Bone density variation and its effects on risk of vertebral deformity in men and women studied in thirteen European centers: the EVOS Study

Abstract

In Europe there is a 3-fold variation, according to geographical center, in risk of vertebral deformity in men and women over the age of 50. We investigated the relationship between bone density, as assessed by dual-energy X-ray absorptiometry (DEXA) of the spine and hip and prevalent vertebral deformities in 13 of the 36 centers participating in the European Vertebral Osteoporosis Study (EVOS). Each center recruited an age-stratified sample of men and women aged 50 years and over, and of those who agreed to densitometry, 288/2088 women and 233/1908 men were found to have one or more deformities of the vertebrae between T4 and L4 as assessed by the McCloskey algorithm. DEXA was in each case performed on L2-L4, the proximal femur, or both. Bone densitometry results were cross-calibrated between centers using the European Spine Phantom prototype and results expressed as bone mineral density (BMD, g/cm2). In both genders, subjects with deformities involving loss of anterior vertebral body height alone comprised over 20% of the total with deformities and these related poorly to BMD. Other classes of deformity were found by logistic regression to relate significantly to BMD in one or both genders, with odds ratios for the risk of any of these ranging from 1.67 to 2.11 for a 1 SD reduction in bone density at spine, femoral neck, or trochanter (p < 0.001). Adjusting for anthropometric variables and BMD did not remove the effect of age on risk which rose 1.67- to 1.78-fold per decade according to gender. The greater unadjusted rate of increase in deformity risk with age in women was attributable to their faster rate of bone loss with age; after adjusting for age, body mass index (BMI), and BMD at the trochanter in grams per square centimeter, men had a 2-fold higher risk of deformity than women. Analysis of the relationship between mean bone density and the prevalence of deformity in each center demonstrated no significant differences between centers in either gender, after adjusting for BMD, age, and BMI together with an a posteriori statistical adjustment for imperfect cross-calibration of densitometers. It is concluded that BMD is an important determinant of deformity risk in both genders. Together with age, BMD explains much of the differences in risk both between the sexes and between individual geographical centers in Europe.