Sex differences in atrial fibrillation: The case of testosterone

In this issue of the European Journal of Preventive Cardiology, Zeller et al.1 delve into the FINRISK97 study to deliver impressive data over a 15-year follow-up period. Their take-home message is simple, yet intriguing: serum testosterone levels and the risk of atrial fibrillation and ischaemic stroke are positively associated in women, but negatively associated in men. More simply, an increase of 1 nmol/L is associated with a 2% reduction of risk in men and a 17% increase of risk in women after correction for the most important cardiovascular risk factors. This finding has a profound (and two-fold) clinical implication. The paper adds evidence to the idea that low testosterone should be considered as a risk factor for atrial fibrillation and the most dreaded complication, ischaemic stroke. The role of testosterone deficiency was suggested in the milestone experimental work of Tsuneda et al.,2 which showed that the gonadectomy of male rats increased atrial vulnerability and arrhythmogenicity by inducing ryanodine receptor type 2-dependent intracellular calcium leaks from the sarcoplasmic reticulum. This pathophysiological hypothesis could explain not only the increased incidence of atrial fibrillation in elderly men (which goes hand-in-hand with the reduced levels of testosterone in older people), but also a non-trivial proportion of so-called ‘lone atrial fibrillation’, which is more common in the young adult and could be due to reduced androgen plasma levels. Of course, an indirect association should also be taken into account. Testosterone deficiency has traditionally been associated with an increased incidence of cardiovascular risk factors (e.g. hypertension, diabetes mellitus and obesity) and cardiovascular events (e.g. ischaemic heart disease and heart failure), all contributing to the atrial fibrillation epidemic of recent years.3 The incidence of atrial fibrillation, as is commonly the case, has been assessed by considering only symptomatic episodes. Thus it was probably underestimated by Zeller et al.1 because the brief, asymptomatic episodes characterizing silent atrial fibrillation4 were not detected in the FINRISK study. Body mass index is already known as one of the greatest modifiable risk factors for atrial fibrillation. A large European cohort (including patients from the FINRISK study) has shown that body mass index interacts with sex, its weight towards the incidence of atrial fibrillation being significantly greater in men than in women (hazards ratio 1.31 versus 1.18; relative risk reduction 0.89).5 Obesity is associated with physiological changes in the plasma levels of sex hormones6 because the increased aromatase activity, mainly mediated by adipocytes, leads to an increase in the peripheral conversion of androgens to oestrogens.7 The same adipose tissue is also responsible for leptin secretion, which, among its many effects, inhibits Leydig cells.7 Therefore it is feasible to hypothesize that the reduced levels of bioavailable testosterone in obese men adds to the intrinsic risk of obesity itself, whereas in women low levels of androgens lower the risk due to a higher fat mass, leading to a difference in the overall risk and a 11% lower relative risk reduction. Data for more precise adiposity measurements, such as the waist-to-hip ratio, could confirm this explanation and potentially expand this topic. The paper by Zeller et al.,1 unfortunately, also leaves many open questions. The most important one is clear: why are low levels of testosterone beneficial to women? According to the FINRISK data, women in the upper testosterone quartiles were at higher risk of developing atrial fibrillation and ischaemic stroke, exactly the opposite to their male counterparts. Why is this? Why do all the pathophysiological mechanisms summarized so far not apply to half the human population? The gap in evidence-based medicine regarding cardiovascular prevention in women has become obvious in recent years, mainly driven by the underrepresentation of female patients in randomized clinical trials.8 However, the lack of evidence for the effects of testosterone in women is especially embarrassing9 and the authors do not comment on this issue in their discussion. Their data are solid, however, and should provide the right push to focus resources and efforts aimed at closing the gap. For now, only a ‘usual suspect’ comes to mind. The renin–aldosterone–angiotensin system, the activation of which has traditionally been associated with an increased risk of atrial fibrillation and other cardiovascular risk factors,10 is regulated by sex hormones in many ways.11 In particular, aldosterone promotes atrial fibrosis and extracellular conduction disturbances, leading to a direct proarrhythmic effect in atrial cells.12 Female sex hormones work together in lowering the effect of aldosterone on the cardiovascular system: oestrogens by reducing angiotensin-converting enzyme activity and angiotensin II-mediated aldosterone secretion, and progesterone by binding to the mineralocorticoid receptor.11 It could be postulated that, in women, low testosterone levels could be seen as a marker of high plasma oestrogens (through aromatase activity), which can then synergize with progesterone in lowering the aldosterone response and its proarrhythmic effect. Of course, this explanation is merely speculative and should be tested in clinical practice. This and many other unanswered question on cardiovascular prevention in women should be priorities in the near future. The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. The author(s) received no financial support for the research, authorship, and/or publication of this article.