Glycometabolic burden and myocardial involvement: a relational two-axis hypothesis

Diabetic cardiomyopathy traditionally described myocardial abnormalities in diabetes without major alternative explanations, such as coronary artery disease or hypertension. The 2024 Heart Failure Association of the European Society of Cardiology consensus broadened this concept to diabetic myocardial disorder, recognizing that diabetes rarely acts alone. Obesity, hypertension, chronic kidney disease, and coronary artery disease often coexist and may jointly contribute to myocardial impairment (1). Community-based longitudinal research also linked persistent or worsening diastolic dysfunction to subsequent heart failure, supporting attention to myocardial abnormalities before clinical presentation (2).Diabetes is defined by clinically useful glycemic thresholds, but these thresholds need not correspond to discrete myocardial transitions. The ACE 1950 study identified associations between glycated hemoglobin (HbA1c) and subclinical cardiac abnormalities, including below the diagnostic threshold for diabetes (3). We propose a relational two-axis hypothesis that treats paired metabolic-myocardial history as a prespecified target of longitudinal inference. It asks whether preceding paired trajectories retain reproducible information about subsequent myocardial progression beyond contemporaneous measurements and established clinical factors.Longitudinal evidence supports a connection between metabolic exposure and myocardial disease. In the Atherosclerosis Risk in Communities (ARIC) study, prediabetes and diabetes were associated with incident high-sensitivity cardiac troponin T elevation over six years (4). The Coronary Artery Risk Development in Young Adults (CARDIA) study related long-term glycemic abnormalities and insulin resistance trajectories to midlife ventricular structure and function (5). These findings support exposure-related myocardial involvement rather than biological independence.Within conventional glycemic categories, however, cardiac phenotype and prognosis remain heterogeneous. A pooled analysis stratified heart failure risk among adults with prediabetes or diabetes using a multimarker profile (6). An analysis of the Systolic Blood Pressure Intervention Trial found that prediabetes combined with elevated or rising cardiac biomarkers was associated with higher incident heart failure risk among adults with hypertension without diabetes (7). (11).These studies do not establish independent trajectories, but they motivate direct testing of temporal metabolic-myocardial relationships.We conceptualize glycometabolic burden and myocardial involvement as separately characterized, continuous dimensions. Glycometabolic burden encompasses the severity and duration of dysglycemia and insulin resistance, rather than diagnostic category alone. Prediabetes and diabetes remain useful clinical anchors, but exposure within each category is heterogeneous. Myocardial involvement comprises objectively detectable injury, stress, and structural or functional impairment.These manifestations are overlapping indicators, not a fixed sequential cascade. Neither axis constitutes a disease stage. Their joint position describes the paired metabolic-myocardial phenotype.The hypothesis is that these biologically connected dimensions need not maintain a fixed relationship across individuals or over time. Separate characterization refers to measurement, not statistical or biological independence. The terms myocardial-predominant and metabolic-predominant are illustrative relational descriptors, not validated phenotypes, clinical categories, or mechanistic endotypes (Figure 1). Myocardial positioning does not itself establish diabetic causality.The proposal builds on existing frameworks for cardiometabolic and heart failure assessment.Cardiovascular-kidney-metabolic (CKM) staging integrates multiorgan risk and disease, including pre-heart failure, without requiring synchronous metabolic and cardiac progression (12). Pre-heart failure, or Stage B heart failure, identifies asymptomatic structural disease, increased filling pressures, or qualifying biomarker abnormalities (13). Diabetic myocardial disorder recognizes heterogeneous myocardial dysfunction in diabetes and contributions from coexisting conditions (1,14). The distinction is therefore analytical rather than diagnostic. The framework prespecifies a relational-history hypothesis for testing within, rather than instead of, existing clinical frameworks.Cross-sectionally, individuals with similar glycometabolic burden but different myocardial involvement, or similar myocardial involvement but different glycometabolic burden, can be compared within and across clinical categories. Such comparisons provide a starting point for investigating cumulative exposure, accompanying conditions, myocardial susceptibility, and measurement factors without establishing causality. They also permit comparisons across the diagnostic threshold for diabetes without treating a change in diagnostic category as evidence of a myocardial transition.Longitudinally, repeated paired assessment tracks the relationship itself, including changes within an unchanged CKM stage. Glycometabolic measurements might improve while myocardial abnormalities persist, or myocardial involvement might worsen despite apparently stable metabolic measurements. Detailed CKM assessment with biomarkers and imaging can capture the same observations. The proposed contribution is therefore not exclusive analytical capability, but prespecification of the paired trajectory as an object of inference. Whether this improves biological understanding, prediction, or research design remains empirical.The mechanisms considered here do not validate a distinct two-axis construct. Their narrower role is to provide biological plausibility for why myocardial response may not be an instantaneous or uniform function of current metabolic measurements. Heterogeneity may reflect exposure history, myocardial susceptibility or adaptive capacity, and additional cardiovascular stressors.Current glycemia does not fully represent the duration or consequences of prior metabolic stress.Ding et al. identified an endothelial CaMK2α/O-GlcNAcylation feedback mechanism that sustained injurious small extracellular vesicle release after glucose normalization in experimental diabetes (15).This finding supports persistence of prior exposure effects, not independence from glycometabolic burden. Apparent discordance between improving glycemic measurements and persistent myocardial abnormalities may therefore reflect metabolic memory. Current glycemia must be distinguished from cumulative exposure when interpreting paired trajectories.Tissue-specific signaling may also modify the myocardial response to systemic metabolic disturbance. Battiprolu et al. showed that cardiomyocyte-specific FoxO1 deletion protected cardiac structure and function in high-fat-fed mice despite persistent systemic metabolic abnormalities (16).Protection was accompanied by preserved myocardial insulin responsiveness and reduced cardiac lipid accumulation. The experiment supports a modifiable myocardial response, but does not establish a naturally occurring protective phenotype in humans.Additional cardiovascular stressors may amplify or alter the myocardial consequences of metabolic disturbance. Schiattarella et al. induced an experimental heart failure phenotype with preserved ejection fraction through combined metabolic and hypertensive stress in mice (17). Inducible nitric oxide synthase impaired IRE1α/XBP1 signaling, and targeted intervention improved the phenotype.In patients evaluated for suspected coronary disease, Taqueti et al. linked coronary microvascular dysfunction with myocardial injury, diastolic dysfunction, and subsequent heart failure with preserved ejection fraction (18). This observational cohort excluded flow-limiting coronary disease and reduced ejection fraction, but did not test two-axis trajectories.Relatively greater myocardial involvement for a given measured metabolic exposure could reflect accumulated exposure, heightened myocardial susceptibility, or additional cardiovascular stressors.Less detectable involvement despite substantial metabolic exposure could reflect shorter exposure, preserved adaptation, fewer additional stressors, or abnormalities below current detection thresholds.These explanations make such configurations starting points for mechanistic investigation rather than causal endotypes. The biological proposition is therefore a history-dependent and context-dependent relationship, not independence between metabolic and myocardial disease.Neither conceptual axis currently has a validated comprehensive score. One illustrative minimal implementation in people without clinical heart failure pairs time-weighted mean HbA1c over a defined historical window with standardized left ventricular global longitudinal strain (GLS). They are selected for feasibility, continuity, interpretability, and serial assessment, not because they are optimal or comprehensive. The HbA1c mean would equal the area under its time curve divided by the window duration, while GLS could be reported as absolute magnitude. The central empirical hypothesis is that, after accounting for current and cumulative glycometabolic exposure, treatment, measurement variability, and major competing cardiac causes, preceding paired metabolic-myocardial trajectories retain reproducible information about subsequent myocardial progression beyond contemporaneous measurements alone. This proposition is distinct from simply plotting two dimensions and can be falsified.For a prespecified population, measurement pair, follow-up interval, comparator model, and minimum meaningful effect, the central hypothesis would be falsified in that setting if, after prespecified accounting for measurement error, cumulative glycometabolic exposure, expected response delay, treatment, and competing cardiac disease, preceding paired trajectories show no reproducible incremental contribution beyond contemporaneous measurements in a conventional model using the same information, with confidence intervals excluding an effect at or above the prespecified minimum meaningful magnitude. An imprecise null result would not suffice. A twodimensional description could remain, but it would no longer have a demonstrated rationale as an additional explanatory construct.Predictive evaluation must distinguish the value of the underlying measurements from that of the relational representation. Comparators should include an established benchmark such as PREVENT (19), a conventional model containing all axis inputs, and a prespecified two-axis implementation.The latter two must use identical information and comparable complexity, with appropriate nonlinear relationships, interactions, and longitudinal terms allowed in the conventional model. External validation should assess discrimination, calibration, and prediction error. If the conventional approach answers the same question equally well and the relational representation provides no gain in reproducibility, interpretation, or prospective study design, the framework would have no demonstrated added value.The rationale for proposing the hypothesis before longitudinal Abbreviations: CKM, cardiovascular-kidney-metabolic; G, glycometabolic burden; M, myocardial involvement.