Sagittal digital study of severe adolescent idiopathic scoliosis: a comparative study with healthy adolescents

Retrospective study. To identify associations between multiple spinal sagittal parameters and severe adolescent idiopathic scoliosis (AIS) and to provide a theoretical foundation for preoperative assessment. We collected 551 sagittal full-spine X-ray images of adolescents; 446 that did not meet the inclusion criteria were excluded. We then included and analyzed, data from 51 healthy controls and 54 adolescents with severe AIS who had complete radiographic data. The sagittal parameters of lumbar lordosis (LL), pelvic tilt (PT), spinal tilt (ST), thoracic kyphosis (TK), pelvic incidence (PI), spinosacral angle (SSA), sacral slope (SS), and spinopelvic angle (SPA) from raw imaging data were compared between healthy adolescents and those with severe AIS. Associations between sagittal parameters and disease status were analyzed using binary logistic regression. Relationships among variables were analyzed using receiver operating characteristic (ROC) curves. Only LL and ST significantly differed between adolescents with severe AIS and healthy controls ( Z = − 2.003, P = 0.048; t = 1.835, P = 0.002). None of the other parameters significantly differed ( P > 0.05). TK moderately correlated positively with LL ( r = 0.419, P = 0.002). Lumbar lordosis weakly correlated positively with ST and PI ( r = 0.302, P = 0.027; r = 0.337, P = 0.013), whereas LL was significantly and positively correlated with SSA and SS, respectively ( r = 0.761, P < 0.001; r = 0.758, P < 0.001). Weak, moderate, and strong positive correlations were found between SSA and ST, PI, and SS, respectively ( r = 0.350, P = 0.009; r = 0.493, P < 0.001; r = 0.838, P < 0.001). The SPA moderately and significantly correlated with PI and PT ( r = − 0.539, P < 0.001; r = − 0.719, P < 0.001). Pelvic incidence was significantly and positively correlated with SS and PT ( r = 0.640, P < 0.001; r = 0.641, P < 0.001). The dependent variable significantly changed in relation to both LL and ST, and the results of ROC curves confirmed the significant predictive value of ST. ST may be regarded as a key factor in the preoperative evaluation of patients with severe AIS. Attention should also be paid to the interactive relationship between spinal tilt and the parameters of different segments to formulate effective treatment strategies.