Abstract P020: 1 H Nmr Metabonomic Profiling Identifies Novel Biomarkers for Cardiovascular Mortality

Background Cardiovascular disease (CVD) is the leading cause of death world-wide. Existing tools for identification of CVD risk have modest predictive power and discrimination. New biomarkers for CVD prediction are urgently needed. Hypothesis We have used 1 H NMR metabonomic profiling of serum to identify novel biomarkers for incident CVD. Methods We investigated 9,179 men and women participating in the London Life Sciences Population study. Participants were recruited between 2002-2008 from the lists of 58 GPs in London UK, and assessed for baseline CVD risk factors, including smoking, body mass index, waist circumference, blood pressure, fasting glucose and lipid profile. Participants were followed for CVD mortality to July 2011 (mean 5.5 years per person). 1 H NMR metabonomic profiling was done on baseline serum sample by 12T Bruker spectrometer, with quantification of 44 fatty acid and low molecular weight markers. Results There were 161 CVD deaths. Compared to survivors, people with CVD death were older, had higher prevalence of type-2 diabetes and cigarette smoking, higher blood pressure, body mass index, glucose, and lower total and HDL cholesterol (Table). Framingham risk scores were higher in cases with CVD death than survivors (18.7±0.9% vs. 11.4±0.7%, P=10 -31 ). We found six NMR metabolites associated with incident CVD at P<0.05; odds ratios (95%CI) for CVD per 1SD increase in biomarker ranged from 1.19 (1.07-1.33, P=0.007) to 1.65 (1.44-1.89, P=3x10 -13 ). The associations of NMR metabolites with CVD were not materially changed by adjustment for age, gender and Framingham risk score. The AUC for prediction of incident CVD was 0.74 using Framingham risk score alone, and 0.77 with incorporation of metabonomic measures. Conclusions We identified six novel biomarkers for incident CVD; these are independent of known CVD risk factors and may improve CVD risk prediction. Our findings demonstrate the potential utility of metabonomic profiling for biomarker discovery and risk stratification. CVD deaths Survivors P Total no of people 161 9018 Age (years) 61.6 (0.5) 53.4 (0.4) 0.001 Male gender (%) 98.2 85.8 <0.001 Smoking (%) 73 49 <0.001 Diabetes (%) 49 20 <0.001 Systolic blood Pressure (mmHg) 144.7 (22.7) 134.0 (19.2) <0.001 Body mass index (kg/m 2 ) 28.4 (5.6) 27.3 (4.3) 0.002 Total cholesterol (mmol/L) 4.8 (1.2) 5.2 (1.1) <0.001 HDL cholesterol (mmol/L) 1.20 1.25 0.064 Glucose (mmol/L) 7.1 5.9 <0.001 Results provided as mean or %.