The impact of blood pressure and treatment on long-term survival in hypertensive patients with high risk for cardiovascular disease using data from the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT)

The ASCOT randomised factorial trial compared calcium channel blocker (CCB) based-therapy versus beta blocker (BB) based-therapy and statin versus placebo. 19,342 hypertensive patients were recruited between 1998 and 2001 and followed for a median of 5.5 years. Primary results were published in 2003 and 2005. A total of 8,580 British ASCOT patients were followed-up for a median of 17.4 years to the end of January 2018, by which time 4040 deaths had occurred, 1,402 from cardiovascular (CV) causes. This thesis analysed the impact of randomised treatment and blood pressure on long-term mortality in this subset of patients and consists of three main sections. The effect of randomised treatment on long-term survival was assessed, taking into consideration potentially non-proportional treatment effects over time and competing risks from different causes of death. Results showed that statin-therapy reduced coronary heart disease (CHD) mortality compared to placebo (hazard ratio [HR]=0.76, p=0.018) and CCB-based treatment reduced stroke mortality compared to BB-based treatment (HR=0.73, p=0.011). Several alternative components of blood pressure recorded at baseline were compared for their ability to predict long-term CV mortality. Each was strongly associated with CV mortality and their relative association attenuated with age. While systolic and pulse pressure (PP) were the strongest single predictors, PP had the clearest continuous monotonic relationship with risk, and was the stronger predictor in older subjects. Repeated blood pressure measurements collected during the trial were used to investigate how features of blood pressure profiles relate to and predict CV-related mortality, e.g. within-subject mean blood pressure, variability and rate of change over time. Factors influencing blood pressure level and variation were investigated. Landmark survival analyses showed again that PP was the most useful summary measure, and both its mean and its variability were independently associated with risk of CV mortality. A clinically useful risk score model was developed containing mean PP and the coefficient of variation (COV) for PP, along with key risk factors. Overall, this thesis provides useful insights into the impact of treatments on CV mortality risk in the long-term and how blood pressure relates to CV mortality risk in the long-term.