Biochemical prediction of changes in spinal bone mass in juvenile chronic (or rheumatoid) arthritis treated with glucocorticoids.

OBJECTIVE: To identify biochemical predictors of spinal bone mineral growth and the development of spinal osteoporosis in children with juvenile chronic arthritis (JCA) treated with glucocorticoids. METHODS: Bone mass measurements were made at 3 monthly intervals for one year in 31 children. At each visit, blood and urine were obtained for assessment of laboratory indices related to the acute phase response and bone remodelling rates. Assessments were also made of joint inflammation (simple joint count). RESULTS: Plasma albumin and C-reactive protein (CRP) concentrations contributed independently of height velocity to the prediction of lumbar spinal bone mineral growth, but only when averaged over the year of observation. The simple joint count did not usefully predict spinal bone mineral changes in the individual patient, nor did any measured index normally related to bone turnover (plasma osteocalcin, 25 (OH) vitamin D, urinary hydroxyproline). Mean values of the simple joint count were predicted by mean CRP and CRP trends. Joint count trends were predicted by hemoglobin trends. None of these relationships, although statistically significant, was strong enough to predict individual outcomes precisely. CONCLUSIONS: Failure of spinal bone mineral growth is related to failure of growth in height and weight but also to biochemical markers for the activation of the acute phase response. Failure of bone growth to correlate with increased hydroxyprolinuria or plasma osteocalcin concentrations may be attributed to the confounding effect of glucocorticoid treatment on plasma osteocalcin levels in children whose bone resorption is little changed from normal levels despite their reduced growth. Biochemical measurements are weak substitutes for bone densitometry in monitoring spinal growth in these children.