Development of normalised quantitative measures of lumbar disc degeneration

Background: Lumbar disc degeneration (DD) is widely regarded as a likely contributor to low back pain (LBP), but the association between DD and LBP outcomes is relatively weak. There is strong evidence that DD progresses with age, regardless of LBP; however, DD is more common in people with LBP than those without LBP. These findings suggest that DD likely has a non-pathological component due to normal ageing, but also a pathological component that may be associated with LBP outcomes. It is therefore important to separate pathological from non-pathological DD. Previous studies have used quantitative measures of the disc height (DH) and disc signal intensity (DSI) from MR images to rate absolute DD severity, but no studies have attempted to systematically separate pathology from non-pathology by rating DD severity relative to factors such as age. Normalised quantitative measures (z-scores) of DSI and DH should better represent relative severity of DD, so are more likely to represent pathology and hence may be more strongly associated with LBP outcomes than existing (unnormalised) quantitative measures. Objective: This project aimed to develop normalised quantitative measures (z-scores) of DSI and DH to rate relative severity of DD and assess construct validity of the resulting normalised quantitative measures. Methods: Raw (unnormalised) quantitative measures of DSI and DH alongside the potential normalisation variables (variables unlikely to represent pathology) were acquired from the MRI scans and clinical data of 76 patients. The univariable and multivariable relationships between raw quantitative DSI/DH and the potential normalisation variables were tested to determine the final normalisation variables, from which the normalised quantitative measures of DSI and DH were developed. The construct validity of these measures was assessed by comparing them to an experienced radiologist’s subjective measures of relative severity of DSI and DH loss. Results: The cerebrospinal fluid signal intensity, age, and disc level were significantly associated with raw DSI (𝑅² = 0.06, 𝑅² = 0.25, and 𝑅² = 0.09, respectively), so were the final normalisation variables for DSI. The total lumbar height and disc level were significantly associated with raw DH (𝑅² = 0.13 and 𝑅² = 0.31, respectively), so were the final normalisation variables for DH. Normalised DSI and DH had stronger relationships (𝑅² = 0.37 and 𝑅² = 0.39, respectively) than raw DSI and DH (𝑅² = 0.31 and 𝑅² = 0.24, respectively) when compared to the radiologist’s subjective measures. Normalised DSI and DH were both normally distributed (𝑝 = 0.320 and 𝑝 = 0.115, respectively). Conclusions: Construct validity suggested that normalised quantitative measures of DSI and DH are better than existing measures of DSI and DH at rating relative DD severity and hence may provide more precise estimates of the association with LBP outcomes. Clinical validation of the normalisation models needs to be tested before the models are recommended in clinical practice.