The relationship between blood pressure and left ventricular mass index depends on an excess adiposity

Abstract

To determine whether blood pressure (BP)-LVM relationships depend in-part on the influence of an excess adiposity and whether this translates into a greater effect of hypertension on LVM in obese as compared with lean people. In 399 randomly recruited participants from a general population with a high prevalence of excess adiposity ( approximately 68%), we assessed whether the relationships between conventional blood pressure (BP) and LVM indexed for height (LVMI) (determined from echocardiography) are influenced by adiposity. We confirmed these outcomes using 24-h ambulatory measurements in 297 participants; and carotid-femoral pulse wave velocity (PWV) (applanation tonometry) in 328 participants and from plasma leptin concentrations, we assessed whether leptin could mediate this effect. After adjustments for appropriate confounders, including the individual terms for adiposity and BP, interactions between adiposity indices (either waist circumference or the mean of subscapular and triceps skin-fold thickness) and either conventional systolic BP (SBP), 24-h SBP, PWV, conventional pulse pressure (PP), or 24-h PP were independently associated with LVMI (P < 0.001 for interactions). The adiposity index-haemodynamic interaction translated into a steeper slope of the BP-LVMI and PWV-LVMI relations in obese as compared with lean participants. Every one SD increase in conventional SBP ( approximately 22 mmHg) was associated with a 1.61 g/m increase in LVMI in participants with a normal waist circumference, in comparison to a 5.24 g/m increase in LVMI in those with an increased waist circumference (P < 0.0001). Furthermore, the adiposity index-haemodynamic interaction resulted in an increased LVMI in never-treated hypertensives with central obesity (LVMI, normotensives = 45.6 g/m, hypertensives = 51.0 g/m, P < 0.02), but not in participants with a normal waist circumference (LVMI, normotensives = 43.4 g/m, hypertensives = 45.0 g/m). Significant plasma leptin concentration-haemodynamic interactions were also associated with LVMI independent of confounders and adiposity indices, an effect that translated into a steeper slope of the haemodynamic factor-LVMI relations in participants with a plasma leptin concentration above as compared with those below the median for the group. Adiposity-induced increases in LVM reflect an enhanced effect of BP on LV growth, an effect that may be mediated by leptin. This translates into an impact of never-treated hypertension on LVMI in centrally obese, but not in lean people in groups of African descent in South Africa.