FMD is widely used to quantify endothelial function. FMD is calculated from the artery’s baseline diameter and peak diameter during reactive hyperemia. To this day, many published FMD results are based on values assessed at predefined time points, e.g., 60 s after start of hyperemia, rather than at the time of MDP. This may limit the accuracy and reproducibility of FMD. We hypothesized that FMD values at MDP would be higher and more reproducible than at predefined times, resulting in a lower number of subjects needed to show significance for a given difference in FMD. FMD was measured in subjects with and without T2D by ultrasound (12 MHz transducer; GE Logiq P7, GE Healthcare, Germany). In a subset of subjects, FMD was measured again after 30 days. All measurements were performed by certified physicians and evaluated by three certified FMD analysts using Brachial Analyzer 5 (MIA LLC, U.S.). FMD values at 60 s and 90 s after start of hyperemia were compared to FMD at MDP using ANOVA. In this study, 100 healthy middle-aged subjects and 52 subjects with T2D (mean ± SD age 56 ± 5.4 years, T2D HbA1c 7.2 ± 0.7%) underwent FMD. FMD was measured again in 25 healthy subjects after 30 days. Overall, FMD was lower at predefined time points compared to MDP (least square mean difference (95% CI)) 60 s vs. MDP 1.12% ( 1.56; -0.68); p<0.0001 and 90 s vs. MDP 1.82% ( 2.27; 1.37); p<0.0001). The intra-subject variability was lowest for MDP compared to 60 s and 90 s (15% vs. 36% vs. 51%). Assuming 80% power and alpha at 0.05, the MDP requires 15 people to detect a 1% difference in FMD versus 26 and 31 subjects if determined at 60 s and 90 s. This study demonstrated significantly higher FMD values and superior reproducibility of the individually determined maximum dilatory peak compared to peaks at predefined time points. This roughly halves the number of study participants needed to detect a 1% difference in FMD.


T. Herbrand: None. H. Coester: None. J. DeVries: Advisory Panel; Self; Novo Nordisk A/S. Employee; Self; Profil Institute for Metabolic Research. Research Support; Self; Medtronic. Speaker's Bureau; Self; Novo Nordisk A/S. C. Heiss: None. T. Heise: Advisory Panel; Self; Mylan. Research Support; Self; ADOCIA, Boehringer Ingelheim International GmbH, Dance Biopharm Holdings Inc., Eli Lilly and Company, Gan & Lee Pharmaceuticals, Johnson & Johnson, MedImmune, Mylan, Nordic Bioscience, Novo Nordisk A/S, Pfizer Inc., Poxel, Saniona, Sanofi, Wockhardt, Zealand Pharma A/S. Speaker's Bureau; Self; Eli Lilly and Company, Novo Nordisk A/S. M. Kelm: None. R. Sansone: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at