Reproducibility of DXA-based bone strain index and the influence of body mass: an in vivo study

Objectives Bone strain index (BSI) is a dual-energy X-ray absorptiometry (DXA)-derived index of bone strength obtained from lumbar densitometric scan. We estimated the reproducibility of BSI in healthy women with different body mass index. Methods We enrolled postmenopausal women (mean age +/- SD: 66 +/- 10 years) divided into three groups (A, B and C) according to body mass index (BMI: < 25; 25-29.9; >= 30 kg/m(2)) and two groups (D and E) according to waist circumference (WC: <= 88; > 88 cm), each of 30 subjects. They underwent two DXA examinations with in-between repositioning, according to the International Society for Clinical Densitometry guidelines for precision estimation. Bone mineral density (BMD) and BSI were expressed as g/cm(2) and absolute value, respectively. The coefficient of variation (CoV) was calculated as the ratio between root-mean-square standard deviation and mean; least significant change percentage (LSC%) as 2.77 x CoV; reproducibility as the complement to 100% LSC. Results BSI increased proportionally to BMI and WC and significantly in group C compared to B and A (p = 0.032 and 0.006, respectively). BSI was significantly higher in E compared to D (p = 0.017), whereas no differences were observed in BMD. Although BSI reproducibility was slightly lower in group C (89%), the differences were not significant between all groups. BMD reproducibility did not significantly differ between all groups. Conclusions BSI reproducibility was significantly lower than that of BMD and decreased proportionally to BMI and WC increase. This reduction of BSI reproducibility was more pronounced in patients with BMI >= 30 and WC > 88, as expected, being BSI a parameter sensible to weight.