Blood pressure, atrial fibrillation and diabetes: evidence from observational and randomized analyses

The subgroup-specific relationships between blood pressure (BP), risk of atrial fibrillation (AF) and risk of diabetes are unclear. These two chronic diseases have key epidemiologic and clinical similarities, are difficult to examine through traditional cohort studies, and are of high clinical and public health importance. Using observational analyses of electronic health records and meta-analyses of randomized trials, the relationships between blood pressure, atrial fibrillation and diabetes are investigated in this thesis. First, I show that blood pressure is modestly associated with risk of atrial fibrillation in a population free of cardiovascular disease [HR 1.21 95% confidence interval 1.19, 1.22 per 20 mm Hg higher systolic blood pressure (SBP)]. While a non-significant effect of BP lowering on AF in non-heart failure populations is observed in a meta-analysis of randomized trials, this randomized estimate is similar in magnitude to the observational association, suggesting that BP lowering may modestly lower the risk of AF. Second, I show that elevated BP is associated with a higher risk of incident diabetes (HR 1.76 CI 1.56, 198 per 20 mm Hg higher SBP), although the causality of this finding is unclear. Third, I demonstrate in both observational and randomized analyses that the relationship between blood pressure and cardiovascular events appears to be modified by achieved BP in type 2 diabetes, with attenuated proportional reductions at achieved SBP less than 130 mm Hg. However, I also show that BP lowering to levels lower than those recommended in current guidelines (<130 mm Hg SBP) may lower the risk of stroke, renal disease and retinopathy. These findings are of clinical and public health importance, as they suggest that BP lowering therapy can be used to reduce the risk of AF, that elevated BP may be a risk factor for incident diabetes, and that BP lowering in type 2 diabetes to lower levels than currently recommended in guidelines may further reduce the burden of cardiovascular disease in diabetic populations.