrail	191 (64.1)	150 (67.0)	341 (65.3)
Robust	33 (11.1)	24 (10.7)	57 (10.9)
Fracture type, n (%)				0.003
Femoral neck	135 (45.3)	134 (59.8)	269 (51.5)
Intertrochanteric	150 (50.3)	81 (36.2)	231 (44.3)
Subtrochanteric	13 (4.4)	8 (3.6)	21 (4.0)
Femoral head	0 (0)	1 (0.4)	1 (0.2)
Preoperative antihypertensive medication administration on the day of surgery				0.013
None	136 (45.6)	129 (57.6)	265 (50.8)
1 drug	106 (35.6)	69 (30.8)	175 (33.5)
> 1 drug	56 (18.8)	26 (11.6)	82 (15.7)
Anesthetic technique, n (%)				0.009
General anesthesia	81 (27.2)	37 (16.5)	118 (22.6)
Spinal anesthesia	217 (72.8)	187 (83.5)	404 (77.4)
Anesthetic drugs
Propofol use, n (%)	141 (47.3)	72 (32.1)	213 (40.8)	< 0.001
Age-adjusted MAC (desflurane), median (IQR)	1 (0.9,1.1)	1.1 (1.0,1.2)	1.0 (0.9,1.1)	0.004
Age-adjusted MAC (sevoflurane), median (IQR)	0.9 (0.7,1)	1 (1,1.08)	0.9 (0.8,1)	0.009
Sensory block level, n (%)				< 0.001
Above T5	57 (29.5)	21 (12.2)	78 (21.4)
T5 to T10	127 (65.8)	139 (80.8)	266 (72.9)
Below T10	9 (4.7)	12 (7.0)	21 (5.8)
Operation time (min), median (IQR)	175 (150,205)	165 (140,195)	170 (145,200)	0.03

Values are presented as n (%) unless otherwise indicated

MAC values represent age-adjusted end-tidal MAC automatically calculated by the anesthesia monitoring system. Sensory block level data were available for 365 of 404 patients who received spinal anesthesia

IOH intraoperative hypotension, MAC minimum alveolar concentration, IQR interquartile range

Among anesthetic-related factors, propofol use, lower median age-adjusted MAC values for desflurane and sevoflurane, and higher sensory block levels were significantly associated with IOH (all p < 0.01).

In multivariable logistic regression analysis, frailty was not independently associated with IOH. Compared with robust patients, the adjusted odds ratios for IOH were 0.78 (95% CI 0.42–1.43, p = 0.419) for prefrail patients and 0.77 (95% CI 0.38–1.53, p = 0.453) for frail patients (Table 3).

Table 3.

Univariable and multivariable logistic regression analyses of factors associated with IOH

Factors	crude OR (95%CI)	crude P value	Adjusted OR (95% CI)	P value
Frailty status (ref.: Robust)				0.680
Prefrail	0.93 (0.53,1.64)	0.795	0.78 (0.42,1.43)	0.419
Frail	1.06 (0.56,2.01)	0.854	0.77 (0.38,1.53)	0.453
Age (per year increase)	1.01 (0.99–1.04)	0.258	1.01 (0.98–1.04)	0.436
Female sex (ref: male)	0.89 (0.61–1.31)	0.562	0.94 (0.62–1.42)	0.737
Fracture type (ref.: Femoral neck)				0.027
Intertrochanteric	1.84 (1.28,2.64)	0.001	1.69 (1.15,2.47)	0.007
Subtrochanteric	1.61 (0.65,4.02)	0.305	1.04 (0.38,2.88)	0.936
Femoral-head	0 (0, Inf)	0.98	0 (0, Inf)	0.980
Anesthetic technique (ref.: Spinal anesthesia)				0.008
General anesthesia	1.83 (1.18,2.83)	0.007	1.83 (1.16,2.89)	0.009
Preoperative antihypertensive medication administration on the day of surgery (ref: None)				0.013
1 drug	1.45 (0.98,2.13)	0.062	1.61 (1.06,2.45)	0.025
> 1 drug	1.95 (1.15,3.31)	0.013	2.05 (1.17,3.57)	0.012
Operation time (per hour increase)	1.37 (1.08,1.74)	0.01	1.32 (1.01,1.72)	0.038
Colloid infusion (Yes vs. No)	2.31 (1.1,4.85)	0.027	2.3 (1.06,4.98)	0.027

Values are presented as crude and adjusted odds ratios (ORs) with 95% confidence intervals (CIs)

The multivariable model included variables selected based on clinical relevance and/or statistical significance in univariable analysis. Age and sex were forced into the multivariable model regardless of statistical significance

IOH intraoperative hypotension, OR odds ratio, CI confidence interval, ref. reference

Intertrochanteric fracture, general anesthesia, preoperative antihypertensive medication administration on the day of surgery, longer operative duration, and colloid infusion were associated with IOH in the multivariable model (Table 3).

The discriminative ability of the multivariable model was modest, with an area under the receiver operating characteristic (ROC) curve of 0.665 (95% CI 0.618–0.713) (Fig. 2).

Fig. 2.

Receiver operating characteristic curve of the multivariable model for predicting IOH

To further evaluate the association between frailty and cumulative duration of IOH, an adjusted quantile regression analysis was performed (Table 4 and Fig. 3). After adjustment for age, sex, fracture type, anesthetic technique, preoperative antihypertensive medication administration on the day of surgery, and operative time, frail patients demonstrated a longer adjusted median cumulative duration of IOH than prefrail patients (p = 0.045). In contrast, the difference between frail and robust patients did not reach statistical significance (p = 0.084), and no difference was observed between prefrail and robust patients (p = 0.918).

Table 4.

Adjusted quantile regression analysis for cumulative duration of IOH among patients who developed IOH

Factors	β coefficient (minutes)	95% CI	P value
Frailty status (ref: Robust)
Prefrail	-0.48	-9.45 to 8.48	0.918
Frail	10.51	-1.44 to 22.46	0.084
Age (per year increase)	0.61	0.10 to 1.13	0.023
Female sex (ref: Male)	0.34	-7.75 to 8.44	0.935
Fracture type (ref: Femoral neck)
Intertrochanteric	0.91	-6.17 to 7.99	0.805
Subtrochanteric	20.28	-3.60 to 44.16	0.120
General anesthesia vs. Spinal anesthesia	-0.15	-7.33 to 7.03	0.967
Preoperative antihypertensive medication administration on the day of surgery (ref: None)
1 drug	5.57	-2.17 to 13.32	0.142
>1 drug	14.69	-0.92 to 30.31	0.064
Operative time (per minute increase)	0.009	-0.068 to 0.086	0.822

Fig. 3.

Adjusted median cumulative duration of IOH according to frailty status. Points represent adjusted medians and error bars represent interquartile ranges (IQRs). Estimates were derived from quantile regression adjusted for age, sex, fracture type, anesthetic technique, preoperative antihypertensive medication administration on the day of surgery, and operative time. Analysis was restricted to patients who developed IOH (n = 298). IOH, intraoperative hypotension

The overall incidence of postoperative complications was 20.1%, with no significant differences among frailty groups (p = 0.342). Stroke was observed only in frail patients (3 events, 2.4%; p = 0.023), although the number of events was very small. Other complications, including major adverse cardiovascular events (MACE) and acute kidney injury (AKI), did not differ significantly among groups (Table 5).

Table 5.

Postoperative complications and length of hospital stay according to frailty status

Outcome	Frail (n = 124)	Prefrail (n = 341)	Robust (n = 57)	P value
Complications, n (%)	29 (23.4)	68 (19.9)	8 (14.0)	0.342
Stroke	3 (2.4)	0 (0)	0 (0)	0.023
MACE	4 (3.2)	8 (2.4)	1 (1.8)	0.908
AKI	3 (2.4)	7 (2.1)	2 (3.5)	0.674
Length of hospital stay, days, median (IQR)	5 (4,9)	5 (4,8)	4 (3,6)	0.053

Values are presented as n (%) unless otherwise indicated

MACE major adverse cardiovascular events, AKI acute kidney injury, IQR interquartile range

The median length of hospital stay tended to be longer in frail patients (5 [IQR 4–9] days) than in robust patients (4 [IQR 3–6] days), although the difference did not reach statistical significance (p = 0.053) (Table 5).

Discussion

In this retrospective cohort study, frailty assessed using the mFI-11 was not independently associated with the occurrence of intraoperative hypotension (IOH) after adjustment for confounders. However, frail patients experienced longer cumulative duration of IOH, and adjusted analyses demonstrated longer cumulative duration of IOH among frail patients, although not all pairwise comparisons reached statistical significance.

Our findings differ from several previous studies reporting a positive association between frailty and IOH. For example, postinduction hypotension has been shown to increase with frailty in older patients undergoing non-cardiac surgery under general anesthesia, and similar associations have been reported in other cohorts [16–18]. In contrast, our results are consistent with studies in elective non-cardiac surgery demonstrating no independent association between frailty and IOH [19]. These discrepancies may be explained by differences in patient populations, surgical urgency, definitions of IOH, and frailty assessment tools.

The definition of IOH may also have influenced our findings. We used a composite definition based on either an absolute MAP threshold (< 65 mmHg) or a relative reduction of ≥ 20% from baseline MAP in patients with hypertension, consistent with previous perioperative studies [13–15]. This approach was intended to account for both universally accepted perfusion thresholds and the possibility that chronically hypertensive patients may require higher perfusion pressures to maintain organ blood flow. However, alternative definitions of IOH have been used in the literature, including different absolute MAP thresholds and relative reductions from baseline blood pressure [14, 15]. Such variability may influence the reported incidence of IOH and could partly explain differences between studies.

Several factors may explain the absence of an independent association in our study. First, the mFI-11 primarily reflects comorbidity burden and functional deficits and may not fully capture physiological vulnerability, particularly impairments in autonomic and cardiovascular reserve [7, 10, 20]. Second, our definition of IOH was based on a blood pressure threshold, while vasopressor administration was recorded as an indicator of clinically significant hypotension. This approach may have underestimated transient hypotensive episodes [13–15]. Third, the high prevalence of antihypertensive medication use in our cohort raised the possibility of attenuation of the frailty effect because antihypertensive medication administration on the day of surgery was independently associated with IOH. However, sensitivity analyses excluding antihypertensive medication use from the multivariable model yielded similar effect estimates, suggesting that overadjustment was unlikely to fully explain the observed findings. In addition, standardized anesthetic management and close intraoperative monitoring at our institution may have mitigated frailty-related hemodynamic instability.

Consistent with previous literature, preoperative antihypertensive medication administration on the day of surgery, intertrochanteric fracture, general anesthesia, and longer operative duration were independently associated with IOH [21–24]. These factors likely reflect both physiological mechanisms and procedural complexity. For example, general anesthesia may contribute to hypotension through vasodilation and myocardial depression, whereas prolonged surgery increases cumulative anesthetic exposure and may be associated with greater blood loss and fluid shifts [23, 24]. The observed association between colloid infusion and IOH should be interpreted cautiously, as colloid administration may have occurred in response to IOH or hemodynamic instability rather than representing a causal risk factor. Therefore, this finding may reflect reverse causation rather than a causal effect. The modest discriminative performance of the multivariable model (AUC 0.665, 95% CI 0.618–0.713) further supports the multifactorial nature of IOH. Important determinants of intraoperative blood pressure regulation, including preoperative volume status, autonomic dysfunction, cardiac function, anesthetic depth, intraoperative blood loss, and fluid management, were not available in the present dataset and may have contributed to the model’s limited predictive performance and residual confounding.

Importantly, frail patients experienced a longer cumulative duration of IOH despite a similar incidence of IOH. Although the association was attenuated after adjustment for potential confounders, frail patients continued to demonstrate a longer cumulative duration of IOH than prefrail patients, while the comparison with robust patients did not reach statistical significance. These duration findings, derived from a secondary analysis, should be interpreted as hypothesis-generating and require confirmation in prospective studies.

Frailty is characterized by reduced physiological reserve, including impaired autonomic regulation and diminished cardiac responsiveness [5, 20]. These factors may limit compensatory responses once hypotension occurs, resulting in sustained reductions in organ perfusion. This finding is clinically relevant, as prolonged intraoperative hypotension has been associated with adverse outcomes such as myocardial injury, acute kidney injury, and stroke [11, 13, 25].

The choice of frailty assessment tool may have influenced our findings. Although the mFI-11 is a validated and widely used frailty instrument [6, 9], it primarily reflects the accumulation of comorbid conditions rather than broader domains of frailty. Functional impairment, cognitive dysfunction, physical performance, and physiological reserve are not directly assessed by the mFI-11, despite their potential relevance to perioperative cardiovascular regulation and autonomic compensation [7, 8, 20]. Consequently, the absence of an independent association between frailty and the occurrence of IOH in our study may partly reflect limitations of the frailty instrument rather than the absence of a true physiological relationship. In particular, omission of the cognitive domain may be important, as cognitive impairment is increasingly recognized as a key component of frailty [7, 8, 20]. As cognitive dysfunction may contribute to reduced physiological resilience and impaired responses to perioperative stress, omission of this domain may have resulted in underestimation of frailty severity in some patients and attenuated the observed association between frailty and IOH.

Although overall postoperative complication rates did not differ significantly among frailty groups, stroke occurred exclusively in frail patients. This finding should be interpreted with considerable caution because only three stroke events were observed. Therefore, the study was not adequately powered to evaluate associations between frailty and postoperative complications, and this observation should be considered hypothesis-generating only. Length of hospital stay was also numerically longer in frail patients, consistent with previous studies showing that frailty is associated with prolonged hospitalization after orthopedic surgery [10].

Taken together, our findings suggest that frailty may not be associated with the occurrence of IOH but may be associated with longer cumulative duration of IOH once it develops. These results highlight the importance of early detection and prompt management of hypotension in frail older adults. Strategies such as individualized anesthetic titration, careful control of neuraxial block height, and proactive hemodynamic monitoring may be particularly beneficial in this population.

Strengths and limitations

This study has several strengths, including a relatively large sample size, comprehensive perioperative data, and multivariable adjustment for clinically relevant confounders. However, several limitations should be acknowledged. First, the retrospective design precludes causal inference and is subject to residual confounding. Second, frailty was assessed using a single instrument (mFI-11), which primarily reflects comorbidity burden and does not capture important domains such as cognitive function, physical performance, functional reserve, and autonomic vulnerability. These factors may influence perioperative hemodynamic responses and could potentially explain the lack of an independent association between frailty and the occurrence of IOH observed in our study. Third, the definition of IOH combined an absolute threshold (MAP < 65 mmHg) with a relative threshold (≥ 20% reduction from baseline MAP in patients with hypertension). Because hypertensive patients were eligible to meet either criterion, whereas non-hypertensive patients were classified primarily using the absolute threshold, differential ascertainment may have occurred. As hypertension was substantially more prevalent among frail patients, this methodological consideration may have influenced the observed association between frailty and the incidence of IOH and should be considered when interpreting the results. Fourth, several potentially important determinants of IOH, including preoperative volume status, cardiac function, autonomic dysfunction, intraoperative blood loss, fluid administration, and anesthetic depth, were not available in this retrospective dataset, which may have contributed to the modest discriminative ability of the multivariable model. Fifth, blood pressure measurements obtained at 3–5-minute intervals may not capture brief hypotensive episodes occurring between recordings, potentially resulting in underestimation of the cumulative duration of IOH. Although this limitation reflects routine clinical monitoring practices, it should be considered when interpreting the association between frailty and cumulative duration of IOH. Finally, this was a retrospective study conducted at a single tertiary-care center. Differences in patient populations, anesthetic practices, monitoring protocols, and perioperative management strategies across institutions may limit the generalizability of our findings. Therefore, our results should be interpreted within the context of the study setting and warrant external validation in multicenter prospective studies.

Despite these limitations, our findings provide clinically relevant information regarding the relationship between frailty and intraoperative hemodynamics in older adults undergoing hip fracture surgery. Future studies incorporating continuous hemodynamic monitoring and more comprehensive frailty assessments, including functional and cognitive domains, may further clarify this relationship and improve perioperative risk stratification.

Conclusions

In older adults undergoing hip fracture surgery, frailty as assessed by the mFI-11 was not independently associated with the occurrence of IOH. However, frail patients tended to experience longer cumulative duration of IOH once it developed. These findings suggest that frailty may be more closely related to the persistence rather than the occurrence of intraoperative hemodynamic instability and highlight the importance of vigilant hemodynamic monitoring and individualized perioperative management in frail older adults.

Abbreviations

AKI

Acute kidney injury

ASA

American Society of Anesthesiologists

AUC

Area under the curve

CI

Confidence interval

IOH

Intraoperative hypotension

IQR

Interquartile range

MAC

Minimum alveolar concentration

MAP

Mean arterial pressure

MACE

Major adverse cardiovascular events

mFI-11

11-item modified frailty index

NIBP

Non-invasive blood pressure

OR

Odds ratio

ROC

Receiver operating characteristic

SD

Standard deviation

Authors’ contributions

KS contributed to study conception and design, interpretation of data, and critical revision of the manuscript. NL contributed to data collection, statistical analysis, interpretation of data, and drafting of the manuscript. PB and MS contributed to interpretation of data and critical revision of the manuscript. All authors read and approved the final manuscript.

Funding

The authors received no specific funding for this work.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Institutional Ethics Committee of the Faculty of Medicine, Prince of Songkla University, Hat Yai, Thailand (REC.67-121-8-1). The requirement for informed consent was waived owing to the retrospective nature of the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.