Background Heart Failure (HF) is a global epidemic with multiple prevalent risk factors, resulting in a large at-risk population. Within this population, prevention programmes must be targeted to those at highest risk of HF development. However, the incidence of HF within a prevention service remains unknown, as does the phenotype of those at-risk patients most likely to develop HF.

Aim This study aimed to determine the incidence of stage C HF in an asymptomatic population with risk factors (stage A) and/or structural/functional cardiac abnormalities (stage B) in the STOP-HF prevention programme (Ledwidge et. al, 2013), and, within this population, to characterise the phenotype of patients most likely to progress to stage C (symptomatic) HF.

Methods 2037 patients in the STOP-HF service were characterised as stage A or stage B HF. Baseline echocardiographic parameters, BNP, blood pressure, weight and renal function, among other indices, were compared between the cohort that progressed to stage C and the population that did not.

Results 86 patients developed Stage C HF during a median follow up of 4.8 years (overall incidence of 4.2%) with an incidence rate of 8.3 per 1000 patient years. Strong clinical univariate predictors of progression-risk included older age (69.5±8.9 vs 63.4±10.6 years, p<0.001), male gender (59.3 vs 44.9%, p=0.012), higher body mass index (BMI) (29.8 vs 28.3, p<0.001), higher BNP (134.6±183.6 vs 39.6±66 pg/ml, p<0.001) and higher creatinine (95.7±34.7 vs 84.6±21.3 umol/L, p=0.01) at baseline. Echocardiographic features included lower baseline EF (57.8±13.9 vs 66.1%±7.8%, p<0.001), higher left atrial volume index (40.6±14.1 vs 28±9.6, p<0.001), left ventricular mass index (118.1±27.2 vs 95±46.7, p<0.001), E/E’ ratio (10.5±4.1 vs 8.1±3.1, p<0.001) and left atrial diameter (44±5.9 vs 37.2±5.7 mm, p<0.001) in patients who subsequently transitioned to stage C. The prevalence of diabetes (47.7 vs 33.5%, p=0.01), peripheral vascular disease (PVD) (9.3% vs 2.3%, p<0.001), valvular heart disease (VHD) (4.7 vs 0.82%, p=0.003), atrial fibrillation (17.4 vs 4%, p<0.001) and prior myocardial infarction (36 vs 7.2%, p<0.001) were significantly higher at baseline in patients who subsequently transitioned to stage C than in those that did not. Multivariate analysis identified baseline BMI, BNP, and reduced EF as independent predictors of the development of stage C HF. Baseline BNP level was the strongest predictor of transition to Stage C, with a value of 60.3 pg/ml determined as the optimal level by which to differentiate high- and low-risk patients.

Conclusions The incidence rate of stage C HF in the STOP-HF prevention programme is 8.3 per 1000 patient years. Among a number of independent predictors identified, the strongest indicator of transition risk is baseline BNP. These findings support the use of BNP measurement to more tightly define HF risk in an asymptomatic at-risk population.