A case-control pilot study of stress fracture in adolescent girls: the discriminative ability of two imaging technologies to classify at-risk athletes

Summary Since stress fractures are common among adolescent athletes, it is important to identify bone assessment tools that accurately identify risk. We investigated the discriminative ability of two imaging technologies to classify at-risk athletes. Findings suggested that peripheral quantitative computed tomography (pQCT) has the ability to distinguish differences in bone structure in injured vs. uninjured limbs. Introduction Given the high stress fracture (SFX) prevalence among adolescent girls, an understanding of the most informative assessment tools to identify SFX risks are required. We investigated the discriminative ability of pQCT vs. dual-energy X-ray absorptiometry (DXA) to classify athletes with or without SFX. Methods Twelve adolescent athletes diagnosed with a lower-extremity SFX were compared with 12 matched controls. DXA measured areal bone mineral density (aBMD) and content of the total body, and lumbar spine. Bilateral tibiae were assessed with pQCT. At the metaphysis (3%), total density (ToD), trabecular density (TrD), trabecular area (TrA), and estimated bone strength in compression (BSIc), and at the diaphysis (38% and 66%), total bone area (ToA), cortical density (CoD), cortical area (CoA), estimated bone strength in torsion (SSIp), and peri- and endocortical and muscle area (MuA) were obtained. Cortical bone mass/density around the center of mass and marrow density (estimate of adiposity) were calculated using ImageJ software. General estimated equations adjusting for multiple comparisons (Holm-Bonferroni method) were used to compare means between (1) injured limb of the case athletes vs. uninjured limb of the control athletes and (2) uninjured limb of the case athletes vs. uninjured limbs of the controls and injured vs. uninjured limb of case athletes with a SFX. Results aBMD and content showed no significant differences between cases and controls. When comparing the injured vs. uninjured leg in the case athletes by pQCT at the 3% tibia, unadjusted TrD, total density, and BSIc were significantly lower ( p