MMP-2 Associates With Incident Heart Failure and Atrial Fibrillation: The ARIC Study

Abstract

Background: MMP (matrix metalloproteinase)-2 participates in extracellular matrix regulation and may be involved in heart failure (HF), atrial fibrillation (AF), and coronary heart disease.

Methods: Among the 4693 ARIC study (Atherosclerosis Risk in Communities) participants (mean age, 75±5 years; 42% women) without prevalent HF, multivariable Cox proportional hazard models were used to estimate associations of plasma MMP-2 levels with incident HF, HF with preserved ejection fraction (≥50%), HF with reduced ejection fraction (<50%), AF, and coronary heart disease. Mediation of the association between MMP-2 and HF was assessed by censoring participants who developed AF or coronary heart disease before HF. Multivariable linear regression models were used to assess associations of MMP-2 with measures of left ventricular and left atrial structure and function.

Results: Compared with the 3 lower quartiles, the highest MMP-2 quartile associated with greater risk of incident HF overall (adjusted hazard ratio, 1.48 [95% CI, 1.21-1.81]), incident HF with preserved ejection fraction (1.44 [95% CI, 1.07-1.94]), incident heart failure with reduced ejection fraction (1.48 [95% CI, 1.08-2.02]), and incident AF (1.44 [95% CI, 1.18-1.77]) but not incident coronary heart disease (0.97 [95% CI, 0.71-1.34]). Censoring AF attenuated the MMP-2 association with HF with preserved ejection fraction. Higher plasma MMP-2 levels were associated with larger left ventricular end-diastolic volume index, greater left ventricular mass index, higher E/e' ratio, larger left atrial volume index, and worse left atrial reservoir and contractile strains (all P<0.001).

Conclusions: Higher plasma MMP-2 levels associate with diastolic dysfunction, left atrial dysfunction, and a higher risk of incident HF and AF. AF is a mediator of MMP-2-associated HF with preserved ejection fraction risk.

Keywords: atrial fibrillation; echocardiography; epidemiology; heart failure; matrix metalloproteinases.

Figure 1.

Incidence of heart failure and heart failure subtypes by quartile of matrix metalloproteinase-2 level The incidences of heart failure overall (Panel A), HFrEF (Panel B) and HFpEF (Panel C) by quartiles of MMP-2 level. HFpEF = heart failure with preserved ejection fraction; HFrEF = heart failure with reduced ejection fraction; IR = incidence rate; MMP-2 = matrix metalloproteinase-2; Q = quartile

Figure 2.

Risk of heart failure and heart failure subtypes associated with the highest quartile of matrix metalloproteinase-2 level quartiles without and with censoring for incident myocardial infarction and atrial fibrillation The associations between MMP-2 and any HF, HFpEF and HFrEF (black circles) and after censoring for incident atrial fibrillation (pink squares) and incident myocardial infarction (green triangles) are summarized. Follow-up for incident atrial fibrillation ended on December 31, 2017. Hazard ratios are adjusted for demographics and clinical risk factors. AF = atrial fibrillation; aHR = adjusted hazard ratio; CI = confidence interval; HFpEF = heart failure with preserved ejection fraction; HFrEF = heart failure with reduced ejection fraction; MI = myocardial infarction; MMP-2 = matrix metalloproteinase-2

Figure 3.

Restricted cubic splines demonstrating the continuous associations between standardized matrix metalloproteinase-2 level and select measures of cardiac structure and function This figure depicts non-linear associations between standardized MMP-2 levels and select measures of cardiac structure and function after adjustment for demographics and clinical risk factors. The number of knots was chosen to minimize the Bayesian information criterion. LA = left atrial; LAVI = left atrial volume index; LVMI = left ventricular mass index; MMP-2 = matrix metalloproteinase-2; PASP = pulmonary artery systolic pressure