Abstract: BACKGROUND: Lumbar spinal stenosis is the most common degenerative disease of the spine in the elderly, and its symptoms and postoperative efficacy are related to both local sequence and overall balance. However, the reasonable range of spinal-pelvic sagittal parameters in this population is unknown.  
OBJECTIVE: To determine and validate the acceptable interval of sagittal lumbo-pelvic parameters for lumbar spinal stenosis patients.
METHODS:  A total of 376 patients with lumbar spinal stenosis who were admitted to Peking University People's Hospital from June 2016 to December 2019 were selected. According to the scoring principle of propensity matching and the ratio of (2-3):1, they were divided into a derivation cohort group (n=276) and a validation cohort group (n=100). Imaging examinations were performed preoperatively and at the final postoperative follow-up to obtain sagittal spino-pelvic parameters, containing thoracolumbar kyphosis, lumbar lordosis, pelvic incidence, pelvic tilt and sagittal vertical axis. | Pelvic incidence - lumbar lordosis |, pelvic tilt and sagittal vertical axis were seen as the main parameters. Quality-of-life was evaluated by the Oswestry disability index, which contained four quarters orderly: mild, sub-mild, sub-severe and severe dysfunction. The reasonable threshold of parameters corresponding to Oswestry disability index was determined by both linear regression and logistic regression. Cases in the validation cohort were used to verify the results concluded from derivation cohort.  
RESULTS AND CONCLUSION: (1) Oswestry disability index got improvement in the final follow-up. At baseline, Oswestry disability index was positively correlated to | pelvic incidence - lumbar lordosis | (P=0.011) while at final Oswestry disability index was positively correlated to | pelvic incidence - lumbar lordosis | and pelvic tilt (P=0.029, P=0.008) but not correlated to sagittal vertical axis in the derivation cohort. (2) Mild, sub-mild, sub-severe and severe dysfunctions of Oswestry disability index were 15-35, 36-39, 40-43, and 44-50 preoperatively, but 0-4, 5-8, 9-12, and 13-40 in the final follow-up, respectively in the derivation cohort. Multiple regression analysis showed that the independent factor affecting preoperative Oswestry disability index was 
| pelvic incidence - lumbar lordosis |, with Oswestry disability index =0.19×| pelvic incidence - lumbar lordosis |+36.9 and the mean threshold of preoperative | pelvic incidence - lumbar lordosis | was 10.7°. At final, pelvic tilt was the influencing factor of Oswestry disability index and | pelvic incidence - lumbar lordosis | was the influencing factor of pelvic tilt with Oswestry disability index =0.21× pelvic tilt +3.16 and pelvic tilt =0.60×| pelvic incidence - lumbar lordosis |+12.22. (3) The mean threshold of postoperative | pelvic incidence - lumbar lordosis | was 16.0° and pelvic tilt was 23.1° by both linear regression and logistic regression. (4) In validation cohort, Oswestry disability index in patients less than the threshold of | pelvic incidence - lumbar lordosis | was superior to cases with over-larged | pelvic incidence - lumbar lordosis | (P=0.046). (5) It is concluded that | pelvic incidence - lumbar lordosis |=16.0° was probably the reasonable sagittal range for lumbar spinal stenosis with posterior surgery and we should be tolerate to spino-pelvic matching for this population.

Key words: lumbar spinal stenosis, sagittal parameter, Oswestry disability index, spino-pelvic matching, sagittal vertical axis, pelvic tilt, pelvic incidence