This investigation elucidates a robust and independent relationship between the combined TyG and FI indices and the risk of both CVD and stroke. Utilizing two large, nationally representative cohorts—CHARLS and NHANES—our analysis demonstrates that elevated TyGFI levels are consistently associated with increased odds of adverse cardiovascular outcomes. Furthermore, this association exhibits a clear dose–response pattern, as participants within the highest TyGFI quartiles experienced significantly greater risk elevations. Importantly, these findings remained stable after rigorous adjustment for a wide array of demographic, clinical, and lifestyle confounders. The observed associations were further corroborated through subgroup analyses, which revealed consistent effects across diverse demographic and clinical strata. Collectively, these results emphasize the potential clinical relevance of the TyGFI as an integrative risk marker for cardiovascular disease prevention and management. The association between higher TyGFI levels and increased risks of CVD and stroke likely reflects the combined effects of metabolic dysfunction and frailty on vascular health. Individuals in higher TyGFI quartiles had worse metabolic profiles, including higher BMI, glucose, HbA1c, and unfavorable lipid levels—all known risk factors for atherosclerosis. The clear dose–response pattern in both CHARLS and NHANES supports TyGFI as a reliable tool for cardiovascular risk stratification, even after adjusting for various confounders. Subgroup analyses further show that TyGFI can amplify risk in the presence of traditional factors like hypertension, diabetes, and dyslipidemia. The consistent findings across two large and diverse cohorts highlight TyGFI’s broad applicability and potential clinical value.
Extensive research has established the TyG as a reliable surrogate marker for insulin resistance and a significant predictor of adverse cardiovascular and cerebrovascular outcomes. Prior studies have consistently demonstrated that elevated TyG levels are associated with increased risks of ASCVD, stroke, and mortality [9, 31,32,33,34]. Ding et al. (2021) confirmed that individuals with higher TyG levels were more likely to develop cardiovascular events, even in the absence of baseline CVD [35]. Cui et al. further highlighted the prognostic value of TyG, particularly in populations with compromised renal function [36]. In addition, Chen et al. reported a strong association between TyG and both all-cause and cardiovascular mortality [33]. In the domain of stroke research, elevated TyG has also been linked to increased incidence, recurrence, and poor prognosis. Yang et al. identified a significant association between higher TyG levels and the risk of ischemic stroke and its recurrence [37], while Cai et al. demonstrated that TyG was independently associated with in-hospital and ICU mortality in patients with critical stroke [31]. Moreover, longitudinal analyses conducted by Wu et al. and Huang et al. emphasized that persistent or increasing TyG trajectories over time were significantly correlated with elevated stroke risk [17, 38, 39]. These findings underscore the importance of TyG as a long-term indicator of metabolic risk. Emerging evidence has also begun to clarify the potential mechanisms behind these associations. Huo et al. showed that TyG mediated a substantial proportion of the relationship between body mass index and stroke, suggesting its involvement in key pathophysiological pathways linking obesity to vascular injury [19].
Mechanistically, the TyGFI index reflects a synergistic interplay between metabolic dysfunction and systemic physiological decline, both of which are central drivers of cardiometabolic vulnerability. The TyG index, a validated surrogate marker of insulin resistance, is associated with chronic low-grade inflammation, vascular endothelial dysfunction, and mitochondrial impairment. These pathological processes are likely mediated through the activation of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) and Janus kinase/signal transducer and activator of transcription (JAK/STAT) signaling pathways, which induce the expression of pro-inflammatory cytokines such as tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), and C-reactive protein (CRP). Simultaneously, insulin resistance downregulates key mitochondrial regulators including peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α) and sirtuin 1 (SIRT1), leading to impaired oxidative phosphorylation and cellular energy homeostasis [40,41,42,43]. Simultaneously, the FI has been mechanistically linked to dysfunction of cluster of differentiation 4–positive (CD4⁺) T cells, reduced levels of insulin-like growth factor 1 (IGF-1) and dehydroepiandrosterone sulfate (DHEA-S), as well as the accumulation of mitochondrial DNA (mtDNA) damage—all of which contribute to diminished physiological reserve and impaired stress adaptation capacity [44,45,46,47]. Moreover, emerging evidence underscores vascular endothelial dysfunction—defined by impaired nitric oxide (NO) bioavailability, increased arterial stiffness, and premature vascular aging—as a convergent pathological mechanism underlying both insulin resistance and frailty [48]. This mechanistic overlap provides a strong biological rationale for the multiplicative TyG × FI interaction: insulin resistance may amplify the vascular and inflammatory vulnerabilities conferred by frailty, while frailty may, in turn, exacerbate the systemic effects of metabolic stress. Meanwhile, recent findings by He et al. demonstrated that progression in frailty status significantly increased the risk of cardiovascular events, independent of traditional metabolic risk factors [17]. However, few investigations have attempted to combine metabolic and functional indicators into a single index.
Overall, this study exhibits multiple methodological and conceptual strengths that reinforce the robustness, generalizability, and clinical relevance of its findings. Notably, the utilization of two nationally representative and demographically distinct cohorts—CHARLS from China and NHANES from the United States—ensures broad external validity across populations with diverse sociocultural, genetic, and healthcare backgrounds. The consistency of results across these cohorts enhances the credibility of the observed associations and addresses an important limitation in prior cardiovascular research [34, 36, 49, 50].
Moreover, the study advances the field by proposing a novel composite risk indicator—TyGFI—that integrates the well-validated TyG, a surrogate marker of insulin resistance, with FI, a widely accepted measure of cumulative physiological deficits. While both TyG and FI have been independently associated with cardiovascular and cerebrovascular outcomes [17, 51, 52], their combination into a single metric represents a conceptual innovation. This dual-domain approach captures both metabolic and functional deterioration, enabling a more holistic and sensitive assessment of cardiovascular risk, particularly among aging individuals. In addition, the analytical framework is characterized by rigorous adjustment for a comprehensive set of covariates, including demographic, socioeconomic, behavioral, and clinical factors. The application of multivariable logistic regression, restricted cubic spline modeling, and stratified subgroup analyses ensures statistical robustness and allows for the detection of both linear and nonlinear relationships. Besides, the study also contributes to the literature by empirically demonstrating the added predictive value of TyGFI over its individual components. In doing so, it addresses a critical gap identified in previous studies that evaluated isolated metabolic indicators, such as TyG-BMI and TyG-WHtR, without accounting for functional health status [19, 53]. The integrative nature of TyGFI offers a more refined stratification tool for identifying individuals at elevated risk of cardiovascular and cerebrovascular events. Therefore, these strengths underscore the originality and applicability of TyGFI as a multidimensional risk marker, offering significant potential for incorporation into precision prevention strategies for cardiovascular and cerebrovascular disease.
However, despite the methodological rigor and cross-cohort validation, several limitations of this study should be acknowledged, particularly in the context of temporal dynamics and evolving risk factors. First, the analysis was based on baseline measurements of the TyG and FI, which are both time-sensitive indicators. However, growing evidence emphasizes the importance of longitudinal changes and cumulative exposure to these markers. For instance, Wu et al. and Huang et al. demonstrated that persistent elevation or upward trajectories in TyG over time were significantly associated with increased stroke risk in middle-aged and older adults [38, 39]. Similarly, He et al. showed that progression in frailty status over time was closely linked with elevated CVD incidence [17]. Second, although the study employed multivariable adjustment to control for confounding factors, its observational nature precludes definitive causal inference. While the associations identified are robust and consistent across two nationally representative cohorts, unmeasured confounding remains a possibility. Incorporating analytical approaches such as Mendelian randomization or instrumental variable analysis could strengthen causal interpretations, as illustrated by Jiang et al. [54]. Third, although CHARLS and NHANES represent different sociocultural and healthcare contexts, generalizability beyond Chinese and U.S. populations may be limited. Diverse populations with varying genetic backgrounds, dietary habits, and healthcare access may present distinct cardiometabolic trajectories [55,56,57,58,59,60]. Furthermore, different versions of the FI were applied in CHARLS and NHANES, which may introduce measurement variability and limit cross-cohort comparability. Nonetheless, although our primary aim was not to compare absolute frailty levels between different populations, we acknowledge that this heterogeneity may affect the broader generalizability of our findings. To strengthen external validity, future research should prioritize the use of harmonized frailty assessment protocols and consider pooled individual-level data to further validate and refine the TyGFI framework across diverse populations. Lastly, while our findings support the predictive value of the TyGFI as a composite indicator, its integration into clinical practice requires further validation. The multiplicative formulation (TyG × FI) is based on the hypothesis that metabolic and functional impairments interact synergistically to increase cardiovascular risk beyond additive effects. Although supported by epidemiological evidence, this approach remains exploratory. To advance the clinical implementation of the TyGFI index, two complementary methodological directions warrant further exploration. On one hand, modeling the longitudinal trajectories of TyGFI may offer nuanced insights into the temporal dynamics of cardiometabolic risk, thereby facilitating earlier identification of subclinical deterioration and enabling more timely evaluation of therapeutic responses. On the other hand, comparative analyses with additive or weighted composite indices are necessary to determine whether the multiplicative structure of TyGFI delivers superior prognostic utility in real-world settings. In particular, benchmarking TyGFI against widely used cardiovascular risk prediction models—such as the Framingham Risk Score and the ASCVD Risk Calculator—is an important next step to establish its relative performance and clinical relevance. Although such direct comparisons were beyond the scope of the present study, we recognize this as a key limitation and are planning follow-up analyses to evaluate the incremental value of TyGFI relative to these established tools. While the current study does not define prespecified clinical cut-off values, the consistently graded associations observed across TyGFI quartiles provide a practical basis for provisional risk stratification. Building upon this gradient, subsequent studies should employ bootstrap-based receiver operating characteristic (ROC) analyses to derive clinically meaningful thresholds. Furthermore, the added value of TyGFI in risk prediction models should be systematically evaluated through net reclassification improvement (NRI) and decision curve analysis (DCA), both of which quantify improvements in model discrimination and clinical benefit.
In summary, despite certain limitations, this study highlights the TyGFI as a novel and clinically accessible composite marker that integrates metabolic dysfunction and frailty. It demonstrates strong potential for improving cardiovascular and stroke risk stratification, particularly in aging populations.
