488 Diabetes Care Volume 37, February 2014

Concetta Irace, Claudio Carallo, Blood Viscosity in Subjects With Faustina Scavelli, Maria Serena De Franceschi, Teresa Esposito, and Normoglycemia and Prediabetes Agostino Gnasso

OBJECTIVE Blood viscosity (BV) is higher in diabetic patients and might represent a risk factor for the development of insulin resistance and type 2 diabetes. However, data in subjects with normal glucose or prediabetes are missing. In the current study, we evaluated the relationship between BV and blood glucose in subjects with normal glucose or prediabetes.

RESEARCH DESIGN AND METHODS Enrolled subjects were divided into three groups according to blood glucose: group A (n = 74), blood glucose <90 mg/dL; group B (n = 96), blood glucose ranging from 90 to 99 mg/dL; and group C (n = 94), blood glucose ranging from 100 to 125 mg/dL. BV was measured at 378C with a cone-plate viscometer at shear rates ranging from 225 to 22.5 s21.

RESULTS Blood pressure, blood lipids, fibrinogen, and plasma viscosity were similar in the three groups. BMI and waist circumference were significantly increased in group C. Hematocrit (P < 0.05) and BV (P between 0.01 and 0.001) were significantly higher in groups B and C compared with group A. Blood glucose was significantly and

PATHOPHYSIOLOGY/COMPLICATIONS inversely correlated with HDL cholesterol and directly with BMI, waist, hematocrit (r = 0.134), and BV (from 225 s21 to 22.5 s21; r ranging from 0.162 to 0.131). BV at shear rate 225 s21 was independently associated with blood glucose.

CONCLUSIONS The current study shows a direct relationship between BV and blood glucose in nondiabetic subjects. It also suggests that, even within glucose values considered completely normal, individuals with higher blood glucose levels have increased BV Dipartimento di Medicina Sperimentale e Clinica “G. Salvatore,”“Magna Græcia” University, comparable with that observed in subjects with prediabetes. Catanzaro, Italy – Diabetes Care 2014;37:488 492 | DOI: 10.2337/dc13-1374 Corresponding author: Agostino Gnasso, [email protected]. Received 10 June 2013 and accepted 13 Blood viscosity (BV) is the force that counteracts the free sliding of the blood layers September 2013. within the circulation and depends on the internal cohesion between the molecules C.I. and C.C. contributed equally to this article. and the cells. Abnormally high BV can have several negative effects: the heart is © 2014 by the American Diabetes Association. overloaded to pump blood in the vascular bed, and the blood itself, more viscous, See http://creativecommons.org/licenses/by- can damage the vessel wall. Furthermore, according to Poiseuille’s law (1), BV is nc-nd/3.0/ for details. care.diabetesjournals.org Irace and Associates 489

inversely related to flow and might According to blood glucose levels, 22.5 s21. For plasma viscosity, the therefore reduce the delivery of insulin participants were divided into three average of measurements at shear rates and glucose to peripheral tissues, groups: group A, blood glucose ,90 mg/dL; of 225 and 90 s21 was calculated. The leading to insulin resistance or diabetes group B, blood glucose between coefficient of variation for blood and (2–5). 90 and 99 mg/dL; and group C, blood plasma viscosity was ,3%. It is generally accepted that BV is glucose between 100 and 125 mg/dL. Microhematocrit was measured increased in diabetic patients (6–8). The ethics committee of Azienda without correction for plasma trapping. fi Although the reasons for this alteration Policlinico Mater Domini approved the The coef cient of variation for ; are still under investigation, it is study. All recruited subjects gave microhematocrit was 1%. believed that the increase in osmolarity informed consent. Erythrocyte rigidity (Tk) was evaluated causes increased capillary permeability All subjects included in the study according to the Dintenfass formula: m 0.4 2 · m 0.4 · 21 and, consequently, increased hematocrit underwent a complete clinical Tk = ( r 1) ( r h) ,where m and viscosity (9). It has also been examination and blood withdrawal. h is hematocrit and r is the relative suggested that the osmotic diuresis, Standing height without shoes was BV (BV/plasma viscosity). consequence of hyperglycemia, could measured to the nearest 0.5 cm. Weight Statistical Analysis contribute to reduce plasma volume was measured to the nearest 0.1 kg in and increase hematocrit (10). All statistical analyses were performed ordinary street clothes. BMI was by SPSS 17.0 for Windows. All studied Cross-sectional studies have also computed as weight (in kg) divided by variables except triglycerides had 2 supported a link between BV, hematocrit, height (in m ). Waist circumference was normal distribution. Triglycerides were – and insulin resistance (11 17). Recently, measured midway between the lower log transformed before analyses. One- a large prospective study has rib margin and the iliac crest. Systolic way ANOVA and LSD post hoc test were demonstrated that BV and hematocrit and diastolic blood pressure was used to test the difference in studied are risk factors for type 2 diabetes. measured, on the right arm, after the variables between groups with different Subjects in the highest quartile of BV participant had been resting for at blood glucose concentrations. Pearson . were 60% more likely to develop least 5 min, with a standardized correlation coefficient was used to test diabetes than their counterparts in the sphygmomanometer. Cigarette the correlation between continuous fi fi lowest quartile (18). This nding con rms smoking and ongoing drug therapies variables. Multiple linear regression previous observations obtained in smaller were investigated by questionnaire. analysis was performed to evaluate the or selected populations, in which the Venous blood for routine and viscosity independent association of clinical and association between hemoglobin or analyses was collected in the morning biochemical variables with blood hematocrit and occurrence of type 2 before the breakfast after overnight glucose. diabetes was investigated (19–22). fasting in order to avoid the influence of These observations suggest that the postprandial lipid increase on RESULTS elevation in BV may be very early, well hemorheological parameters (25). The clinical, biochemical, and before the onset of diabetes, but hemorheological characteristics of the Attention was paid to avoid venous definite data in subjects with normal 264 participants are shown in Table 1. stasis, and the hemostatic loop, when glucose or prediabetes are missing. In Seventy-four subjects had blood glucose used, was immediately removed after the current study, we evaluated the ,90 mg/dL (group A), 96 subjects had cannulation of the vein. Blood glucose relationship between BV and blood blood glucose ranging from 90 to 99 and lipids were measured by routine glucose in subjects with normal glucose mg/dL (group B), and 94 had blood methods. Subjects with blood glucose or prediabetes in order to verify whether glucose ranging from 100 to 125 mg/dL ,100 mg/dL were classified as alterations in viscosity are appreciable (group C). Blood pressure, blood lipids, nondiabetic, those with values between in these subjects and at which blood fibrinogen, and plasma viscosity were 100 and 125 mg/dL were classified as glucose concentration they appear. similar in the three groups. BMI and prediabetic, and those with blood waist were significantly increased in . fi RESEARCH DESIGN AND METHODS glucose 125 mg/dL were classi ed as subjects with prediabetes (group C) Subjects were enrolled among free- diabetic and excluded from the present compared with the other two groups living participants in a cardiovascular analysis. (P , 0.05). Subjects with high normal disease prevention program between Blood and plasma viscosity were blood glucose were younger (P , 0.05) February 2011 and December 2012 measured within 2 h from blood compared with the other two groups. (23,24). In order to reduce the influence withdrawal; the blood specimen was Hematocrit (P , 0.05) and BV of confounding factors, cigarette added with heparin (35 IU/mL). (P between 0.01 and 0.001) were smoking, diabetes, plasma triglycerides Viscosity measurement was performed significantly higher in subjects with .400 mg/dL, female sex before at 378C with a cone-plate viscometer prediabetes and in those with blood menopause, and drug use (chronic (Wells-Brookfield DV-III; Wells- glucose ranging from 90 to 99 mg/dL treatment and any drug in the week Brookfield, Stoughton, MA) equipped compared with subjects with blood before blood withdrawal) were exclusion with a cp-40 spindle. BV was recorded glucose ,90 mg/dL. Tk was similar in criteria for the present analysis. at shear rates ranging from 225 to the three groups. 490 Blood Viscosity and Blood Glucose Diabetes Care Volume 37, February 2014

Table 1—Clinical, biochemical, and rheological characteristics of subjects [0.355], and Tk [0.377]). The results Group A Group B Group C P indicate that hematocrit, plasma viscosity, and erythrocyte deformability n 74 96 94 are the only variables independently 6 6 6 # Age (years) 56.2 12.4 52.8 10.8 57.2 10.0 0.02 associated with the viscosity in these Cholesterol (mg/dL) 209 6 41 220 6 40 213 6 40 #0.23 subjects. Triglycerides (mg/dL) 116 6 51 132 6 64 141 6 74 #0.09 HDL cholesterol (mg/dL) 53 6 17 51 6 15 49 6 12 #0.27 CONCLUSIONS Glucose (mg/dL) 85 6 5956 3 108 6 7 d BMI (kg/m2)28.36 4.1 28.0 6 4.5 29.9 6 4.9 #0.01 The current study shows, for the first Waist (cm) 91 6 10 91 6 13 96 6 10 #0.01 time, a direct relationship between BV SBP (mmHg) 139 6 22 136 6 19 139 6 24 #0.60 and blood glucose in nondiabetic DBP (mmHg) 86 6 10 85 6 10 87 6 10 #0.59 subjects. It also suggests that, even within glucose values considered Hematocrit (%) 42.6 6 3.9 44.0 6 3.5 43.8 6 3.4 #0.05 completely normal, individuals with BV (cP) 4.58 6 0.60 4.98 6 0.73 4.90 6 0.72 #0.01 225 higher blood glucose levels have BV (cP) 5.38 6 0.76 5.95 6 1.07 5.79 6 0.99 #0.01 90 increases in BV comparable with those BV (cP) 6.08 6 1.01 6.94 6 1.82 6.78 6 1.58 #0.01 45 observed in subjects with prediabetes. BV22.5 (cP) 7.40 6 1.35 8.61 6 2.87 8.41 6 2.51 #0.01 Plasma viscosity (cP) 1.48 6 0.15 1.54 6 0.16 1.55 6 0.16 #0.06 Many studies thus far have investigated Fibrinogen (mg/dL) 323 6 86 337 6 85 318 6 82 #0.33 BV in patients with diabetes (26). Tk 0.85 6 0.06 0.85 6 0.08 0.83 6 0.07 #0.18 Usually, however, these studies were performed in small groups of patients, , – Group A: blood glucose 90 mg/dL; group B: blood glucose 90 99 mg/dL; group C: blood often only at few shear rates, and glucose 100–125 mg/dL. DBP, diastolic blood pressure; SBP, systolic blood pressure. sometimes did not demonstrate any difference between diabetic and control fi In simple correlation analysis in the one at a time.) Only BV at shear rate subjects (27,28). Some ndings have whole population, blood glucose was 225 s21 resulted independently associated suggested a role for the increase in BV significantly and inversely correlated with blood glucose, while BV at lower as a pathogenetic factor for the development of microvascular with HDL cholesterol (r = 20.123) and shear rates was not (Table 2). complications (29,30). Important directly with BMI (r = 0.142), waist Two additional stepwise multiple increases in BV have been reported in (r = 0.180), hematocrit (r = 0.134), and 2 regression analyses were performed to diabetic retinopathy, and it has been BV (from shear rate 225 s 1 to shear 2 identify the variables independently hypothesized that these changes lead rate 22.5 s 1; r ranging from 0.162 to associated with the viscosity at a high to a prolonged reduction in the supply of 0.131). No correlation was found with 21 shear rate (225 s ) and low shear rate oxygen and nutrients to the capillaries, age, total cholesterol, triglycerides, 21 fi (22.5 s ). All variables signi cantly causing the development of angiopathy plasma viscosity, fibrinogen, or Tk. All correlated to these two viscosities were (31,32). Once retinopathy has the above variables significantly included as independent variables (for developed, its progression may be correlated with blood glucose, as well as 21 BV at shear rate 225 s [correlation favored by a reduction in BV and fi fi 2 age signi cantly different between coef cients], age [ 0.178], total hemoglobin (33). Overall, changes in groups at ANOVA, were tested for cholesterol [0.258], triglycerides viscosity in diabetic patients are – 2 independent association with blood [0 185], HDL cholesterol [ 0.202], blood accepted as common and as a result of glucose in stepwise multiple regression glucose [0.162], hematocrit [0.440], the disease. However, the relationship analyses. (Age, HDL cholesterol, BMI, plasma viscosity [0.438], and Tk [0.294]; between blood glucose, diabetes, and 21 waist, and hematocrit were constantly for BV at shear rate 22.5 s , age viscosity may be much more complex. 2 in the model, while the viscosities at [ 0.148], total cholesterol [0.219], First, it is useful to underline that the different shear rates were introduced hematocrit [0.207], plasma viscosity blood is a non-Newtonian fluid, in that its viscosity varies with the flow velocity. Specifically, when the flow velocity is Table 2—Stepwise multiple regression analysis high, the cellular component of blood b fi -Coef cient tP(red cells, white cells, and platelets) is 2 BV 225 s 1 2.421 2.179 0.031 concentrated at the center of the vessel, Variables not in the model while the plasma flows in the periphery. Age 0.144 1.870 0.063 In conditions of low flow velocities, the 2 2 HDL cholesterol 0.109 1.386 0.168 cellular component occupies the entire BMI 0.055 0.718 0.474 fl Waist 0.096 1.256 0.211 column of owing blood, causing an Hematocrit 0.061 0.711 0.478 increase in viscosity. BV at high shear rates ($90 s21, corresponding with in Dependent variable: blood glucose. vivo systolic flow condition) is strongly care.diabetesjournals.org Irace and Associates 491

influenced by erythrocyte The observation that the change in viscosity fibrinogen and other plasma proteins deformability, while at lower shear rates is already evident at blood glucose and, to a lesser extent, by obesity and (representing in vivo diastolic flow values .90 mg/dL seems very blood lipids (38,39). In the current condition) red cell aggregation plays the important. This allows understanding of study, subjects with high normal blood most important role (34). The why in some studies, no difference was glucose or prediabetes have higher deformability and the aggregation of observed between patients with values of BMI and slightly of blood erythrocytes are, in turn, driven by diabetes and control subjects (27,28). lipids, and this may explain why different characteristics of the Probably, the control subjects had high plasma viscosity tends to higher values, erythrocytes and the plasma. normal blood glucose levels or were thus contributing to the increase in Furthermore, factors such as cigarette prediabetic. In a previous study, whole BV. smoking, hyperlipidemia, and conducted in a small population of The current study has, in our opinion, hyperfibrinogenemia, which are very diabetic patients and a control group, we some limitations. Insulin resistance has frequent in patients with diabetes, may found no difference in BV measured at not been evaluated, since insulin was 21 alter BV. In light of this, it is evident that shear rate of 225 s (37). The control not measured. Similarly, data on further studies are needed to clarify the subjects, however, had a mean blood inflammation and plasma proteins are relationship between diabetes and BV. glucose of .95 mg/dL, and although lacking, although the careful selection of nondiabetic, they probably already had The importance of BV was recently the subjects, with the exclusion of those increased BV. Furthermore, the value of highlighted after the observation that it taking drugs, may have limited the 90 mg/dL would suggest a greater increases the risk of developing diabetes impact of these factors. The subjects attention in subjects who exceed this in normoglycemic subjects participating were classified on the basis of a single threshold in terms of correction of the in the Atherosclerosis Risk in fasting glucose, and HbA1c was not conditions that can cause increased Communities (ARIC) study (18). In that available in this population; therefore, it viscosity. study, the authors did not directly is possible that a few subjects were measure BV but used two validated In this regard, it is intriguing to try to classified inappropriately. We believe formulas: one for the shear rate 208 s21 identify the factors that influence the that this potential bias might have and the other for the shear rate 0.5 s21 viscosity. Changes in osmolality or affected only a few subjects and (35). The findings of our study are in line osmotic diuresis have been suggested as therefore does not invalidate the with the conclusions of the ARIC Study factors able to modify BV (9,10). While results. Finally, the erythrocyte and further demonstrate that the BV of these factors may play a role at very high deformability was calculated and not the subjects with prediabetes or high blood glucose concentrations, it seems directly measured. Therefore, the data normal blood glucose is significantly unlikely that they may affect BV within on erythrocyte deformability should be higher compared with that of subjects normal or slightly increased blood interpreted with caution. with low normal blood glucose at all glucose values. Among the variables In conclusion, the main finding of the considered shear rates. This supports considered in our study, hematocrit, study is that BV significantly increases the hypothesis that the alterations are plasma viscosity, and erythrocyte already at high-normal blood glucose very early and present in any flow deformability are the factors levels, independently of other common condition. Furthermore, these findings independently associated with high and determinants of hemorheology. arealsoinlinewiththeresultsofa low shear rates BV, confirming recently Intervention studies are needed to previous study, which showed that the published data (38). The gradual verify whether changes in BV can BV in patients with prediabetes was increase in plasma viscosity, hematocrit, influence the development of type 2 comparable with that of diabetic and Tk, even within levels considered diabetes. subjects with complications (36) and completely normal, leads to increased significantly higher than that of healthy BV. It is possible to hypothesize that control subjects. If increased viscosity gradually increasing concentrations of Duality of Interest. No potential conflicts of somehow alters the supply of oxygen blood glucose can directly affect the interest relevant to this article were reported. and nutrients to the tissues, then the erythrocytes, possibly resulting in Author Contributions. C.I., C.C., and A.G. observed changes may contribute to the impaired deformability and researched data and wrote the manuscript. F.S. fi and T.E. researched data. M.S.D.F. contributed development of peripheral insulin aggregation. If con rmed in larger and prospective studies, these results could to the discussion and reviewed the manuscript. resistance (by reduced glucose A.G. is the guarantor of this work and, as such, utilization in the muscle) and, in the long indicate the need for very early had full access to all the data in the study and term, diabetes. It cannot be excluded intervention on these factors in an takes responsibility for the integrity of the data that a slight increase in blood glucose, attempt to prevent the development of and the accuracy of the data analysis. secondary to other conditions of insulin diabetes. References resistance such as aging, inflammation, The plasma viscosity was slightly higher 1. Guyton A, Hall J. Textbook of Medical physical inactivity, etc., can cause in groups B and C compared with Physiology. 10th ed. Philadelphia, PA, WB increased BV, thereby creating a vicious group A but only close to statistical Saunders Company, 2000, p. 212–213 circle. Clearly, this requires further significance. The plasma viscosity is 2. Stuart J, Kenny MW. Blood rheology. J Clin study for a more exact definition. strongly influenced by the levels of Pathol 1980;33:417–429 492 Blood Viscosity and Blood Glucose Diabetes Care Volume 37, February 2014

3. Yang YJ, Hope ID, Ader M, Bergman RN. men and women. Metabolism 1998;47: insulin, and glucagon on blood viscosity in Insulin transport across capillaries is rate 831–835 vitro in healthy humans and patients with limiting for insulin action in dogs. J Clin 16. Capoglu˘ I, Unuvar¨ N, Bektas¸Y,YilmazO, diabetes mellitus. Clin Hemorheol – – Invest 1989;84:1620 1628 Kaya MD. The effects of high haematocrit Microcirc 2001;24:65 74 4. Schultz TA, Lewis SB, Westbie DK, Wallin JD, levels on glucose metabolism disorders. 28. Rimmer T, Fleming J, Kohner EM. Hypoxic Gerich JE. Glucose delivery: a modulator of J Int Med Res 2002;30:433–437 viscosity and diabetic retinopathy. Br J – glucose uptake in contracting skeletal 17. Perez-Martin´ A, Dumortier M, Pierrisnard E, Ophthalmol 1990;74:400 404 – muscle. Am J Physiol 1977;233:E514 E518 Raynaud E, Mercier J, Brun JF. Multivariate 29. Merimee TJ. Diabetic retinopathy. A 5. Baron AD, Clark MG. Role of blood flow in analysis of relationships between insulin synthesis of perspectives. N Engl J Med the regulation of muscle glucose uptake. sensitivity and blood rheology: is plasma 1990;322:978–983 – Annu Rev Nutr 1997;17:487 499 viscosity a marker of insulin resistance? Clin 30. Williamson TH, Rumley A, Lowe GD. Blood – 6. Skovborg F, Nielsen AV, Schlichtkrull J, Hemorheol Microcirc 2001;25:91 103 viscosity, coagulation, and activated Ditzel J. Blood-viscosity in diabetic patients. 18. Tamariz LJ, Young JH, Pankow JS, et al. protein C resistance in central retinal vein Lancet 1966;1:129–131 Blood viscosity and hematocrit as risk occlusion: a population controlled study. – 7. Dintenfass L. Blood viscosity factors in factors for type 2 diabetes mellitus: the Br J Ophthalmol 1996;80:203 208 severe nondiabetic and diabetic atherosclerosis risk in communities (ARIC) 31. Chmiel B, Cierpka L. Organ preservation – retinopathy. Biorheology 1977;14:151–157 study. Am J Epidemiol 2008;168:1153 1160 solutions impair deformability of 8. Schmid-Schonbein¨ H, Volger E. Red-cell 19. Wannamethee SG, Perry IJ, Shaper AG. erythrocytes in vitro. Transplant Proc 2003; – aggregation and red-cell deformability Hematocrit and risk of NIDDM. Diabetes 35:2163 2164 – in diabetes. Diabetes 1976;25(Suppl. 2): 1996;45:576 579 32. Linderkamp O, Poschl¨ J, Ruef P. Blood cell 897–902 20. Tulloch-Reid MK, Hanson RL, Saremi A, deformation in neonates who have sepsis. – 9. Meiselman HJ, Merrill EW, Gilliland ER, et al. Hematocrit and the incidence of type NeoReviews 2006;7:e517 e523 Pelletier GA, Salzman EW. Influence of 2 diabetes in the pima indians. Diabetes 33. Irace C, Scarinci F, Scorcia V, et al. – plasma osmolarity on the rheology of Care 2004;27:2245 2246 Association among low whole blood human blood. J Appl Physiol 1967;22: 21. Medalie JH, Papier CM, Goldbourt U, viscosity, haematocrit, haemoglobin and 772–781 Herman JB. Major factors in the diabetic retinopathy in subjects with type 2 – 10. Lowe GD, Lowe JM, Drummond MM, et al. development of diabetes mellitus in 10,000 diabetes. Br J Ophthalmol 2011;95:94 98 – Blood viscosity in young male diabetics with men. Arch Intern Med 1975;135:811 817 34. Kim S, Popel AS, Intaglietta M, Johnson PC. and without retinopathy. Diabetologia 22. Wilson PW, McGee DL, Kannel WB. Obesity, Effect of erythrocyte aggregation at normal 1980;18:359–363 very low density lipoproteins, and glucose human levels on functional capillary density 11. Høieggen A, Fossum E, Moan A, Enger E, intolerance over fourteen years: The in rat spinotrapezius muscle. Am J Physiol – Kjeldsen SE. Whole-blood viscosity and the Framingham Study. Am J Epidemiol 1981; Heart Circ Physiol 2006;290:H941 H947 – insulin-resistance syndrome. J Hypertens 114:697 704 35. de Simone G, Devereux RB, Chien S, 1998;16:203–210 23. Gnasso A, Calindro MC, Carallo C, et al. Alderman MH, Atlas SA, Laragh JH. Relation 12. Nordby G, Moan A, Kjeldsen SE, Os I. Awareness, treatment and control of of blood viscosity to demographic and Relationship between hemorheological hyperlipidaemia, hypertension and physiologic variables and to cardiovascular factors and insulin sensitivity in diabetes mellitus in a selected population risk factors in apparently normal adults. – normotensive and hypertensive of southern Italy. Eur J Epidemiol 1997;13: Circulation 1990;81:107 117 – premenopausal women. Am J Hypertens 421 428 36. Richards RS, Nwose EU. Blood viscosity at 1995;8:439–444 24. Irace C, Cortese C, Fiaschi E, et al. different stages of diabetes pathogenesis. – 13. Catalano C, Muscelli E, Natali A, et al. Components of the metabolic syndrome Br J Biomed Sci 2010;67:67 70 Reciprocal association between insulin and carotid atherosclerosis: role of 37. Irace C, Carallo C, Crescenzo A, et al. NIDDM sensitivity and the haematocrit in man. Eur elevated blood pressure. Hypertension is associated with lower wall shear stress of – J Clin Invest 1997;27:634–637 2005;45:597 601 the common carotid artery. Diabetes 1999; – 14. Smith S, Julius S, Jamerson K, Amerena J, 25. Stamos TD, Rosenson RS. Low high density 48:193 197 Schork N. Hematocrit levels and physiologic lipoprotein levels are associated with an 38. Irace C, Carallo C, Scavelli F, et al. Influence factors in relationship to cardiovascular risk elevated blood viscosity. Atherosclerosis of blood lipids on plasma and blood – in Tecumseh, Michigan. J Hypertens 1994; 1999;146:161 165 viscosity. Clin Hemorheol Microcirc. 12:455–462 26. Cho YI, Mooney MP, Cho DJ. 7 February 2013 [Epub ahead of print] Hemorheological disorders in diabetes 15. Facchini FS, Carantoni M, Jeppesen J, 39. VayaA,Sim´ o´ M, Santaolaria M, Carrasco P, Reaven GM. Hematocrit and hemoglobin mellitus. J Diabetes Sci Tech 2008;2: Corella D. Plasma viscosity and related – are independently related to insulin 1130 1138 cardiovascular risk factors in a Spanish resistance and compensatory 27. Schnyder L, Walter R, Rohrer A, Contesse J, Mediterranean population. Thromb Res hyperinsulinemia in healthy, non-obese Reinhart WH. No influence of C-peptide, 2007;120:489–495