A radiographic comparative study.
To investigate the influence of radiographic lower limb-spinal length proportion on sagittal radiographic parameters.
Summary of Background Data.
Although lordotic realignment of the lumbar spine is a well-established surgical strategy, its ideal target has not been fully understood. The widely used pelvic incidence-lumbar lordosis discrepancy (PI-LL) method to guide lordotic restoration of the lumber spine in the standing posture, may be further refined using the novel, radiographic lower limb-spinal length proportion parameter in selected subjects.
A 100 healthy subjects were imaged in the standing posture using EOS imaging to obtain whole-body lateral radiographs for the measurement of sagittal radiographic parameters. Univariate analyses were performed to compare radiographic parameters between groups with different radiographic lower limb-spinal length proportion. Multivariate analyses were performed to identify the associations between lower limb-spinal length proportions and other radiographic parameters.
Regardless of lower limb-spinal length proportion (mean = 1.4), global lumbar angle (GLA) differed from spinal lordosis (SL), with the absolute means of SL and GLA larger and smaller than pelvic incidence (PI) respectively. Univariate analysis showed that patients with proportionately larger lower limb-spinal length proportion are more likely to have larger mean T1-slope, global thoracic angle (GTA), spinal kyphosis (SK), GLA, and SL. Multivariate analysis showed that a larger lower limb length-spinal length proportion is predictive of larger GLA is less than –47.69° (Odds Ratio (OR) 2.660, P = 0.026), and larger T1-slope of more than 18.84° (OR 3.695, P = 0.012).
Larger radiographic lower limb-spinal length proportion results in naturally accentuated spinal curves. These patients balance with a larger lumbar lordosis that is closer to the PI and a higher T1-slope which should be considered for spinal realignment. SL differs from GLA and should be separately assessed.
Level of Evidence: 3