Insulin Resistance Indices and Subclinical Left Ventricular Systolic Dysfunction in Adults: A Systematic Review and Meta-Analysis.
Source: PubMed, NCBI / U.S. National Library of Medicine
Insulin resistance (IR) is a key characteristic of cardiometabolic disease and may contribute to subclinical ventricular dysfunction before the development of overt heart failure. Studies have reported heterogeneous results regarding the associations between IR indices and myocardial strain. Thus, this review synthesized the available evidence on the relationships between IR indices and subclinical left ventricular systolic dysfunction in adults. We systematically searched PubMed, Embase, Web of Science Core Collection, and the Cochrane Library from inception to 1 December 2025 for observational studies involving adults that reported associations between IR indices and speckle-tracking echocardiography-derived strain parameters. Eligible exposures included commonly used IR indices, such as the homeostatic model assessment of insulin resistance (HOMA-IR), the triglyceride-glucose (TyG) index, the metabolic score for insulin resistance (METS-IR), and lipid-based ratios. The primary outcome was global longitudinal strain (|GLS|). Right ventricular strain was summarized narratively because only one study reported this condition. Adjusted odds ratios (ORs) for subclinical LV systolic dysfunction per unit increase in an IR index and mean differences (MDs) in |GLS| between the highest and lowest IR categories were synthesized using random effects models with Hartung-Knapp adjustment. Four studies reporting fully adjusted ORs per unit increase in an IR index showed a consistent posit
Abstract
Insulin resistance (IR) is a key characteristic of cardiometabolic disease and may contribute to subclinical ventricular dysfunction before the development of overt heart failure. Studies have reported heterogeneous results regarding the associations between IR indices and myocardial strain. Thus, this review synthesized the available evidence on the relationships between IR indices and subclinical left ventricular systolic dysfunction in adults. We systematically searched PubMed, Embase, Web of Science Core Collection, and the Cochrane Library from inception to 1 December 2025 for observational studies involving adults that reported associations between IR indices and speckle-tracking echocardiography-derived strain parameters. Eligible exposures included commonly used IR indices, such as the homeostatic model assessment of insulin resistance (HOMA-IR), the triglyceride-glucose (TyG) index, the metabolic score for insulin resistance (METS-IR), and lipid-based ratios. The primary outcome was global longitudinal strain (|GLS|). Right ventricular strain was summarized narratively because only one study reported this condition. Adjusted odds ratios (ORs) for subclinical LV systolic dysfunction per unit increase in an IR index and mean differences (MDs) in |GLS| between the highest and lowest IR categories were synthesized using random effects models with Hartung-Knapp adjustment. Four studies reporting fully adjusted ORs per unit increase in an IR index showed a consistent positive association with subclinical LV systolic dysfunction. The pooled OR per unit increase in the IR index was 1.36 (95% confidence interval (CI): 1.20 to 1.53; I= 0%). Five studies comparing GLS across IR quartiles reported that participants in the highest quartile had worse LV systolic function than those in the lowest quartile. The random-effects pooled MD in GLS (Q4 vs. Q1) was -1.58 percentage points (95% CI: -2.22 to -0.93), indicating a lower strain magnitude (worse subclinical systolic function) in the highest IR group. Higher IR indices were independently associated with lower GLS values, indicating worse subclinical LV systolic function. The association was directionally consistent across IR metrics, but the supportingevidence remains limited, prospective validation and interventional studies are needed. CRD420251251563, https://www.crd.york.ac.uk/PROSPERO/view/CRD420251251563.
