Cardiac troponin T predicts future lower left ventricular systolic function and synchrony assessed by cardiac magnetic resonance

Abstract Funding Acknowledgements Type of funding sources: Other. Main funding source(s): Akershus University Hospital Trust and Vestre Viken Hospital Trust. Roche Diagnostics provided reagents for high sensitivity cardiac troponin T analyses free of charge. Background In the general population, higher concentrations of cardiac troponin T (cTnT) are associated with risk of cardiovascular (CV) disease even within the normal range. Cardiac magnetic resonance (CMR) offers state-of-the-art assessment of the left ventricle with recently developed techniques allowing assessment of myocardial strain and synchrony. There is little prospective observational data describing the association between cTnT and future left ventricular (LV) function assessed by CMR. Purpose To investigate whether cTnT is predictive of future lower LV systolic function and synchrony in the general population. Methods The cohort included 186 participants from the general population who underwent cTnT measurements at baseline followed by CMR feature-tracking 4–7 years later. All participants were born in year 1950, had no overt coronary artery disease and normal kidney function at inclusion. Outcomes were LV global longitudinal, circumferential and radial strain (GLS, GCS and GRS), as well as mechanical dispersion (MD). MD was calculated as the standard deviation of time to peak longitudinal strain of 16 left ventricular segments adjusted for heart rate. Associations between cTnT concentrations and outcomes were evaluated using linear regression with cTnT modeled as (1) a continuous variable and (2) dichotomized at the level of detection (≥ 3.0 ng/L). Models adjusted for a priori selected confounders influencing CV risk and cTnT concentrations were obtained (model 1: sex, age and estimated glomerular filtration rate; model 2: additionally, hypertension, body mass index, diabetes mellitus and daily smoking). An extended adjustment model included metrics affecting LV loading conditions (model 3: systolic blood pressure, heart rate and end diastolic volume indexed by body surface area at the time of CMR), as well as QRS duration for MD. Results Participants with detectable cTnT had significantly lower absolute GLS, GCS and GRS and higher MD than those with undetectable cTnT (Table 1). The detectable cTnT group had a greater proportion of males, a higher prevalence of hypertension, lower kidney function and was slightly younger at inclusion. Higher cTnT concentrations were independently associated with lower absolute GLS, GCS and GRS and higher MD in both risk factor adjusted models (table 2). For detectable cTnT, similar associations were found in model 1 but the association between detectable cTnT and MD was attenuated in model 2. In the fully adjusted model 3 both detectable and continuous cTnT concentrations remained associated with lower absolute GLS, whereas only detectable cTnT was associated with lower absolute GCS and GRS. Conclusion Higher cTnT, largely within the normal range, is predictive of future lower LV systolic function and synchrony assessed by CMR in a general middle-aged population.