Objectives: to gain information on bone architecture by performing bon
e texture analysis in a sample of women covering a broad age range, Pa
tients and methods: We studied 29 healthy women aged 23 to 80 years (5
5 +/- 18 years), 19 of whom were postmenopausal, None was taking drugs
known to influence bone mass or bone metabolism, Computed tomography
of the nondominant-distal radius was performed with 1 mm slice thickne
ss and 1 mm gap, Four consecutive coronal and four consecutive axial s
ections were selected for each patient and entered into a PC-type comp
uter, Bone texture was evaluated using grey level run length analysis
(five parameters), differential local variation analysis (four paramet
ers), fractal analysis (two parameters), trabecular network extraction
and three-dimensional relief characterization. The mean of each study
parameter for the four coronal sections and for the four axial sectio
ns was calculated, Absorptiometry was done in 16 patients, Results : L
inear correlations with age were strongest (P < 0.001) for parameters
measured on coronal sections by trabecular network extraction, i.e., t
rabecular bone volume (r = -0.68), trabecular plate separation (r = -0
.65), total skeletal length (r = -0.71), number of nodes (r = 0.73), n
umber of node-node segments (r = -0.74) and trabecular bone pattern fa
ctor (r = 0.71), Also, these parameters were significantly different b
etween premenopausal women (33 +/- 9 years) and postmenopausal women (
67.3 +/- 9 years), Correlations between bone mineral density and textu
re parameters were few in number and modest in strength, suggesting th
at the parameters measured may reflect bone structure rather than bone
mass, Conclusion: Bone texture in women undergoes changes with advanc
ing age that may reflect alterations in bone microarchitecture.