OSTEOARTHRITIS AND CHRONIC BACK PAIN ARE ASSOCIATED WITH LATERAL SPINE SHAPE: A STUDY USING THE UK BIOBANK

INTRODUCTION

Chronic back pain is very common and affects over 600 million adults worldwide and has been partly attributed to OA. We have previously shown that the lateral spine has an intrinsic shape and that specific shapes have been shown to be associated with back pain in early old age. However, there is little evidence in the literature that directly links lateral spine shape with OA.

OBJECTIVE

To explore the relationships between OA, chronic back pain and lateral spine shape in a sub-cohort of the UK Biobank.

METHODS

Lateral spine iDXA scans (n=4784) from the UK Biobank imaging enhancement study were used. The cohort was 52.1% female, and the mean age was 62.2±7.5 years (Table 1). Images were annotated semi-automatically using a 143-point template encompassing the vertebral bodies from T7 to the superior margin of L5 using custom software (The University of Manchester). The points were subjected to Procrustes transform and then Principal Component Analysis to build a statistical shape model (SSM). Self-reported OA and chronic back pain (greater than 3 months duration) were taken from the questionnaire data provided at the imaging centre visit. Binary logistic regression was used to explore the associations between self-reported OA, chronic back pain, and the first 10 modes of variation. The model was adjusted for age, sex, height, weight and total spine BMD. We report odds ratios (OR) with 95% confidence intervals (CI) for each standard deviation change in mode.

RESULTS

537 participants reported OA (not site specific) and 630 reported chronic back pain. The first 10 SSM modes accounted for 88.9% of the total model variation. We found that three modes were associated with self-reported OA (modes 3,9 & 10) and a single mode was associated with chronic back pain (mode 3). It was observed that mode 3 (6.5% total model variation; Fig 1.), describing vertebral height and decreased vertebral column height was negatively associated with both self-reported OA [OR 0.88 95% CI 0.8-0.97, p=0.007] and chronic back pain [OR 0.81 95% CI 0.70-0.94, p=0.005]. Mode 3 also described a loss of spinal curvature (Fig. 1). Mode 9 (0.7% of total model variation), describing narrowing of the lumbar vertebrae) and mode 10 (0.5% of total model variation), describing a disconnect between lumbar and thoracic sections of the vertebral column were associated with an increased risk of OA [mode 9 OR 1.11 95% CI 1.01-1.022, p=0.031; mode 10 OR 1.12 95% CI 1.02-1.23, p=0.011].

CONCLUSION

We found that loss of spinal curvature and decreased vertebral body height were negatively associated with OA. Our data indicated that there was an increased risk of OA with rotation of the spine.