Altered bone composition in children with vertebral fracture

Abstract

In [1], the objective was to determine the differences in bone composition in fracture-prone children with and without vertebral fractures, as assessed by Fourier transform infrared spectroscopic imaging (FTIRI) and bone histomorphometry. We collected iliac crest biopsies (n = 24) from children who were suspected of primary osteoporosis. In Materials and Methods in the subsection “Bone biopsies,” we stated, “The iliac crest biopsies were taken mainly within 6 months of the fractures (n = 20), but in some cases up to 12 months had passed since the fracture (n = 4).” However, this statement was unfortunately incorrect. It should say “The iliac crest biopsies were taken mainly within 12 months of the fractures (n = 19), but in some cases up to 18 months had passed since the fracture (n = 5).” As stated above, the duration from the last fracture until the biopsy was longer than reported in the original study in some children. The time between fracture and biopsy is, of course, critical. However, significant differences between children with and without vertebral fractures were still found, and we do not believe this has affected our findings or the conclusions of the study. In Results in the subsection “Bone histomorphometry-based data,” we stated, “Although children with low BV/TV tended to have lower carbonate-to-phosphate ratios, this relationship was not statistically significant (p = .07, Fig. 3A).” After re-checking our data, this is a significant change (p = .033). Therefore, it should be corrected as “Children with low BV/TV had lower carbonate-to-phosphate ratios than children with normal BV/TV (p = .033, Fig. 3A).” A The children with low cancellous bone volume in biopsy (BV/TV Z-score below –1 SD; n = 7) showed lower carbonate-to-phosphate ratio (p < .05) than the children with normal cancellous bone volume (BV/TV Z-score ≥−1 SD; n = 17). The figure presents mean ± SD, and statistical difference (Mann-Whitney U test, *p < .05) are indicated. Further, in the second last paragraph of Results we stated “When the samples were grouped based on turnover rate, carbonate-to-phosphate and carbonate-to-amide I ratios were lower (p < .05, ANOVA) among children with vertebral fracture than among children without vertebral fracture”. Unfortunately this statement is incorrect (p = .056–.106) and should be omitted. These statements have been discussed in the fourth paragraph of Discussion. The statement “In the pooled sample set, the children with low cancellous bone volume in the biopsy tended to show a lower level of carbonate substitution in crystals than children with normal bone volume (p = .07)” should be corrected as “In the pooled sample set, the children with low cancellous bone volume in the biopsy showed a lower level of carbonate substitution in crystals than children with normal bone volume.” The conclusions remain unchanged. The authors regret the mistakes.