Introduction:​ It is important to identify possible pathological mechanisms that underlie the known sexual dimorphism in bone fragility in old age. In this cross-sectional population-based study, we use data from three different skeletal sites to examine sex differences in volumetric bone density, geometry and strength indices and determine whether sex differences in these bone strength measures continue to increase into very old age.

Materials and methods:​ A total of 1715 elderly individuals (807 men and 908 women) age 67–93 years, participants in a population-based study, the Age, Gene/Environment Susceptibility-Reykjavik Study (AGES-REYKJAVIK) and not taking medications affecting bone metabolism, were studied. Quantitative computed tomography (QCT) was performed in the lumbar spine, hip and mid-femoral shaft to estimate volumetric trabecular, cortical and integral BMD, bone geometry and bone strength indices. Regression models were used to assess the effects of age and gender-adjustment for standing midlife height and current weight.

Results:​ At age 67–69 years, men had 24.9–31.7% larger cross-sectional bone size at measured sites than women. At all bone sites, women had two- to fivefold diminution in net bone mass with age compared to men but had comparable increments in bone size (1.8–6.0% per 10 years). This was reflected in significantly worse (more than twofold) bone strength measures with age in women, including compressive strength indices at the spine, femoral neck and trochanter and bending strength indices at the femoral neck.

Conclusion:​ With the limitations of a cross-sectional study, our data support the hypothesis that sex differences in bone strength continue into old age. These sex differences appear to be due to greater net bone loss in women rather than due to greater bone gain in men.