Postoperative bone loss after posterior lumbar interbody fusion is a risk factor of cage subsidence independently of preoperative CT-derived attenuation: a retrospective study based on hounsfield unit

This study aims to explore the relationship between postoperative bone loss and cage subsidence (CS), as well as paraspinal muscle state and CS. The difference in vertebral Hounsfield Units (HU) between preoperative and postoperative CT scans was used to assess bone loss. Baseline characteristics of CS and non-CS groups were compared, followed by propensity score matching (PSM) and inverse probability treatment weighting (IPTW) for balance. Potential predictors from univariate analysis were adjusted for sex, age, and BMI in multivariate analysis to identify independent risk factors for CS. A total of 129 participants were included in the final analysis. The subsidence rate was 34.9%. The CS group had lower HU values. PSM successfully matched 36 pairs of patients, and the CS group exhibited more pronounced bone loss (CS 12.04 (11.07) vs. non-CS 5.64 (11.50), p = 0.016). IPTW analysis confirmed the robustness of the results. Univariate analysis indicated association between age, paraspinal muscle density, cage height, HU, bone loss, and CS. In multivariate analysis, only bone loss (OR: 1.96, 95% CI: 1.20–3.18, p = 0.01), and low HU (OR: 0.43, 95% CI: 0.22–0.84, p = 0.01) were identified as independent risk factors. Sensitivity analysis results were consistent with the main analysis. Preoperative vertebral HU, and postoperative bone loss were identified as independent risk factors for CS. Our study suggested that spine surgeons should pay attention to the preoperative bone status of lumbar fusion patients and emphasize postoperative anti-bone loss treatments to avoid implant complications.