Abstract
Pelvic incidence and lumbar lordosis have only normative values for spines comprising five lumbar and five sacral vertebrae. However, it is unclear how pelvic incidence and lumbar lordosis are affected by the common segmentation anomalies at the lumbo-sacral border leading to lumbosacral transitional vertebrae, including lumbarisations and sacralisations. In lumbosacral transitional vertebrae it is not trivial to identify the correct vertebral endplates to measure pelvic incidence and lumbar lordosis because ontogenetically the first sacral vertebra represents the first non-mobile sacral segment in lumbarisations, but the second segment in sacralisations. We therefore assessed pelvic incidence and lumbar lordosis with respect to both of these vertebral endplates. The type of segmentation anomaly was differentiated using spinal counts, spatial relationship with the iliac crest and morphological features. We found significant differences in pelvic incidence and lumbar lordosis between lumbarisations, sacralisations and the control group. The pelvic incidence in the sacralised group was mostly below the range of the lubarisation group and the control group when measured the traditional way at the first non-mobile segment (30.2°). However, the ranges of the sacralisation and lubarisation groups were completely encompassed by the control group when measured at the ontogenetically true first sacral vertebra. The mean pelvic incidence of the sacraliation group thus increased from 30.2° to 58.6°, and the mean pelvic incidence of the total sample increased from 45.6° to 51.2°, making it statistically indistinguishable from the control sample, whose pelvic incidence was 50.2°. Our results demonstrate that it is crucial to differentiate sacralisations from lumbarisation in order to assess the reference vertebra for pelvic incidence measurement. Due to their significant impact on spino-pelvic parameters, lumbosacral transitional vertebrae should be evaluated separately when examining pelvic incidence and lumbar lordosis.