Spine Journal - 2026-07-01 - Journal Article
The mediating role of segmental lordosis changes in the relationship between surgical approach and the incidence of adjacent segment disease in patients with degenerative spinal disorders: a retrospective cohort study.
Liu H, Huang D, Chen Z, Wang Z, Fang M, Zhang N
Topics
Key Takeaway
Segmental lordosis change (∆SL) mediated 64.6% of the ASD risk difference between MIS-LLIF and TLIF (23.2% vs 9.5%), with each additional degree of ∆SL reducing ASD odds by 47% (OR 0.53).
Summary Depth
Choose how much analysis to show on this article page.
Summary
This study asked whether ∆SL mediates the relationship between surgical approach (MIS-LLIF vs TLIF) and ASD incidence in single-level lumbar fusion. Using Baron-Kenny/Sobel and counterfactual bootstrap mediation on 143 patients, MIS-LLIF produced less segmental lordosis restoration (1.50° vs 2.60°, p<0.001) and higher crude ASD odds (OR 2.78). After adjusting for ∆SL, approach was no longer independently associated with ASD (OR 1.78, p=0.300), while ∆SL remained a significant protective factor (OR 0.53/degree, p=0.010), accounting for 64.6% of the total effect.
Key Limitation
The 1.1° mean difference in ∆SL between groups (1.50° vs 2.60°) is small relative to measurement error on plain radiographs, raising the possibility that the mediation effect is partially attributable to radiographic noise rather than true biomechanical differences.
Original Abstract
BACKGROUND CONTEXT
Minimally invasive lateral lumbar interbody fusion (MIS-LLIF) and transforaminal lumbar interbody fusion (TLIF) are widely used for lumbar degenerative disease. However, their comparative risk for adjacent segment disease (ASD) remains controversial. Changes in segmental lordosis (∆SL) may play a mediating role in ASD development, but this pathway has not been rigorously quantified.
PURPOSE
This study aimed to compare the risk of ASD following MIS-LLIF versus TLIF and to evaluate whether ∆SL mediates this relationship.
STUDY DESIGN/SETTING
An exploratory retrospective cohort study of patients who underwent single-level lumbar fusion at a single institution.
PATIENT SAMPLE
We reviewed 143 patients who underwent single-level MIS-LLIF or TLIF between January 2017 and December 2022.
OUTCOME MEASURES
The primary outcome measure was the incidence of radiographically confirmed ASD with a minimum 2-year follow-up. The mediating variable was the change in segmental lordosis (∆SL).
METHODS
Baseline demographics, surgical parameters, and radiographic outcomes were collected. The mediating effect of ∆SL was assessed using Baron and Kenny with Sobel testing, adjusting for covariates. A counterfactual-based mediation analysis with bootstrap confidence intervals (1,000 samples) was also performed to validate the findings.
RESULTS
The overall incidence of ASD was 16.1% (23.2% in MIS-LLIF vs 9.5% in TLIF; p=.045). MIS-LLIF resulted in less segmental lordosis improvement (∆SL) than TLIF (1.50° vs 2.60°; p<.001). Initially, MIS-LLIF was associated with higher ASD odds (odds ratio [OR] 2.78, 95% confidence interval [95% CI:] 1.12-7.45; p=.027). Mediation analysis (α=0.10) identified ∆SL as a mediator, accounting for 64.6% of the total effect (indirect effect: 0.083, 95% CI: 0.01-0.20, p=.028). After adjusting for ∆SL, the surgical approach was no longer significantly associated with ASD (OR 1.78, 95% CI: 0.56-5.95; p=.300), whereas ∆SL remained an independent protective factor (OR 0.53 per degree, 95% CI: 0.31-0.87; p=.010).
CONCLUSIONS
In this exploratory analysis, ∆SL statistically explained a substantial proportion of the association between surgical approach and ASD risk. Optimizing segmental lordosis restoration may be a critical and modifiable factor for mitigating ASD, warranting prospective validation.