Short-term risk of fracture is increased by deficits in cortical and trabecular bone microarchitecture independent of DXA BMD and FRAX: bone microarchitecture international consortium (BoMIC) prospective cohorts
Marine Sarfati,Roland Chapurlat,Alyssa B Dufour,Elisabeth Sornay-Rendu,Blandine Merle,Steven K Boyd,Danielle E Whittier,David A Hanley,David Goltzman,Pawel Szulc,Andy Kin On Wong,Eric Lespessailles,Sundeep Khosla,Serge Ferrari,Emmanuel Biver,Claes Ohlsson,Mattias Lorentzon,Dan Mellström,Maria Nethander,Elizabeth J Samelson,Douglas P Kiel,Marian T Hannan,Mary L Bouxsein
DOI: https://doi.org/10.1093/jbmr/zjae143
2024-09-05
Abstract:Identifying individuals at risk for short-term fracture is essential to offer prompt beneficial treatment, especially since many fractures occur in those without osteoporosis by DXA-aBMD. We evaluated whether deficits in bone microarchitecture and density predict short-term fracture risk independent of the clinical predictors, DXA-BMD and FRAX. We combined data from eight cohorts to conduct a prospective study of bone microarchitecture at the distal radius and tibia (by HR-pQCT) and 2-year incidence of fracture (non-traumatic and traumatic) in 7327 individuals (4824 women, 2503 men, mean 69 ± 9 years). We estimated sex-specific hazard ratios (HR) for associations between bone measures and 2-year fracture incidence, adjusted for age, cohort, height and weight, and then additionally adjusted for femoral neck (FN) aBMD or FRAX for major osteoporotic fracture. Only 7% of study participants had FN T-score ≤ -2.5, whereas 53% had T-scores between -1.0 to -2.5 and 37% had T-scores ≥-1.0. Two-year cumulative fracture incidence was 4% (296/7327). Each SD decrease in radius cortical bone measures increased fracture risk by 38%-76% for women and men. After additional adjustment for FN-aBMD, risks remained increased by 28%-61%. Radius trabecular measures were also associated with 2-year fracture risk independently of FN-aBMD in women (HRs range: 1.21 per SD for trabecular separation to 1.55 for total vBMD). Decreased failure load was associated with increased fracture risk in both women and men (FN-aBMD ranges of adjusted HR = 1.47-2.42). Tibia measurement results were similar to radius results. Findings were also similar when models were adjusted for FRAX. In older adults, failure load and HR-pQCT measures of cortical and trabecular bone microarchitecture and density with strong associations to short-term fractures improved fracture prediction beyond aBMD and FRAX. Thus, HR-pQCT may be a useful adjunct to traditional assessment of short-term fracture risk in older adults, including those with T-scores above the osteoporosis range.