Abstract 4144972: Predictors of cardiovascular disease risk and total mortality: Findings from the UK Biobank.

Background: Cardiovascular disease (CVD) continues to remain a major health challenge with several physiological and lifestyle markers associated with its incidence, burden and mortality. Purpose: We aimed to investigate diet and other lifestyle variables in combination with CVD risk markers (clinical and laboratory measurements) to estimate the risk of CVD events, all-cause Mortality, and survival rates over an 18-year period. Methods: Data from the UK biobank was utilised. Participants recruited between 2006 and 2010. Variables included predictor (sex, age, smoking history, and Townsend deprivation), Clinical (BMI, SBP, DBP, HBA1C, triglycerides, TC/HDL ratio, and total cholesterol (TC)) and behavioural (diet, physical activity). Deaths were coded using the ICD, 10th revision. COX proportional hazard models used to analyse each outcome for every predictor variable and models adjusted for confounders. Survival analysis was done with logistic regression models. Outcomes included all-cause mortality and incidence of fatal cardiovascular disease. Results: Over 502,621participants aged between 40-69 years with 54.4% males. 4.6% developed CVDs. High (>8) and low (<7) sleep hours, High BMI, HBA1C, SBP, triglycerides and TC/HDL had a higher CVD incidence risk. Age, Sex, smoking, and Townsend deprivation were independent predictors of mortality with sex and smoking carrying the highest risk. Vegetarians and participants on combined special diets had a lower total mortality. Physical activity of 2-4 hrs and 4-6 hrs weekly had lower mortality compared to no physical activity. However, light or vigorous physical activity had no significant effect on total mortality. Similarly, individuals with no metabolic equivalent task (MET) had an increased mortality compared to those with 12-14 hrs, 14-16hrs, and 16-84hrs MET. High BMI, triglycerides, HBA1C and TC/HDL ratio were associated with significantly decreased survival rates while High HDL and high TC was associated with increased survival rates. Conclusion: Lifestyle factors including physical activity, sleep, and diets were identified as important predictors of all-cause mortality and CVD risk events. The findings further strengthen the evidence on the role of other known modifiable and non-modifiable factors in predicting CVD incidence risk and total mortality. The identified predictors would be useful in the development of risk scoring systems which incorporate lifestyle factors not factored into most CVD risk prediction models.