Knee osteoarthritis (OA) is one of the most prominent forms of the degenerative joint disease. Due to changes in lifestyle, it affects patients of increasingly younger age with higher levels of physical activity, justifying less invasive treatment options than total knee replacement (TKR) surgery. Unicompartmental knee replacement (UKR) surgery has been used for over 40 years to treat late-stage OA locally limited to one compartment of the knee. However, within months of the operation pain occurs in the antero-medial region of the proximal tibia in a substantial number of patients. The change in load distribution after implant insertion is thought to affect the remodelling process of the bone. As the structure within the bone changes, so does the bone marrows response to loading. Previous work has shown that the perceived location coincides with a local von Mises strain maximum in the trabeculae. As nociceptors in the bone are located on the border of the bone marrow, the current study investigated the effect of changes in bone microstructure on the pressure within the bone marrow. A fluid-structure interaction (FSI) microstructure model representative of the trabecular bone structure and bone marrow was generated using the open-source software OpenFOAM. Data generated in experiments conducted on a 3D printed specimen representing the model were used to analyse the pressure distribution throughout the structure and to validate the trabecular bone and bone marrow model. The overall pressure showed good agreement between the experiment and the model, which raised confidence in the modelling technique. The local pressure distribution, however, was not the same. Models of simple structures were generated to isolate and analyse those aspects affecting the fluid response. Alterations to the simple structure provided further insight into the effect of structural changes on the pressure within the bone marrow. The generated FSI model demonstrated the highly non-linear behaviour of the fluid and solid domains and helped to identify aspects influencing the pressure response. While an overall larger fluid domain decreased the pressure, it was not possible to identify a single predominant factor able to predict the change in pressure distribution. The choice of the fluid model had a large effect on the fluid response.