Dear Editor, I was interested to read the paper by Abderhalden and colleagues published in the March 2017 issue of Osteoporosis International [1]. Persons with a spinal cord injury (SCI)… Click to show full abstract
Dear Editor, I was interested to read the paper by Abderhalden and colleagues published in the March 2017 issue of Osteoporosis International [1]. Persons with a spinal cord injury (SCI) have substantial morbidity and mortality following osteoporotic fractures. The aim of the authors was to determine whether dual-energy X-ray absorptiometry (DXA) measurements predict osteoporotic fractures in this population. Additionally, the authors suggested that osteopenia (OR = 4.7) or osteoporosis (OR = 4.3) compared with normal BMD was significantly associated with fractures and higher T scores at the hip were inversely associated with fractures (OR = 0.7). There was no significant association of Tscores orWorldHealth Organization (WHO) classification with incident fractures in those with complete SCI (p > 0.15 for both). The study suggested that DXA-derived measurements at the hip predict fracture risk in persons with a SCI [1]. However, this result has nothing to do with prediction. First, significant association does not necessarily mean prediction. Moreover, for prediction studies, we need data from two different cohorts or at least from one cohort divided into two to first to develop a prediction model and subsequently validate it. Misleading results are generally the main outcome of research that fails to validate its prediction models [2–7]. Finally, in prediction studies, we must assess the interactions between important variables. Final results can be impacted dramatically when qualitative interactions are present [2–7]. This means that most of the time, without assessing the interaction terms, prediction studies will mainly produce misleading messages.
               
Click one of the above tabs to view related content.