Associations between parameters of peripheral quantitative computed tomography and bone material strength index.

Abstract

Background: Bone material strength index (BMSi) is measured in vivo using impact microindentation (IMI). However, the associations between BMSi and other bone measures are not clear. This study investigated whether bone parameters derived by peripheral quantitative computed tomography (pQCT) are associated with BMSi.

Methods: Participants were men (n = 373, ages 34-96 yr) from the Geelong Osteoporosis Study. BMSi was measured using an OsteoProbe (Active Life Scientific, USA). Bone measures were obtained at both the radius (n = 348) and tibia (n = 342) using pQCT (XCT 2000 Stratec Medizintechnik, Germany). Images were obtained at 4% and 66% of radial and tibial length. Associations between pQCT parameters and BMSi were tested using Spearman’s correlation and multivariable regression used to determine independent associations after adjustment for potential confounders. Models were checked for interaction terms.

Results: Weak associations were observed between total bone density (radius 4%; r = +0.108, p = 0.046, tibia 4%; r = +0.115, p = 0.035), cortical density (tibia 4%; r = +0.123, p = 0.023) and BMSi. The associations were independent of weight, height, and glucocorticoid use (total bone density: radius 4%; ß = 0.020, p = 0.006, tibia 4%; ß = 0.020, p = 0.027 and cortical density: radius 4%; ß = 4.160, p = 0.006, tibia 4%; ß = 0.038, p = 0.010). Associations with bone mass were also observed at the 66% radial and tibial site, independent of age, weight, and glucocorticoid use (ß = 4.160, p = 0.053, ß = 1.458, p = 0.027 respectively). Total area at the 66% tibial site was also associated with BMSi (ß = 0.010, p = 0.012), independent of weight and glucocorticoid use. No interaction terms were identified.

Conclusion: There were weak associations detected between some pQCT-derived bone parameters and BMSi.