Causal relationship between blood pressure and AF risk

Introduction and methods

Atrial fibrillation (AF) is the leading type of cardiac arrythmia and is associated with increased risk for hospitalization, HF, stroke and death [1,2]. Large longitudinal cohort studies have identified several putative risk factors for incident AF, including aging, smoking, alcohol abuse, hypertension, obesity, diabetes, MI, and HF [3-6]. Especially the severity and duration of hypertension have been identified as important risk factors for new-onset AF.

Based on observational studies, meta-analyses and secondary analyses of RCTs in patients with hypertension, guidelines recommend blood pressure (BP) lowering therapy for the prevention of AF [1,2]. But it is difficult to confirm causality between hypertension and AF, as they might be prone to systemic biases and confounded by several factors [6-8]. For example, elevated BP is related to aging, but aging is also an important risk factor for the development of AF [6-8]. These confounding factors impede epidemiological studies in assessing causal effects between BP and AF.

Mendelian randomization (MR) has emerged as a reliable method to assess causal effects. This approach reduces confounding, as random assortment of alleles at conception ensures a balanced distribution of genotypes. In addition, MR is less prone to reverse causation, as genotypes are not affected by the presence of a disease. This study performed large-scale MR analyses of functional genetic variants related to BP to assess a potential causal relationship between BP levels and risk of developing AF.

Novel and previously published (confirmed and independent) genetic variants associated with BP traits were retrieved from the UK Biobank and International Consortium of Blood Pressure Genome Wide Association Study (ICBP-GWAS) [10-13]. From 901 associated variants, 894 SNPs were available on a GWAS of AF genetics that included 60,620 cases with AF of European descent and 970,216 controls [14]. The genetic variants (266 SNPs for SBP, 345 SNPs for DBP, and 283 SNPs for pulse pressure [PP]) were used in three separate two-sample MR analyses. This was conducted to test the potential causal relationship between DBP, SBP, and PP with risk of AF by estimating the association results in two non-overlapping populations. The three key assumption underlying these two-sample MR analyses are: 1) the SNPs must be strongly associated with the BP trait, 2) the genetic variants must affect the outcome only through their effect on BP, and 3) variants must be independent of any confounders of the association between BP and AF [15]. The estimated causal effects of BP for AF development were only retrieved from the ICBP-GWAS to avoid possible sample overlap. The MR-Egger regression method and MR pleiotrophy residual sum and outlier (MR-PRESSO) test were used to test robustness of the results.

Main results

• An increase in 1 mm Hg SBP was causally associated with a 1.8% increase in risk for AF (OR 1.018, 95% CI: 1.012-1.024, P <0.001).

• MR analysis showed that 1 mm Hg increase in DBP was causally associated with an 2.6% increased risk for AF (OR 1.026, 95% CI:1.016-1.035, P<0.001).

• 1 mm Hg increase in PP was causally associated with an 1.4% increased risk for AF of (OR 1.014, 95% CI:1.001-1.028, P=0.033).

• Sensitivity analysis revealed robustness in the findings.

• When correcting for SNPs that were associated with ischemic heart disease and obesity, the causal relationship between AF and BP were comparable to the main MR analysis results (SBP: OR 1.020, 95% CI: 1.014-1.026, n=251 SNPs; DBP: OR 1.029, 95% CI: 1.019-1.039, n=332 SNPs; and PP: OR 1.015, 95% CI: 1.001-1.029, n=272 SNPs).

Conclusion

References

Find this article online at Eur J Prev Cardiol