Artificial Sweeteners: A New Dietary Environmental Risk Factor for Atrial Fibrillation?

In this issue of Circulation: Arrhythmia and Electrophysiology Sun et al. study the influence of sweetened beverages on incident atrial fibrillation (AF) in the UK Biobank cohort¹. Three types of beverages were studied in relationship to incident AF: Sugar-sweetened beverages (SSB), artificially sweetened beverages (ASB), and pure fruit juices (PJ). They found that processed drinks in the form of SSB and ASB associated with increased risk of AF after multivariate adjustment (HR (95% CI); SSB 1.10 (1.01-1.20); ASB 1.20 (1.10-1.31) > 2 liters/week after full model adjustment). In contrast, PJ consumption appeared to be protective when consumed at low volumes (HR (95% CI); 0.92 (0.87-0.97) ≤ 1 liters/week after full model adjustment) and neutral to the incidence of AF when consumed at greater volumes.

These findings are consistent with other large European and American cohort studies where added sweeteners (both for natural sugar and artificial sweeteners) in our diets increase the risk for type 2 diabetes, insulin resistance, total mortality, and cardiovascular disease^2-6. Thus, the finding that SSB and ASB have an association with incident AF is not surprising. Somewhat surprising was that subjects with increased consumption of ASB were at the highest risk of incident AF. These findings raise the possibility that artificial sweeteners may have a more important role and contribute to the pathogenesis of AF.

Environment and AF

The role of acquired environmental factors in pathogenesis of AF is well established. Modifiable environmental risk factors, such as smoking, hypertension, coronary artery disease, alcohol intake, diabetes, and obesity, have been cited in the initiation and progression of AF⁷. Although there are examples of monogenic forms of AF, these cases tend to be in younger healthier individuals, and this genetic predisposition still only accounts for a minority of cases⁸. However, a larger fraction of AF heritability is accounted for by common genetic variants, which can be used collectively to calculate polygenic risk scores (PRS). Those individuals inheriting the top quintile of PRS for AF, compared to the bottom 4 quintiles, have an AF odds ratio (OR) of 2.43, while those in the PRS top percentile, compared to the remaining 99%, have an AF OR of 5.23⁹. Like other complex metabolic diseases, such as coronary artery disease, heart failure, hypertension, and type 2 diabetes, environmental exposures over time, combined with common genetic variants, play a dominant role in AF pathogenesis and progression. A common major exposure, either by direct or indirect influence, is our diet.

Diet alone has not been cited as a major risk factor for AF, but it is a known contributor to many of the risk factors that are associated with AF. Diets that are high in sodium and low in potassium and that are not rich in fruits and vegetables are known major risk factors for hypertension. Likewise, dietary intervention alone can significantly reduce blood pressure to a similar effect as mono drug therapy for hypertension^10,11. In the original Dietary Approaches to Stop Hypertension (DASH) clinical trial, subjects on the DASH diet, a diet rich in fruits, vegetables, and low-fat dairy products, reduced systolic blood pressure by 11.4 mmHg and the diastolic blood pressure by 5.5 mmHg in study subjects with hypertension¹⁰. A follow up study demonstrated that using the DASH diet with concomitant sodium restriction was more effective in lowering blood pressure than either intervention alone¹¹.

Obesity from excessive caloric intake is a major risk factor for AF. In the Women’s Health study a 4.7% increase in AF was observed per 1 kg/m2 increase in BMI ¹². The mechanisms by which obesity may be causal for AF has been linked to autonomic dysfunction, inflammation, and fibrosis, but obesity may also contribute to other known risk factors for AF, such as diabetes, dyslipidemia, obstructive sleep apnea, and hypertension^13,14. In addition, weight loss through bariatric surgery has been associated with a lower incidence new onset AF and the risk of AF recurrence post ablation^15,16.

A western diet has long been a known risk factor for the development of coronary artery disease and acute coronary syndromes. Numerous studies have demonstrated that a plant-based diet reduces cardiovascular risk and mortality^17-19. In more recent studies a mechanistic link between a Western diet and cardiovascular disease has been identified via the ingestion of trimethylamine (TMA)-containing dietary nutrients enriched in animal products, such as choline, phosphatidylcholine, and carnitine^20-22. Dietary TMAs are substrate for the generation of an atherogenic metabolite, trimethylamine N-oxide (TMAO) via a two-step conversion, first by the gut microbiota and then the host liver²⁰. Animal models, microbial transplantation and cell-based studies have suggested a mechanistic link between TMAO production and atherosclerotic disease^20-24.

Beyond coronary artery disease, TMAO has also been shown to independently associate with incident AF in two large Norwegian cohorts with long term follow up²⁵. Additionally, TMAO and the gut microbiome have also been linked to platelet hyperreactivity, coagulopathy, and stroke. These data suggest that the diet and gut microbiome may also have a role in the pathogenesis and sequelae of AF^26,27. Indeed, even amongst individuals with an optimal risk profile (no history of smoking, hypertension, obesity, heart failure, myocardial infarction, or excessive alcohol intake), the lifetime risk of AF remains 23%²⁸ . This suggests that there are other risk factors yet to be elucidated contributing to the pathogenesis of AF. One such risk factor could be artificial sweeteners.

Artificial sweeteners are widely used as sugar substitutes to reduce total calorie intake while maintaining or enhancing the taste in food²⁹. However, there is little evidence to suggest that these sugar substitutes confer any benefit in weight loss and decreasing insulin resistance²⁹. Despite this rapid incorporation into the food chain, the long-term safety of using such substitutes has not been well established. In fact, numerous epidemiological studies now demonstrate that artificial sugars associate with increased cardiovascular risk and all-cause mortality ^3-6,30. In a recent publication, Hazen and colleagues demonstrate plasma erythritol, a frequently used sugar substitute, not only is an independent predictor of major adverse cardiovascular events, but also increases thrombosis potential³⁰. At physiological relevant concentrations, erythritol increased in vitro platelet reactivity and thrombosis risk. Additionally, in a small clinical study erythritol supplementation was shown to enhance platelet responsiveness and in vivo thrombosis risk in healthy subjects, suggesting a mechanistic link to the reported poor cardiovascular outcomes³⁰.

Here, Sun et al. extend a link of artificial sweeteners to arrhythmogenesis and AF¹. In the large well characterized UK Biobank population, ASBs were independently associated with incident AF in a dose dependent manner. This risk was persistent despite adjustments for patient characteristics, risk factors for AF, including a genetic predisposition, and dietary habits. The major strength of this paper is the first-time association of ASB, and to a lesser extent SSB, with incident AF. This raises many hypothesis-generating questions. Amongst ASB, what artificial sweetener(s) are associated with AF? What underlies the pathogenesis of this observation? Are there indirect or direct effects of artificial sweeteners on the myocardium? And finally, should longer term studies be required before introduction of artificial nutrients into the food?

There are some weaknesses of the findings. First, this is an observational study and there may be unaccounted subject characteristics that prompted individuals to consume more ASB that could bias these results. Additionally, the consumption of beverages was self-reported and based on mostly one or two 24-hour dietary recall periods over a 4–5 year study period. It is therefore unclear whether a couple of single day dietary recalls is representative of dietary habits over the study period. The exact components of the drinks are also not completely characterized. The types of artificial sweeteners, the caffeine content or other additives are not known. Finally, the analyses by genetic PRS demonstrates the strong influence of genetics with AF that appears to have a much higher impact on AF risk than the intake of SSB, ASB or PF. Despite these drawbacks, this study in the first to demonstrate a link of high consumption of SSB and ASB to AF. Further epidemiological and mechanistic studies are needed to confirm and extend these findings.