Lumbar spine bone mineral density and trabecular bone score-adjusted FRAX, but not FRAX without bone mineral density, identify subclinical carotid atherosclerosis

Purpose Osteoporosis and atherosclerosis share common risk factors. Aim of this study was to test if FRAX (which is an algorithm that can identify subjects at risk of fracture), without or with BMD values, also adjusted for trabecular bone score (TBS) was able to identify subclinical atherosclerosis, evaluated by measurement of carotid intima media thickness (cIMT >= 0.9 mm) as compared to DXA values. Methods Ninety postmenopausal women underwent DXA measurement and cIMT evaluation. For each patient, the FRAX algorithm for major osteoporotic fracture (M) and for hip fracture (H) without BMD was computed, together with FRAX with BMD and TBS-adjusted FRAX. Serum levels of osteoprotegerin, sRANKL, and interleukin-6 were also measured. Results There were no differences in anthropometric parameters and cardiovascular risk factors between subjects with cIMT >= 0.9 mm (35% of subjects, group A) compared to those with cIMT < 0.9 mm (group B). The prevalence of osteoporosis and FRAX BMD, TBS-adjusted FRAX both for M and H were higher in group A compared to group B. The best ROC curves to identify subjects with a cIMT >= 0.9 mm were: lumbar spine T-score, with a threshold of - 2.5 SD (area under the curve, AUC 0.64; p = 0.02) with a sensibility of 50% and a specificity of 76%; TBS-adjusted FRAX H with a sensibility of 50% and a specificity of 72% (AUC 0.64; p = 0.01 with a threshold of 3%). Interleukin-6 positively correlated with FRAX BMD H and M. Conclusions FRAX without BMD does not identify subclinical carotid atherosclerosis, while lumbar spine T-score and TBS-adjusted FRAX H similarly detected it with higher specificity for T-score.