To establish a causative relationship between a given risk factor and an outcome, several conditions must be met. Among them, temporal relationship, strength and dose-response of association, consistency, specificity, and, very importantly, biological plausibility. A statistical association alone is certainly insufficient,[1] especially when it comes to relationships between risk factors and outcomes that have not been already observed in randomized clinical trials, but are based solely on observational studies. This is the case with the longstanding debate on the relationship between alcohol consumption and health, an issue for which there is a lack of large randomized clinical trials because they are difficult to implement.[2] Therefore, the available scientific evidence is presently based on a great number of generally well-performed observational studies that have assessed both the harmful effects of nonmoderate alcohol consumption and the protective effects of low consumption, especially in relation to cardiovascular health.[3,4]
To shed light on the relationship between alcohol consumption and health, it is crucial to elucidate the underlying mechanisms (biological plausibility) that determine how alcohol may affect our health. One specific question of interest is how light/moderate alcohol consumption might exert the cardiovascular protection that many, although not all, observational studies have shown. Identified mechanisms, including increase of high-density lipoprotein, anti-inflammatory and antioxidant effects, and improved blood flow, do not seem to be entirely exhaustive or convincing.
