(Economist) - FEW things arouse such warm adulation and sharp denunciation as alcohol. It is beloved by some and despised by others, and its consumption is governed by legal and religious rules. Wine is central to Christian rites but is widely considered by Muslims to be forbidden by the Quran. It is also the subject of debate within the scientific community: some researchers contend that alcohol, particularly wine, has health benefits, but others disagree.

Pro-oenological forces point to a large body of evidence demonstrating wine’s positive effect on both the cardiovascular system and longevity. This viewpoint was given additional support this week by a new study in mBio led by Ming-liang Chen and Man-tian Mi of the Third Military Medical University in China. Using mice, the team showed that resveratrol, a molecule found in grapes and berries, reduced the formation of plaques in arteries—a cardiovascular condition known as atherosclerosis that limits blood flow and can trigger heart attacks and strokes.

The authors found that resveratrol acts in a surprising and indirect way. Mice fed a diet supplemented with resveratrol experienced a dramatic shift in the composition of their gut bacteria. Simultaneously, production of trimethylamine-N-oxide, a risk factor for atherosclerosis, was reduced. The team suspect that the remodelling of the gut microbiota was the cause for this drop. Indeed, when mice were also given antibiotics, the benefits of resveratrol disappeared. Thus, the study provides a plausible biological mechanism for how wine and other resveratrol-containing foods might promote cardiovascular health.

But anti-alcohol advocates can claim a victory of their own in the Journal of Studies on Alcohol and Drugs. Tim Stockwell of the University of Victoria in Canada and Tanya Chikritzhs of the National Drug Research Institute at Curtin University in Australia carried out a meta-analysis of 87 epidemiological investigations. They concluded that so-called moderate drinkers do not benefit from a reduction in mortality compared to abstainers. This finding strikes a blow at the very heart of the imbibers’ claim.

Dr Stockwell and Dr Chikritzhs contend that the inconsistency is due to a common misclassification error. Many of the studies the team examined had classified former drinkers as abstainers. This is incorrect, the authors argue, because former drinkers often give up alcohol for health reasons. Those who become teetotallers due to doctors’ orders are probably less healthy than the general population. So including them in a study creates a bias that artificially inflates the health benefits of alcohol. When the authors restricted their analysis to high-quality studies that included only lifetime abstainers and properly controlled for other confounding variables, the health benefits of moderate drinking vanished.

But there might be a way to reconcile these contradictory findings. Might moderate drinkers be swilling too many martinis and not enough red wine? Alas, a 2014 study in JAMA Internal Medicine concluded that resveratrol obtained from the Western diet had no impact on either the health or mortality of elderly people.