Trabecular bone score: why bone density alone underestimates fracture risk

Bone density tells you how much mineral is packed into a given area of bone. It does not tell you how that mineral is arranged, and arrangement is a large part of what makes bone strong.

Trabecular bone score, usually shortened to TBS, is a measure of that arrangement. It is derived from the same lumbar spine image produced during a standard DEXA scan, at no additional scanning time or radiation.

Bone quality and bone density are not the same thing

The interior of a vertebra is not solid. It is a three-dimensional lattice of fine struts called trabeculae, arranged along the lines of the loads the bone habitually carries.

That lattice can lose strength in two different ways. The struts can become thinner while the structure stays intact, or the struts can break and the connections between them disappear. Bone density measurement records the mineral lost in both cases. It cannot distinguish between them.

Structurally, they are not equivalent. A lattice with fewer connections is disproportionately weaker than one that is simply thinner, in the same way that removing cross-braces from a structure weakens it more than shaving each brace slightly.

TBS analyses the texture of the DEXA image to infer which situation is present. A high score indicates a dense, well-connected structure. A low score indicates a degraded one.

Where it changes the picture

• Type 2 diabetes. This is the clearest example. People with type 2 diabetes have a higher fracture rate than their bone density predicts — in some analyses their bone density reads as normal or even above average while fractures occur more often. TBS is typically reduced in this group, and helps explain a discrepancy that density alone cannot [1,2].
• People in the middle of the osteopenia range. For someone with a T-score around −1.8, the question of whether to treat is genuinely uncertain. A degraded trabecular score moves the assessment towards higher risk; a preserved one is reassuring. It is precisely in this ambiguous band that additional information has most value.
• Corticosteroid treatment. Steroids affect bone structure early, sometimes before density changes substantially.
• Hyperparathyroidism and other secondary causes, where the architectural effect can precede the density effect.

TBS can be incorporated into FRAX to adjust the calculated ten-year fracture probability, so it feeds into a risk figure rather than sitting alongside it as a separate number. A meta-analysis of 17,809 men and women across 14 prospective cohorts found TBS predicted major osteoporotic fracture independently of FRAX probability [1].

What it does not do

TBS does not replace bone density. It is an addition to it, and it is interpreted alongside the T-score rather than instead of it. It is measured at the lumbar spine only, so it does not directly assess the hip.

And like bone density, it is affected by body composition in the region being imaged, which is accounted for in the calculation but is a reason the analysis should be done by a service familiar with it.

Why it is not routine

TBS requires a software licence and a scanner capable of producing images at the necessary resolution. Many private scanning services, particularly those operating on price, do not offer it. NHS availability varies considerably by centre.

The consequence is that a substantial number of people are assessed on density alone, in circumstances where structure would have changed the answer.