Association Between Antecedent Myocardial Infarction and Heart Failure with Preserved Versus Reduced Ejection Fraction

Traditionally, the link between myocardial infarction (MI) and heart failure (HF) was based on the understanding that the size of an infarct and the concomitant loss of left ventricular function lead to left ventricular adverse remodelling, thereby resulting in HF with reduced ejection fraction (HFrEF).1 However, in the modern era, timely state-of-the-art reperfusion strategies have considerably limited infarct sizes and the effects thereof, and yet, the risk of post-MI HF remains substantial.1 Simultaneously, recent endeavours have demonstrated an increase in the prevalence of HF with preserved ejection fraction (HFpEF), suggesting that it may become the most common HF subtype in the future.2 Therefore, a new paradigm has been put forth in which MI patients who undergo timely reperfusion may develop HFpEF years after developing MI.1 However, it is still unclear whether HFpEF is indeed the result of MI, or simply a by-product of the conglomeration of risk factors and comorbidities prevalent in old age. Using a large community-based cohort with 25-year follow-up, this study aimed to (i) investigate the association between previous MI and the risk of developing of HFpEF; (ii) elucidate the duration between MI and HFpEF; and (iii) compare these findings to HFrEF. The study was conducted using the PREVEND (Prevention of Renal and Vascular End-stage Disease) cohort.3, 4 A total of 8592 subjects completed the screening programme, which consisted of a baseline visit in 1997, followed by 5 screening rounds separated by 3–4 year intervals up until 2011. Further details regarding the PREVEND cohort have been described previously.3, 4 MI diagnosis was defined by Minnesota codes 1.1 and 1.2 during the first screening visit, and by the International Classification of Diseases criteria, 10th revision (ICD-10), thereafter.5 Details regarding HF data in the PREVEND cohort up until 2011 have been previously described.3 After 2011, and until January 2022, all patient files were manually reviewed, and HF diagnosis was made using the 2021 European Society of Cardiology HF guidelines.6 In instances of doubt, the case was adjudicated by two HF specialists. Left ventricular ejection fraction (LVEF) at the time of HF diagnosis was then used to allocate each HF case into the different subtypes: HFrEF (LVEF ≤40%), HF with mildly reduced ejection fraction (LVEF 41–49%) and HFpEF (LVEF ≥50%).3 The relationship between antecedent MI and HFpEF development was investigated utilizing a nested case-control design. Clinical and laboratory parameters were extracted from the PREVEND screening visit, most recent to the date of HF diagnosis. Patients with HF (HFpEF or HFrEF) were considered cases. Controls were age- and sex-matched at a 1:4 ratio from the population that was free of HF at the time each case developed HF. The PREVEND study was conducted in compliance with the Declaration of Helsinki and was approved by the relevant ethics committee. Follow-up data were available until January 2022. In this 25-year follow-up study, 273 patients developed HFpEF and 361 patients developed HFrEF. A total of 2451 PREVEND subjects without HF served as controls. For HFpEF cases and their controls, the mean age at inclusion was 74 years (51% female; 96% Caucasians). The two subgroups showed significant differences at baseline in terms of comorbidities, medication and laboratory markers; HFpEF patients had higher rates of obesity, hypertension and type 2 diabetes (42% vs. 16%, 82% vs. 57%, and 21% vs. 12%, respectively; p < 0.001). The mean age of HFrEF patients and their controls was 71 years (30% female). Similarly to HFpEF, HFrEF patients had higher rates of obesity, hypertension and type 2 diabetes when compared to their controls, in addition to higher rates of smoking and hypercholesterolaemia. The proportion of antecedent MI was higher in patients with HFpEF when compared to HF-free controls (21% vs. 8%; p < 0.001). This was also the case for HFrEF patients (36% vs. 9%; p < 0.001). Moreover, the time between MI diagnosis and subsequent development of HF was considerably shorter in the HFrEF group relative to the HFpEF group (4.1 vs. 9.5 years; interquartile range 0.4–9.3 vs. 4.9–16.0 years; p < 0.001). Unconditional logistic regression analysis revealed that antecedent MI was associated with subsequent HFpEF development, even after correcting for age, sex, atrial fibrillation, obesity, hypertension, type 2 diabetes, hypercholesterolaemia and estimated glomerular filtration rate (odds ratio OR 2.44; 95% confidence interval CI 1.48–4.00; p < 0.001) (Figure 1).6-9 Antecedent MI was also associated with HFrEF, even after correcting for smoking status alongside the aforementioned confounders (OR 5.11; 95% CI 3.53–7.38; p < 0.001) (Figure 1). To test whether the association between MI and HFrEF was stronger than that between MI and HFpEF, a new set of models using HFpEF as a reference was used, which showed that the difference was significant (p = 0.009) (Figure 1). Using data from a carefully characterized population-based cohort across a time span of 25 years (1997–2022), the present nested case-control study demonstrates that 21% of HFpEF patients have had previous MI. Furthermore, it provides evidence that antecedent MI is associated with a significantly increased risk of subsequent HFpEF development, even after correcting for important confounders. Additionally, our present investigation shows that the median time between index MI and HFpEF diagnosis (9.5 years) is more than double that between index MI and HFrEF development (4.1 years). However, several limitations warrant consideration. First, although several methods were employed to control for confounding, we cannot rule out the possibility of unmeasured confounding. Second, because the definition of MI changed over time, we had to apply different definitions of MI over the whole study period. Third, we did not record parameters that could gauge the size of infarcts, collect information regarding the reperfusion strategy used to treat MI nor did we account for recurrent MI after index MI. Fourth, clinical and laboratory parameters were extracted from the PREVEND visit most recent to the date of HF diagnosis. Therefore, these values do not perfectly represent true values at the time of HF development. This being said, our findings should prompt further research on the potential relationship between MI and subsequent HFpEF development. Use of artificial intelligence was limited to the conceptualization of writing and for the identification of coding errors in statistical analysis. The PREVEND study was financially supported by the Dutch Kidney Foundation (grant number E0.13). Conflict of interest: J.T. is supported by the National University of Singapore (Start-up grant), the (tier 1 grant) from the Ministry of Education and the (CS-IRG New Investigator Grant) from the National Medical Research Council; has received research support from AstraZeneca and consulting or speaker fees from Daiichi-Sankyo, Boehringer Ingelheim, Roche diagnostics and Us2.ai, and owns patent US-10702247-B2 unrelated to the present work. K.D. reports speaker/consultancy fees to his institution by AstraZeneca, Abbott, Novartis, Boehringer Ingelheim, Echosense and FIRE1. R.A.d.B. received support from the Netherlands Heart Foundation (grant numbers 2017-21, 2017-11); the institutions of R.A.d.B. has received research grants and/or fees from are AstraZeneca, Abbott, Bristol Myers Squibb, Cardior Pharmaceuticals GmbH, Novo Nordisk, and Roche; R.A.d.B. has had speaker engagements with and/or received fees from and/or served on an advisory board for Abbott, AstraZeneca, Bristol Myers Squibb, Cardior Pharmaceuticals GmbH, Novo Nordisk, and Roche; and received travel support from Abbott, Cardior Pharmaceuticals GmbH, and Novo Nordisk. E.L. has received an institutional educational grant from Abbott Medical Nederland B.V. All other authors have nothing to disclose.