Cardiac remodelling detected by imaging and circulating biomarkers in the general population associated with systolic and diastolic blood pressure

Background/Introduction Hypertension can cause end-organ damage. Limited large-scale evidence is available on subclinical cardiac alterations in the community. Purpose We aimed to understand early cardiac changes associated with systolic blood pressure (SBP) and diastolic blood pressure (DBP) in the general population, detected by dedicated cardiac imaging and circulating biomarkers. Methods Participants from a population-based cohort study in Germany underwent standardized blood pressure measurements, in-depth transthoracic echocardiography and blood sampling. Linear regression models were performed, with standardized continuous dependent variables and multivariable adjustment for age, sex, smoking status, educational level, body mass index, logarithmic (log) alcohol intake, heart rate, diabetes, total cholesterol, kidney disease, and a composite self-reported health status score derived from five domains. Results In n=9,589 participants (median age 63 years, 51% women) median SBP and DBP were 137 and 82 mmHg, respectively (Figure 1). Age had different non-linear effects on SBP and DBP (Figure 2). In multivariable-adjusted linear regression analysis significant associations between SBP and 23 of 28 tested variables could be observed (Figure 2), including left ventricular (LV) mass (Standard coefficient [β] 0.0051, p<0.001), aortic root diameter (β -0.0014, p=0.022), interventricular septum diameter end diastolic (IVSd) (β 0.0063, p<0.001), LV cardiac index (β 0.0053, p<0.001), tricuspid annular peak systolic excursion (TAPSE) (β 0.0030, p<0.001), left atrial (LA) systolic volume (β 0.0033, p<0.001), right atrial reservoir strain (β -0.0021, p=0.019), E/E’ (mean of lateral and medial) (β 0.0072, p<0.001), mitral valve E/A (β -0.0045, p<0.001), log NTproBNP (β 0.0049, p<0.001) and log high-sensitive Troponin I (hsTnI). Numerically, there were slightly less significant linear associations with DBP in 20 of 28 variables, including IVSd (β 0.0125, p<0.001), LV mass (β 0.0070, p<0.001), aortic root diameter (β 0.0056, p<0.001), E/E’ (mean of lateral and medial) (β 0.0028, p=0.018), TAPSE (β -0.0038, p=0.006), mitral valve E/A (β -0.0138, p<0.001), LA global peak strain (β -0.0033, p=0.024), log NTproBNP (β 0.0066, p<0.001) and log hsTnI (β 0.0082, p<0.001). For both medial and lateral E-wave velocity and aortic root diameter, the effect of SBP and DBP was statistically significant diverging in direction (Figure 2). Spline analyses revealed non-linear associations for both SBP and DBP and aortic root diameter, A-wave velocity, mitral valve E/A, log NTproBNP and log hsTnI, and additionally between DBP and LV mass. Conclusions This is one of the largest population-based cohorts to date to demonstrate the association of SBP and DBP with a wide range of distinct subclinical cardiac remodelling including elevated biomarkers as well as cardiac dimensions and function.