13 The association between cardiovascular risk factors and left atrial structure and phasic function

Introduction

Changes in left atrial (LA) phasic function – defined as reservoir function (filling), conduit function (passive emptying), booster function (active emptying) (figure 1A) – are thought to precede structural changes (dilatation) of the LA and therefore may be a sensitive biomarker in heralding cardiovascular disease. This study is the first to examine the association between traditional cardiovascular risk factors and LA structure and function in a large population.

Methods

LA volumes were calculated from four- and two-chamber CMR cines performed as part of the UK Biobank imaging substudy using the biplane method. To determine phasic function, the LA was automatically contoured in the four-chamber views in all fifty frames of the cardiac cycle using a three-dimensional sparse active shape model (SPASM) with parameters estimated using a deep neural network. A time-volume curve across the cardiac cycle was plotted with a smooth line of best fit (figure 1B). Quality control was performed manually on an initial subset of cases, following which an automated tool was implemented. Utilising the presence of local maxima and minima, LA maximal volume (LAmax), LA minimal (LAmin) volume and LA pre-atrial contraction (LApre) were determined, allowing derivation of total emptying fraction, conduit function and booster function. Cardiovascular risk factors considered were: age, sex, ethnicity, height, BMI, hypertension, diabetes, hyperlipidaemia, physical activity (metabolic equivalent [MET] minutes/week) and smoking. Multivariable linear regression models were fitted, adjusting for all risk factors with LA parameters as dependent variables.

Results

After quality control, 44,957 participants were included in the analysis. The mean age was 64±8, 51.8% of the cohort were female. Regression analysis is detailed in table 1. Hypertension and increasing BMI were significantly associated with lower total emptying fraction, conduit function and booster pump function. Cigarette smoking also resulted in reduced conduit function and booster pump function. Hypertension and BMI were associated with larger LAmax volumes. Beta coefficients, 95% confidence intervals and p-values estimating association of risk factors with LA parameters in multivariable linear regression models. Co-variates included in models are as follows: age, sex, ethnicity, height, BMI, hypertension, diabetes, hyperlipidaemia, total MET minutes/week, cigarette smoking.

Conclusion

This is the first study to calculate time-volume curves to derive markers of atrial function in a large cohort. We demonstrate that hypertension and increasing BMI – risk factors that are heavily implicated in diastolic dysfunction – are most frequently associated with alterations in LA phasic function.