Study Population and Ethical Oversight

This hospital-based case–control study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki (1975, as revised in 2003) and was reviewed and approved by the Institutional Ethics Committee of Genetika, Centre for Advanced Genetic Studies (Approval Ref. No.: 15/2023/IECG; Dated: 03/08/2023). Prior to participant enrollment, written informed consent was obtained from all participants or their legally authorized representatives. Participant confidentiality, anonymity, and data privacy were strictly maintained throughout the study.

Clinical data and biological specimens were prospectively collected from patients presenting to tertiary cardiac centers across Kerala, India, between January 2023 and January 2025. The clinical diagnosis of HF was independently confirmed by two or more experienced cardiologists based on comprehensive clinical history, physical examination (identifying classic symptoms and physical signs of congestion or low output), 12-lead electrocardiography (ECG), and transthoracic echocardiography according to American Heart Association (AHA) and European Society of Cardiology (ESC) guidelines. Patients were excluded if they met any of the following criteria: (1) presence of acute or chronic systemic infections; (2) end-stage renal disease (estimated glomerular filtration rate < 15 mL/min/1.73 m or ongoing dialysis) or advanced hepatic failure; (3) active malignancy; (4) congenital heart disease; (5) systemic autoimmune or rheumatic immune disorders; or (6) incomplete clinical or laboratory records.

A total of 150 patients with clinically confirmed, stable HF who met all eligibility criteria were consecutively enrolled in the case group. Simultaneously, 150 age- and sex-matched healthy control individuals (n = 150) were recruited from community health screening programs conducted in the same geographic region during the identical time period. All potential control subjects underwent comprehensive clinical history taking, physical examination, 12-lead ECG, and baseline biochemical evaluation to rigorously exclude cardiovascular disease, hypertension, diabetes mellitus, renal dysfunction, or other major systemic comorbidities prior to enrollment. The final study cohort comprised 300 participants (150 HF cases and 150 healthy controls). Figure 1 outlines the participant screening, recruitment, and enrollment flow.

Sample Size Determination

The required sample size was calculated using the standard formula for comparing two independent means:9

where Z = 1.96 corresponds to a two-sided significance level (α) of 0.05, Z = 0.84 provides 80% statistical power (β = 0.20), σ represents the pooled standard deviation, and Δ is the expected clinically meaningful difference between group means. Given the paucity of published baseline data on peripheral blood LIF mRNA expression in human HF cohorts at the time of study initiation, a conservative moderate effect size (Cohen’s d = 0.5) was assumed. While myocardial tissue studies have reported large effect sizes for LIF expression (Cohen’s d > 0.8),10 peripheral blood transcriptomic measurements exhibit greater biological and inter-individual variability. Based on these parameters (α = 0.05, power = 80%, Cohen’s d = 0.5), a minimum sample size of 63 participants per group was required. To ensure robust statistical power for multivariable logistic regression modeling, subgroup comparisons, and receiver operating characteristic (ROC) analyses, the final cohort was expanded to 150 HF patients and 150 matched healthy controls (total N = 300).

Diagnostic Criteria and Definitions

The following operational definitions were applied throughout the investigation:

Heart Failure Cases: Defined as adult patients with clinically documented, chronic stable HF diagnosed in accordance with standard AHA/ESC guidelines, supported by documented clinical signs and symptoms of HF, echocardiographic evidence of cardiac structural or functional abnormalities, and elevated biomarkers.

Healthy Controls: Defined as apparently healthy individuals with no prior history, symptoms, or clinical signs of cardiovascular disease. All control participants demonstrated normal 12-lead ECG recordings, normal blood pressure, normal fasting glucose, and absence of systemic hypertension, diabetes mellitus, chronic kidney disease, chronic inflammatory disorders, or oncological history.

Clinical and Laboratory Data Collection

Standardized clinical proformas were used to record baseline demographic parameters, including age, sex, body mass index (BMI, kg/m), smoking history, and pre-existing comorbidities (hypertension, type 2 diabetes mellitus, and coronary artery disease). Following an 8- to 12-hour overnight fast, 6 to 8 mL of peripheral venous blood was drawn from the antecubital vein of each participant using sterile, vacuum-assisted venipuncture under resting conditions.

Blood samples were collected into two specialized Vacutainer tubes: (1) clot activator tubes for serum separation and biomarker estimation (NT-proBNP, IL-6, MDA, and sST2), and (2) sterile ethylenediaminetetraacetic acid (EDTA) tubes for total RNA extraction and subsequent transcriptomic analysis. Clot activator tubes were allowed to clot completely at room temperature (22–25 °C) for 30 minutes, followed by centrifugation at 3,000 rpm for 15 minutes at 4 °C. The separated serum was aliquoted into sterile cryovials and stored at −80 °C until batch ELISA analysis. EDTA whole blood samples were processed promptly for RNA isolation to prevent enzymatic degradation.

Circulating Biomarker Quantification by ELISA

Serum concentrations of cardiac, inflammatory, oxidative, and remodeling biomarkers were quantified using commercially available, high-sensitivity enzyme-linked immunosorbent assay (ELISA) kits in strict accordance with the manufacturers’ standard operating protocols:

• NT-proBNP: Measured using a quantitative sandwich ELISA kit (Origin Diagnostics and Research, India; Cat. No.: OPK1212). Standards and serum samples were added to microplate wells pre-coated with a monoclonal anti-NT-proBNP capture antibody. Following incubation and washing, a biotinylated secondary detection antibody and streptavidin–horseradish peroxidase (HRP) conjugate were added. Chromogenic development was achieved using 3,3′,5,5′-tetramethylbenzidine (TMB) substrate and halted with sulfuric acid stop solution. Optical density (OD) was measured at 450 nm (with wavelength correction at 630 nm) on a Multiskan FC microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). Absolute concentrations were calculated using a four-parameter logistic (4-PL) standard curve.

• Interleukin-6 (IL-6): Quantified using a high-sensitivity sandwich ELISA kit (Origin Diagnostics and Research, India; Cat. No.: OPK1156). Serum samples were incubated in capture antibody–coated wells, followed by HRP-conjugated detection antibody chemistry, TMB color development, and absorbance measurement at 450 nm. Concentrations were derived from standard calibration curves.5

• Soluble ST2 (sST2): Determined using a sandwich ELISA kit (Origin Diagnostics and Research, India; Cat. No.: OPK9271) specific for human sST2. Absorbance was read at 450 nm, and concentrations were interpolated from a validated 4-PL standard curve.

• Malondialdehyde (MDA): Quantified as an index of lipid peroxidation using a competitive ELISA kit (Origin Diagnostics and Research, India; Cat. No.: OPK8428). In this assay format, endogenous serum MDA competes with pre-coated MDA–protein conjugate for specific antibody binding sites, resulting in an inverse relationship between absorbance at 450 nm and sample MDA concentration. Final concentrations were determined from an inhibition standard curve fitted with a 4-PL algorithm.

All ELISA assays were performed in duplicate. Internal quality-control standards were included in each run, and an intra-assay coefficient of variation (CV) of < 10% was maintained across all plates.

Leukemia Inhibitory Factor (LIF) Gene Expression Analysis

Total RNA was extracted from whole blood collected in EDTA tubes using a silica column-based RNA isolation kit (Origin Diagnostics and Research, India; Cat. No.: ODP419). Briefly, 250 μL of whole blood was lysed with three volumes of lysis buffer (Buffer RZ), followed by phase separation with 150 μL of analytical-grade chloroform and centrifugation at 12,000 rpm for 10 minutes at 4 °C. The upper aqueous phase was transferred into an RNase-free microcentrifuge tube, mixed with an equal volume of absolute ethanol (96%–100%), and applied to a silica-membrane spin column (CR3). Spin columns were washed sequentially with Buffer RP and Buffer RW to remove protein and salt impurities. After centrifugation to remove residual ethanol, purified total RNA was eluted in 30–100 μL of nuclease-free water. RNA yield and purity were determined spectrophotometrically using A260/A280 absorbance ratios on a NanoDrop spectrophotometer.

Complementary DNA (cDNA) was synthesized from 50 ng of purified total RNA using a specialized reverse transcription kit (Origin Diagnostics and Research, India; Cat. No.: ODR41). Each 20 μL reaction mixture contained RNA template, 0.5 μL of oligo(dT) primer, 0.5 μL of random hexamer primers, 10 mM dNTP mix, 5× RT buffer, and reverse transcriptase enzyme. Reverse transcription was performed on a thermal cycler under the following conditions: primer annealing at 25 °C for 5 minutes, enzymatic cDNA synthesis at 50 °C for 60 minutes, and enzyme inactivation at 95 °C for 5 minutes. The resulting cDNA was stored at −20 °C until real-time amplification.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR) was performed using SYBR Green chemistry on a Bio-Rad CFX Opus 96 Real-Time PCR System (Bio-Rad Laboratories, Hercules, CA, USA). Each 20 μL reaction volume contained 10 μL of 2× SYBR Green Master Mix, 1 μL of forward primer (10 μM), 1 μL of reverse primer (10 μM), 2 μL of cDNA template, and 6 μL of nuclease-free water. Target-specific oligonucleotide primers for human LIF (RefSeq: NM_002309.5) and the reference housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH; RefSeq: NM_002046.7) were designed using the NCBI Primer-BLAST tool and commercially synthesized by Eurofins Genomics India Pvt. Ltd. (Bangalore, India).11 Primer sequences are summarized in Table 1.

Thermal cycling parameters consisted of an initial denaturation step at 94 °C for 1 minute, followed by 40 amplification cycles of denaturation at 94 °C for 15 seconds, primer annealing at 58 °C for 1 minute, and a final extension at 72 °C for 10 minutes. A continuous dissociation (melt-curve) analysis was conducted from 65 °C to 95 °C with increments of 0.5 °C to confirm reaction specificity and verify the absence of primer dimers or non-specific amplicons.

Data Analysis and Expression Quantification

Fluorescence quantification cycle (C) values were exported from the Bio-Rad CFX Maestro software. Mean C values from duplicate wells were used for all quantitative calculations. Relative fold changes in LIF mRNA expression were calculated using the comparative 2 method, where ΔC = C − C, and ΔΔC = ΔC − mean ΔC. Melting curve profiles confirmed a single, distinct peak for all target and reference reactions (Figure 2).

Statistical Analysis

Statistical computations were performed using Stata statistical software version 17.0 (StataCorp LLC, College Station, TX, USA). Continuous variables were assessed for distributional normality using the Shapiro–Wilk test; non-normally distributed continuous data are presented as mean ± standard deviation (SD) and were compared between groups using the non-parametric Mann–Whitney U test. Categorical variables are reported as frequencies and percentages (n (%)) and were compared using Pearson’s Chi-square (χ) test.

Diagnostic performance and discriminative accuracy of LIF gene expression and circulating biomarkers were evaluated using receiver operating characteristic (ROC) curve analysis. The area under the ROC curve (AUC) and 95% confidence intervals (CIs) were computed, and optimal cut-off values were identified using the Youden index (J = sensitivity + specificity − 1) to determine corresponding sensitivities, specificities, positive likelihood ratios (LR), and negative likelihood ratios (LR). Bivariate relationships between LIF gene expression and circulating biomarkers (NT-proBNP, IL-6, sST2, and MDA) were evaluated using Spearman rank correlation coefficients (ρ) and linear regression coefficients of determination (R).

To examine the independent association between LIF gene expression and HF status, sequential multivariable binary logistic regression models were constructed: Model 1 was unadjusted; Model 2 was adjusted for age, sex, and BMI; Model 3 was additionally adjusted for hypertension; and Model 4 (fully adjusted model) incorporated age, sex, BMI, hypertension, NT-proBNP, IL-6, sST2, and MDA. Results are presented as odds ratios (ORs) with corresponding 95% CIs. Potential multicollinearity among regression covariates was formally evaluated using variance inflation factors (VIFs), with a VIF threshold of < 5 indicating acceptable collinearity. All statistical tests were two-tailed, and a p-value < 0.05 was considered statistically significant.

Demographic and Baseline Clinical Characteristics

A total of 300 participants (150 HF cases and 150 healthy controls) were evaluated in this study. The baseline demographic and clinical features of the study cohort are presented in Table 2. The mean age was 46.6 ± 9.3 years in the HF group and 45.6 ± 9.8 years in the control group, demonstrating no statistically significant difference (p = 0.254). The sex distribution was also well-balanced between groups (females: 53.3% in cases vs. 52.0% in controls; males: 46.7% in cases vs. 48.0% in controls; p = 0.817), confirming successful matching. Conversely, HF patients exhibited a significantly higher prevalence of obesity (62.0% vs. 11.3%, p < 0.01) and clinical hypertension (40.7% vs. 0.0%, p < 0.01) compared with healthy controls.

Circulating Biomarker Concentrations and LIF Transcriptional Upregulation

As summarized in Table 3, patients with HF demonstrated marked alterations across all evaluated circulating biomarkers and gene expression levels compared with healthy controls. Circulating concentrations of NT-proBNP were significantly higher in HF patients than in controls (142.3 ± 41.0 pg/mL vs. 85.3 ± 34.7 pg/mL, p < 0.01). Similarly, the pro-inflammatory cytokine IL-6 (7.5 ± 3.3 pg/mL vs. 4.3 ± 2.5 pg/mL, p < 0.01), the myocardial remodeling marker sST2 (20.9 ± 9.2 ng/mL vs. 13.1 ± 6.6 ng/mL, p < 0.01), and the oxidative lipid peroxidation marker MDA (4.3 ± 2.7 ng/mL vs. 2.1 ± 1.0 ng/mL, p < 0.01) were significantly elevated in HF cases.

Quantitative RT-PCR analysis demonstrated a significant upregulation of peripheral blood LIF mRNA expression in HF patients compared with healthy controls (mean 2 fold change: 2.1 ± 1.2 vs. 1.1 ± 0.2, p < 0.01). These results establish that active transcriptional induction of LIF is detectable in the peripheral circulation of HF patients alongside elevated systemic markers of myocardial strain, inflammation, and oxidative stress.

Diagnostic Performance of LIF Gene Expression and Biomarkers

The diagnostic utility of peripheral blood LIF gene expression and circulating biomarkers was evaluated using ROC curve analysis (Figure 3). LIF gene expression demonstrated robust diagnostic discriminative ability, yielding an AUC of 0.784 (95% CI: 0.732–0.836). Utilizing the optimal cut-off value determined by the Youden index, LIF expression exhibited a diagnostic sensitivity of 58.0% and an outstanding specificity of 98.0%, resulting in a high positive likelihood ratio (LR) of 29.0.

Among circulating protein biomarkers, IL-6 demonstrated comparable diagnostic accuracy with an AUC of 0.782, a sensitivity of 68.0%, and a specificity of 74.0%. MDA displayed moderate diagnostic performance, with an AUC of 0.765, a specificity of 83.0%, and an LR of 3.63. For sST2, the optimal diagnostic cut-off determined by the Youden index was ≥ 14.91 ng/mL, which yielded an AUC of 0.748 (95% CI: 0.692–0.804), a sensitivity of 72.0%, and a specificity of 70.7%. Collectively, these findings indicate that while established protein biomarkers provide good sensitivity, peripheral blood LIF mRNA expression offers superior diagnostic specificity for identifying heart failure.

Correlation Between LIF Expression and Biomarker Profiles

Bivariate correlation analysis revealed significant positive associations between peripheral blood LIF gene expression and key biomarkers reflecting hemodynamic stress, inflammation, oxidative damage, and cardiac remodeling (Figure 4). Specifically, LIF gene expression showed a moderate, statistically significant positive correlation with NT-proBNP (ρ = 0.5146, R = 0.3189, p < 0.001), indicating that elevated transcriptional levels of LIF track with increasing myocardial wall stress in HF patients.

A significant positive correlation was likewise observed between LIF expression and IL-6 (ρ = 0.4425, R = 0.1895, p < 0.001), consistent with their shared membership in the gp130 cytokine family and cooperative roles in inflammatory cascade activation. Furthermore, LIF mRNA expression correlated positively with sST2 (ρ = 0.5321, R = 0.2598, p < 0.001), suggesting a close link between LIF transcription and ongoing extracellular matrix remodeling and myocardial fibrosis. Finally, a significant correlation was observed between LIF expression and MDA levels (ρ = 0.4371, R = 0.1636, p < 0.001), indicating that systemic oxidative injury and lipid peroxidation parallel the upregulation of LIF-associated transcriptional pathways in heart failure.

Multivariable Logistic Regression Analysis

To determine whether peripheral LIF gene expression serves as an independent predictor of HF, sequential multivariable binary logistic regression models were evaluated (Table 4). In the unadjusted model (Model 1), elevated LIF gene expression was strongly associated with HF status (OR = 15.123, 95% CI: 6.300–36.303, p < 0.01). This association remained robust and statistically significant after adjusting for age, sex, and BMI in Model 2 (OR = 6.220, 95% CI: 2.417–16.009, p < 0.01) and following additional adjustment for hypertension status in Model 3 (OR = 3.248, 95% CI: 1.271–8.300, p = 0.014), although the effect magnitude progressively attenuated.

However, in the fully adjusted model (Model 4), which incorporated NT-proBNP, IL-6, MDA, and sST2 alongside demographic and clinical covariates, the association between LIF gene expression and HF was markedly attenuated and no longer statistically significant (adjusted OR = 0.625, 95% CI: 0.202–1.935, p = 0.415). In this fully adjusted model, NT-proBNP remained a strong, independent predictor of HF (OR = 1.025, 95% CI: 1.008–1.042, p = 0.003). These findings demonstrate that the association of peripheral LIF expression with HF is largely mediated through or co-linear with shared downstream inflammatory and myocardial stress pathways represented by NT-proBNP, IL-6, sST2, and MDA.

Multicollinearity Diagnostics

To ensure that the attenuation of LIF in the fully adjusted model was not artifactually caused by severe collinearity among predictor variables, variance inflation factors (VIFs) were calculated for all independent variables included in Model 4 (Figure 5). The resulting VIF values ranged from 1.46 to 4.56 (with LIF exhibiting a VIF of 3.56). All values remained well below the conservative empirical threshold of 5.0, confirming the absence of problematic multicollinearity and verifying the mathematical stability and reliability of the regression coefficients.

LIF Gene Expression and Transcriptional Activation in Heart Failure

Previous experimental studies have consistently highlighted the involvement of LIF in cardiac physiology and pathology. In canine models of experimental congestive heart failure, Jougasaki et al. demonstrated marked increases in myocardial LIF immunoreactivity and tissue gene expression, suggesting a localized compensatory or stress-activated response.13,14 Similarly, Eiken et al. reported enhanced myocardial mRNA and protein expression of gp130 and its cognate ligands, including LIF and IL-6, in failing human hearts obtained at transplantation.10 Further experimental studies by Berry et al. showed that targeted cardiac overexpression of LIF attenuates post-infarction remodeling and promotes cell survival pathways.15 Our study significantly extends these tissue-based and animal observations by demonstrating that transcriptional upregulation of LIF is readily detectable in human peripheral blood. LIF is rapidly induced by cellular stretch, mechanical strain, inflammatory cytokines, and hypoxia.16 Because total RNA was extracted from peripheral whole blood, the elevated LIF transcript levels observed here likely reflect systemic transcriptional activation across circulating leukocytes and vascular endothelial interactions rather than isolated myocardial spillover. Although peripheral blood profiling does not pinpoint single-cell origins, it offers an accessible, minimally invasive modality for monitoring systemic cytokine pathway activation in HF patients.

Diagnostic Utility and Complementary Performance of LIF Expression

Receiver operating characteristic analysis revealed that peripheral blood LIF gene expression achieved an AUC of 0.784, with an exceptional diagnostic specificity of 98% and a high positive likelihood ratio of 29.0 (Figure 3). While natriuretic peptides such as NT-proBNP remain the clinical cornerstone for HF diagnosis, their circulating levels can be confounded by non-cardiac conditions such as advanced age, female sex, renal insufficiency, and obesity.17 In contrast, transcriptional regulation of peripheral cytokines may be less susceptible to acute volumetric shifts, potentially providing distinct biological information. The high specificity of LIF expression indicates that marked transcriptional elevation is rarely present in healthy individuals. However, its modest diagnostic sensitivity (58%) precludes its use as an isolated rule-out test. Instead, these results support the integration of LIF gene expression into a multimarker diagnostic panel, where it can provide high-specificity molecular confirmation alongside highly sensitive protein biomarkers such as NT-proBNP and IL-6.

Interplay with Inflammation, Oxidative Stress, and Myocardial Remodeling

Our analysis demonstrated robust positive correlations between peripheral LIF gene expression and circulating NT-proBNP, IL-6, sST2, and MDA (Figure 4). Both LIF and IL-6 signal through receptor complexes containing the common gp130 signal-transducing subunit, which initiates downstream Janus kinase/signal transducer and activator of transcription (JAK/STAT), mitogen-activated protein kinase (MAPK), and phosphatidylinositol-3 kinase/protein kinase B (PI3K/Akt) signaling cascades.18,25,26,27,28 In cardiac myocytes and fibroblasts, activation of the LIF–LIFR–gp130–STAT3 axis regulates genes governing cellular survival, hypertrophy, extracellular matrix synthesis, and cytoprotection against oxidative injury.27,30

The positive correlation between LIF expression and MDA—a validated marker of lipid peroxidation—aligns with evidence that reactive oxygen species activate redox-sensitive transcription factors, such as nuclear factor-κB (NF-κB) and STAT3, which directly drive the transcription of IL-6 family cytokines, including LIF.19,29 Furthermore, the correlation with sST2—an established marker of mechanical stretch and myocardial fibrosis20,21—corroborates earlier work by Vasan et al., Polidori et al., and Weinberg et al., who documented the coordinated upregulation of inflammatory cytokines, oxidative markers, and ST2 in progressive heart failure.22,23,24 Together, these correlations indicate that peripheral blood LIF transcription is tightly integrated into the broader pathophysiological network of oxidative stress, neurohormonal strain, and adverse matrix remodeling in HF.

It is noteworthy that obesity and hypertension were significantly more prevalent in the HF group than in the control group (Table 2). Both conditions represent states of chronic low-grade systemic inflammation and endothelial dysfunction, which may independently stimulate IL-6 and LIF transcription. Although multivariable regression adjusted for these clinical covariates, residual confounding related to metabolic syndrome cannot be entirely ruled out and should be considered when interpreting these molecular profiles.

Multivariable Modeling, Shared Signaling Variance, and Mechanistic Implications

A central finding of this study is that although LIF gene expression demonstrated a strong association with HF in unadjusted (OR = 15.123) and partially adjusted models (OR = 3.248 in Model 3), this relationship was attenuated to non-significance (OR = 0.625, p = 0.415) following full adjustment for NT-proBNP, IL-6, sST2, and MDA (Table 4). Collinearity diagnostics confirmed that this attenuation was not an artifact of severe multicollinearity, as all VIF values were < 5.0 (Figure 5).

From a biological standpoint, because LIF and IL-6 share the gp130 signaling receptor and converge upon identical intracellular cascades (JAK/STAT and MAPK),18,31 adjusting for circulating IL-6 and downstream markers of cardiac stress (NT-proBNP) and remodeling (sST2) partitions out the shared biological variance. In statistical terms, because these biomarkers lie on the same overarching pathophysiological pathway, adjusting for downstream effectors may introduce overadjustment bias. Thus, the loss of independent statistical significance in Model 4 does not diminish the biological importance of LIF; rather, it confirms that peripheral LIF transcription is an integral component of the coordinated gp130 cytokine and stress-response network in heart failure.

Clinical Implications and Future Research Directions

The quantification of peripheral blood mRNA transcripts via RT-qPCR highlights the feasibility of non-invasive molecular profiling in clinical cardiovascular research. Although peripheral LIF gene expression is not intended to replace established biomarkers like NT-proBNP, its high diagnostic specificity suggests potential utility in multimarker risk stratification algorithms or as a pharmacodynamic marker for emerging therapies targeting gp130/STAT3 signaling pathways. Future research should combine transcriptomic assays with high-sensitivity protein quantification (e.g., serum LIF ELISA) to establish whether peripheral mRNA abundance directly translates into elevated circulating cytokine concentrations. Additionally, prospective cohort studies are needed to evaluate the prognostic value of LIF expression for predicting long-term adverse events, such as HF rehospitalization and cardiovascular mortality.

Study Strengths and Methodological Limitations

The strengths of this investigation include a well-characterized case–control cohort with robust age and sex matching, comprehensive simultaneous evaluation of multiple biomarker domains (cardiac wall stress, inflammation, lipid peroxidation, and remodeling), and rigorous RT-qPCR and ELISA methodologies. Nevertheless, several methodological limitations must be acknowledged:

1. Study Design and Causality: The cross-sectional case–control design precludes any determination of temporality or causal inference regarding LIF expression and HF pathogenesis.

2. Geographic and Phenotypic Generalizability: Participants were recruited from tertiary centers in southern Kerala; thus, validation in diverse multi-ethnic cohorts is warranted. Furthermore, detailed clinical stratification by left ventricular ejection fraction phenotypes—specifically HF with preserved (HFpEF), mildly reduced (HFmrEF), and reduced ejection fraction (HFrEF)—and New York Heart Association (NYHA) functional class was not available for all subjects, precluding phenotype-specific subanalyses.

3. Circulating Protein Concordance: Circulating LIF protein levels were not measured in parallel; given post-transcriptional and translational regulation, mRNA expression cannot be assumed to be identical to circulating protein concentrations.

4. Cellular Origin and Globin Transcripts: Total RNA was extracted from unseparated whole blood without globin mRNA depletion or peripheral blood mononuclear cell (PBMC) sorting; therefore, the precise cellular origin of circulating LIF transcripts could not be resolved.

5. Methodological Considerations: Formal assessment of RNA Integrity Numbers (RIN) and multi-gene reference stability algorithms was not performed under complete MIQE guidelines. Furthermore, detailed information on guideline-directed medical therapies (e.g., ACE inhibitors/ARBs, ARNIs, beta-blockers, SGLT2 inhibitors) was not universally available, and their potential modulating effects on cytokine transcription could not be adjusted for.

6. Statistical Validation: The diagnostic ROC metrics were derived from the discovery cohort without external cross-validation, and multiple comparisons were performed without formal alpha adjustments (e.g., Bonferroni correction), which may increase the risk of Type I error.