Investigating the level of agreement of two positioning protocols when using dual energy X-ray absorptiometry in the assessment of body composition
Flinn, S, Persson, C, Simas, V, Furness, J, Schram, B & Climstein, M 2017, 'Investigating the level of agreement of two positioning protocols when using dual energy X-ray absorptiometry in the assessment of body composition', PeerJ, vol. 5, pp. e3880.
Dual energy X-ray absorptiometry (DXA) is a commonly used instrument for analysing segmental body composition (BC). The information from the scan guides the clinician in the treatment of conditions such as obesity and can be used to monitor recovery of lean mass following injury. Two commonly used DXA positioning protocols have been identified—the Nana positioning protocol and the National Health and Nutrition Examination Survey (NHANES). Both protocols have been shown to be reliable. However, only one study has assessed the level of agreement between the protocols and ascertained the participants’ preference of protocol based upon comfort. Given the paucity of research in the field and the growing use of DXA in both healthy and pathological populations further research determining the most appropriate positioning protocol is warranted. Therefore, the aims of this study were to assess the level of agreement between results from the NHANES protocol and Nana protocol, and the participants’ preference of protocol based on comfort.
Thirty healthy participants (15 males, 15 females, aged 23–59 years) volunteered to participate in this study. These participants underwent two whole body DXA scans in a single morning (Nana positioning protocol and NHANES positioning protocol), in a randomised order. Each participant attended for scanning wearing minimal clothing and having fasted overnight, refrained from exercise in the past 24 h and voided their bladders. Level of agreement, comparing NAHNES to Nana protocol was assessed using an intra-class correlation coefficient (ICC), concordance correlation coefficient (CCC) and percentage change in mean. Limit of agreement comparing the two protocols were assessed using plots, mean difference and confidence limits. Participants were asked to indicate the protocol they found most comfortable.
When assessing level of agreement between protocols both the ICC and CCC scores were very high and ranged from 0.987 to 0.997 for whole body composition, indicating excellent agreement between the Nana and NHANES protocols. Regional analysis (arms, legs, trunk) ICC scores, ranged between 0.966 and 0.996, CCC ranged between 0.964 and 0.997, change in mean percentage ranged between −0.58% and 0.37% which indicated a very high level of agreement. Limit of agreement analysis using mean difference ranged between −0.223 and 0.686 kg and 95% CL produced results ranging between −1.262 kg and 1.630 kg. The majority (80%) of participants found the NHANES positioning protocol more comfortable.
This study reveals a strong level of agreement as illustrated by high ICC’s and CCC’s between the positioning protocols, however systematic bias within limit of agreement plot and a large difference in 95% confidence limits indicates that the protocols should not be interchanged when assessing an individual. The NHANES protocol affords greater participant comfort.