Sarcopenia as a Predictor of Mortality Following Vertebral Compression Fractures Treated With Vertebroplasty

Abstract

Study Design

Retrospective cohort study.

Objectives

To evaluate the impact of sarcopenia on the timing of recurrent vertebral compression fractures (VCFs) and mortality following vertebroplasty (VP).

Methods

A total of 162 patients who underwent VP between 2015 and 2021 were included. Sarcopenia was assessed using the psoas lumbar vertebral index (PLVI). Patients were categorized into sarcopenia and non-sarcopenia groups. Outcomes included second VCF incidence, time to second fracture, and overall mortality, analyzed using t-tests, chi-square tests, and Kaplan–Meier survival analysis.

Results

Sarcopenia was identified in 26 patients (19.1%). While the incidence of second VCFs was similar between groups (38.5% vs 41.9%, P = .743), the sarcopenia group experienced significantly shorter intervals to second fracture (13.6 vs 31.3 months, P < .05) and higher mortality (61.5% vs 11.1%, P < .05).

Conclusions

Sarcopenia may not increase the overall rate of recurrent VCFs but is associated with earlier recurrence and significantly elevated mortality following VP.

Introduction

Osteoporotic vertebral compression fractures (VCFs) represent a common fragility fracture and are frequently associated with severe, localized spinal pain. When conservative management fails, vertebroplasty is a well-established intervention for pain relief and functional improvement. ¹

Sarcopenia, characterized by progressive loss of skeletal muscle mass, strength, and physical function, has recently been identified as an independent risk factor for the occurrence of VCFs. ² However, evidence regarding the impact of sarcopenia on post-vertebroplasty outcomes—particularly recurrent or adjacent-level fractures—remains limited and inconclusive.

Therefore, the purpose of this study was to investigate the association between sarcopenia and clinical outcomes following vertebroplasty in patients with vertebral compression fractures.

Methods

Patient

This retrospective cohort study reviewed 482 patients who underwent percutaneous vertebroplasty at Taipei Veterans General Hospital between 2015 and 2021. All data were obtained from the institutional database and approved by the Institutional Review Board (IRB). We included patients diagnosed with vertebral compression fractures who underwent vertebroplasty due to persistent, intolerable pain unrelieved by conservative treatments, such as pharmacologic or physical therapy. Patients were excluded if they had pathological compression fractures, fractures at the L4 level, multiple compression fractures at the time of presentation, or incomplete clinical or imaging data required for sarcopenia assessment or outcome evaluation, or if they were lost to follow-up. Surgical intervention was generally recommended within 4 months for patients with symptomatic vertebral body fractures. ¹ We reviewed the patients’ medical records and extracted the following data: age, gender, body mass index (BMI), dual-energy X-ray absorptiometry (DXA) results, underlying medical conditions, and time to re-fracture.

Image Evaluation

Vertebral compression fractures were diagnosed based on imaging studies, including X-rays, stress views, and MRI. All fracture images were reviewed by the senior author, a board-certified orthopedic surgeon. Patient medical histories and prior imaging were carefully evaluated to differentiate between pathologic and benign vertebral compression fractures.^3,4

Sarcopenia was assessed using the Psoas Lumbar Vertebral Index (PLVI), defined as the ratio of the mean cross-sectional area of the psoas muscles (PCSA) to the cross-sectional area of the L4 vertebral body for each patient. ⁵ The formula was presented as follows:

Psoas: L4 vertebral index

= ([right PCSA ( ${mm}^2$ ) + left PCSA ( ${mm}^2$ )]/2)/L4 vertebral CSA ( ${mm}^2$ )

All patients undergoing vertebroplasty received lumbar spine or abdominal/pelvic magnetic resonance imaging (MRI) prior to surgery to enable accurate measurement of the cross-sectional area (CSA) of the bilateral psoas muscles and the L4 vertebral body. Based on a previous study, patients with a PLVI ≥ 0.84 were classified as non-sarcopenia group, whereas those with a PLVI < 0.84 were classified as sarcopenia group. ⁵ (Figure 1).

Figure 1.

Cross-sectional area (CSA) measurement of bilateral psoas muscle and the vertebral body of 4th lumbar

Postoperative Follow-Up and Rehabilitation Protocol

Bracing was recommended for the first 3 weeks during daily activities and ambulation. Patients were followed up in the outpatient department at 2, 4, and 12 weeks postoperatively, and subsequently every 6 months. Osteoporotic medications were prescribed at the physician’s discretion, provided there were no contraindications.

Outcomes Measurement

The primary outcome measure was the mortality rate. Secondary outcomes included the incidence of a second vertebral compression fracture (VCF), as detected on follow-up radiographs, and the time interval between the first and second fracture episodes.

Statistics

Statistical analyses were performed using IBM SPSS Statistics version 25. Continuous variables are presented as mean, range, and standard deviation (SD), while categorical variables are expressed as number and percentage. Differences between the sarcopenia and non-sarcopenia groups for continuous variables were compared using independent t-tests, and categorical variables were compared using chi-square tests. The time interval between the first and second vertebral compression fractures was analyzed using Kaplan–Meier survival curves and compared with the log-rank test. A P-value of less than 0.05 was considered statistically significant.

Results

A total of 482 patients were initially screened. Patients were excluded for the following reasons: compression fractures caused by pathological conditions such as metastasis or infection (n = 7), absence of MRI at the L4 level (n = 49), initial presentation with L4 compression fracture (n = 58), initial presentation with multiple compression fractures (n = 35), prior surgical interventions (eg, instrumentation, discectomy, fusion, vertebroplasty) (n = 14), and loss to follow-up or incomplete data (n = 157). After applying these exclusion criteria, 163 patients were included in the final analysis, as illustrated in the CONSORT diagram. (Figure 2).

Figure 2.

Consort diagram

The cohort included 72 males (44.4%) and 90 females (55.6%), with a mean age of 78.1 ± 7.43 years. Sarcopenia was defined as a PLVI < 0.84, according to a previous study. ⁵ The sarcopenia group consisted of 26 patients (19.1%), while the non-sarcopenia group included 136 patients (80.9%). There were no significant differences between the two groups in terms of age, sex, DXA results, vertebral body cross-sectional area (CSA), or bilateral psoas muscle CSA. However, body mass index (BMI) was significantly lower in the sarcopenia group compared to the non-sarcopenia group (22.91 ± 2.61 vs 23.55 ± 3.93, P = .038). (Table 1).

Table 1.

Patient Demographic Data and Characteristic

	Total	PLVI < 0.84	PLVI > 0.84	P-value
Number of patients	162	26 (16%)	136 (84%)
Mean age (years)	78.1 ± 7.425 (55-94)	77.69 ± 7.478 (66-92)	80 ± 6.957 (55-94)	0.578
Sex				0.063
Male	72 (44.4%)	10 (38.5%)	62 (45.6%)
Female	90 (55.6%)	16 (61.5%)	74 (54.4%)
BMI	23.45 ± 3.75	22.91 ± 2.61	23.55 ± 3.93	0.038
Diabetes mellitus	162			0.849
With	27 (16.7%)	5 (19.2%)	22 (16.2%)
Without	135 (83.3%)	21 (80.8%)	114 (83.8%)
Smoking				0.907
Yes	12 (7.4%)	2 (7.7%)	10 (7.4%)
No	150 (92.6%)	24 (92.3%)	126 (92.6%)
CV risk				0.205
Yes	99 (61.1%)	13 (50%)	86 (63.2%)
No	63 (38.9%)	13 (50%)	50 (36.8%)
ASA				0.545
2	58 (35.8%)	10 (38.5%)	48 (35.3%)
3	98 (60.5%)	16 (61.5%)	82 (60.3%)
4	6 (3.7%)	0 (0%)	6 (4.4%)
DXA	−2.529 ± 0.888	−2.585 ± 0.978	−2.518 ± 0.873	0.898
Left psoas CSA	7.651 ± 2.272	6.336 ± 2.124	7.902 ± 2.220	0.575
Right psoas CSA	7.454 ± 2.290	5.969 ± 1.856	7.738 ± 2.260	0.330
L4 body CSA	12.398 ± 1.867	13.114 ± 2.207	12.262 ± 1.771	0.338
1st fracture at T-L junction				0.983
Yes	62 (38.3%)	16 (61.5%)	84 (61.8%)
No	100 (61.7%)	10 (38.5%)	52 (38.2%)
Refracture				0.743
Yes	67 (41.6%)	10 (38.5%)	57 (41.9%)
No	95 (58.6%)	16 (61.5%)	79 (58.1%)
Interval to second episode	28.8 (months)	13.6 (months)	31.5 (months)	0.038
Mortality				<0.05
Yes	31 (19.1%)	16 (61.5%)	15 (11.1%)
No	131 (80.9)%	10 (38.5%)	121 (88.9%)

Age was expressed as mean ± standard deviation; Categorical variables are expressed as number and percentage; CV, cardiovascular; ORIF, open reduction and internal fixation; IQR, interquartile range; PLVI, psoas lumbar vertebral index. Bold values indicate statistical significance with p < 0.05.

The incidence of second fractures was comparable between the two groups (38.5% in the sarcopenia group vs 41.9% in the non-sarcopenia group, P = .743). However, the interval between the index fracture and the subsequent second fracture was significantly shorter in the sarcopenia group (13.6 months vs 31.5 months, P < .05) (Figure 3). Additionally, the mortality rate was significantly higher in the sarcopenia group (16/26 [61.5%] vs 15/136 [11.1%], P < .05).

Figure 3.

Comparison of vertebral body second fracture interval among sarcopenia group and non sarcopenia group measured by psoas lumbar vertebral index (PLVI). The Kaplan-Meier survival analysis showed the sarcopenia group had shorter *interval from first vertebroplasty to second episode of VCF compared to the non- sarcopenia group with statistic significant (P = .038)

Discussion

This study is the first large cohort investigation to evaluate the impact of sarcopenia on patients with vertebral compression fractures (VCF) following vertebroplasty. Unlike unmodifiable factors such as age, hypertension, or diabetes mellitus, sarcopenia is a potentially treatable condition. Involving a total of 162 patients and defining sarcopenia using the PLVI, we found that sarcopenia was significantly associated with increased mortality following vertebral compression fractures treated with vertebroplasty.

The consensus for diagnosing sarcopenia includes assessment of skeletal muscle strength, skeletal muscle mass, and physical performance. Muscle strength and physical performance can be evaluated through tests such as grip strength, the chair rise test, and gait speed. Several methods have been established to measure skeletal muscle mass, including appendicular skeletal muscle mass (ASMM) assessed by dual-energy X-ray absorptiometry (DXA) or bioelectrical impedance analysis (BIA), whole-body skeletal muscle mass (SMM), and the Psoas Lumbar Vertebral Index (PLVI).^2,5–7

Among the aforementioned methods, PLVI has been shown to be negatively associated with post-traumatic morbidity and can be easily calculated using MRI. In our study, patients with vertebral compression fractures routinely underwent MRI examinations to exclude pathological fractures or soft tissue lesions. ³ Furthermore, PLVI has been reported as a valid and reliable measure for assessing central sarcopenia. It serves as an ideal objective method for patients with vertebral compression fractures, enabling simultaneous evaluation of skeletal muscle mass and identification of sarcopenia.

Studies have demonstrated that sarcopenia is an independent risk factor for hip and ankle fractures in men, vertebral and hip fractures in women, as well as for prosthetic joint infections.^6,8,9 Additionally, numerous studies have highlighted the impact of sarcopenia on mortality rates. For example, Arango-Lopera et al 1 ¹⁰ reported that general surgery patients in the lowest quartile of core muscle size had a significantly lower three-year survival rate compared to those in the highest quartile (75% vs 91%). This effect was even more pronounced in liver transplant patients, where the three-year survival rate was 26% in the lowest quartile vs 77% in the highest quartile. ¹⁰ Sarcopenia has also been shown to predict 1-year mortality after acetabular fractures and 5-year mortality in patients with hip fractures.^11,12 In our study, sarcopenia was identified as a significant predictor of mortality (P < .05) in patients with vertebral compression fractures following vertebroplasty, with mortality rates of 61.5% in the sarcopenia group compared to 11.1% in the non-sarcopenia group.

The incidence of new vertebral compression fractures (VCF) was 10/26 (38.5%) in the sarcopenia group compared to 57/136 (41.9%) in the non-sarcopenia group (P > .05). Although the incidence of new VCF was comparable between groups, the interval from the first to the second episode was significantly shorter in the sarcopenia group (13.6 months vs 31.5 months, P < .05). Previous studies have reported a 1-year mortality rate after VCF ranging from 2% to 42%, 1 ¹³ indicating that some patients may have died before experiencing a second fracture.

Frailty has been consistently identified as a strong predictor of mortality and adverse outcomes in older adults and reflects the cumulative burden of multisystem health deficits. Sarcopenia is increasingly recognized as a central biological substrate of physical frailty, sharing common pathophysiological mechanisms including chronic inflammation, hormonal dysregulation, and diminished physiological reserve. Although comprehensive frailty indices demonstrate robust prognostic performance, their application in retrospective spine cohorts is frequently limited by incomplete clinical data and the need for extensive deficit documentation. In contrast, sarcopenia provides an objective, imaging-based, and reproducible parameter that can be readily assessed in patients undergoing vertebroplasty.

Our findings suggest that sarcopenia may function as an independent and pragmatic predictor of mortality following vertebral compression fractures treated with vertebroplasty, potentially capturing a clinically meaningful component of frailty-related vulnerability. Future prospective studies integrating validated frailty indices alongside sarcopenia assessment are warranted to further elucidate their relative and additive prognostic value.

There are several limitations to our study. Firstly, its retrospective design may introduce selection bias and makes the study more susceptible to data loss and confounding factors compared to prospective studies. We endeavored to follow all patients meeting our inclusion criteria; however, most patients required a caregiver and had impaired comprehension or cognitive status, which precluded the collection of functional outcome measures and frailty indices. Consequently, a validated frailty index was unavailable due to the retrospective design and incomplete health deficit documentation, and frailty may therefore represent an unmeasured confounder. Nevertheless, sarcopenia provides an objective, imaging-based, and clinically feasible surrogate of physical vulnerability, and future prospective studies incorporating both sarcopenia and validated frailty indices are warranted. Another limitation is that sarcopenia was defined solely using an imaging-based parameter (PLVI), without assessment of muscle strength or physical performance as recommended by current consensus definitions. Therefore, the sarcopenia definition used in this study may not fully capture the functional aspects of the condition. Despite these limitations, a key strength of this study is the large consecutive patient cohort, providing sufficient analytical power.

Conclusion

Sarcopenia in patients with vertebral compression fractures (VCF) following vertebroplasty (VP) may be associated with a shorter time interval before a subsequent fracture occurs. While it may serve as a predictive factor for mortality, it does not appear to significantly affect the overall incidence rate of second VCF events.

ORCID iDs

Hung-Kai Liao https://orcid.org/0000-0002-5985-3118

Yu-Cheng Yao https://orcid.org/0000-0001-7555-4321

Po-Hsin Chou https://orcid.org/0000-0001-5899-1124

Shih-Tien Wang https://orcid.org/0000-0002-0942-8332

Hsi-Hsien Lin https://orcid.org/0000-0003-0679-5222

Ethical Considerations

The study protocol was approved by the IRB of Taipei Veterans General Hospital (IRB No. 2015-08-006ACF)

Consent to Participate

Written informed consent was obtained from all the participants. All procedures were performed in accordance with relevant guidelines.

Data Availability Statement

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.*