Metacarpal bone mineral density by radiographic absorptiometry predicts fracture risk in patients undergoing maintenance hemodialysis

2020 
Abstract The 2017 Kidney Disease: Improving Global Outcomes (KDIGO) guideline update suggests bone mineral density testing to assess fracture risk in patients with chronic kidney disease, but dual-energy X-ray absorptiometry is not available in most dialysis facilities. Radiographic absorptiometry is an inexpensive and quick method for evaluating bone mineral density. Therefore, we analyzed a historical cohort of 456 maintenance hemodialysis patients to determine whether metacarpal bone mineral density measured by digital image processing, a computer-assisted radiographic absorptiometry technique, predicts fracture risk. At baseline, the median metacarpal bone mineral density T-score was -2.05 (interquartile range, -3.35 to -0.99). During a mean follow-up of 5.3 years, there were 16 clinical fractures and 11 asymptomatic vertebral fractures as estimated by height loss. Metacarpal bone mineral density T-score was significantly lower in patients who sustained a clinical fracture than in those remaining event-free. Decreasing metacarpal bone mineral density T-score was significantly associated with increased risk of clinical fracture (hazard ratio, 1.41 per 1 standard deviation decrease in bone mineral density T-score [95% confidence interval, 1.09 to 1.83]; the hazard ratio for lowest versus highest tertile was 4.86 [1.03 to 22.92]. Similar associations were observed between metacarpal bone mineral density T-score and vertebral fracture or any fracture. The results were robust to different analysis strategies and were consistent across different subgroups. Thus, radiographic absorptiometry could be a useful tool for primary screening of hemodialysis patients at high risk for fracture. Additional studies are required to determine the predictive ability of radiographic absorptiometry techniques compared to dual-energy X-ray absorptiometry or other established methods.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    29
    References
    4
    Citations
    NaN
    KQI
    []