Intensive systolic blood pressure control and prevention of new onset atrial fibrillation in the SPRINT study: is the association really controversial?

An apparent controversy has arisen about whether intensive systolic blood pressure (BP) control prevented new onset atrial fibrillation (AF) in the Systolic Blood Pressure Intervention Trial (SPRINT). A report based on the publically available SPRINT dataset [1] could not find a statistically significant reduction in AF with intensive systolic BP control, whereas a report from SPRINT investigators at the ECG reading centre who had access to the raw data and ECG recordings [2] did find significantly less new onset AF with intensive systolic BP control.

SPRINT tested the hypothesis that intensive systolic BP control would reduce cardiovascular disease and mortality to greater extent than treatment to the standard systolic BP target in people with hypertension [3]. Because of benefit in the intensive treatment arm, SPRINT was stopped after a median follow-up of 3.26 years [4]. Intensive lowering of systolic BP reduced the rates of the primary outcome, a composite of acute myocardial infarction, acute coronary syndrome not resulting in myocardial infarction, stroke, acute decompensated heart failure and cardiovascular death, by 25% and the risk of all cause death by 27%. We have discussed the SPRINT findings in previous issues of Blood Pressure [5,6].

Hypertension, due to its high prevalence in the general population, is the most important risk factor for the development of AF. Elevated BPs, even in the high normal or pre-hypertensive range, are also associated with increased risk of AF [7]. Left ventricular hypertrophy, serum creatinine, body mass index and history of hypertension, stroke, and coronary disease have all been shown to be independent and highly significant predictors of new onset AF in people with high cardiovascular risk. An analysis of participants in two large outcome trials of similar design (n = 30,424), the Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial (ONTARGET) and Telmisartan Randomised AssessmeNt Study in ACE iNtolerant subjects with cardiovascular Disease studies (TRANSCEND), documented the association between hypertension, or hypertension mediated organ damage, and the risk of new onset AF [8]. The relationship between the history of hypertension and risk of AF persisted despite confounding by coexisting vascular disease or complications of diabetes that were highly prevalent in the patient populations. It is likely that these comorbidities may have escalated the relationship between hypertension and AF.

A working group of the European Society of Hypertension has reviewed the role of high BP in the aetiology, pathophysiology and prediction of AF [9]. They found that up to 90% of patients with AF who participated in large randomised clinical trials of new anticoagulant or antiarrhythmic medications for AF treatment had a history of hypertension. This may have even been an underestimate of the prevalence of hypertension because the study participants did not undergo 24-h ambulatory BP measurements in order to diagnose masked hypertension.

Left ventricular hypertrophy is a strong predictor of AF. In the Losartan Intervention For Endpoint Reduction in Hypertension (LIFE) study, in which all participants had left ventricular hypertrophy (n = 9193) and were randomised to treatment with the angiotensin receptor blocker losartan compared to the beta-blocker atenolol, BP, left ventricular hypertrophy on electrocardiogram (ECG) by Cornell voltage-duration product, heart rate, age, and male gender were all predictors of AF [10]. In LIFE, a decrease in incidence of new onset AF was positively correlated with regression of left ventricular hypertrophy during antihypertensive treatment [11]. A difference in stroke drove the difference in the primary endpoint in favour of losartan over atenolol [10,11]. A reduction in new-onset AF during treatment with losartan compared to atenolol contributed to a reduction in incident stroke, and thus to the difference in the primary composite endpoint in LIFE [12].

Increased left atrial size in the setting of hypertension, as seen in the LIFE participants [13], involves stretching of the atrial myocardial fibres, which leads to reduced electrical contact between atrial myocardial cells, a mechanism that disturbs in electrical impulse transmission and may provoke AF. Conversely, reduced left atrial size during antihypertensive therapy, also observed in LIFE [14], reduces the risk of new-onset AF in hypertensive patients with left ventricular hypertrophy. Pulse pressure in LIFE was equivalent to systolic and diastolic BPs combined in predicting new-onset AF, but when forced into the same statistical model, the pulse pressure was by far the strongest predictor of new-onset AF among the various BP components [15]. The strong association of new-onset AF with pulse pressure suggests that in advanced hypertensive disease, the stiff arteries with increased afterload and stretch of atrial walls drive atrial chamber dilation and retrograde pressure back into the pulmonary veins, and that this may be a key mechanism for promoting the unstable electrical properties that lead to AF, as reported from the Framingham Heart Study [16].

The presence of left ventricular hypertrophy by ECG, age, and coronary disease were all significant predictors of new onset AF in the Valsartan Antihypertensive Long-term Use Evaluation (VALUE) trial. In VALUE, patients with high risk hypertension (n = 15,245) were randomised to treatment with the angiotensin receptor antagonist valsartan compared to the calcium channel blocker amlodipine for prevention of cardiac disease. VALUE was neutral for the primary endpoint which was a composite of cardiac morbidity and mortality. But VALUE also included serial ECGs for detection of new-onset AF and found that new-onset AF increased the risk of congestive heart failure [17]. One explanation for this finding is that people with the advanced cardiac disease cannot tolerate the 20–25% reduction of their cardiac output that results from loss of the atrial component of left ventricular filling. Another explanation is that increased vascular stiffness in the setting of hypertension may lead to increased left ventricular filling pressures, increased left atrial volumes and increased left atrial pressure, resulting in increased risk of congestive heart failure, where AF is the clinical phenotype of this condition. New-onset AF was strongly associated with incident heart failure in these high-risk hypertensive participants in VALUE, whether or not they had diabetes, although those with both hypertension and diabetes had twice the risk of new onset heart failure compared to those without diabetes. As in LIFE [12], angiotensin receptor blocker treatment in VALUE prevented new-onset AF better than the comparator, amlodipine [18].

With respect to the apparent controversy of new onset AF in SPRINT, an accompanying editorial discussed differences between the two reports [1,2] and concluded that whether or not intensive control of systolic BP prevented new onset AF depended on patient selection for the analysis [19]. Consistent with analyses from the other large clinical trials that included serial ECGs discussed above (LIFE, ONTARGET, TRANSCEND, VALUE), the SPRINT investigators identified new-onset AF, by excluding from analysis patients with AF on ECG at baseline and those with missing ECGs at baseline or during follow-up [2]. They did not exclude patients with a history of AF but with sinus rhythm on ECG at baseline. In contrast, the other analysis excluded both patients with a history of AF and those with AF on ECG at baseline. This led to a much lower number of cases of new-onset AF in the latter analysis than in the analysis by the SPRINT investigators, resulting in different conclusions about the effect of intensive BP lowering in prevention of new-onset AF [19]. However, when the new-onset AF results of SPRINT [2] were combined with similar data from two other outcome trials with intensive BP control arms, the Action to Control Cardiovascular Risk in Diabetes (ACCORD) [20] and the Systolic Hypertension in the Elderly Programme (SHEP) [21], the pooled risk ratio reached statistical significance for prevention of new onset AF by intensive BP control in both SPRINT analyses [19]. These supplementary data included in the accompanying editorial [19] showed that prevention of new-onset AF with intensive BP control compared to standard treatment had a hazard ratio of 0.80 (95% CIs, 0.61–0.98, p = 0.047) (1) and of 0.77 (95% CIs, 0.63–0.95, p = 0.01) (2), respectively.

The finding that intensive treatment of SBP prevented new onset AF in SPRINT [2] is consistent with the observations of reduced new-onset AF with systolic BP control in LIFE patients with left ventricular hypertrophy [22] and in high-risk hypertensive patients in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) [23]. The participants in LIFE with the highest risk of new onset AF, those with left ventricular hypertrophy combined with isolated systolic hypertension, had the greatest benefit of systolic BP control in preventing new-onset AF [24].

In conclusion, there should be no doubt that intensive systolic BP control prevents new onset AF in patients with hypertension. Prevention of new onset AF plays an important role in preventing hypertensive complications such as stroke and heart failure. The most recent European and American hypertension guidelines addressed prevention of new onset AF and recommended use of a blocker of the renin-angiotensin-aldosterone system (RAAS) in patients at high risk of new onset AF [25] and treatment of hypertension with an angiotensin receptor blocker (ARB) for prevention of recurrence of AF [26].

Disclosure statement

SEK, KN, MB and SO are editors of Blood Pressure and report no relevant conflicts of interest to disclose related to this editorial.