Validated formulae for estimation of whole blood viscosity underestimate the influence of erythrocyte aggregation and deformability

Dear Editor,

We enjoyed reading the recently published cross-sectional study by Gori and colleagues promoting the use of validated formulae to indirectly correlate low- and high-shear whole blood viscosity (WBV) with a variety of cardiovascular risk factors [Gori et al. 2015]. Both formulae, validated in healthy subjects, include hematocrit and plasma protein concentrations, two key determinants of blood viscosity [De Simone et al. 1990]. These formulae may be adequate for correlating blood viscosity in clinical conditions strongly influenced by these factors; however, they fall short, underestimating the impact of erythrocyte aggregation and deformability. Indeed, in their validation of these formulae, De Simone and colleagues found an 8% and 23% contribution of red blood cell aggregation to the high- and low-shear WBV, respectively (p < 0.00001 for both) [De Simone et al. 1990]. Rigidity showed a 3% contribution to high-shear WBV (p < 0.00001). This fact, though mentioned as a limitation by the authors, was not adequately discussed, an issue that would have been helpful to the reader. We believe that this limitation, and the lack of direct viscosity measurements, may account for some of the study’s results which are in disagreement with prior published data.

These observations that reveal no associations between WBV and stroke or myocardial infarction are contrary to other researchers. Antanova and Velcheva reported the presence of chronic hyperviscosity as a risk factor for stroke in patients with chronic cerebral infarctions and transient ischemic attacks [Antonova and Velcheva, 1999]. Velcheva and colleagues also reported a positive association between transient ischemic attacks and unilateral cerebral infarctions [Velcheva et al. 2008]. While hematocrit and fibrinogen are significantly elevated in cerebrovascular patients (p < 0.001), red blood cell aggregation (p < 0.05) and red blood cell deformability (p < 0.01) are also significantly impaired in stroke patients [Szapary et al. 2004]. Meanwhile, the Edinburgh Artery Study followed 1592 middle-aged adults for an average of 5 years and found that high-shear WBV was significantly higher in patients experiencing myocardial infarction and strokes than those who were event-free (p = 0.0003) [Lowe et al. 1997]. This positive correlation with cardiovascular events was as strong as diastolic blood pressure and low-density lipoprotein (LDL)-cholesterol, and even demonstrated a stronger correlation than smoking. Overall, the lack of correlation between calculated WBV and history of stroke or myocardial infarction reported by Gori and colleagues may be explained by pathological impairments in erythrocyte properties [Gori et al. 2015]. Coull and colleagues found that the severity of blood hyperviscosity but not hematocrit varies with the severity of cerebrovascular disease. Furthermore, the most significant elevations in blood viscosity were seen at low-shear rates, indicating erythrocyte aggregation and deformability as key contributors to blood viscosity in those with cerebrovascular disease [Coull et al. 1991]. Indeed, Dintenfass attributed a 10-fold increase in low-shear blood viscosity to increased erythrocyte aggregability in patients experiencing thrombosis and coronary occlusion [Dintenfass, 1963].

The negative associations between low- and high-shear WBV and female gender are consistent with previous reported findings, likely due to higher erythrocyte deformability and lower hematocrit, respectively. Kameneva and colleagues compared biophysical parameters of blood from 47 premenopausal women and 50 age-matched men [Kameneva et al. 1999]. Premenopausal women had significantly lower hematocrit (45.8 ± 2.7% versus 40.0 ± 2.4%, p < 0.001) and higher erythrocyte deformability (erythrocyte rigidity index: 3.1 ± 0.1 versus 2.8 ± 0.1, p < 0.001) than age-matched males, reflecting monthly physiologic bleeding and erythrocyte renewal [Kameneva et al. 1999]. We agree with the findings reported by Gori and colleagues and believe that these findings were explained by the significantly larger difference of hematocrit between genders as opposed to differences in erythrocyte deformability [Gori et al. 2015]. Also consistent with previous findings [Smith et al. 1992], the authors reported significant positive associations between WBV and systolic blood pressure, diastolic blood pressure, and mean arterial blood pressure (p < 0.0001 for all). These findings were found to be similar in both groups with and without cardiovascular risk factors or established cardiovascular disease and are likely attributed to the influence of hematocrit. We believe the study authors correctly interpreted these findings, suggesting that the effect of blood viscosity on systemic vascular resistance are more important than endothelial nitric oxide production in determining blood pressure. While endothelial nitric oxide release reduces systemic vascular resistance acutely and locally in response to elevated viscosity, its effects are potentially exhaustible and even toxic when present in the systemic circulation [Sloop et al. 2015c]. Further, bioenergetics for systemic vascular resistance control favor a decrease in red cell mass rather than upregulating nitric oxide synthesis [Sloop et al. 2015c].

While the use of these validated formulae may be appropriate when direct measurement of viscosity is not possible, they have definite limitations. The present study reported a negative relationship between body mass index (BMI) and WBV, likely due to the undetected influence of erythrocyte aggregation and other hemorheologic factors. Overall adiposity has been associated with increased aggregation, abdominal adiposity with increased hematocrit, and overall BMI with increased plasma viscosity and red cell rigidity, all of which increase WBV [Brun et al. 2011; Guiraudou et al. 2013].

We applaud the vision of Gori and colleagues for bringing attention to WBV, a neglected risk factor in atherosclerotic cardiovascular disease. However, we think their statement ‘The impact of blood viscosity on cardiovascular homeostasis remains incompletely understood’ is bold and misleading. The Hemorheologic-Hemodynamic Theory adequately explains the natural history of atherogenesis, including potential complications, localization, and associations with risk factors. Furthermore, previous publications have presented additional factors that can contribute to and affect blood viscosity, such as BMI, diabetes, smoking, coronary artery disease, including myocardial infarction [Sloop, 1996; Jeong et al. 2010; Sloop et al. 2015a]. This theory predicts the opposite effects of LDL-cholesterol and high-density lipoprotein (HDL)-cholesterol on blood viscosity, and provides an explanation for the protective effect for HDL-cholesterol in atherogenesis [Sloop et al. 2015b]. The shortcomings of mainstream atherogenesis theory, including the reverse transport and oxidative stress theories, have received published recognition [Sloop, 1999; Sloop et al. 2000, 2002; Scott, 2002]. Naturally, we appreciate and recognize the author’s point of view; however, at the same time, an overly cautious approach may lead to a repeat episode like the Helicobacter pylori story. More precisely, this bacterium was discovered and rediscovered several times and even successfully treated with antibiotics long before Warren and Marshall were successful in enacting a paradigm shift. The delay in recognition of the true causative relationship of this bacterium to peptic ulcer disease and subsequent mistreatment undoubtedly cost lives.

To educate the reading audience is paramount. We applaud the authors for their insight in bringing this correlation of low and high-shear whole blood viscosity with a variety of cardiovascular risk factors; however, additional factors appear to also play a meaningful role in influencing this relationship. The role of erythrocyte morphology and its functional aspects should also be given due consideration for a more full explanation.