Cardiovascular disease (CVD) is one of the leading causes of mortality worldwide, with an incidence that continues to rise. This not only poses a serious threat to individuals’ health and quality of life but also presents a major challenge to public health systems (1–3). Its etiology is multifactorial, encompassing lifestyle behaviors, metabolic abnormalities, genetic predisposition, and underlying comorbidities (4–7). Cardiovascular disease and kidney disease are closely interrelated (8, 9), They exhibit a high degree of interdependence; cardio-renal interactions can precipitate a vicious spiral of organ dysfunction, marked by stepwise worsening of disease severity and a sharp increase in the risk of adverse outcomes. Accordingly, the American Heart Association (AHA) has integrated the cardiovascular and renal domains and introduced the concept of cardiovascular–kidney–metabolic (CKM) syndrome (10), The introduction of this concept has played an exceptional role in preventing disease onset, mitigating or slowing disease progression, and improving clinical outcomes. It underscores the necessity of investigating cardiovascular disease in conjunction with renal and metabolic disorders.

The triglyceride–glucose index (TyG) is widely regarded as a simple surrogate marker of insulin resistance (11); Research on the TyG index in cardiovascular and cerebrovascular diseases has expanded rapidly in recent years, confirming its important role in disease onset, progression, and prognosis (12–14). Compared with the euglycemic–hyperinsulinemic clamp or the homeostatic model assessment of insulin resistance (HOMA-IR), the TyG index can be calculated in virtually all clinical settings at minimal additional cost, and it has been closely associated with the incidence of cardiovascular and renal events. Therefore, the TyG index and TyG-based composite indicators offer a practical approach for detecting subclinical insulin resistance not only in patients with established type 2 diabetes, but also in apparently non-diabetic individuals who may already carry an excessive cardio-renal risk burden.The neutrophil-to-lymphocyte ratio (NLR) reflects systemic inflammation and immune status (15). The TyG-NLR, formed by the combination of the two, holds significant importance for evaluating metabolic disorders and inflammatory responses. Several studies have shown that metabolic imbalance and inflammatory responses act in concert to exacerbate atherosclerosis, myocardial remodeling, and renal injury (16). Research on composite indices incorporating the TyG index has deepened in recent years. Combinations of TyG with measures such as waist circumference (WC), body mass index (BMI), and waist-to-height ratio (WHtR) are widely used to predict and evaluate the onset and prognosis of cardiovascular disease, and these combined indices provide greater discriminative value than the TyG index alone. The C-reactive protein–triglyceride–glucose index (CTI) integrates information on lipid status, glucose metabolism, and inflammation, and shows potential predictive value for cardiovascular events and deterioration of renal function (17, 18).

Most existing studies have focused on binary outcomes (“presence or absence” of disease), while investigations addressing the ordered progression of disease severity remain limited. Moreover, comparative studies of the two composite indices—TyG-NLR and CTI—within the same population, using *disease severity* as a shared clinical endpoint, are scarce. This gap has hindered their broader application in practical risk stratification and model enhancement evaluation.

Based on this evidence gap, the present study aims to systematically assess the associations of TyG-NLR and CTI with the severity of cardiorenal disease in a real-world clinical population. By employing “Severity of Disease” as an outcome more closely aligned with clinical decision-making, this study seeks to validate and compare the value of these two easily accessible composite indices in cardiorenal risk assessment, thereby laying a methodological and evidential foundation for future multicenter and population-based studies.

This study stratified type 2 diabetes mellitus (T2DM) patients according to the CKM staging system to further categorize their cardiorenal status and investigate the effects of the TyG-NLR and CTI indices on cardiorenal function. Using real-world inpatient data in a retrospective design, we found that composite indicators reflecting systemic inflammation and insulin resistance were independently associated with the severity of cardiorenal impairment in this population. Notably, TyG-NLR exhibited a relatively steeper risk gradient and modest improvements in discrimination and calibration, and it yielded slightly higher net clinical benefit across a range of decision thresholds. Given that the absolute AUC remained in the moderate range and the increase in Nagelkerke R² was small, the added value of TyG-NLR should be interpreted as modest but potentially useful for early risk assessment.

Inflammatory markers play a crucial role in cardiovascular disease, a concept that has been well established in numerous previous studies, For example, Kaptoge S and colleagues found that C-reactive protein (CRP) concentrations are consistently associated with the development and prognosis of cardiovascular disease (19). Xia and colleagues likewise reported that the systemic immune-inflammation index (SII) and the systemic inflammation response index (SIRI) are closely associated with cardiovascular mortality and all-cause mortality (20); In addition, the study by Haffner SM also substantiated the impact of inflammatory status on cardiovascular outcomes (21). In a prognostic study of heart failure, the NLR was identified as a practical tool for risk stratification and screening of high-risk patients in routine clinical practice (22). The mechanisms by which inflammation leads to cardiac injury are likely multifactorial. Inflammatory cytokines can induce cardiomyocyte apoptosis and even contribute to ventricular remodeling (23);In addition, inflammatory responses can disrupt lipid metabolism for example, by impairing reverse cholesterol transport thereby diminishing anti-atherosclerotic capacity and further accelerating the onset and progression of cardiovascular disease (24).

Insulin resistance (IR) also exerts an indispensable impact on cardiorenal disease. The onset of IR is often accompanied by dysregulation of metabolic factors, such as blood lipids and glucose that are established risk factors for cardiorenal disorders (25);Concurrently, IR impairs vascular vasodilation and exacerbates arteriosclerosis (26);IR also activates inflammatory responses (27), which, together with its vascular effects, promotes plaque formation. Insulin resistance (IR) is not only a metabolic abnormality but also a direct driver of cardiovascular and renal injury. First, IR induces endothelial dysfunction by reducing nitric oxide bioavailability, increasing oxidative stress, and enhancing endothelin-1 activity, thereby impairing vasodilation (28, 29). These alterations accelerate arterial stiffness and the formation of atherosclerotic plaques. Second, IR disrupts reactive oxygen species (ROS) homeostasis in the myocardium, leading to cumulative superoxide-mediated injury or cellular dysfunction. This imbalance contributes to mitochondrial stress, cardiomyocyte hypertrophy, and interstitial fibrosis, ultimately resulting in adverse cardiac remodeling and a heart failure phenotype (30). In the kidney, insulin and hyperinsulinemia increase tubular sodium reabsorption and glomerular hyperfiltration, creating intraglomerular hypertension and mesangial expansion, which in turn promote proteinuria and progressive loss of renal function (31).Moreover, IR activates the sympathetic nervous system and the renin–angiotensin–aldosterone system, causing elevated blood pressure, increased volume load, and further hemodynamic stress on both the heart and kidneys. IR is also accompanied by low-grade chronic inflammation; adipose-derived cytokines such as TNF-α and IL-6 impair insulin signaling while fostering a pro-inflammatory and pro-thrombotic microenvironment. In recent years, research on the surrogate index TyG has deepened, and its combinations with other risk factors have markedly improved discrimination for diagnosis, incidence, and prognosis—e.g., TyG-BMI (32), TyG- waist circumference (33), TyG-ABSI (34). Accordingly, TyG-NLR integrates metabolic and immuno-inflammatory dimensions into a single construct. We found that higher TyG-NLR quartiles were closely associated with increased disease severity (Q4 vs Q1 OR = 2.14), supporting the notion that synergistic metabolic–inflammatory stress may accelerate injury to cardiac and renal tissues. This composite mechanism may partly explain the observed nonlinear exposure–response curve: risk rises steeply up to approximately TyG-NLR ≈ 16.6 and then plateaus, suggesting saturation of pathophysiological burden. By contrast, CTI exhibited an approximately linear dose–response relationship, indicating a more stable—and less dynamic—reflection of systemic inflammation. Although CTI was also significantly associated with disease severity (Q4 vs Q1 OR = 1.59), its overall discriminative ability was inferior to that of TyG-NLR.

Compared with previous studies, our findings extend the existing evidence linking TyG and NLR separately to cardiovascular and renal outcomes (15, 35–37). By evaluating the composite indices TyG-NLR and CTI against an ordered severity endpoint and comparing them within a unified analytic framework, we show that TyG-NLR may outperform CTI in quantifying the integrated metabolic inflammatory burden associated with cardiorenal outcomes.

Our decision curve analysis demonstrated that TyG-NLR provided meaningful net clinical benefits within a threshold probability range of 0.05–0.35—a range commonly encountered in real-world screening and early intervention decision-making scenarios. Therefore, incorporating TyG-NLR into electronic medical record (EMR) systems may assist clinicians in identifying high-risk hospitalized patients with diabetes who would benefit from intensified cardiorenal protective strategies. Moreover, the incremental improvement of the TyG-NLR model (ΔNagelkerke R² = 0.02, LR χ² = 50.8, P < 10^-10) suggests that this index offers additional prognostic information beyond traditional factors such as hypertension, albumin, and uric acid. These findings support integrating TyG-NLR into comprehensive risk prediction frameworks.

Several limitations should be acknowledged. First, the retrospective single-center design precludes causal inference and may introduce residual confounding, although we adjusted for key demographic and metabolic covariates. Therefore, we avoided drawing strong causal conclusions and interpreted the findings primarily as epidemiological associations. Second, Second, we were unable to include several important factors, such as glycated hemoglobin (HbA1c), physical activity, dietary habits, and medication adherence, which may influence both inflammation–metabolic indices and cardio-renal outcomes. The absence of these variables should be regarded as a potential source of unmeasured confounding, and future studies incorporating these parameters are warranted to verify the robustness of our findings. Although estimated glomerular filtration rate (eGFR) was collected during data acquisition and used to verify the accuracy of chronic kidney disease (CKD)–related diagnoses in our cohort, it was not entered as an independent covariate in the final statistical models. The main reason is that eGFR constituted part of the disease classification framework rather than an exposure or confounder to be adjusted for, and including it could raise concerns about model overadjustment and collinearity with the outcome definition. Third, the cross-sectional assessment of biomarkers and disease severity restricted longitudinal inference; prospective validation is needed to confirm their temporal predictive value. Future research should explore whether dynamic changes in TyG-NLR or CTI during hospitalization or follow-up could predict the onset of cardiovascular or renal events.

This study demonstrates that both TyG-NLR and CTI are independently associated with the severity of cardio-renal impairment among patients with type 2 diabetes. Among the two, TyG-NLR exhibited a steeper risk gradient and provided a modest incremental predictive value. Prospective, multicenter longitudinal studies are needed to further validate their predictive utility and potential therapeutic implications across diverse populations.