Whole blood viscosity is associated with reduced myocardial mechano‐energetic efficiency in nondiabetic individuals

This cross‐sectional study aimed to investigate the association between myocardial mechano‐energetic efficiency (MEE) and whole blood viscosity (WBV) in nondiabetic adults participating in the CATAnzaro MEtabolic RIsk factors (CATAMERI) study.


| INTRODUCTION
[3][4][5][6][7][8][9][10] Cardiac work relies on energy derived almost entirely from aerobic oxidation, and requires the coupling between myocardial oxygen consumption (MVO 2 ) and LV function. 8[3]5 Specific assessment of MVO 2 requires invasive techniques such as coronary sinus catheterization 11 or expensive and time-consuming techniques such as noninvasive positron emission tomography, 12 two methods unsuitable in large epidemiological studies based on routine examinations. 8To overcome this problem, a simple, noninvasive, ultrasound-based method to measure myocardial MEE has been validated.This method is based on the calculation of the ratio between LV stroke work, assessed as stroke volume (SV) × systolic blood pressure (SBP), and MVO 2 estimated by the double product, that is, SBP multiplied by heart rate (HR). 2,3athophysiological mechanisms causing impaired myocardial MEE have not been elucidated so far.Amongst others, whole blood viscosity may be a plausible candidate.Whole blood viscosity is a measure of the intrinsic resistance of blood to flow and is generated by the frictional interactions between the main blood components, such as plasma, plasma proteins and red blood cells. 135][26][27][28][29][30][31][32][33][34][35] However, the impact of increased whole blood viscosity on myocardial MEE has not been explored yet.To this end, we examined the association between myocardial MEE per gram of left ventricular mass (MEEi) and whole blood viscosity in a cohort of nondiabetic adults participating in the CATAnzaro MEtabolic RIsk factors (CATAMERI) study.

| Study population
The CATAMERI study is an ongoing observational study consecutively recruiting Caucasian adults who are screened for one or more cardio-metabolic risk factors including dysglycaemia, overweight/obesity, dyslipidaemia and hypertension at the 'Magna Graecia' of Catanzaro University hospital as previously described in details. 36,37 In the present study, 1143 participants were analysed after the exclusion of subjects with type 1 and type 2 diabetes, a history of malignant or autoimmune diseases, end-stage renal disease, hepatic failure, anaemia, hemoglobinopathies, haemolytic disease, gastrointestinal affections associated with bleeding or malabsorption, eating disorders or any other cause of malnutrition, acute or chronic infectious diseases, chronic or acute pancreatitis, accumulation diseases such as amyloidosis and hemochromatosis, self-report drug abuse or alcohol consumption of >20 g/ day, positivity for antibodies to hepatitis C virus (HCV) or hepatitis B surface antigen (HBsAg), and treatment with corticosteroids, antiplatelet and anticoagulant drugs.A study flowchart describing enrolment is reported in the (Figure S1).All participants were visited in the morning, after an overnight fast.During the first visit, information regarding medical history, smoking status, drug use and alcohol consumption was gathered, anthropometrical parameters including body mass index (BMI), waist circumference and clinic blood pressure were recorded, and a sample of venous blood was drawn for laboratory measurements.A 75 g oral glucose tolerance test (OGTT) was performed for glucose tolerance assessment.
The study was approved by the local ethics committee (Comitato Etico Azienda Ospedaliera 'Mater Domini'), and every participant provided written informed consent in accordance with the principles of the Helsinki Declaration.

| Echocardiographic measurements
All participants underwent echocardiographic investigation performed by a single skilled examiner using a VIVID-7 Pro ultrasound machine (GE Technologies, Milwaukee, WI) with an annular phased array 2.5-MHz transducer.Measurements of interventricular septal (IVS) thickness and LV internal diameter were done at end diastole.LV end-diastolic (LVEDV) and end-systolic volume (LVESV) were assessed according to the Simpson method and indexed for body surface area (BSA). 38Left ventricular mass (LVM) was calculated using the Devereux formula 39 and normalized by BSA [LVMI]). 40SV was calculated as the difference between LVEDV and LVESV.External myocardial work was evaluated as stroke work calculated as SBP × echocardiographic SV.MVO2 was measured by the 'double product' of HR × SBP.Therefore, MEE was calculated as SBP × SV/SBP × HR = SV/HR where HR was expressed in seconds (HR/60).Because MEE is highly related to LVM, MEE was normalized to LVM to attain a measure of the amount of blood ejected in 1 second by each gram of LVM (i.e., indexed MEE, MEEi, ml/s per g). 2,3,41

| Calculations
The homeostasis model assessment index of insulin resistance (HOMA-IR) was defined as fasting insulin × fasting glucose/22.5. 42hole blood viscosity at 208 s −1 of shear rate was calculated by a previously validated equation that takes into account haematocrit and plasma proteins 13 : whole blood viscosity = [0.12× h] + [0.17 × (p-2.07)],where h is haematocrit (%) and p is plasma protein levels (g/dl).
Estimated glomerular filtration rate (eGFR) was calculated by using the CKD-EPI equation 43 where Scr is serum creatinine, k is 0.7 for females and 0.9 for males, α is −0.329 for females and −0.411 for males, min indicates the minimum of Scr/k or 1, and max indicates the maximum of Scr/k or 1).

| Statistical analysis
Normality of continuous variables was assessed by kurtosis and skewness measures and normal probability plots.Non-normally distributed variables including triglycerides, HOMA-IR and hsCRP were log transformed to achieve normality before statistical analyses.Continuous data are expressed as means ± SD.As for study design, only subjects with all data available were included in the analysis.Correlations between normally distributed variables were determined by Pearson's correlation coefficient.Correlations between categorical variables were determined b Spearman's correlation coefficient.The independent association between myocardial MEEi and whole blood viscosity was tested by multiple regression analysis, including the variables significantly associated with myocardial MEEi in a univariate analysis.Variables that are incorporated into the formulas of the whole blood viscosity such as haematocrit and plasma protein levels and variables used to calculate myocardial MEEi were not computed in the regression model to avoid potential collinearity.Variation inflation factor (VIF <2) and condition index and (<15) were applied to demonstrate the absence of multicollinearity between variables comprised in the multiple linear regression model. 45A p-value ≤0.05 was considered statistically significant.All statistical analyses were performed by SPSS software program version 27 for Windows (IBM CORP).
The main anthropometric and metabolic characteristics of the study population are shown in Table 1.Six-hundred and thirty (55.1%) of the 1143 participants, are categorized as never smokers, 264 (23.1%) as current smokers and 249 as ex-smokers.Individuals with NGT were 63% of the whole population, those with IFG were 14.3%, while those with IGT or IFG/IGT combined were 13% and 9.7%, respectively (Table 1).In the whole study group, mean haematocrit was 42 ± 4%, mean total plasma proteins concentration was 7.26 ± 0.44 g/dL, and mean whole blood viscosity was 5.96 ± 0.44 cP.Echocardiographic parameters are shown in Table 2.
Next, we sought to examine the association between myocardial MEEi and whole blood viscosity.As shown in Table 3, in a univariate analysis, myocardial MEEi was significantly associated with sex, age, smoking status, BMI, waist circumference, total cholesterol, HDL cholesterol, triglycerides, fasting glucose, fasting insulin, HOMA-IR, eGFR, glucose tolerance status, hsCRP, haematocrit, haemoglobin, total plasma protein levels and whole blood viscosity.
In addition, whole blood viscosity was significantly associated with sex, age, smoking status, waist circumference, total cholesterol, HDL cholesterol, triglycerides, fasting glucose, fasting insulin, HOMA-IR, eGFR, glucose tolerance status, haemoglobin, systolic and diastolic blood pressure, heart rate, stroke work and myocardial oxygen consumption.
To estimate the independent contribution of whole blood viscosity to myocardial MEEi, we built a multivariable regression model including variables that were significantly associated with MEEi in a univariate analysis, that is, sex, age, smoking status, BMI, total and HDL cholesterol, triglycerides, hsCRP, eGFR, glucose tolerance status and HOMA-IR.Comparison of standardized coefficients allowed the determination of the relative strength of each trait association with MEEi (listed from strongest to weakest): age (β = −0.146,p < .001),HOMA-IR (β = −0.145,p < .001),whole blood viscosity (β = −0.130,p < .001)and glucose tolerance status (β = −0.064,p = .04)(Table 4).The independent association between whole blood viscosity and MEEi remained statistically significant (β = −0.125,p < .001)when HOMA-IR was replaced by fasting plasma glucose and insulin levels in the regression model (Model 2; Table 4).Moreover, the independent association between whole blood viscosity and MEEi remained statistically significant (β = −0.122,p < .001)when antihypertensive therapy and lipid-lowering therapy were included in the regression model (Model 3; Table 4).Finally, the independent association between whole blood viscosity and MEEi remained statistically significant (β = −0.133,p = .006)when the regression analysis was restricted to the individuals having a value of HOMA-IR ≥2.5 (n = 616).

| DISCUSSION
The main result of the present study is that myocardial MEEi, as estimated by an indirect validated method, 2 is significantly associated with whole blood viscosity in a large cohort of nondiabetic individuals.This finding was strengthened by results of a multivariate linear regression that investigated whether whole blood viscosity was associated with reduced myocardial MEEi independently of well-established cardio-metabolic risk factors including age, sex, smoking status, BMI, total cholesterol, HDL, triglycerides, glucose tolerance status, hsCRP and HOMA-IR index of insulin resistance.1][52][53] We found that the association between whole blood viscosity and myocardial MEEi remained statistically significant even after adjustment for antihypertensive therapy.This makes unlikely that antihypertensive therapy contributed to the observed association between whole blood viscosity and MEEi.There are several potential pathophysiological mechanisms underpinning the relationship between elevated whole blood viscosity and impaired capability of the left ventricle to convert chemical energy into stroke work.Stroke volume is determined by the interaction between the preload, the contractile state of the heart and the postload forces.According to Poiseuille's law, a rise in whole blood viscosity, causing a decrease in blood flow rate results in a physiologic compensatory increase in blood pressure or vasodilation. 54Thus, we found that whole blood viscosity was positively associated with systolic and diastolic blood pressure suggesting a compensatory blood pressure increase in response to raised blood viscosity.This compensatory elevation in blood pressure may constitute an extra circulatory workload for the heart.A cartoon displaying the assumed pathophysiological mechanisms of the association between MEEi and whole blood viscosity is reported in Figure 1.
Impaired insulin sensitivity and compensatory hyperinsulinemia may represent additional pathophysiological mechanisms linking increased whole blood viscosity and depressed MEEi.It has been reported that a higher degree of insulin resistance, estimated by the HOMA-IR index, is associated with a reduction in myocardial MEEi, 41 and that an impairment in insulin-stimulated myocardial glucose metabolism, measured by dynamic positron emission tomography with 18F-fluorodeoxyglucose combined with euglycemic-hyperinsulinemic clamp, is associated with reduced myocardial MEEi in individuals with different degrees of glucose tolerance. 55There is also substantial T A B L E 3 Correlations between myocardial MEEi, whole blood viscosity and anthropometric and metabolic variables.7][58][59] An increase in blood viscosity is associated with decreased flow, which, in turn, may lead to a decreased delivery of glucose to the skeletal muscle. 60These changes lead to an increase in blood glucose levels which promotes insulin secretion and compensatory hyperinsulinemia.Moreover, hyperinsulinemia may induce vasoconstriction via sympathetic neural activation, which, in turn, would lead to increased blood pressure and cardiac workload. 61Taken together, these data support the idea that insulin resistance and compensatory hyperinsulinemia may represent a pathophysiological mechanism linking whole blood viscosity to myocardial MEEi.Although we observed that whole blood viscosity was positively correlated with both HOMA-IR index, and fasting plasma glucose and insulin levels, the findings that whole blood viscosity remained associated with myocardial MEEi even after adjustment for fasting plasma glucose and insulin levels or HOMA-IR index argue against the possibility that these factors might have contributed to the association of whole blood viscosity with myocardial MEEi.

Variables
5][26][27][28][29][30][31][32][33][34][35] The present finding of an independent association between whole blood viscosity and decreased myocardial  coupled with results of previous studies showing the role of impaired myocardial energetics in the development of CV disease 1-3,10 may be viewed as evidence that reduced myocardial MEEi may represent one of the physio-pathologic mechanisms contributing to the increased risk of CV disease observed in individuals with elevated blood viscosity. 28,33trengths of the current study include the large sample size equally comprising men and women, a careful clinical characterization with anthropometric and metabolic features gathered according to a standardized protocol that allowed us to adjust for multiple confounders, the measurement of biochemical and haematological parameters in fresh, rather than stored, blood samples thus preventing possible problem of degradation and cell lysis, the exclusion of individuals with disorders potentially affecting haemoglobin concentration or treated with corticosteroids, antiplatelet and anticoagulant drugs.
Nevertheless, some limitations should be considered.Firstly, MEEi was evaluated by indirect measures rather than by coronary sinus catheterization 11 or noninvasive positron emission tomography, 12 two techniques unsuitable in large epidemiological studies based on routine examinations.Secondly, we did not perform direct assay of blood viscosity by capillary viscometry.Nonetheless, assessments of whole blood viscosity were based on a prediction equation that has been validated in previous studies. 13,62,63urthermore, our large sample size counterbalances the lower accuracy of blood viscosity measures, and the consistency of the observed associations using routine haematological parameters may have useful implications for the clinical practice.Moreover, the cross-sectional nature of the present study precludes us from drawing any firm conclusion on the role of increased whole blood viscosity in inducing depressed MEEi, and therefore, it is not possible to discern cause-effect relationship using this design.Additionally, this study is not a randomized controlled trial, but rather an observational investigation, and thus, the results may be subject to residual unknown confounding factors.Finally, the results are based only on Caucasian outpatients with one or more risks for cardio-metabolic disease attending a referral university hospital, and thus, it is not possible to generalize the findings to the general population and other ethnic groups.

| CONCLUSION
The current study suggests that whole blood viscosity is associated with reduced myocardial MEEi in a large cohort of nondiabetic Caucasian individuals independently of other cardio-metabolic risk factors, including hypertension and insulin resistance.This finding supports the idea that a decrease in cardiac energetics, measured by myocardial MEEi, may be one of the key mechanisms leading to increased CV risk in individuals with elevated blood viscosity.However, other studies are required to confirm the presence of a cause-effect relationship of these results in the general population.

F I G U R E . 1
Pathophysiological mechanism of the association between MEEi and whole blood viscosity.HR, heart rate; SBP, systolic blood pressure.
Anthropometric and metabolic characteristics of the study population (N = 1143).Data are means ± SD.Triglycerides, HOMA-IR and hsCRP were natural log transformed for statistical analyses, but values in the table represent back transformation to the original scale.Left ventricular geometry, systolic and diastolic function and mechano-energetic performance in the study subjects (N = 1143).
T A B L E 1Note:T A B L E 2 Multiple regression analysis evaluating the association between whole blood viscosity, anthropometric and metabolic variables and myocardial MEEi as dependent variable.