Allometric scaling of flow-mediated dilation: is it always helpful?

Flow-mediated dilation (FMD) is calculated as the greatest percent change in arterial diameter following an ischaemic challenge. This Traditional %FMD calculation is thought to have statistical bias towards baseline diameter (Dbase ), which is reduced by allometric scaling. This study examined whether allometric scaling FMD influenced the difference between a group of healthy young and older adults compared to the Traditional %FMD, and to determine whether a New (allometric) scaling %FMD improved the ability to obtain individually scaled FMD. Popliteal artery FMD was assessed in 18 young (26 ± 3 years) and 17 older adults (77 ± 5 years). 'Corrected' mean FMD was generated from a log-linked ANCOVA model. Individual %FMD was evaluated using three calculations: (1) Traditional %FMD calculation; (2) Atkinson (allometric) scaling %FMD (peak diameter (Dpeak)/(Dbasescalingexponent)); and (3) New scaling %FMD ((Dpeak-Dbase)/(Dbasescalingexponent)). Traditional %FMD was significantly larger in young (5·82 ± 2·58%) versus old (3·72 ± 1·26%). 'Corrected' FMD means (Y: 5·97 ± 2·12%; O: 3·98 ± 2·06%) were similar to Traditional %FMD; however, the logarithmic transformation prevents statistical interpretation of group differences. Individually scaled %FMD using the Atkinson scaling resulted in values that were corrected for variations in Dbase but that were twofold to threefold larger than those of the Traditional calculation. New scaling %FMD resulted in values that were similar to values expected (Y: 6·21 ± 2·75%; O: 3·98 ± 1·36%); however, it did not effectively correct for variation in Dbase . Recommendations regarding the advantages of allometrically scaling %FMD should be made with caution until research clearly establishes the benefits of this approach.