Allometric scaling of diameter change in the original flow-mediated dilation protocol

The negative correlation between percent flow-mediated dilation (FMD%) and baseline diameter (D(base)) has only recently been recognised as a fundamental ratio-scaling problem, which is not resolved by significance-testing of D(base) inequality between groups/conditions, nor by covariate-adjusting FMD% itself by D(base). It is resolved appropriately by allometric scaling of the relationship between peak diameter (D(peak)) and D(base) using statistical models. Therefore, we extracted data from a seminal study on FMD%, and re-analysed it using allometric methods. We found that D(peak) did not increase as a constant proportion of D(base), rendering FMD% a biased estimator of differences in endothelial function between artery sites (brachial vs femoral) and age-groups (children vs. adults). The allometric expression was D(peak)/D(base)(≈ 0.90), rather than a simple ratio. In agreement with our previous research, a proper allometric perspective on FMD led to unbiased estimates of endothelial function, with full adjustment for the influence of baseline diameter.