Background The STOP HF cohort is an ongoing longitudinal study of 1,040 individuals with cardiovascular risk factors and no known ventricular dysfunction at baseline. The study is located in South Dublin and the Southeast of Ireland. This is an analysis of the deaths reported at the interim analysis conducted in January 2012 (mean follow-up, 4 years, range 3.5–7.0 years).

Methods Univariable analyses compared the profiles of survivors to those deceased. Logistic regression identified the foremost associates of death. Cox’s proportional hazard models estimated the risk of death over time, considering the presence of certain factors.

Results Between January 2005 and April 2008 1,040 individuals were recruited into the STOP HF study (Mean age 65, 49% male, hyperlipidemia 53%, hypertension 47%, CAD 16%, MI 8%, arrhythmia 8%, diabetes 12%, smoker 14%). At the interim analyses 64 deaths were reported, (6.1% of the cohort). Univariable analyses indicated those deceased were older (65 vs. 71 years, p<0.0001) with lower diastolic blood-pressure (p<0.005) and a 17% greater prevalence of diabetes (p<0.0001). Those deceased had biomarker levels more frequently ≥75th percentile for gender in BNP (p<0.005), hsCRP (p<0.0001), Urea (p<0.005), Creatinine (p<0.0001) and ALP (p<0.05) and were more likely to be taking ACE inhibitor (p<0.02) and anti-diabetic therapy (p<0.0001). The final multivariable model identified age (ExpB=1.07, OR 1.03–1.11, p<0.0001), diabetes (ExpB=3.44, OR 1.96–7.00, P=0.001), BNP (ExpB=1.00, OR 1.000–1.007, p=0.075) and hsCRP (ExpB=1.07, OR 1.04–1.11, p<0.000) as associates of mortality (model χ2(4, N=1041)=50.240, p<0.001). Controlling for age, cox’s proportional hazard models indicated that those with diabetes had a 3 times higher risk of death (ExpB=3.117, OR 1.801–5.304, p<0.0001); those with a hsCRP level ≥75th percentile for gender were almost 3.5 times more likely to die (ExpB=3.439, OR 1.995–5.498, p<0.0001) and those with ALP ≥75th percentile for gender were also more likely to die durin the follow-up period (ExpB=1.919, OR 1.096–3.359).

Conclusion Diabetes is associated with in excess of a 3 times increase in the likelihood of death over 4 years follow-up in this population, while other morbid conditions such as coronary disease and atrial fibrillation show no significant association. Despite aggressive cardiovascular risk management through primary and secondary prevention strategies, those with diabetes continue to have a poor prognostic outlook. Observation of biomarkers such as hsCRP and ALP may be of added value in monitoring cohorts with existing cardiovascular risk factors and identifying those that may be sub-clinically manifesting heightened risk of mortality in this setting.