LAUSR

BACKGROUND A reason for concentric left ventricular (LV) remodelling predicting cardiovascular outcomes independent of conventional risk factors and LV mass (LVM) has not been provided. We hypothesized that independent of LVM, concentric LV remodelling is associated with inflammatory changes rather than a pressure load on the LV. METHODS In 764 randomly selected community participants, we assessed relations between several inflammatory markers (ELISA) and LV relative wall thickness (RWT) (echocardiography), LV mass index (LVMI), and indexes of diastolic function. RESULTS No independent relations were noted between circulating concentrations of inflammatory markers and LVM index (LVMI) (p > 0.13 for all). However, independent of confounders including LVMI and blood pressure (BP), circulating tumour necrosis factor-α (TNF-α) (partial r = 0.14, p < 0.0005) and to a lesser degree interleukin-6 (partial r = -0.09, p < 0.02) were associated with RWT. The impact (standardized β-coefficient) of TNF-α on RWT (0.12 ± 0.03, p < 0.0005) was at least as strong as age (0.13 ± 0.05, p < 0.005), and second only to LVMI (0.27 ± 0.04, p < 0.0001), whilst neither office, 24-hour, central aortic BP, nor aortic stiffness were associated with RWT independent of LVMI. With adjustments, as compared to participants with a normal LVMI and geometry (12.7 ± 0.8), circulating TNF-α concentrations (pg/ml) were increased as much in participants with concentric LV remodelling (16.8 ± 1.5, p < 0.05) as in those with concentric LV hypertrophy (LVH) (17.0 ± 1.3, p < 0.005), whilst eccentric LVH (13.7 ± 0.9) was not. No independent relations between inflammatory markers and LV diastolic function (trans-mitral and tissue Doppler) were noted. CONCLUSIONS Independent of LVMI, a pro-inflammatory state rather than BP load is strongly associated with LV concentric remodelling.