Although it is recognized that both hypertension and obesity are associated with increased left ventricular mass, the relative impacts of obesity, arterial hypertension, and gender on the prevalence of ventricular hypertrophy remain uncertain. Accordingly, echocardiographic left ventricular mass normalized for height to the power of the allometric or growth relation between ventricular mass and height was compared in 164 normotensive subjects (85 Gender-specific upper normal limits were used to identify ventricular hypertrophy. Left ventricular mass/height 17 was higher in obese than normal-weight normotensive subjects (P<.004) independently of the level of blood pressure and identified a higher prevalence of hypertrophy (mainly eccentric) in obese than in normal-weight normotensive subjects (14% versus 5%, P<.04), a difference that was not detected by left ventricular mass/body surface area. Left ventricular mass/ height 2 -7 identified hypertrophy in 52% of obese and 30% of normal-weight hypertensive patients (/V.0001) because of T he association between obesity and hypertension and its effect on cardiac structure and function have been extensively studied.1 " 5 The obesityhypertension association might have a multiplicative effect on the risk of cardiac morbidity, 6 although it does not explain completely the independent adverse effect of obesity that is detectable in some but not all epidemiologic studies. 67The combination of obesity and arterial hypertension is more consistently associated with left ventricular (LV) hypertrophy than either stimulus alone. 5 In view of evidence that LV hypertrophy is an extremely strong risk factor for cardiovascular morbidity and mortality, 812 it can be speculated that LV hypertrophy stimulated by the combination of obesity and hypertension might predict particularly high risk. However, available evidence that obesity stimulates LV hypertrophy independently of obesity-associated changes in arterial blood pressure may have been partially confounded by the difficulty of correctly accounting for the effect of body size on LV mass in comparing overweight with normal-weight persons.Received September 30, 1993; accepted in revised form January 10, 1994.From the Division of Cardiology, The New York HospitalCornell Medical Center, New York, NY.Correspondence to Dr Richard B. Devereux, Division of Cardiology, Box 222, The New York Hospital-Cornell Medical Center, 525 E 68th St, New York, NY 10021.higher prevalences in obese than normal-weight patients of both eccentric (34% versus 20%) and concentric ventricular hypertrophy (18% versus 10%). The increase in left ventricular mass was independent of blood pressure values in obese normotensive women (but not men), and the prevalence of supranormal left ventricular mass/height 17 was even higher in hypertensive obese women (58%) than men (49%). Logistic analysis showed that body mass index was the main predictor of left ventricular hypertrophy in the entire population or hypertensive patients only (P<.0001 in both genders) ...

Although increased blood viscosity occurs in several cardiovascular diseases, little is known of factors influencing blood rheology in normal adults. Accordingly, we examined the relations of whole blood viscosity (WBV) to its rheologic determinants (hematocrit level, plasma viscosity, protein concentration, and red cell aggregability and rigidity), to demographic and laboratory variables, and to cardiovascular risk factors in 128 normotensive employed adults. Hematocrit levels accounted for 67-84% of variability of WBV at shear rates from 208 to 0.1 sec`with lesser contributions from plasma viscosity, red cell aggregability, and rigidity (multiple r=0.95-0.97); WBV was predicted accurately from standard measurements of hematocrit and total plasma protein levels (multiple r=0.78-0.92 in "learning" and "test" analysis). Male sex, obesity, dietary Na+ intake, and increasing age had additive effects on WBV (multiple r.0.59, p<0.00001); the last three of these factors and black race independently predicted plasma viscosity (multiple r=0.36, p<0.001). Among regulators of plasma volume, plasma renin activity and urinary Na+ excretion bore independent positive relations to WBV. Diastolic and mean blood pressures were independent predictors of WBV and hematocrit levels (all p <0.05). Conventional risk factors (e.g., triglycerides, obesity, and cholesterol levels) were positively related to WBV or plasma viscosity. Thus, in apparently normal adults, 1) WBV or plasma viscosity are increased by male sex, obesity, high sodium intake, aging, and black race, 2) WBV is positively related to plasma renin activity, 3) WBV or plasma viscosity are related to diastolic and mean blood pressures, triglycerides and cholesterol concentrations, and 4) WBV can be predicted from simple measurements of hematocrit and total plasma protein levels. (Circulation 1990;81:107-117) Changes in whole blood viscosity (WBV) have been reported in several human cardiovascular diseases,'-22 indicating that blood viscosity may be a major cardiovascular risk factor. A positive relation between blood pressure and blood viscosity or All editorial decisions for this article, including selection of reviewers and the final decision, were made by a guest editor. This procedure applies to all manuscripts with authors from the

To investigate alterations of magnesium metabolism in Type 2 (non-insulin-dependent) diabetes mellitus, we utilized a new magnesium-specific selective ion electrode apparatus to measure serum ionized magnesium (Mg-io) in fasting subjects with and without Type 2 diabetes, and compared these values to levels of serum total magnesium, and of intracellular free magnesium (Mgi) analysed by 31P-NMR spectroscopy. Both Mg-io (0.630 +/- 0.008 vs 0.552 +/- 0.008 mmol/l, p < 0.001) and Mgi (223.3 +/- 8.3 vs 184 +/- 13.7 mmol/l, p < 0.001), but not serum total magnesium, were significantly reduced in Type 2 diabetes compared with non-diabetic control subjects. Furthermore, a close relationship was observed between serum Mg-io and Mgi (r = 0.728, p < 0.001). We suggest that magnesium deficiency, both extracellular and intracellular, is a characteristic of chronic stable mild Type 2 diabetes, and as such, may predispose to the excess cardiovascular morbidity of the diabetic state. Furthermore, by more adequately reflecting cellular magnesium metabolism than total serum magnesium levels, Mg-io measurements may provide a more readily available tool than has heretofore been available to analyse magnesium metabolism in a variety of diseases.