One of the many arguments used to try and persuade us to drink less is that the calorific content of alcohol contributes to obesity. For example, this article from the UK NHS Choices website argues, among many other things:

Regularly drinking more than the NHS recommends can have a noticeable impact on your waistline as well as cause less obvious but more serious health problems. Many women don’t realise that two large glasses of white wine not only puts them over the recommended daily limit for regular alcohol consumption, but also provides them with nearly 20% of their recommended daily calorie intake, at approximately 370kcal in total.

Most people would baulk at consuming a full glass of single cream, but wouldn’t think twice about the calorie content of a couple of pints. But the calorie content is similar and, over time, excess alcohol intake can easily contribute to gaining weight.

Wine, beer, cider, spirits and all our favourite drinks are made from natural starch and sugar. Fermentation, and distillation for certain drinks, is used to produce the alcohol content. This helps explain why alcohol contains lots of calories – seven calories a gram in fact, almost as many as a gram of fat. And, of course, additional calories can be present in added mixer drinks.

But a new book argues that this is just wrong empirically.

Here are a some of the claims as summarised in an Australian news article (I'm sure some will say they would say that):

A six-year study of 43,500 people by the University of Denmark. Key findings: teetotallers and infrequent drinkers ended up with the biggest waistlines, daily drinkers had the smallest.

An eight-year study of 49,300 women by University College Medical School, London. Key findings: women who drank below 30 grams a day (around two medium glasses of wine) were up to 24 per cent less likely to put on weight than teetotallers.

A ten-year study of 7,230 people by the U.S. National Center for Disease Control. Key findings: drinkers gained less weight than non-drinkers. Alcohol intake did not increase the risk of obesity.

And a possible explanation is that the metabolic routes for alcohol digestion don't involve the production of glucose (though, obviously, some booze also contains high glycaemic index carbohydrate so will cause peaks in blood sugar).

Is the claim true? Does the alcohol content of drinks make no observed contribution to excess weight?

  • @Articuno Your first point is sound and could be the basis of a good answer. Your second point is an over literal interoperation of the language in the quotes: the clear argument is that alcohol content of drinks doesn't contribute to weight gain.
    – matt_black
    Commented Nov 28, 2013 at 22:47
  • A question I asked on bio.SE biology.stackexchange.com/questions/23362/… Commented Nov 5, 2014 at 7:43

1 Answer 1


The studies quoted in the question are observational, not interventional, and thus are not suited to address the question: "does the alcohol content of drinks make you fat". To answer that question, we need an interventional study to look at the counterfactual.

This has been done in the case of alcoholics (Lieber, 1991).

To investigate whether ethanol contributes to the body’s energy balance according to its caloric value of 7.1 kcal/g, the effects of ethanol on body weight were assessed in 14 [alcoholic] subjects given ethanol under metabolic-ward conditions.

Isocaloric substitution (replacing calories they would have been consuming with ethanol calories) of up to 50% of the daily calorie intake resulted in a 0.9kg decrease in weight over 16 days, compared to +0.04kg increase in weight in a control group. This was statistically significant (p < 0.001). This experiment used 12 subjects.

In a separate experiment, 2 subjects were given 2000kcal per day of additional calories in ethanol form. They gained 0.19kg over 30 days (subject 1), and 0.45kg over 11 days (subject 2).

Lieber's study concludes:

Weight watchers are usually advised to restrict their alcohol consumption. This advice is based on the assumption that the calories derived from ingested ethanol are approximately the same as those liberated when ethanol is combusted in a bomb calorimeter, namely 7.1 kcal/g. This advise is still wise and pertinent in the case of the social drinker because all of the evidence related to the above-concerning energy deficit pertains to the chronic abuser of large amounts of alcohol. The energy loss seems to result either from the liver damage produced by alcohol and the consequent inefficient utilization of fat as well as other nutrients, and/or the energy wastage associated with the induction of microsomal pathways that results from the chronic consumption of substantial amounts of ethanol.

(Suter and Angelo, 2005) examine the broader question as it effects non-alcoholics.

They say:

Epidemiologic data showed a positive, negative, or no relationship between alcohol intake and body weight. Despite the difficulty in assessing alcohol intake as well as controlling for different confounders of the energy-balance equation, the conflicting epidemiologic data can be explained in most instances.

The studies you quote would fall under that description. They are three studies that show a negative relationship between alcohol intake and body weight. However, they are only part of a larger body of work that presents conflicting epidemiologic data.

Here is their summary of epidimiological studies, organized into whether they found a positive, negative, or no relationship between alcohol intake and body weight:

enter image description here

Regarding the moderate drinker:

The experimental metabolic evidence suggests that the consumption of moderate amounts of alcohol has to be accounted for in the energy-balance equation and may represent a risk factor for the development of a positive energy balance and thus weight gain.


Experimental data in combination with epidemiologic findings suggest that alcohol energy counts more in moderate nondaily alcohol consumers than in some moderate daily and all heavy consumers.

There is also a large individual variation in response to alcohol, strongly influenced by genetic factors (Suter and Angelo, 2005).

Interventional studies would be ideal, but "due to ethical reasons and the typical effects of alcohol, controlled long-term intervention studies on the effect of alcohol on body weight are difficult to perform. Accordingly, very few long-term intervention studies address the effect of alcohol on body weight regulation as the primary aim" (Suter and Angelo, 2005).

They conclude that the question should be rephrased:

Accordingly the question is not “Whether alcohol calories do count” but “How much do alcohol calories count?”

How much alcohol calories count is highly dependant on the individual. Factors that conspire together to affect the energy balance of an individual include "body weight, substrate composition of the diet, substrate oxidation capacity, genetic background, gender, and physical activity as well as the drink (i.e., frequency and amount of consumption, pattern of drinking regarding concomitant food intake)". (Suter and Angelo, 2005)

"The alcohol calories count more in an overweight person, with a high-fat diet, low levels of physical activity, and a positive family history of obesity as well as a low rate of ethanol degradation." (Suter and Angelo, 2005)


Lieber, Charles S. "Perspectives: do alcohol calories count?." The American journal of clinical nutrition 54, no. 6 (1991): 976-982.

Suter, Paolo M., and Angelo Tremblay. "Is alcohol consumption a risk factor for weight gain and obesity?." Critical reviews in clinical laboratory sciences 42, no. 3 (2005): 197-227.

  • Why are the studies contradictory (positive and negative)? And the abstract of the one study is confusing or non-obvious, "Isocaloric substitution of carbohydrates by ethanol results in weight loss, and addition of ethanol to an otherwise normal diet does not produce the expected weight gain": it supports the claim ("does not produce the expected weight gain") in some circumstances, but the abstract isn't clear to me about what those circumstances are.
    – ChrisW
    Commented Nov 29, 2013 at 10:10
  • 1
    The circumstances were in chronic abusers of large amounts of alcohol. The conclusion and methodology sections are clearer about that.
    – user5582
    Commented Nov 29, 2013 at 10:20
  • Suter and Angelo hypothesize that the discrepancy between the epidimiological studies is because "most are opportunistic, i.e., their primary target was an issue other than the relationship of body weight to alcohol. Because the initial design and methodology are of crucial importance, these studies have to be interpreted with caution."
    – user5582
    Commented Nov 29, 2013 at 10:25
  • They were most collecting data for another purpose and later, that data was used for an epidimiological study examining alcohol consumption and weight gain. Whatever the inclusion criteria were for the original collection (a particular morbidity, or age/weight combination) can affect the outcome of the later study.
    – user5582
    Commented Nov 29, 2013 at 10:27
  • Another possible cause is that these are necessarily post-hoc analyses. The researchers could be exploiting researcher degrees of freedom (either consciously or subconsciously) in their methodological decisions that give a particular result. I haven't read through that long list of studies, so this is just speculation, but accidental exploitation of researcher degrees of freedom would leave signals in the publications that we could look for.
    – user5582
    Commented Nov 29, 2013 at 11:04

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .