Shear stress‐induced flow‐mediated dilatation: is it up to haemoglobin?

Importantly, FMD is a prognostic indicator of future cardiovascular disease (Areas et al. 2019) and is quantified clinically during ischaemia–reperfusion testing. The magnitude of FMD (i.e. percentage change in arterial diameter) is often normalized to to account for inter-individual differences in the magnitude of post-ischaemic reperfusion (i.e. shear stress stimulus), under the assumption that blood viscosity does not change. However, blood is a non-Newtonian fluid, meaning viscosity will vary in response to changes in . Leo et al. recently demonstrated this concept by showing that blood viscosity decreased in response to increased during forearm exercise (Leo et al. 2020). Consequently, these findings have significant implications for clinical tests of FMD, because if changes in blood viscosity affect shear stress during ischaemia–reperfusion testing, then may be an inappropriate surrogate for normalizing FMD responses to τ. To address this gap in the literature, in a recent issue of The Journal of Physiology, Hoiland et al. examined post-ischaemic FMD and reactive hyperaemia in the brachial artery of 12 healthy young men before and after isovolumic haemodilution (Hoiland et al. 2020). Briefly, the protocol involved two ischaemia–reperfusion trials consisting of 1 min baseline, 5 min ischaemia and 3 min of recovery. The cuff was positioned just distal to the elbow joint and Doppler ultrasound was used to measure changes in brachial artery blood velocity and diameter. Following the first ischaemia–reperfusion trial, subjects then underwent an isovolumic haemodilution protocol in which ∼20% of the subject's whole blood was replaced with a 5% human serum albumin solution. The success of the haemodilution protocol was confirmed via repeated testing of haematocrit. Moreover, echocardiograms and blood draws were performed before each ischaemia–reperfusion test to examine changes in cardiac function and blood haemoglobin and nitric oxide levels, respectively. Although the isovolumic haemodilution protocol significantly reduced blood viscosity by ∼20%, shear stress during the post-ischaemic reperfusion period was not different between the control and haemodilution trials. However, shear rate was greater during the haemodilution trial due to greater rates of brachial artery blood flow. Interestingly, despite the similar shear stress stimuli, FMD was augmented following haemodilution compared to the control trial (+9.7% vs. +3.8%). Given that haemodilution reduced blood viscosity, which should reduce shear stress (eqn (1)) and FMD, the authors concluded that the augmented FMD response following haemodilution was due to the reduced scavenging of nitric oxide by haemoglobin. These data are further supported by a marked reduction in nitric oxide metabolites following haemodilution. Despite these compelling results, additional figures illustrating how brachial artery diameter, blood velocity and blood flow changed over time during reperfusion would have aided the interpretation of data, especially given that the temporal changes in shear stress were similar between the control and haemodilution trials, but that peak FMD was achieved ∼30 s later. Perhaps shear rate, on its own, is more important to stimulating FMD than shear stress, which would suggest that the augmented FMD observed during the haemodilution trial was a function of prolonged/augmented shear rates. Nevertheless, the strengths of this study far outweighed the minor consideration described above. The isovolumic replacement of whole blood with human serum albumin is indeed a unique and innovative strategy to experimentally manipulate blood viscosity without affecting cardiac output or blood pressure (as evidenced by the data presented). Moreover, differences in FMD were clearly not a function of the long supine rest period between haemo-conditions, as FMD did not change in a cohort of 11 young men who underwent the same protocol without isovolumic haemodilution (i.e. time control experiment). That Hct, [Hb], viscosity and NO levels were measured before each trial, allowed explicit demonstration of isovolumic haemodilution. The results and experimental approach described by Hoiland et al. (2020) stimulate several new and exciting lines of research. Indeed, the most interesting observation in the present study is the apparent increase in FMD following isovolumic haemodilution, despite near identical estimates of shear stress throughout the reperfusion period between the control and haemodilution trials. These similar shear stress profiles persisted regardless of isovolumic haemodilution reducing blood viscosity, which should reduce shear stress (eqn (1)). While reduced NO scavenging by haemoglobin is a very plausible explanation, as put forth by the investigators, it is interesting to consider the role of reduced arterial oxygen content (and oxygen delivery) on the present results. Indeed, our laboratory has recently highlighted the importance of the ‘ischaemic stimulus’ on reactive hyperaemia (Rosenberry et al. 2019). While the between condition difference in peak reactive hyperaemia did not reach statistical significance, there was a clear trend favouring the haemodilution condition. Whether this was the result of a greater ischaemic stimulus to vasodilate, and the extent to which it influenced the present results, should be considered in future experiments. It is also interesting to consider the role of resistive forces on fluid dynamics. Future studies will indeed need to delineate whether isovolumic haemodilution augments FMD purely as a function of reduced NO stimulation, or if reductions in the blood internal flow resistance also stimulate greater FMD through greater shear rates. To deductively separate the two, new experimental designs will be necessary. For instance, repeating the isovolumic condition with varying degrees of nitric oxide synthase inhibition and/or physical manipulation of shear rate (e.g. limb position changes, arterial compression). The results of this study also have important clinical implications. For example, brachial artery FMD is an indicator of future cardiovascular disease due to its correlation with coronary artery FMD (Takase et al. 1998). Given that Hoiland et al. demonstrated lower haematocrit and haemoglobin concentrations were associated with greater FMD, we would expect these factors should also improve coronary endothelial function. However, reduced haemoglobin is associated with increased risk of hospitalization and all-cause mortality in chronic heart failure patients (Tang & Katz, 2006). More work is therefore needed before we can fully integrate the present results into the clinical setting. Perhaps the biggest question raised from this important work is: is it necessary to measure haemoglobin concentration when performing routine ischaemia–reperfusion testing? At present, this is not typically performed. Hoiland et al. sought to assess the influence of haematocrit/haemoglobin on FMD in healthy humans via isovolumic haemodilution. Their results suggest that haemoglobin is an important factor in regulating FMD by constraining endothelium-dependent NO-mediated vasodilatation. The data also raise several new important questions, stimulating an existing area of future research. No competing interests declared. All authors have read and approved the final version of this manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed. None.