Background: Lumbar disc degeneration (DD) is commonly associated with low back pain (LBP); however, the relationship remains uncertain, potentially due to differences in the way DD has been measured across studies. A valid measure of DD is needed to accurately measure the extent of DD. This study aimed to assess the predictive validity of different magnetic resonance imaging-based (MRI) methods of measuring DD including qualitative, quantitative, and new normalized measures.
Methods: The study used de-identified data from 76 participants who had recovered from LBP. Magnetic resonance imaging scans of the lumbar spine were performed and assessed at baseline and participants were followed for one-year to assess if a recurrence occurred. Multivariate Cox regression survival analysis evaluated the predictive validity of five MRI-based grading systems, using separate cox regression models for both the average and worst summary measures.
Results: The analysis demonstrated no differences in the predictive validity of the five DD grading systems in regard to LBP recurrence; however, variations in the point estimates suggested that the grading components, normalization, and method of analysis appeared to influence the direction and magnitude of effect between DD and LBP recurrence.
Conclusion: No difference in the predictive validity of different grading systems for DD was observed. However, the components used to grade DD, normalization, and how the grading system was scored for analysis may influence the measurement of associations with clinical outcomes of LBP. Standardization of a grading system measure for DD is therefore recommended.