Molecular multi-omics signatures of autonomic imbalance using the LF/HF ratio in individuals with heart failure
S Zeid, G Buch, T Koeck, D Velmeden, J Soehne, A Kerber, F Rausch, J H Prochaska, P Lurz, P S WildAbstract
Background
Heart rate variability (HRV) from 24-hour Holter ECG is a well-established non-invasive measure of autonomic function, with the ratio of low-frequency to high-frequency power (LF/HF ratio) widely interpreted as a measure of sympatho-vagal balance. Although alterations in LF/HF are prognostic in heart failure, the molecular mechanisms driving these changes remain largely unknown. We aimed to identify transcriptomic and proteomic signatures associated with the LF/HF ratio and to evaluate their prognostic relevance across heart failure phenotypes.
Methods
In a deeply phenotyped heart failure cohort (MyoVasc Study, N=3,289), 24-hour Holter ECG recordings were analyzed to derive LF, HF, and the LF/HF ratio. Blood samples underwent whole-blood RNA sequencing and targeted proteomic analysis (Target 96 panels covering 538 proteins). Elastic net regression identified transcripts and proteins associated with the LF/HF ratio independently of age and sex. Weighted transcriptomic and proteomic LF/HF scores were constructed and evaluated in the full cohort. Pathway and network analyses (STRINGdb) were used to characterize biological processes underlying autonomic dysfunction. Associations with mortality and worsening of heart failure were assessed using multivariable Cox regression models adjusting for cardiovascular risk factors, comorbidities, and medication intake.
Results
Holter ECG and proteomic data were available for 974 individuals (age: 64.4±10.4 years; women: 34.1%), with transcriptomic data in 60.7%. Elastic net regression identified 20 proteins and 49 transcripts independently associated with the LF/HF ratio. Top ranked proteins included NT-proBNP, FABP4, NOS3, FGF-23, and TNF, while transcripts were enriched for pathways related to autonomic signaling, ion channel regulation and cardiac remodeling (e.g. LPAR5, SCN3A, DAAM1, GRK5). Both the proteomic and transcriptomic LF/HF scores predicted adverse heart failure outcomes, independently of age, sex, cardiovascular risk factors, comorbidities and medication use. The proteomic LF/HF score was strongly associated with a lower risk of worsening of heart failure (N=3,188; HR per SD 0.55, 95% CI 0.48; 0.63; p<0.0001). This association was more pronounced in HFpEF (HR 0.44, 95% CI 0.33; 0.60) than in HFrEF (HR 0.63, 95% CI 0.49; 0.82; p for interaction <0.05). The transcriptomic LF/HF score also remained independently prognostic (N=1,644; HR 0.81, 95% CI 0.72–0.90; p=0.0002).
Conclusions
This first integrative multi-omics investigation of Holter ECG-derived LF/HF ratio identifies distinct molecular signatures underlying autonomic dysfunction in heart failure. LF/HF-associated proteomic and transcriptomic profiles are strongly associated with clinical outcomes, with greater prognostic relevance in HFpEF, highlighting autonomic regulation-related molecular mechanisms and phenotype-specific potential therapeutic targets.