ObjectiveECG markers of heart failure (HF) with preserved ejection fraction (HFpEF) are lacking. We hypothesised that the Cornell product (CP) is a risk marker of HFpEF and has prognostic utility in HFpEF.MethodsCP =[(amplitude of R wave in aVL+depth of S wave in V3)×QRS] was measured on baseline 12-lead ECG in a prospective Asian population-based study of 606 healthy controls (aged 55±10 years, 45% men), 221 hypertensive controls (62±9 years, 58% men) and 242 HFpEF (68±12 years, 49% men); all with EF ≥50% and followed for 2 years for all-cause mortality and HF hospitalisations.ResultsCP increased across groups from healthy controls to hypertensive controls to HFpEF, and distinguished between HFpEF and hypertension with an optimal cut-off of ≥1800 mm*ms (sensitivity 40%, specificity 85%). Age, male sex, systolic blood pressure (SBP) and heart rate were independent predictors of CP ≥1800 mm*ms, and CP was associated with echocardiographic E/e′ (r=0.27, p<0.01) and left ventricular mass index (r=0.46, p<0.01). Adjusting for clinical and echocardiographic variables and log N-terminal pro B-type natriuretic peptide (NT-proBNP), CP ≥1800 mm*ms was significantly associated with HFpEF (adjusted OR 2.7, 95% CI 1.0 to 7.0). At 2-year follow-up, there were 29 deaths and 61 HF hospitalisations, all within the HFpEF group. Even after adjusting for log NT-proBNP, clinical and echocardiographic variables, CP ≥1800 mm*ms remained strongly associated with a higher composite endpoint of all-cause mortality and HF hospitalisations (adjusted HR 2.1, 95% CI 1.2 to 3.5).ConclusionThe Cornell product is an easily applicable ECG marker of HFpEF and predicts poor prognosis by reflecting the severity of diastolic dysfunction and LV hypertrophy.