Limitation of the Augmentation Index for Evaluating Arterial Stiffness

Li-Tao Cheng,Li-Jun Tang,Lei Cheng,Hai-Yan Huang,Tao Wang
DOI: https://doi.org/10.1291/hypres.30.713
2007-01-01
Hypertension Research
Abstract:Although the augmentation index (AIx) is widely used to evaluate arterial stiffness in clinics and research, some conflicting data exist in regard to its validity. We therefore performed a series of studies to test the validity of AIx. The first study in 196 peritoneal dialysis patients showed that AIx in diabetics was lower than that in non-diabetic patients (p<0.05), which was in contradiction with the previous studies. Further analysis showed that AIx was just weakly correlated with pulse pressure (PP)—a known index of arterial stiffness. We also found that the increase of augmentation pressure (AP) was usually accompanied with increased central PP (C-PP). As AP and C-PP are used as the numerator and denominator in the AIx formula, an increase in the numerator (AP) would not necessarily result in an increase of the quotient (AIx) unless the denominator (C-PP) was stable. We then conducted a second study trying to test the validity of AIx through mathematical ratiocination. The increases in the central second peak (P2) and AP were assumed to represent increased arterial stiffness. Different values of AIx were obtained by varying the central initial systolic peak (P1) and diastolic pressure (DP). Mathematical ratiocination showed that AIx was dependent on multiple factors, F=(ΔSP−ΔDP)×(P1−P2)+(Δ P2−Δ P1)×(SP–DP), which suggested that a change of AIx would not always be attributable to changes in P2 and AP. This speculation was further proved by clinical data in our third study. In conclusion, through a series of studies and ratiocination, we showed that the augmentation index (AIx or AIx@75bpm) might not be a sensitive surrogate for a change in central pressure waveforms, which is a manifestation of change in large artery function. The limitation of AIx as an index of arterial stiffness is rooted in its formula, which has a clear mathematical flaw.
What problem does this paper attempt to address?