Correlation of pelvic orientation with adult scoliosis

Prospective analysis of elderly volunteers with adult scoliosis.

To analyze the correlation between the sagittal pelvic parameters and adult scoliosis.

There are little data on the relationship between the sagittal pelvic parameters and adult scoliosis.

The study group comprised 108 elderly volunteers (21 men and 87 women). All underwent anteroposterior and lateral radiographs of whole spine including hip joints. The participants were classified into 3 groups: normal (Cobb angle <10 degrees), low grade (Cobb angle 10 to 19 degrees), and high grade (Cobb angle ≥20 degrees). The radiographic parameters were magnitude, location and direction of curves, sacral slope, pelvic tilting, pelvic incidence, S1 overhang, thoracic kyphosis, lumbar lordosis, and C7 plumbline. Statistical analysis was performed to identify the significant differences between the groups. Analysis of variance and Pearson correlation coefficient were used to determine the significance.

The mean age and Cobb angle of the participants was 73 years (range: 61 to 87 y), and 11.2 degrees (range: 0.3 to 36.3 degrees), respectively. There were no significant differences in the sacral slope, thoracic kyphosis, and lumbar lordosis between the 3 groups (P>0.05, analysis of variance). The mean pelvic tilting of the high-grade or low-grade groups was significantly higher than that of the normal group (P<0.0001, P=0.003, Tukey HSD post hoc test). The high-grade group had significantly higher pelvic incidence than either the low-grade or normal groups (P=0.016, P<0.0001), and the low-grade group had higher pelvic incidence than the normal group (P<0.0001). The high-grade or low-grade groups had a significantly higher S1 overhang than the normal groups (P<0.0001, P=0.002). Comparing the C7 plumbline, high-grade or low-grade groups had a significantly higher value than normal groups (P<0.0001, P=0.012). The pelvic incidence, pelvic tilting, and S1 overhang were significantly correlated with the magnitude of curvature (r=0.592, P=0.0001; r=0.434, P=0.0001; and r=0.461, P=0.0001, respectively).

There was a significant correlation between the sagittal pelvic parameters and adult scoliosis in elderly volunteers. There was a significant difference in the pelvic incidence between the normal, low-grade, high-grade adult scoliosis groups. The pelvic tilting and S1 overhang of the high-grade or low-grade groups were significantly higher than the normal group. These 3 parameters were also associated with the magnitude of curvature.