Skip to Main Content


October 29, 2007
Volume 85, Number 44
p. 48

Science Friction with Bob Wolke

Heather Mull
Bob Wolke

A few months ago, C&EN's Newscripts (Aug. 6, page 48)—not written by me—quoted a paper published in the British Medical Journal (1999, 319, 1600). It reported that a shaken martini contains almost twice as many healthful antioxidants than if it had been merely stirred. This conclusion, which has generated unquenchable interest among chemists for perhaps more than professional reasons, was also the subject of a recent letter to the editor (C&EN, Oct. 8, page 6).

The research was done by C. C. Trevithick, et al., of the department of biochemistry, faculty of medicine and dentistry, University of Western Ontario. Upon its publication in BMJ, it made headlines around the world, and its findings are still reverberating, unchallenged, eight years later.

But it's time to blow the whistle, because that research was fatally flawed. Shaking and stirring have no effect on the level of antioxidants in a martini.

When I originally read about this discovery, I was troubled. Why? Because it has been my custom for many years to self-administer the medically recommended 24–30 g of ethanol per day (that's for men; 12–15 g for women) to lower my risk of cardiovascular disease. My chosen dosage form of this medication is a stirred martini made of 80 mL gin plus a spritz of dry vermouth. And now they tell me that I am not getting the full health benefit if I don't shake it?

I was shocked as a consumer and skeptical as a chemist. I simply had to examine the validity of that Canadian study, if I had to go into the lab and repeat it myself.

(For those who may wish to avail themselves of my health-giving protocol, the number of grams of ethanol in any drink, whether beer, wine, or spirits, is given by: grams of ethanol = ounces of drink ?? percent alcohol by volume ?? 0.23. Thus, 80 mL or 2.7 oz of 80-proof or 40% gin contains 25 g of ethanol.)

I obtained a copy of the BMJ paper, studied it carefully, and found several potential flaws. Let's see how many of my readers can spot them as I describe the work.

• Of the six authors of the paper, the first four were undergraduates on a summer work-study program and the last two were professors. The first-named student, who conceived the project, was Christine Trevithick, daughter of the last-named author, John Trevithick. Not that there's anything wrong with that, of course.

• The students were partially supported by Labatt Breweries. Not that there's anything wrong with that, either.

• But the "martinis" were a mixture of 6 mL gin and 3 mL vermouth. No ice. While a warm, 2-to-1 mixture may be what an undergraduate assistant thinks is a martini, one might expect that the two professors know better. Most of the professors I know do.

• The shaken samples were shaken for one minute, which any bartender will tell you is an unreasonably long time.

• Antioxidant power was determined by the sample's ability to diminish the chemiluminescence of a luminol-hydrogen peroxide solution. After being shaken or stirred (for an unspecified time), 70-µL aliquots of the samples were added to a solution of 20 µL of luminol-albumin solution plus 10 µL of hydrogen peroxide, and the resulting luminosity was measured. I do have to wonder, however: Would these microliter amounts of reagents simulate the shaking and stirring of full-sized, real-life martinis? I know of no one who makes martinis with micropipettes.

Now take your choice of the possible flaws that would discredit this research. But whichever one you choose, you're wrong. The fatal flaw is that the whole thing was a hoax, a put-on.

In the prestigious, peer-reviewed BMJ? How could that happen?

What most Americans don't know is that BMJ's last issue of the year is traditionally what we would call an April Fool's issue, with offbeat "studies" masquerading as legitimate research. One paper in the December 1999 issue reported the relative frequencies of swearing by surgeons during operations (orthopedic surgeons cussed the most), while other mock studies were undoubtedly funny to doctors but cannot be described in a general-circulation magazine.

The time is long overdue to stop this "shaken, not stirred" nonsense, whether based on the "authority" of a fictitious James Bond or on bogus chemical research.

Please tell every martini drinker you know: The antioxidant experiment was a hoax.

I'll keep stirring my martinis.

Bob Wolke can be reached at

Save/Share »

Chemical & Engineering News
ISSN 0009-2347
Copyright © 2009 American Chemical Society


Save/Share »

Adjust text size:

A- A+

Articles By Topic