Are the Relationships of Lean Mass and Fat Mass With Bone Microarchitecture Causal or Due to Familial Confounders? A Novel Study of Adult Female Twin Pairs

It is not known whether the relationships of lean mass (LM) and fat mass (FM) with bone microarchitecture and geometry are causal and/or are because of confounders, including familial confounders arising from genetic and environment effects shared by relatives. We tested the hypotheses that: (i) LM is associated with cortical bone traits, (ii) FM is associated with trabecular bone traits, and (iii) these relationships of LM and FM with bone microarchitecture and geometry have a causal component. Total body composition was quantified for 98 monozygotic (MZ) and 54 dizygotic (DZ) white female twin pairs aged 31 to 77 years. Microarchitecture at the distal tibia and distal radius was quantified using HRpQCT and StrAx software. We applied the Inference about Causation through Examination of FAmiliaL CONfounding (ICE FALCON) method. Within-individuals, distal tibia total bone area, cortical area, cortical thickness, and trabecular number were positively associated with LM (standardized regression coefficient (β) = 0.13 to 0.43; all p < 0.05); porosity of the inner transitional zone (ITZ) was negatively associated with LM (β = −0.22; p < 0.01). Trabecular number was positively associated with FM (β = 0.40; p < 0.001), and trabecular thickness was negatively associated with FM (β = −0.27; p < 0.001). For porosity of ITZ and trabecular number, the cross-pair cross-trait association with LM was significant before and after adjustment for the within-individual association with LM (all ps < 0.05). For trabecular number, the cross-pair cross-trait association with FM was significant before and after adjustment for the within-individual association with FM (p < 0.01). There were no significant changes in these cross-pair cross-trait associations after adjustment for the within-individual association (p = 0.06 to 0.99). Similar results were found for distal radius measures. We conclude that there was no evidence that the relationships of LM and FM with bone microarchitecture and geometry are causal; they must in part due to by familial confounders affecting both bone architecture and body composition.