Sagittal balance in the conservative treatment of pathological kyphosis: a retrospective observational study

Methods Study Design: Retrospective observational study. Population: A total of 23 patients (15 males; 8 females), aged 11 to 17 years were included. Each patient had at least two clinical evaluations and spinal X rays (lateral projection) at the time of therapy start (T0) and stop (T1). Methods: 20 patients presented thoracic hyperkyphosis and three presented thoracolumbar kyphosis. All of the patients were treated conservatively; specifically, 17 were treated with brace plus specific exercises and 6 were treated with only specific exercises. For all of the patients we measured the sagittal parameters of thoracic kyphosis (TK), lumbar lordosis (LL), pelvic incidence (PI), pelvic tilt (PT), sacral slope (SS) and spinosacral angle (SSA). Statistical analysis: paired Wilcoxon test and correlation (Pearson). Results Mean and standard deviation for each parameter were measured in T0: TK, LL, PI 43.83± 7.95, PT 4.83± 6.96, SS 38.91± 7.67, SSA; and in T1: TK, LL, PI 41.87± 9.62, PT 5.48± 7.04, SS 36.39± 7.80, SSA. Highly significant improvements were found for the main spinal parameters from T0 to T1: TK reduced from 54.22± 13.58 to 44.48± 12.80 (p= 0.0006), LL from 57.43± 11.16 to 51.13± 10.28 (p= 0.0077) and SSA from 132.87± 9.80 to 127.09± 9.24 (p= 0.0036); no significant differences were found for the other parameters. Looking at sagittal parameter correlations,very few were found at the start of treatment: SS correlated with LL (R=0.651; p=0,007), SSA (R=0.600; p=0.0025) and PI (R=0.596; p=0.0027) which also correlated with PT (R=0.515; p=0.0119). Contrarily, at the end of treatment, most parameters were correlated: LL with PI ( R=0.514; p=0.0122), SS (R=0.499; p=0.0152), and SSA (R=0.743; p<0.0001), PI with PT(R=0.600; p=0.0024), SS (R= 0.691; p=0.0003) and SSA (R=0.628; p=0.0013) and SS with SSA (R=0.650; p=0.0008).