Highlight
- Bayesian non-parametric modeling identified six distinct cardiac function trajectories integrating systolic (LVEF) and diastolic (E/A ratio) measures in late-life individuals.
- Specific trajectories differentially associate with risks of heart failure subtypes: HFrEF and HFpEF.
- Trajectory membership improved prediction of incident heart failure beyond conventional risk factors.
- Mendelian randomization linked 13 plasma proteins causally to LVEF and cardiac volume changes, suggesting novel therapeutic targets.
Study Background and Disease Burden
Heart failure (HF) remains a major global health challenge, with an increasing incidence particularly among older adults. Heart failure with preserved ejection fraction (HFpEF) and heart failure with reduced ejection fraction (HFrEF) represent clinically and pathophysiologically distinct phenotypes that require differentiated diagnostic and management approaches. While left ventricular ejection fraction (LVEF) assesses systolic function, the E/A ratio measured via echocardiography reflects diastolic filling patterns, providing complementary insight into cardiac function. Integrating these longitudinal systolic and diastolic indices could better characterize cardiac aging and help predict heart failure risk. However, traditional approaches rarely model these parameters jointly over time. Additionally, proteomic profiling holds promise for identifying circulating biomarkers and pathogenic pathways underpinning cardiac dysfunction, which may facilitate precision medicine in HF.
Study Design
This investigation utilized a Bayesian non-parametric trajectory approach to model longitudinal patterns of LVEF and E/A ratio in late-life. The primary modeling cohort comprised 747 participants from the Jackson Heart Study who underwent at least two echocardiograms across three exam periods: 2000–2004 (mean age 65 ± 5 years), 2011–2013 (75 ± 5 years), and 2018–2019 (81 ± 5 years). The approach identified distinct integrated systolic-diastolic trajectories over approximately two decades.
These trajectory models were externally validated by predicting trajectory membership in a larger testing cohort of 4,419 participants from the Atherosclerosis Risk in Communities (ARIC) study, based on a single time-point measurement of LVEF and E/A ratio at around 75 years of age.
Multivariable Cox proportional hazards models analyzed associations between predicted trajectory groups and incident heart failure subtypes—HFpEF and HFrEF. Moreover, plasma proteomic analysis using SOMAscan assessed associations of 4,877 plasma proteins with trajectory membership. Mendelian randomization was applied to examine potential causal effects of trajectory-associated proteins on ventricular ejection fraction and volume.
Key Findings
Six distinct trajectories were identified:
1. Pink (50% prevalence): Increasing LVEF with decreasing E/A ratio with aging.
2. Light green (17%): Similar to pink, with increasing LVEF and decreasing E/A ratio.
3. Red (22%): No increase in LVEF over time.
4. Dark green (4%): Declining LVEF.
5. Orange (2%): Steeply declining LVEF coupled with rising E/A ratio.
6. Blue (4%): Rising E/A ratio despite increasing LVEF.
Associations with heart failure subtypes were clear in the testing ARIC cohort. Red and dark green trajectories were significantly associated with incident HFrEF only, indicating that patterns of reduced or non-increasing systolic function predispose to systolic heart failure. The blue trajectory, characterized by increased diastolic dysfunction but preserved or improved systolic function, was specifically linked to HFpEF risk. The orange trajectory, showing rapid systolic decline combined with worsened diastolic function, was associated with risk of both HFpEF and HFrEF.
Trajectory status provided significant incremental predictive value for HF and HFpEF beyond known clinical risk factors and echocardiographic indices alone, underscoring its potential utility in risk stratification.
Plasma proteomic analyses revealed 13 proteins strongly associated with trajectory groups. Mendelian randomization suggested potential causal roles for these proteins in modulating LVEF and cardiac volume changes, illuminating biological pathways relevant to HF pathogenesis and novel targets for therapeutic intervention.
Expert Commentary
This pioneering study leveraged an advanced Bayesian non-parametric strategy to jointly capture systolic and diastolic function trajectories over extended periods, reflecting the complex interplay of cardiac aging processes. The validation in a large external cohort bolsters generalizability. Notably, the differential association of trajectories with HF subtypes aligns with clinical observations that HFpEF is predominantly a disease of diastolic dysfunction and preserved or hyperdynamic systolic function, while HFrEF involves progressive systolic decline.
The integration of proteomic findings with trajectory modeling is especially provocative, linking circulating biomarkers with dynamic cardiac function patterns and providing mechanistic insights. Mendelian randomization adds a layer of causal inference, strengthening biological plausibility. However, limitations include reliance on echocardiographic measurements that may be operator-dependent and potential confounding factors in observational data. Further studies are needed to explore whether early identification of high-risk trajectories can influence clinical decision-making and improve outcomes.
Conclusion
This comprehensive study illustrates the value of integrated systolic and diastolic cardiac function trajectories in predicting heart failure subtypes in older adults. Bayesian trajectory modeling offers a nuanced framework capturing cardiac aging heterogeneity beyond static measures. Combined with proteomic correlations, this approach paves the way for enhanced individualized risk stratification and discovery of molecular therapeutic targets. Future research should focus on prospective validation and translation into clinical practice to ultimately mitigate the burden of heart failure across populations.
References
1. Reimer Jensen AM, Ross JC, Arthur V, Hall ME, Matsushita K, Lennep B, Lutsey PL, Biering-Sørensen T, Shah AM. Integrated trajectories of systolic and diastolic function differentially associate with risk for heart failure with preserved and reduced ejection fraction and proteomic profiles. Eur J Heart Fail. 2025 Sep 2. doi: 10.1002/ejhf.70015. Epub ahead of print. PMID: 40891821.
2. Bozkurt B, Edwards BS, Preston IR. Heart failure with preserved ejection fraction: perspectives and priorities. Am Heart J. 2020;230:105-116. doi:10.1016/j.ahj.2020.03.009
3. Shah SJ, Katz DH, Selvaraj S, et al. Phenomapping for novel classification of heart failure with preserved ejection fraction. Circulation. 2015;131(3):269-279. doi:10.1161/CIRCULATIONAHA.114.010637
4. Verma A, Jackson JA, Whelton SP, et al. Prognostic value of integrated echocardiographic indices for cardiovascular outcomes and mortality: a systematic review and meta-analysis. J Am Coll Cardiol. 2023;81(11):1086-1097. doi:10.1016/j.jacc.2022.12.032