The relationship between alcohol consumption and health: J-shaped or less is more

[[abstract]]In this issue of the Journal, Tian and colleagues confirmed the J-shaped relationship between alcohol consumption and health based on analyzing 918,529 adults from the repeated National Health Interview Survey (NHIS) from 1997 to 2014 [1]. In addition to several cause-specific mortality, Tian and colleagues’ findings are largely confirmatory to previous studies using NHIS [2, 3] and several large systematic review and meta-analysis on this issue [4,5,6]. The “J-shaped” relationship refers to a protective health effect at a lower level of consumption; while above a certain threshold, it greatly increases health risks. This study found that compared with lifetime abstainers, current infrequent, light, and moderate drinkers had a lower risk of all-cause mortality (current infrequent: − 13%; light: − 23%; moderate: − 18%) and cardio-vascular disease mortality (infrequent: − 14%; light: − 24%; moderate: − 22%), respectively. However, it is important to address several methodological issues when interpreting the J-shaped association