Skip to main content

Preoperative CT-derived psoas attenuation and correction loss after traumatic thoracolumbar fracture fixation: A construct-dependent association

In brief

Without fracture screws, 10-HU higher psoas attenuation tracked with 2.6 degrees less correction loss

In 78 adults with traumatic thoracolumbar fractures, higher preoperative psoas attenuation was associated with less correction loss when screws did not include the fractured vertebra; no clear association appeared when they did. The subgroup estimate was not conclusive in a robust analysis, and this retrospective, single-center finding needs prospective validation before muscle measurements guide screw placement.

Journal
Injury (Q1)
Published
7 November 2026
Study design
Prospective / inception cohort
Evidence level
Level 2, Moderate (CEBM 2b)
Authors
Murat Özdemir, Fatma Latifoğlu, Osman Talha Uğur, Faruk Peksoy, Kağan Kamaşak
PMID
43014230
DOI
10.1016/j.injury.2026.113893

Why clinicians should know about it

  • Picked for Spine Surgery (top studies of the week, 15 November 2026): Traumatic thoracolumbar fixation and construct-related correction loss

Abstract

BACKGROUND: Postoperative correction loss remains a concern after posterior fixation of traumatic thoracolumbar fractures. Although bone quality and construct characteristics influence mechanical stability, the role of preoperative muscle quality remains poorly defined. We examined whether CT-derived psoas attenuation was associated with correction loss and whether fracture-level screw placement modified this association. METHODS: This retrospective single-center cohort included 78 adults who underwent posterior pedicle screw fixation for traumatic AO Spine A3 thoracolumbar fractures. Preoperative psoas and paravertebral attenuation and cross-sectional area were measured at L3. Correction loss was defined as the difference between final-follow-up and immediate postoperative Cobb angles measured on supine sagittal CT reconstructions. Multivariable linear regression tested the interaction between centered psoas attenuation and fracture-level screw placement, adjusting for age, sex, vertebral attenuation, instrumented levels, and follow-up duration. An extended model added vertebral height loss and canal compromise. HC3 robust variance estimation and sensitivity analyses were performed. RESULTS: Mean correction loss was 4.1 ± 5.4°, and 23 patients (29.5%) experienced loss greater than 5°. The psoas attenuation-by-fracture-level screw interaction was significant (B=0.311°/HU; 95% CI, 0.091-0.530; p = .006; HC3 p = .037). Without fracture-level screws, each 10-HU increase in psoas attenuation corresponded to approximately 2.6° less correction loss (B=-0.255°/HU; 95% CI, -0.440 to -0.071; p = .007), although HC3 inference crossed the null (p = .068). No clear association was observed with fracture-level screws. The interaction persisted after adjustment for vertebral height loss and canal compromise (B=0.342°/HU; 95% CI, 0.127-0.556; p = .002; HC3 p = .016), when short- versus long-segment fixation replaced instrumented-level count, and in patients with at least 12 months of follow-up. Paravertebral attenuation showed a similar pattern, whereas muscle cross-sectional area did not. CONCLUSIONS: Preoperative psoas attenuation showed a construct-dependent association with correction loss after fixation of traumatic AO Spine A3 fractures. Lower attenuation may identify mechanical vulnerability when the fractured vertebra is not incorporated into the construct. These findings are hypothesis-generating and require prospective validation before CT-based muscle assessment can guide construct selection.

Abstract as published, via PubMed.

View on PubMedFull text at the publisherOpen in the app

For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.