Alcohol consumption increases risk of incident AF
Alcohol consumption, cardiac biomarkers, and risk of atrial fibrillation and adverse outcomesLiterature - Csengeri D, Sprünker N, Di Castelnuovo A et al., - Eur Heart J. 2021 Jan 13;ehaa953. doi: 10.1093/eurheartj/ehaa953.
Introduction and methods
There are conflicting data on the relation of alcohol consumption with incident atrial fibrillation (AF) [1-6]. Especially in individuals with low alcohol consumption the association is unclear. This study assessed the relation between alcohol consumption and incident AF in a large European community-based pooled cohort.
Data from 100,092 participants free of AF at baseline from five community-based cohorts (DAN-MONICA, FINRISK, Moli-sani, Tromsø, and Northern Sweden) were used in this analysis. Mean age was 47.8 (range 24-97) years at baseline, 51.7% were women. Participants were asked in local food frequency questionnaires how often they drank wine, beer and spirits. It was assumed that 120mL of wine, 330mL of beer, or 40mL of spirits contained 12 g of alcohol. Hazard ratio’s for incident AF were calculated using continuous alcohol variables (number of drinks/day or amount of alcohol in gram/day). In addition, HR’s for incident AF were determined for categories of alcohol intake (former drinker, never drinker [lifelong abstainers, 0 g/day, reference group], occasional drinker [<1 g/day], 1–12 g/day, 12.1–24 g/day, 24.1–48 g/day and >48 g/day) and by drinking pattern (lifelong abstainer (reference group), ex-drinker, less than once a week, 1–2, 3–5, and 6–7 days/week). In addition, information on classical CV risk factors and incident heart failure (HF) was collected and biomarkers N-terminal pro-B-type natriuretic peptide (NT-proBNP) and serum high-sensitivity troponin (hsTnI) were measured. NT-proBNP was measured in stored blood samples from 23205 participants. Data on hsTnI was available from 31129 participants. Median follow-up was 13.9 years. Primary outcome was diagnosis of AF, defined as either AF or atrial flutter.
- Sex and cohort-stratified Cox regressions showed that alcohol consumption was positively associated with incident AF (HR for one drink/day [12 g alcohol/day]: 1.16, 95% CI 1.11–1.22, P< 0.001). HRs non-linearly increased with alcohol consumption. Using continuous alcohol variables, the cut-off for statistically significantly increased risk for AF was observed with a regular alcohol consumption of 2 g/day, which translates to 0.17 drinks/day.
- The associations did not change markedly after adjustment for classical CV risk factors nor after having accounted for all-cause mortality as competing risk. The observed HRs were similar for men and women. No interaction for alcohol consumption and incident AF by sex was observed (P=0.07).
- When looking at associations of alcohol consumption and incident AF by drinking categories, no statistically significant associations were found in former drinkers, occasional drinkers and individuals drinking up to one alcohol consumption per day. Higher categories of alcohol intake were associated with increased risk of incident AF. No statistically significant association was found with drinking patterns.
- A sex and cohort-stratified Cox regression analysis revealed a J-shaped relationship between alcohol consumption and incident HF, with HRs below 1 with alcohol consumption up to 20 g/day (1.6 drinks/day). The interaction between incident HF and alcohol consumption with AF was not statistically significant (P=0.07).
- Levels of NT-proBNP or hsTnI concentrations did not significantly modify the association between alcohol consumption and incident AF.
This study revealed a non-linearly increasing relationship between alcohol consumption and incident AF. The association was explained neither by concentrations NT-proBNP or hsTnI nor by the development of HF during follow-up. The finding that even low levels of alcohol consumption are associated with incident AF needs to the considered in AF prevention.