Radiographic analysis of spino-pelvic sagittal alignment and prediction of lumbar lordosis in Chinese asymptomatic subjects

English Keywords:Pelvic incidence  Lumbar lordosis  Linear regression

English Abstract:

  【Abstract】 Objectives: To investigate a prediction equation of lumbar lordosis from selective spinal and pelvic sagittal parameters in Chinese asymptomatic subjects. Methods: This was a prospective radiological analysis by using full-spine standing lateral radiographs of Chinese volunteers. One hundred seventy-one volunteers[94 males, 77 females; mean age 23.0±1.8 years(range, 18-28 years)] participated in this series. Pelvic and spinal parameters were measured, including pelvic incidence(PI), pelvic tilt(PT), sacral slope(SS), lumbar lordosis(LL), thoracic kyphosis(TK), cervical lordosis(CL), sagittal vertical axis(SVA). Pearson correlation test was calculated and a multifactor regression analysis was conducted by using the lumbar lordosis(dependent variable) and the other spinal and pelvic parameters(independent variables) to determine the best sets of predictors. According to a formula proposed by Duval-Beaupere, LL＝PI+9, a paired sample t test was conducted between the predicted and measured values. Results: PI was 44.6°±9.5°, LL was 48.4°±10.8°, SS was 34.4°±8.0°, TK was 24.2°±9.6°, CL was 7.9°±9.6°, SVA was -20.5±30.1mm. LL was significantly correlated with PI, SS, TK, CL and SVA(P<0.05). Based on a multifactor regression analysis, LL predictive equation was LL＝2.958+0.760 SS+0.323 TK+0.198 PI+0.122 CL－0.118 SVA. Only based on PI, predictive equation was LL＝0.623 PI+20.611. According to LL＝PI+9, there was a significant difference between measured and predicted lordosis(t=7.025, P<10-6). Conclusions: Using simple linear regression analysis, LL can be precisely predicted, LL＝0.623 PI+20.611. The formula proposed by Duval-Beaupere(LL＝PI+9) can not be applied in Chinese subjects.