Dual X-Ray Absorptiometry (DXA) for measurement of body composition in athletes: Experiences that underpin the importance of optimising the reliability of measurement

Abstract

Over the past decade, DXA has become a “gold standard” for the assessment of body composition in sports nutrition, offering a reasonably precise and practical tool to monitor fat mass and fat free mass, and its changes, in athletic populations. On deciding to invest in such a tool, AIS Sports Nutrition undertook an assessment of the reliability of such measurements, with interest in identifying and minimising the sources of machine, technician and biological error involved. From this, the AIS “Best Practice Protocol” was devised, involving standardisation of athlete presentation (overnight fasted and rested, euhydrated, standard dress), placement on the scanning bed (use of position aids, protocols to address tall and wide athletes), and technician precision. The value of the implementation of this protocol was demonstrated in a research project involving overtraining and recovery in cyclists, where differences in the effects on body composition between the treatment and control groups were detectable from DXA scans using this protocol, but not from additional scans collected randomly over the day. In this case, the reduction in typical error of measurement with the standardised protocol allowed the detection of small but meaningful changes in body composition that were otherwise obscured by the measurement error associated with the intake of food and/or fluid shifts associated with exercise. More recently, we have developed a “correction” factor that can account for the artefact in measurement of fat-free mass which occurs due to changes in muscle glycogen and water content. This allowed us to detect a change in Resting Metabolic Rate, relative to fat free mass, following a low carbohydrate, high fat (LCHF) diet that was otherwise masked by the artefact introduced by the depletion of muscle glycogen and water associated with this diet. Although Best Practice Protocols add a layer of complexity to measurements and may reduce the practicality of a technique, they add value to the interpretation of the results in populations for whom small changes make a big difference.