The calcar femorale or ‘true neck’ of the femur has a role in transmitting load from the cantilevered neck to the femoral shaft (Zhang 2009). It can appear as a distinct condensation in clinical CT images because its structure is very similar to compact bone (Aspden 1998). Harty (1957) proposed that the calcar acts as a ‘spike’ in certain fall situations, contributing to splitting of the trochanter. We hypothesised that among elderly fallers, the size of the calcar would influence whether fractures occurred in the trochanteric (TR) or femoral neck (FN) site. We also asked whether patients who sustained a fracture had more or less calcar bone than frailty-matched controls that fell but didn't fracture.

The FEMCO study is designed to investigate male (M) and female (F) patients with acute hip fracture with multi-detector CT, before they undergo surgery. It includes an age, sex and frailty-matched control group (who have sustained at least one injurious fall without hip fracture). The fractured hip is reconstructed in 3D for classification of fracture type (FN or TR). For the present pilot study, there were 14 cases (5TR, 9FN mean 80+/−8.5yrs. 7M, 7F) and 11 controls (83+/−7.0yrs. 3M, 8F). Axial CT slices where a calcar was visible were opened in Stradwin 4.1 software (Treece 2011). The calcar femorale was semi-automatically selected with the flood fill tool. Each axial image that contained a visible calcar was included in the analysis, so that for each femur a single calcar volume was generated. Results were examined using ANOVA.

Combining male and female results, there was a non-significant trend towards a higher calcar volume in patients sustaining trochanteric rather than femoral neck fractures (0.73cm3 +/− 0.26 vs 0.61cm3 +/−0.14, p=0.27) but no difference between cases and controls. Males had a significantly higher calcar volume than females (mean 0.82cm3 +/− 0.24 vs 0.59cm3 +/− 0.13, p=0.005). Further studies are now planned in larger samples of each sex, to examine the role of the calcar in fracture mechanics. Three-dimensional visualisations provide a novel insight into the damage patterns and resultant fragment locations.

Introduction: Previously, 3D finite element models of the midfemoral neck (MFN) predicted maximal compressive stress in the inferior cortex and trabeculae during the heel-strike of normal gait. We hypothesised that older women (compared with younger women) would have relative preservation of inferior trabecular bone, reflecting continued walking activity into old age. Recently, we identified preservation of the infero-anterior femoral neck cortex in old age despite women at 85 having 55–80% thinner superior cortices than women aged 25. Here, we examined regional trabecular bone mineral density in an extended region of interest (ROI) of the MFN, to establish age-associated differences and their likely contribution to hip fracture in a cross-sectional study of 100 women aged 20–90 years using 64-slice computed tomography (CT).

Methods: 100 healthy women were recruited by decade from age 20 to 90 (inter-quartile range 38–72). Participants consented to an extension of a routine clinical pelvic CT scan (Siemens 64-slice CT) using a fixed threshold of 450mg/cm3 and Mindways Software (BIT-2) to analyse both hips (1mm slice thickness, 0.59 mm voxel size). The starting position for cross sections was a 1mm thick MFN slice where the femoral neck height to width ratio was 1.4, since this location along the neck axis has been shown to be highly reproducible and unaffected by age (mean 51%, SE 0.016%). 5 parallel 1mm thick slices were evaluated towards the midline. Age effects on trabecular BMD (tbBMD mg/cm3.) were evaluated using linear regression (by anatomical quadrant).

Results: Trabecular BMD was significantly lower at 85 than at 25 years old amounting to over 50% difference in 3 out of 4 quadrants: Supero-posterior tbBMD difference between 25–85 years by modelling −146mg.cm3., (absolute %) −53%, r 2 (age) 0.61, p < 0.0001; Infero-anterior −25mg.cm3., −24%, r 2 0.094, p < 0.002. In the infero-anterior quadrants (that receive maximal stress at the heel strike in normal gait) there was relative preservation (r squared only 0.094) and only a 24% difference between 85 and 25. Body weight had no association with tbBMD in any quadrant (unlike cortical thickness).

Conclusion: Trabecular bone was better preserved in the infero-anterior quadrant in elderly women than in the other 3 quadrants. Since tbBMD also makes a small but significant contribution to the reduction of the risk of failure of femoral necks ex-vivo, increasing trabecular bone in the remaining quadrants should be included in strategies to prevent hip fracture, alongside efforts to strengthen the cortex.