Multi-scale modelling of mouse tibia loading reveals the critical role of muscle insertion surfaces.

The definition of the muscle insertion surface (MIS) is often simplified in micro-finite element (micro-FE) models despite its potential impact on the predictions of mechanical properties that are relevant to bone adaptation. This study investigated the effect of MIS size on local strain energy density (SED) in the mouse tibia. Micro-FE models of six mouse tibiae were developed from micro-CT images, and loading conditions were derived from musculoskeletal simulations of trotting gait. For the seven muscles with the highest contributions to tibial loading, three MIS configurations were tested: full anatomical surface, 50% reduced surface, and a single-node application. A total of 198 micro-FE models were run, both for individual muscle applications and multi-muscle configurations at key instants of the stance phase. SED values were compared across MIS conditions using the mean absolute difference, normalised by the median, mean, or 95th percentile of the SED. Significant differences in SED were observed between MIS configurations, with the largest effects near the MIS and boundary conditions. The effect of MIS modification on SED values was heterogeneous across the different muscles, highlighting the importance of anatomical features. Even with multiple muscle loading, SED differences exceeded 20% in most comparisons. These results demonstrate that simplifications of the MIS, especially to single-node applications, can considerably alter local SED predictions and are likely to compromise the accuracy of bone adaptation models. Accurate representation of the MIS is therefore essential for reliable mechanoregulation analysis and should be prioritised in multiscale models of skeletal loading.