The effect of fat on bone mineral measurements in normal subjects with recommended values of bone, muscle and fat attenuation coefficients.

Published data concerning the amount, distribution and composition of bone marrow and soft tissue fat in normal subjects are reviewed. Experimental studies of the effect of fat upon bone mineral measurement have been examined. Values of calculated and measured attenuation coefficients for bone, soft tissue and fat in the energy range of interest for bone mineral measurements have been collected and compared. For most measurement sites and in most normal subjects the thickness of fat within bone marrow is comparable with the thickness of the subcutaneous fat layer, although in obese individuals subcutaneous fat thickness can be twice the marrow fat thicknesses. In specific measurement sites there is also a contribution from deep-seated soft tissue fat which is of a similar magnitude to marrow fat. The presence of bone marrow fat alters measured values to the greatest extent in CT measurements of lumbar spine mineral, partly because the incident photon energy is higher than in single and dual photon absorptiometry but also because the fraction occupied by fat in the volume of tissue examined is larger. For scattering techniques the measured value is reduced by the volume fraction occupied by fat. The normal changes in marrow and soft tissue fat observed with aging have little influence on either scattering or transmission measurement. The values of measured and calculated linear attenuation coefficients given by previous workers show reasonable agreement despite considerable differences between measurement techniques and calculation procedures.