The role of chronic hyperviscosity in vascular disease

Abstract

The pathogenesis of several major cardiovascular diseases, including atherosclerosis, hypertension, and the metabolic syndrome, is not widely understood because the role of blood viscosity is overlooked. Low-density lipoprotein accelerates atherosclerosis by increasing blood viscosity in areas of low flow or shear, predisposing to thrombosis. Atherosclerotic plaques are organized mural thrombi, as proposed by Duguid in the mid-twentieth century. High-density lipoprotein protects against atherosclerosis by decreasing blood viscosity in those areas. Blood viscosity, at the least, contributes to hypertension by increasing systemic vascular resistance. Because flow is inversely proportional to viscosity, hyperviscosity decreases perfusion and glucose utilization by skeletal muscle, contributing to hyperglycemia in the metabolic syndrome. Therapeutic phlebotomy reduces blood pressure and serum glucose levels in the metabolic syndrome by improving blood viscosity.

Background

The pathogenesis of vascular disease is complex and as of yet not totally understood. Conventional wisdom suggests that atherosclerotic cardiovascular disease is caused by the accumulation of lipid molecules in the arterial luminal wall. Various therapeutic modalities have been proposed; however, current treatment with cholesterol-lowering drugs has not completely solved this problem. Risk factors aiding in the development of atherosclerotic disease have been long promoted, such as genetics, dyslipidemia, hypertension, smoking, lack of adequate exercise, metabolic syndrome, and obesity. However, 50% of myocardial infarctions occur in subjects without overt hyperlipidemia, and 20% of myocardial infarctions occur in the absence of any classic risk factors [Ridker and Libby, 2011].

The reason why the pathogenesis of chronic vascular diseases, including atherosclerosis, hypertension, and the metabolic syndrome, is not fully understood by the mainstream is because the role of blood viscosity has been ignored. Most physicians are aware of acute hyperviscosity syndromes as can be seen in Waldenstrom’s macroglobulinemia, polycythemia vera, and leukemia, which require immediate intervention. In these conditions, blood viscosity is in the range of 60 millipoise or greater at a shear rate of 94.5/s. However, chronic lesser elevations of viscosity do occur (roughly, 50–56 millipoise at 94.5/s, normal being 37–49 millipoise at the same shear rate) [Antonova and Velcheva, 1999; Velcheva et al. 2008]. Rosenson and colleagues reported a normal value of 32.6 millipoise at a shear rate of 100/s [Rosenson et al. 1996].

States of chronic, lower elevations of viscosity are clinically less obvious, and left unappreciated, can ultimately shorten one’s lifespan by contributing to cardiovascular disease. In the study of Antonova and Velcheva, viscosities in the range referred to here as ‘chronic hyperviscosity’ were associated with chronic cerebral infarctions, transient ischemic attacks, and risk factors for stroke [Antonova and Velcheva, 1999]. In Velcheva and colleagues’ work, chronic hyperviscosity was associated with transient ischemic attacks and unilateral cerebral infarctions [Velcheva et al. 2008].

It should be noted that the reference ranges for blood viscosity are tentative, and work remains in standardizing the reporting and measuring of blood viscosity. Hopefully, a collaboration similar to the Reference Values for Arterial Stiffness’ Collaboration, which established reference ranges and standardized methodology for pulse wave velocity testing [Reference Values for Arterial Stiffness’ Collaboration, 2010] can provide useful information in the field of hemorheology and widen clinical utilization of blood viscosity measurement.

We see a parallel between blood viscosity and hypertension, which itself may be caused by increased blood viscosity (see below). Acute hyperviscosity is conceptually similar to a hypertensive crisis in that both are extreme elevations and medical emergencies. Chronic hyperviscosity and primary hypertension are both lesser elevations and asymptomatic but can decrease longevity. Like primary hypertension, chronic hyperviscosity should be treated.

The basics of blood viscosity

Viscosity is the resistance to flow in a blood vessel. Increased blood viscosity is seen in association with all of the major risk factors for atherosclerosis: hypercholesterolemia, hypertension, cigarette smoking, diabetes, metabolic syndrome, obesity, aging, hyperfibrinogenemia, and male sex [Sloop, 1996; Sloop et al. 1998]. Blood is a non-Newtonian fluid, meaning that its viscosity decreases as shear rate increases. This is due to both erythrocyte deformability and erythrocyte disaggregation. At high rates of shear or flow, erythrocytes reversibly deform to minimize resistance to flow and viscosity. In addition, erythrocytes ‘entrain’ or organize into columns, which minimizes potential collisions between erythrocytes caused by tumbling and orbiting motions, further decreasing viscosity [Schmid-Schonbein et al. 1969]. In areas of slow flow or low shear, as occur naturally in the vascular tree against the outer wall of vascular branches and inner wall of curves, erythrocytes can aggregate and increase viscosity. Any molecule with a diameter or length large enough to simultaneously bind two erythrocytes, such as fibrinogen, can foster erythrocyte aggregation. These aggregates are weak and break apart with increasing shear force. They do not form against the opposite vessel wall because of higher shear and faster flow.

The role of lipoproteins in atherosclerosis

Lipoproteins, containing both proteins and lipids, are essential to aid in transporting fats in our blood to all bodily tissues. Lipoproteins are classified according to their density, such as very low, intermediate, low density and high density. Low-density lipoprotein (LDL) has a particle diameter of 21.8–27.5 nm [Campos et al. 1992], which is large enough to simultaneously bind two erythrocytes with a separation between the erythrocytes of approximately 18 nm due to their electronegative surface charge [Van Oss and Absolom, 1983], causing erythrocyte aggregation and increasing blood viscosity at low shear rates [Sloop and Garber, 1997]. High-density lipoprotein (HDL), having a particle diameter of 8.6–10.1 nm [Van der Steeg et al. 2008], is too small to simultaneously bind two erythrocytes. Rather, by competing with LDL for erythrocyte binding, HDL decreases blood viscosity, further emphasizing the importance in raising HDL levels, which can help in protecting against vascular atherosclerosis. Clinical investigations into developing pharmaceuticals, such as torcetrapib and dalcetrapib, which inhibit cholesteryl ester transfer protein, have failed because of their untoward effect in increasing the size of HDL particles so that they also cause erythrocyte aggregation and increase blood viscosity [Brousseau et al. 2004; Holsworth et al. 2014].

Atherosclerotic plaques are organized mural thrombi

The development of atherosclerotic plaques appears to have its origin in infancy with fatty streaking. Over time, accumulation of lipids, smooth muscle cells, foamy macrophages, and eventually cholesterol crystals with or without calcium occurs under the endothelium. With continued progression, a plaque is formed, which protrudes into the vessel’s lumen. Unfortunately, if a plaque ruptures, a thrombus is formed.

It is becoming clear that atherosclerotic plaques develop preferentially in areas of oscillatory low shear [Cecchi et al. 2011]. A thrombus can form spontaneously in this low shear environment because of decreased shear-dependent endothelial expression of antiplatelet molecules such as nitric oxide and prostacyclin (prostaglandin I2) [Wentzel et al. 2012]. In arteries, thrombi usually remain localized to one part of the vessel wall, rather than occlusive due to the higher velocity of blood flow against the opposite wall, protecting from a thrombotic condition. Thus, these thrombi are called mural or ‘parietal’ [Majno and Joris, 2004]. Following organization into an atherosclerotic plaque, this partial obstruction causes an increase in peak blood velocity post obstruction, creating an area of flow separation and low shear force distal to the obstruction, much like rapidly flowing water past a rock. This area of relative stasis will limit oxygenation of the underlying endothelium, leading to potential ulceration or erosion and subsequent occlusive thrombosis. This scenario accounts for the common occurrence of thrombosis (79% of patients in one study) [Schoenhagen et al. 2002] occurring on younger, less stenotic plaques, that is, ‘vulnerable’ or ‘unstable’ plaques distal to the more proximal culprit lesion. Regions of slow blood flow, as seen in areas of flow separation with changing arterial geometry or in deep veins during immobility, are more prone to thrombosis, as Virchow noted during the nineteenth century [Del Zoppo, 2008]. In these areas, one can imagine a scenario of unremitting positive feedback, in which erythrocytes aggregate, increase local blood viscosity, decrease local flow, allowing further aggregation, further decreases in flow, ultimately leading to thrombosis. Not surprisingly, hypercholesterolemia is a risk factor for deep venous thrombosis [Kawasaki et al. 1997] by increasing serum LDL and augmenting erythrocyte aggregation.

Once a thrombus is formed, organization occurs. Organization is a process in which capillaries grow into a thrombus, hematoma, or other avascular tissue, allowing influx of fibroblasts and subsequent deposition of collagen. Atherosclerotic plaques are organized mural thrombi, as shown by Duguid in the mid-twentieth century [Duguid, 1960; Sloop et al. 2002].

Recognition that atherosclerotic plaques are organized mural thrombi is crucial because it eliminates the need to force a cause and effect relationship with fatty streaks. Although not widely publicized or even scrutinized, a putative relationship between fatty streaks and atherosclerotic plaques is problematic: fatty streaks are universally present in infancy, regardless of the prevalence of atherosclerotic plaques in the adult population, and routinely resolve without sequelae [Sloop, 1999; Sloop et al. 1998]. In the Pathobiological Determinants of Atherosclerosis in Youth Study in subjects aged 30–34, the percent of intimal surface containing fatty streaks in the abdominal aorta (27.8%) is much greater than in the right coronary (6.9%), and yet the surface occupied by atherosclerotic plaques is nearly equal in the two (8.3% compared with 5.7%). In the thoracic aorta, 20.1% of the intima contained fatty streaks, and only 0.8% contained atherosclerotic plaques [Pathobiological Determinants of Atherosclerosis on Youth Research Group, 1993]. These data should raise doubt about a simple cause and effect relationship of cholesterol and atherosclerosis.

Further, there is biochemical dissimilarity between the lipids in fatty streaks and atherosclerotic plaques. Smith and colleagues found no overlap between the proportions of oleic, linoleic, and eicosatrienoic acids in fatty streaks and fibrous plaques [Smith et al. 1968]. The inconsistent relationship between fatty streaks and atherosclerotic plaques should be impetus to reject a possible relationship, not the stimulus for more research. From the standpoint of the principle of simplicity (Occam’s razor), explaining the entire course of a disease, from lesion to complication (superimposed thrombosis), with a single pathogenic mechanism is desirable.

The idea that an atherosclerotic plaque is an organized mural thrombus was widely accepted prior to the growing awareness that hypercholesterolemia is a risk factor for atherosclerosis [Holsworth et al. 2014]. At that point, the ‘modified response to injury’ paradigm became dominant. Ultimately, the facts became too obvious to ignore and the idea that organization of thrombi can cause an atherosclerotic plaque, particularly of the lipid poor variety (type Vc in the classification of Stary and colleagues) was accepted [Stary et al. 1995]. These lesions are common. In a morphometric analysis of plaque constituents, in lesions which caused less than 51% narrowing, 0% consisted of extracellular lipid [Kragel et al. 1990]. The hemorheologic–hemodynamic theory of atherogenesis [Sloop, 1996, 1999] reconciles the facts that atherosclerotic plaques develop from organized mural thrombi and hypercholesterolemia is a risk factor for atherosclerotic cardiovascular disease.

The Edinburgh Artery Study demonstrated the role of chronic hyperviscosity in atherosclerotic cardiovascular disease. In one cohort of 1592 men and women aged 55–74, patients sustaining a cardiovascular event had significantly higher blood viscosity values (p = 0.0003). Blood viscosity was at least as strong a signal for cardiovascular events as were diastolic blood pressure and LDL cholesterol, and was a stronger signal than cigarette smoking [Lowe et al. 1997]. This is not surprising because increased blood viscosity is one final common pathway by which all risk factors accelerate atherogenesis. The major epidemiologic studies of blood viscosity in cardiovascular disease have recently been reviewed by Holsworth and colleagues [Holsworth et al. 2013].

Hypertension

The precise etiology behind the development of hypertension is unknown; however, this ‘silent killer’ continues to exact a heavy toll on humanity, despite improved pharmacotherapy. Trying to prove a role for dietary salt in hypertension has consumed substantial resources and caused great controversy. The journalist Gary Taubes wrote in Science in 1998: ‘two conspicuous trends have characterized the salt dispute: on the one hand, the data are becoming increasingly consistent—suggesting at most a small benefit from salt reduction—while on the other, the interpretations of the data, and the field itself, have remained polarized’ [Taubes, 1998]. Bearing out Taubes’ comments is the result of a Cochrane meta-analysis published in 2013. Reduction of dietary salt by 4.4 g/day decreased systolic blood pressure by 4.18 mmHg and diastolic blood pressure by 2.06 mmHg. In patients with hypertension, systolic blood pressure dropped 5.39 mmHg and diastolic blood pressure dropped 2.82 mmHg. [He et al. 2013]. Could there be an adjuvant therapy to further aid in blood pressure reduction and its consequences if not controlled? The modest decreases in blood pressure reported in the meta-analysis are small in comparison to the reductions in blood pressure found with therapeutic phlebotomy [Houschyar et al. 2012]. Besides the reduction in systolic blood pressure, therapeutic phlebotomy also offers the added feature of reducing blood viscosity.

Blood pressure is determined by, among other factors, cardiac output and systemic vascular resistance. Systemic vascular resistance is determined by vascular tone and blood viscosity. These relationships suggest a role for blood viscosity in hypertension, which has been previously reported, in which blood viscosity is directly related to blood pressure [Letcher et al. 1981, 1983]. Furthermore, Lominadze and colleagues reported in a rat model of spontaneous hypertension that increased red blood cell aggregation and plasma hyperviscosity were present not only during the established phase of hypertension, but also before blood pressure became elevated [Lominadze et al. 1998]. The strongest evidence for the role of increased blood viscosity in hypertension is the decreased blood pressure in patients with metabolic syndrome following therapeutic phlebotomy [Houschyar et al. 2012].

Blood pressure is related to vascular wall tension as represented by the Young–Laplace equation: tension = blood pressure × vessel radius/wall thickness [Nichols and O’Rourke, 1998]. Because of the direct relationship of blood viscosity to hypertension demonstrated by Letcher and colleagues [Letcher et al. 1981, 1983], chronic hyperviscosity increases the mechanical load on the elastic elements of the vasculature, leading to acceleration of mechanical fatigue, reduced vascular compliance, and ultimately failure of those elements. This results in another example of unremitting positive feedback: decreased arterial compliance results in a widened pulse pressure, increased wall stress, accelerated arterial stiffening, further increasing the pulse pressure, etc.

Increased aortic stiffness increases peak aortic blood velocity, which increases Reynolds number, the parameter describing the likelihood of developing eddy currents and areas of low shear in the vascular tree. Increased peak arterial blood velocity and the development of areas of flow separation and eddy currents in arteries is the second pathway by which risk factors accelerate atherosclerosis [Perret and Sloop, 2000].

Viscosity is inversely proportional to Reynolds number (R_e = blood velocity × vessel diameter × fluid density/dynamic viscosity [Nichols and O’Rourke, 1998]) and should decrease the propensity for development of areas of eddy currents and low shear in the vascular tree. However, in any single subject, increased peak blood velocity will decrease blood viscosity because of the shear-thinning nature of blood viscosity, as well as an increase in the Reynolds number (an increase in the numerator with a decreasing denominator).

Metabolic syndrome and therapeutic phlebotomy

Metabolic syndrome is the constellation of insulin resistance, obesity, hypertension, and hypercholesterolemia. The cause of the metabolic syndrome is obscure; however, with the increasing standard of living worldwide and the attendant epidemic of obesity, insight into the cause of metabolic syndrome is becoming imperative. In 1991, Julius and colleagues proposed that the syndrome is caused by decreased perfusion of skeletal muscle [Julius et al. 1991]. In 1998, Hoieggen and colleagues demonstrated increased blood viscosity in the syndrome [Hoieggen et al. 1998]. Each element or trait in this syndrome has also been shown to be associated with increased blood viscosity [Sloop, 1996]. Theoretically, because flow is inversely proportional to viscosity, reducing blood viscosity should improve perfusion of muscle and increase glucose utilization, decreasing blood glucose levels. This was confirmed by Houschyar and colleagues who reported that when removing between 250 and 500 cm³ of blood (similar to a blood donation) from subjects on two occasions, baseline and 4 weeks later, systolic blood pressure decreased 18.3 mmHg compared with only 0.2 mmHg in controls. In addition, serum glucose levels decreased 12.5 mg/dl in subjects and only 2 mg/dl in controls [Houschyar et al. 2012]. This study demonstrates the causative relationship of blood viscosity to blood pressure and the importance of achieving normal blood flow and viscosity to maintain homeostasis.

The relationship of therapeutic phlebotomy to hemorheologic variables was recently reviewed by Holsworth and colleagues [Holsworth et al. 2014]. Therapeutic phlebotomy results in improving several hemorheologic parameters. In a study of 30 patients who donated 450 ± 30 cm³ every 4 days for a total of up to four donations in anticipation of possible autologous transfusion, hematocrit, the single most powerful determinant of blood viscosity, decreased from 42.1% to 30.9% at the end of the donation period and increased to 38.5% approximately 3 weeks later. Blood viscosity at 75/s decreased from 35.9 to 23.0 millipoise over the donation period and increased to 33.7 millipoise approximately 3 weeks later [Clivillé et al. 1998]. Houschyar and colleagues reported that two phlebotomies 4 weeks apart resulted in a drop in hemoglobin from 14.3 ± 1.2 to 13.3 ± 1.1 mg/dl after 6 weeks [Houschyar et al. 2012].

However, repeated phlebotomy will eventually induce iron deficiency anemia and possibly reactive thrombocytosis, which may be deleterious overall.

The reduction in hematocrit following a single episode of phlebotomy is relatively short lived. However, decreases in blood viscosity are more sustained. Older, less deformable erythrocytes are replaced by younger more deformable cells [Holsworth et al. 2014]. In addition, Houschyar and colleagues found a reduction in LDL/HDL ratio, which can result in decreased erythrocyte aggregability [Clivillé et al. 1998].

Conclusion

Cardiovascular disease is still the leading cause of deaths for both men and women worldwide. Many risk factors have been identified and current therapeutic efforts have been centered on addressing these risk factors. However, as of today, the role that blood viscosity plays in this disease has not yet received its due attention. Viscosity is a fundamental property of any fluid. Its important role in both normal individuals and patients afflicted with cardiovascular disease has been underestimated. Past and current research has reported the benefits in addressing this important factor; however, mainstream medicine has not appreciated or fully accepted this important measurement. With continued research and published, peer-reviewed studies pertaining to the importance of blood viscosity in cardiovascular diseases, this relationship will be recognized, appreciated and will no doubt reveal the positive aspects of hemorheology, which will save lives.