Thyroid hormones directly modulate cardiovascular function by enhancing myocardial contractility while also exerting significant indirect effects through sympathetic nervous system activation (1). Even subtle disturbances in thyroid homeostasis are linked to adverse cardiovascular outcomes (2). Heart failure (HF) patients frequently exhibit alterations in thyroid hormone metabolism, including low T3 syndrome and subclinical hypothyroidism (SCH), both established predictors of poor prognosis (3–5). Furthermore, thyrotropin (TSH) levels above 7.0 mU/L show dose-dependent associations with increased coronary heart disease mortality (6, 7), overt hyperthyroidism elevates cardiovascular mortality (8), and high-normal free thyroxine (FT4) correlates with increased mortality in the elderly (9).

Thyroid hormone homeostasis is regulated by the hypothalamic-pituitary-thyroid (HPT) axis (10). Triiodothyronine (T3), the biologically active hormone, is predominantly generated peripherally via deiodination of thyroxine (T4) (11). The FT3/FT4 ratio quantifies peripheral T4-to-T3 conversion efficiency and tissue-level thyroid hormone bioavailability (12, 13), serving as a surrogate marker of peripheral thyroid sensitivity (14, 15). This ratio may be more sensitive than isolated FT3 or FT4 measurements in detecting subtle thyroid metabolic perturbations (16). Emerging evidence supports the FT3/FT4 ratio’s prognostic value in cardiovascular disease for predicting adverse events across multiple populations, including euthyroid acute coronary syndrome patients (17, 18), general cardiovascular disease cohorts (19), and specific groups such as dilated cardiomyopathy (20).

Despite this, the prognostic significance of FT3/FT4 across the spectrum of thyroid function in HF remains underexplored. Notably, HF has an average 1-year mortality rate of up to 33% (21), underscoring the prognostic usefulness for precise, thyroid status-specific risk stratification tools. Defining thyroid function-dependent FT3/FT4 thresholds could enable personalized risk assessment and guide targeted therapies (22).

To address these gaps, this study aimed to: (1) Elucidate the relationship between the FT3/FT4 ratio and composite endpoint risk, and (2) Characterize the nature (linear vs. non-linear) and magnitude of associations of the FT3/FT4 ratio with adverse outcomes.

This prospective study demonstrates that the FT3/FT4 ratio, a marker of peripheral thyroid hormone sensitivity, exhibits thyroid status-dependent associations with 1-year mortality rehospitalization risk in heart failure patients. Our key novel findings are: (1) An inverse relationship was observed between the FT3/FT4 ratio and composite risk across the overall HF cohort, particularly among patients with SCH; (2) A significant U-shaped relationship is observed in euthyroid HF patients.

HF is characterized by metabolic derangements including reduced nutrient intake (30), chronic inflammation, and oxidative stress (31), which can impair micronutrient absorption (iodine, selenium, zinc, iron) crucial for thyroid hormone synthesis and conversion (32). Thyroid hormones, in turn, profoundly impact cardiac electrophysiology, contractility, and structure (33). Hence, Subtle fluctuations in thyroid hormone bioavailability can therefore significantly influence cardiac function and HF progression (34). The inverse FT3/FT4 ratio-adverse outcome relationship likely involves multifactorial pathways. Reduced ratios indicate impaired peripheral T4-to-T3 conversion, directly contributing to tissue-level hypothyroidism during acute/chronic disease states (18). T3, acting via nuclear receptors (TRα/TRβ) in the heart, enhances the expression of key proteins involved in calcium handling (SERCA2a, RYR2) and mitochondrial function/biogenesis; processes often compromised in HF (35). Lower FT3 levels directly contributes to HF pathogenesis by impairing left ventricular relaxation and increasing myocardial stiffness (36). Furthermore, HF is associated with sympathetic overactivation and Renin-Angiotensin-Aldosterone System (RASS) hyperactivity (37); T3 restores autonomic balance in heart failure by: (1) Attenuating sympathetic overactivation through downregulation of myocardial β adrenergic receptor density and reduction of circulating norepinephrine levels (38); (2) Recovering baroreflex function via upregulation of neuronal nitric oxide synthase (nNOS) in the nucleus tractus solitarius (39), thereby mitigating these detrimental neurohormonal axes (40).

Our findings align with and extend previous research. Studies linked lower FT3/FT4 ratios to increased mortality in dilated cardiomyopathy (20). A study using propensity matching demonstrated that a reduced ratio of FT3/FT4 is a robust predictor of all-cause mortality in heart failure patients (41). Another study discovered that the FT3/FT4 ratio independently predicts all-cause mortality in the general population. In addition, a study noted that in euthyroid patients with type 2 diabetes mellitus, a low FT3/FT4 ratio independently contributed to major adverse cardiac events following acute myocardial infarction (42). However, our study is the first to comprehensively evaluate this ratio across distinct thyroid functional states (euthyroid vs. SCH) within an HF cohort and to identify state-specific risk patterns. The protective association observed in SCH patients is particularly noteworthy and suggests that maintaining adequate peripheral conversion is crucial in this subgroup, potentially outweighing the risks associated with mildly elevated TSH. This finding resonates with experimental data showing T3 reduces infarct size and activates cardio-protection during ischemia/reperfusion (43), and clinical evidence that low-dose T3 replacement improved LV function post-AMI (44). Notably, in patients with SCH, multivariate Cox regression demonstrated an inverse association between the FT3/FT4 ratio and the risk of composite outcomes, whereas KM analysis revealed no significant survival difference among the three groups. This discordance likely reflects the inherent limitation of KM analysis: as a non-parametric method estimating unadjusted survival probabilities from observed events, it does not account for confounding factors.

The U-shaped relationship observed in euthyroid patients is a novel and significant finding. The loss of significance in the fully adjusted multivariate Cox model for the euthyroid subgroup, particularly after accounting for age and NT-proBNP-both strongly associated with the ratio and outcomes, suggesting that the univariate association was partly confounded. This interpretation is primarily informed by the simultaneous measurement of NT-proBNP and thyroid indicators during the same blood draw, which inherently precludes establishing the temporal sequence required for mediation (where exposure must precede the mediator, which subsequently influences the outcome) (45). However, the highly significant U-curve revealed by RCS analysis indicates a complex, non-monotonic relationship. This suggests an optimal range for peripheral thyroid sensitivity in euthyroid HF. Ratios significantly below 0.22 likely indicate impaired conversion and tissue hypothyroidism, increasing risk. Conversely, ratios significantly above this point might reflect excessive peripheral conversion, potentially linked to hypermetabolic states or impaired hormone clearance, which could paradoxically increase cardiovascular strain or indicate other underlying metabolic disturbances detrimental in HF. Compensatory mechanisms preserving tissue-level thyroid hormone action within the normal functional range might also buffer the impact of ratio variations, except at the extremes (46). Potential mechanisms involve: (1) impaired enzymatic efficiency due to the Thr92Ala DIO2 polymorphism at low FT4/FT3 ratios, whereas DIO2 overexpression under high-T4 conditions depletes essential cofactors (e.g., glutathione), thereby exacerbating T4-to-T3 conversion failure (47); (2) pro-inflammatory cytokines (TNF-α, IL-6) suppressing deiodinase activity through NF-κB-mediated downregulation of DIO1 and DIO2 expression, impairing T3 generation and promoting thyroid hormone resistance during chronic inflammation (48). Further research is needed to elucidate the mechanisms underlying this U-shape.

To our knowledge, this is the first study to identify the optimal FT3/FT4 for stratifying mortality/readmission risk in HF patients with varying thyroid states. Limitations should be acknowledged. First, the single-center design necessitates cautious interpretation regarding generalizability. As iodine sufficiency varies significantly across global regions, and Asian populations often exhibit distinct iodine nutritional status compared to Western cohorts, our results should be interpreted within this contextual framework. Second, only baseline thyroid function was assessed; serial measurements might better capture dynamic changes relevant to prognosis. Third, the mechanisms linking the FT3/FT4 ratio to HF outcomes, particularly the U-shape in euthyroidism, require further elucidation. Fourth, In the SCH cohort, the observed hazard ratio (HR) of 0.16 should be interpreted with caution due to the limited sample size. Finally, observational design precludes causal inference.

In conclusion, the FT3/FT4 ratio is a thyroid status-dependent predictor of 1-year mortality and HF rehospitalization risk. Monitoring the FT3/FT4 ratio, offers a valuable tool for risk stratification. Future studies should validate these thresholds in diverse populations and explore whether interventions aimed at optimizing peripheral thyroid hormone sensitivity.