Standard bone densitometry measurements lack accuracy in predicting an individual’s bone fracture risk, partly because these measurements are insensitive to collagen and water in bone, two known determinants of bone’s ability to resist fracture. In this context, promising techniques for improving fracture risk assessment include MR methods, as they are sensitive to the level of hydration of the bony matrix. Precisely, transverse relaxation time constants (T2) can isolate proton signals from water bound to the collagen matrix. Using ¹H NMR, our group has shown that collagen-bound water concentration better explains the age-related decrease in fracture resistance than X-ray-derived measurements.^(1-3) However, these studies were limited to cortical bone while many osteoporotic fractures occur in trabecular bone sites (e.g. vertebral bodies, femoral neck, distal radius). Analyzing trabecular bone specimens from human cadaveric bone using ¹H NMR relaxometry and μCT imaging analysis, we hypothesized that i) as a measure of bone matrix, bound water concentration is independent of bone quantity (i.e. bone volume fraction), and ii) bound water concentration decreases with age in human trabecular bone.Fresh-frozen trabecular bone cores were harvested from the medial femoral condyle of 24 human donors (14 female, 10 male, aged 23-91 years) using a diamond-embedded trephine (Fig.1). Using a water jet, cores (nominal 5 mm (height) x 9 mm (diameter)) were thoroughly cleaned of soft tissue to limit the contribution of lipid protons from bone marrow to the NMR signal. While immersed in phosphate buffered saline, each specimen was imaged using micro-computed tomography scanner (μCT50, Scanco Medical) at a 15 μm voxel size to quantify the volume of bone tissue (BV, mm³) and bone per tissue volume (BV/TV, %). The hydrated specimens were inserted, along with a reference marker, into a custom-built, low proton, loop-gap-style radiofrequency coil and placed in a 4.7-T horizontal-bore magnet (Varian Medical Systems, Santa Clara, CA, USA). Using 90°/180° RF pulses of 9-ms/18-ms duration, Carr-Purcell-Meiboom-Gill measurements with 10,000 echoes were collected at 100-μs spacing, yielding data that were fitted with multiple exponential decay functions to generate a T2 spectrum.⁽⁴⁾ The signal amplitude for bound water was calculated as the integrated T2-spectrum amplitude between 150 μs and 1.5 ms (Fig.2). Given the known proton content in the marker and the volume of bone tissue (obtained from μCT), the bound water signal amplitude was converted into units of proton concentration in bone (mol ¹H/Lbone). As data were normally distributed, Pearson correlation coefficients were used to investigate the relationship between bound water, age and BV/TV.Bound water concentration did not correlate with BV/TV (r²=15.3%, p=0.06), confirming that bound water relates to properties of bone matrix at the material level, independent of bone quantity. While BV/TV did not vary between young and older donors (r²=2.6%, p=0.6), bound water concentration in trabecular bone decreased with age (r² = 32.7%, p = 0.004, Fig. 3). The later result is consistent with observations reported for cortical bone from human cadaveric tissue.⁽⁵⁾ Taken together, these results suggest that measurement of bound water concentration in trabecular bone could be indicative of changes in matrix quality that occur with advanced age, including a shift in collagen crosslinking profile,⁽⁶⁾ decrease in collagen content,⁽⁷⁾ or decrease in non-collagenous proteins.We report for the first time ¹H NMR measurements of bound water in trabecular bone. These results validate the potential of NMR for detecting significant changes in the quality of bone tissue which directly imparts bone its fracture resistance. Future work should focus on translating these measure to clinical MRI, thereby improving the assessment of fracture risk in trabecular sites prone to fracture.