Page last updated: March 8, 2017
Update on the J-shaped Curve for the Relation of Alcohol Intake to Health - Professor R. Curtis Ellison, Boston University School of Medicine, USA

Epidemiologic studies for many decades have consistently shown that, when compared with non-drinkers, light-to-moderate consumers of alcoholic beverages have a lower risk of cardiovascular disease (CVD), and also show a reduced risk of total mortality. In most studies that contain an adequate number of heavy drinkers, such subjects show a greater risk than moderate drinkers as well as non-drinkers; this phenomenon results in what is known as a “J-shaped curve.”

While many of the early studies included ex-drinkers in the non-drinking referent group (that may have increased the risk of disease for “current abstainers”), a similar J-shaped curve has almost always been seen in more recent studies when only lifetime abstainers make up the non-drinking category. Numerous improvements in epidemiologic & statistical techniques, as well as better definition of potential confounders, have provided even clearer results from observational epidemiologic studies (Ronksley et al, 2011; Midlov et al, 2016). Randomized control trials in humans (Brien et al, 2011) and essentially all animal experiments have shown a similar pattern: beneficial effects on risk factors and disease with light-to-moderate alcohol exposure, adverse effects with large amounts. Even when repeated estimates of alcohol intake are used to construct trajectories of consumption, Passos et al (2017) have recently shown a J-shaped curve between alcohol and CVD. And the opposite is seen when studying the quality of life: higher scores for subjects reporting moderate drinking and lower scores among both abstainers and those with heavy drinking (Schrieks et al, 2016).

Differences between a linear and a J-shaped curve

Wikipedia defines a J-shaped curve as “A variety of J-shaped diagrams where a curve initially falls, then steeply rises above the starting point.” For the consumption of wine and other alcoholic beverages, this means a decline in risk (of heart disease, dementia, total mortality, etc.) for intake up to a certain level of drinking, a return to the same risk as non-drinkers with more alcohol, and then an increase in risk for heavy drinking. For most middle-aged or older adults (unless there are contraindications to any alcohol from previous abuse, severe liver or certain other diseases, etc.) current scientific data indicate that the message in terms of health is “No alcohol is not good for you, a little alcohol is good for you, a lot of alcohol is bad for you.”

Seeking a better measure of the “exposure” to alcohol; the importance of the pattern of drinking

It is unfortunate that most epidemiologic studies have been forced to use only the average amount of alcohol consumed (over a week or month) as the measure of exposure. Most have been unable to adequately control for the pattern of drinking, even though it is clear that regular moderate consumers of alcohol have considerable health advantages over binge drinkers of the same average amount. Mukamal et al (2005) found that binge drinking (versus no binge drinking) eliminated the protective effects of alcohol on mortality among subjects who had suffered a myocardial infarction. Piano et al (2017) recently reported that binge drinking was associated with many adverse cardiovascular effects; they found that binge drinking in middle-aged and older adults was associated with a higher risk of hypertension, myocardial infarction, and stroke. Further, these authors found that binge drinking in the young (18-30 years) had adverse effects on blood pressure, endothelial function, and cardiac arrhythmias. Many animal experiments show the same pattern; for example, Liu et al (2011) found completely different effects on coronary atherosclerosis induced among mice when the same amount of alcohol was administered on a daily basis (a decrease in atherosclerosis) versus only on two days of the week (an increase in atherosclerosis).

In epidemiologic studies, adding self-reported information on the frequency of binge drinking to analyses has only partly corrected the problem of mixing regular moderate drinkers and binge drinkers in the same category. Klatsky and Udaltsove (2007) have found from their very large Kaiser-Permanente cohort that a more accurate identification of subjects who are “underreporters” of their alcohol intake improves the precision of estimates of effect. For mortality, these authors report: “The analysis reconfirms that the relation of alcohol drinking to total mortality is J-shaped, with reduced risk (mainly because of less cardiovascular disease) for lighter drinkers and increased risk for persons reporting more than 3 drinks per day. Infrequent (occasional) drinkers have risk similar to that of lifelong abstainers, while former drinkers are at increased risk, especially for non-cardiac death. The general shape of the relation of alcohol to mortality is similar for men and women. Age differences are substantial, with the apparent benefit from light-moderate drinking not seen before middle life. Our data indicate further that the apparent magnitude of benefit of lighter drinking is probably reduced by systematic underreporting.” Their subsequent analyses have strengthened this premise, indicating that underreporting of alcohol intake affects the apparent risk for cancer as well; they found that when recognised underreporters are removed from the analysis, the risk of cancer from moderate drinking is essentially nil (Klatsky et al, 2014).

Does the type of beverage matter?

While most epidemiologic studies indicate that moderate drinking of any type of alcoholic beverage is associated with a lower risk of CVD, an increasing number of both animal experiments and human trials are demonstrating that the polyphenols and other non-alcoholic substances in wine and in some beers provide additional protection against disease. Such effects are especially important for wine when it is consumed with food, as has been recently summarised by Boban et al, 2016. It is anticipated that when more accurate assessments of actual alcohol consumption and drinking patterns become available, even greater protection against disease from moderate drinking will be shown, especially when the beverage of choice is wine.

Public health implications of a J-shaped curve

As of 2017, essentially all epidemiologic studies continue to show a J-shaped curve, especially for CVD and mortality. Some health officials suggest that the public should be advised to focus on the nadir of the J-shaped curve for making decisions about drinking. Instead, many moderate drinkers are inclined to view the point on the curve where the risk of adverse health outcomes exceeds that of abstainers, a point that indicates when the level of drinking may begin to be less healthy than the risk associated with abstinence. Chokshi (2016) has pointed out that the marked differences between a linear curve and a J-shaped curve cause problems for public health messages. “Traditional messages such as restrict, ban, etc. work for linear relations between exposures and health (e.g., for cigarettes, illegal drugs) – ‘Just say no!’ However, they work less well for an exposure with a ‘J-shaped’ relation with health, such as wine.” In can be argued that, on the other hand, most people can appreciate that taking a little of something (e.g., a glass of wine, an aspirin tablet) is different from taking a lot of it (e.g., a bottle or two of wine, 100 aspirin tablets). The J-shaped curve provides an accurate portrayal of the scientific data relating alcohol to disease outcomes.

Focusing on a “Healthy Lifestyle”

Making recommendations to the public regarding alcohol consumption should always include other important lifestyle factors that affect health: smoking, obesity, diet, and exercise. Scientific data show that subjects in large epidemiologic studies who are non-smokers, are not obese, eat a Mediterranean-type diet, and get regular exercise have much lower risk of CVD and total mortality, whether or not they consume alcohol. However, as pointed out by Mukamal et al (2006), regardless of other “healthy” risk factors, the presence of moderate alcohol consumption in subjects, when added to the analysis, provides significantly greater protection against CVD than seen for the other factors alone. Thus, current data indicate that our definition of a “healthy lifestyle” should include, unless contraindicated, small to moderate amounts of alcohol, especially when consumed on a regular (non-binge) basis and with food.


Boban M, Stockley C, Teissedre P-L, Restani P, Fradera U, Stein-Hammer C, Ruf J-C. Drinking pattern of wine and effects on human health: why should we drink moderately and with meals? Food Funct 2016;7:2937–2942. DOI:10.1039/c6fo00218h

Brien SE, Ronksley PE,Turner BJ, Mukamal KJ, Ghali WA. Effect of alcohol consumption on biological markers associated with risk of coronary heart disease: systematic review and meta-analysis of interventional studies. BMJ 2011;342:d636; doi:10.1136/bmj.d636.

Chotsky DA. Viewpoint: J-shaped curves and public health. JAMA 2015;314:1339-1340

Klatsky AL, Udaltsova N. Alcohol drinking and total mortality risk. Ann Epidemiol 2007;17:S63–S67.

Klatsky AL, Udaltsova N, Li Y, Baer D, Tran HN, Friedman GD. Moderate alcohol intake and cancer: the role of underreporting. Cancer Causes Control 2014; on-line publication; DOI 10.1007/s10552-014-0372-8 2014.

Liu W, Redmond EM, Morrow D, Cullen JP. Differential Effects of Daily-Moderate VersusWeekend-Binge Alcohol Consumption on Atherosclerotic Plaque Development in Mice. Atherosclerosis 2011;219:448-454.

Midlöv P, Calling S, Memon AA, Sundquist J, Sundquist K, Johansson S-E. Women’s health in the Lund area (WHILA) - Alcohol consumption and all-cause mortality among women – a 17 year follow-up study. BMC Public Health 2016;16:22. DOI 10.1186/s12889-016-2700-2

Mukamal KJ, Chiuve SE. Rimm EB. Alcohol consumption and risk for coronary heart disease in men with healthy lifestyles. Arch Intern Med 2006;166:2145-2150.

Passos VL, Klijn S, van Zandvoort K, Abidi L, Lemmens P. At the heart of the problem - A person-centered, developmental perspective on the link between alcohol consumption andcardio-vascular events. Int J Cardiol 2017, pre-publication.

Piano MR, Mazzuco A, Kang M, Phillips SA. Cardiovascular Consequences of Binge Drinking: An Integrative Review with Implications for Advocacy, Policy, and Research. Alcoholism: Clinical and Experimental Research 2017; Pre-publication. doi: 10.1111/acer.13329

Ronksley PE, Brien SE, Turner BJ, Mukamal KJ, Ghali WA. Association of alcohol consumption with selected cardiovascular disease outcomes: a systematic review and meta-analysis. BMJ 2011;342:d671; doi:10.1136/bmj.d671

Schrieks IC, Wei MY, Rimm EB, Okereke OI, Kawachi I, Hendriks HFJ, Mukamal KJ. Bidirectional associations between alcohol consumption and health-related quality of life amongst young and middle-aged women. J Intern Med 2016;279:376-387. doi: 10.1111/joim.124532015.


All text and images © 2003 Alcohol In Moderation.