Abstract
OBJECTIVE: To investigate the potential utility of dual x-ray absorptiometry (DXA) in incidence studies of lumbar spinal spondyloarthropathy. METHODS: Fifty-eight women recruited after menopause to a study of spinal bone loss were measured every 2 years for over a decade. Five developed scan image evidence of patchy calcification and 10 developed statistically significant (p < 0.05) nonparallelism of their bone loss (or gain) in L2, L3, and L4. The number of years since menopause at which these abnormal calcification trends (ACT) occurred was made the outcome in Cox proportional hazard modeling. At baseline, diet was assessed twice using 3-day weighed intakes. Nutrients estimated included carbohydrate, fat, protein, fiber, calcium and other minerals, and 6 vitamins. Measurements at baseline of fat mass and other anthropometric variables were made. RESULTS: The best single explanatory variable for developing ACT was whole body fat mass. Dietary fat was also predictive (p = 0.05) and adding dietary vitamin D (obtained substantially from oily fish) as a second predictor improved the diet model further (to p = 0.006 for model). These 2 dietary variables remained significantly predictive when fat mass was adjusted for (p = 0.0003 for model). CONCLUSION: Serial DXA measurements of the lumbar spine have the potential to provide a new, low radiation-dose approach to early identification of localized abnormal spinal calcification in epidemiology and trials. Alongside body fat, dietary fat intake and its components may warrant further investigation as risk factors for incident degenerative disease of the spine.
