Predictive model of global tilt (GT) determined by individual thoracic kyphosis, lumbar lordosis and pelvic incidence in the human degenerative spine.

Abstract

Previous studies only investigated the simple linear relationships of global tilt (GT) with thoracic kyphosis (TK), lumbar lordosis (LL) and pelvic incidence (PI). This study aimed to establish multiple linear regressions of GT based on individual TK, LL and PI in patients with degenerative spines. Four hundred sixty-nine adult patients aged older than 35 years with various degenerative spinal diseases were enrolled in this two-centre study between January 2017 and December 2020. Full-spine X-rays were performed for all the subjects in a standing position. The following regional and global parameters were measured in the sagittal plane: TK, LL, PI, pelvic tilt (PT), sagittal vertical axis (SVA) and GT. The correlations of the GT, PT and SVA with the TK, LL and PI were analysed, and then multiple linear regressions were constructed. GT was statistically correlated with TK, LL and PI. Additionally, TK, LL and PI were significant predictors for the GT, PT and SVA models. The relevant predictive formulae were as follows: GT = -9.60 + 1.09*PI + 0.89*LL + 0.42*TK (R² = 0.935), PT = -4.49 + 0.81*PI + 0.56*LL + 0.24*TK (R² = 0.792) and SVA = -25.68 + 2.98*LL + 2.37*PI + 1.67*TK (R² = 0.416). The specific sagittal spinopelvic morphology, evaluated by GT parameters, should be determined by individual TK, LL and PI values in the degenerative spine. Surgeons can utilize such predictive models to better understand the degenerative evolution of sagittal alignment and recognize the relationships between regional parameters and global sagittal alignment to customize a precise correction strategy.