Hip structure and incident fracture: a time-updating survival analysis over 20 years of data from the Geelong Osteoporosis Study

Hip structural analysis parameters are associated with risk of fracture in women across a long follow-up period, with buckling ratio persisting independent of total hip BMD.

Purpose: Hip structural analysis (HSA) uses dual X-ray absorptiometry (DXA) hip scans to calculate geometries of narrow neck (NN), intertrochanter (IT), and shaft (S), which may complement bone mineral density (BMD) for assessing fracture risk. We aimed to determine whether HSA parameters were associated with fracture.

Methods: Participants were women (n = 986, ages 40-94 year) from the Geelong Osteoporosis Study. HSA was calculated from Lunar DPX-L scans. Low-trauma incident fractures were identified radiologically. Participants were followed from baseline to first fracture, death, or end of follow-up period (31/12/16) (13,487 person-years follow-up). Time-updating Cox-proportional hazards modelling investigated associations between HSA parameters and fracture.

Results: Three hundred thirty-five participants reported fractures (rate: 24.99/1000 person-years [95% CI 22.46-27.80]). Higher NN BMD (HR:0.12, 95% CI:0.05-0.29), cross-sectional area (CSA) (0.37, 0.26-0.52), cross-sectional moment of inertia (CSMI) (0.66, 0.50-0.89), section modulus (SM) (0.40, 0.24-0.68) and cortical thickness (CT) (0.00, 0.00-0.01 due to rounding) were associated with decreased risk. IT BMD (0.08, 0.04-0.20), CSA (0.58, 0.49-0.69), CSMI (0.90, 0.85-0.94), and SM (0.69, 0.59-0.81) were similarly associated. Decreased risk was observed at increased values of S BMD (0.20, 0.10-0.38), CSA (0.60, 0.47-0.76), SM (0.60, 0.43-0.83), and CT (0.03, 0.01-0.14). Higher S endocortical diameter (1.81, 1.29-2.53), and buckling ratio (BR) at all sites (NN: 1.07, 1.04-1.11; IT: 1.08, 1.05-1.11, S: 1.31, 1.19-1.46) were associated with increased risk. After adjustment for total hip BMD, the associations with BR at the shaft (1.14, 1.00-1.30) were sustained. Other associations were attenuated.

Conclusion: A greater shaft buckling ratio was associated with an increased risk for fracture, independent of total hip BMD.