Intended for healthcare professionals


Shaken, not stirred: bioanalytical study of the antioxidant activities of martinis

BMJ 1999; 319 doi: (Published 18 December 1999) Cite this as: BMJ 1999;319:1600
  1. C C Trevithick, research assistant,
  2. M M Chartrand, research assistant,
  3. J Wahlman, research assistant,
  4. F Rahman, research assistant,
  5. M Hirst, professor,
  6. J R Trevithick, professor (trevjohn{at}
  1. Department of Biochemistry, Faculty of Medicine and Dentistry, University of Western Ontario, London, Ontario, Canada N6A 5C1
  1. Correspondence to: J R Trevithick


    Background: Moderate consumption of alcoholic drinks seems to reduce the risks of developing cardiovascular disease, stroke, and cataracts, perhaps through antioxidant actions of their alcohol, flavonoid, or polyphenol contents. “Shaken, not stirred” routinely identifies the way the famous secret agent James Bond requires his martinis.

    Objectives: As Mr Bond is not afflicted by cataracts or cardiovascular disease, an investigation was conducted to determine whether the mode of preparing martinis has an influence on their antioxidant capacity.

    Design: Stirred and shaken martinis were assayed for their ability to quench luminescence by a luminescent procedure in which hydrogen peroxide reacts with luminol bound to albumin. Student's t test was used for statistical analysis.

    Results: Shaken martinis were more effective in deactivating hydrogen peroxide than the stirred variety, and both were more effective than gin or vermouth alone (0.072% of peroxide control for shaken martini, 0.157% for stirred v 58.3% for gin and 1.90% for vermouth). The reason for this is not clear, but it may well not involve the facile oxidation of reactive martini components: control martinis through which either oxygen or nitrogen was bubbled did not differ in their ability to deactivate hydrogen peroxide (0.061% v 0.057%) and did not differ from the shaken martini. Moreover, preliminary experiments indicate that martinis are less well endowed with polyphenols than Sauvignon white wine or Scotch whisky (0.056 mmol/l (catechin equivalents) shaken, 0.060 mmol/l stirred v 0.592 mmol/l wine, 0.575 mmol/l whisky).

    Conclusions: 007's profound state of health may be due, at least in part, to compliant bartenders.


    James Bond, the well known fictional secret agent (“007”) of the British intelligence services, not only is astute in matters of clandestine affairs at a personal and international level but may also possess insights of interest to medical science. Take for example his insistence on having his martini “shaken, not stirred.” Does this straightforward direction to the barman merely yield a crisper drink, more to James Bond's taste, or is there more to it?

    Moderate consumption of alcoholic beverages has been associated with a decreased risk of several age related diseases, including cardiovascular disease, 1 2 stroke,3 and cataract. 4 5 This effect has been tentatively ascribed to the antioxidant activities of alcohol,6 flavonoids, or polyphenols7 in the beverages, since it has been established that the antioxidant vitamin E reduces the risk of cardiovascular disease 8 9 and cataract development.10

    Embedded Image

    Is it martinis that help James Bond stay so healthy?



    Mini-martinis were prepared by mixing two parts (vol/vol) gin (6 ml) with one part vermouth (3 ml). They were either shaken vigorously (9 ml in a 100 ml medicine bottle for one minute), or stirred (9 ml in a 20 ml glass vial, using a vortex mixer). Aliquots of the martinis were then added into a luminescent assay 11 12 to see if they altered the luminescence resulting from the addition of a standard amount of hydrogen peroxide. In both cases, the addition of the martinis decreased the net luminescence to a small percentage of control values. When we analysed the net luminescence statistically, using the t test function of Microsoft's Excel spreadsheet program, the luminescence remaining after we added the shaken mixture to the peroxide was virtually half that afforded by the stirred mixture (table 1, P=0.0057) This indicates that there was twice as much peroxide remaining after treatment with the stirred martinis than after the shaken variety Thus, shaken martinis are better able to “neutralise” peroxide than stirred martinis.

    The normalised luminescent count rate is reduced by the addition of martinis Counts per minute were obtained with a Lumac Biocounter M2010 (Celsis, Chicago) when hydrogen peroxide was incubated with the luminol bound to albumin, as described below. Samples consisted of an aqueous portion (0.3 ml), and a dimethyl sulphoxide portion (0.4 ml). The aqueous portion contained luminol and albumin (0.02 ml, 10 mg/ml of each) prepared as described, 11 12 and phosphate buffered saline (0.11 ml) containing martini (0.07 ml) or distilled water (for control). Hydrogen peroxide (0.1 ml of 1.0%, final concentration 42 mmol/l) was added after the addition of dimethyl sulphoxide (0.4 ml), just before the tube was placed in the counter to begin counting of the emitted light. We determined the probability that the means of each set of samples are identical by using Student's t test in the Excel program.

    We considered that a difference might arise from air oxidation of reactive martini components during the vigorous shaking. This may, however, not be the reason because when, as an alternative to shaking or stirring, we vigorously bubbled air or nitrogen at similar rates through the martinis (for one minute), the two treatments showed no significant difference in net luminescence (table 1) Nevertheless, the martinis bubbled with both air and nitrogen showed net luminescent counts equivalent to the shaken rather than to the stirred martinis.

    Table 1.

    Remaining luminescence after addition of martinis to the luminescent assay containing peroxide

    View this table:


    To ascertain the relative contribution of the gin and vermouth components, both were assayed in a preliminary experiment for their abilities to reduce luminescence produced by the peroxide challenge. Vermouth was much more potent, causing a 98.1% (SE 0.5%) (n=3) decrease in count rate, while gin reduced the count rate after challenge with peroxide by only 41.7% (14.1%) (n=36). This implies that the vermouth contributes more to the antioxidant properties of martinis. Even so, the combination of gin and vermouth is better than either gin or vermouth alone, resulting in a much lower net luminescent count rate (0.072% of peroxide control for the shaken martini) than those found after either gin (58.3% of control) or vermouth (1.9% of control) alone. The remaining luminescence with martinis is substantially lower than that of the vermouth itself.

    Since much of the antioxidant activity of wine and whisky has been ascribed to the polyphenols they contain,7 the polyphenol content in the martinis was investigated using Folin reagent.7 As shown in table 2, the phenolic concentrations in the martinis, in catechin equivalents, were an order of magnitude lower than those in white wine or 12 year old Scotch whisky, and there was no significant difference between the phenolic contents of shaken and stirred martinis.

    Table 2

    Polyphenol content of martinis compared with white wine and blended Scotch whisky

    View this table:

    The martinis, when undiluted, are capable of suppressing counts from 42 mmol/l peroxide by over 99.9%. We calculate that after ingestion an absorbed martini may be able to react with 210 μmol/l of hydrogen peroxide. We have previously determined that 5 mmol/l ethanol, a blood concentration of ethanol found after absorption of one or two typical alcoholic drinks, would eliminate 131 μmol/l peroxide.6 The peroxide concentrations detected in the aqueous humour have ranged from 14 μmol/l to 31 μmol/l: mean 24 (SE 7) μmol/l for normal humans, with higher concentrations in cataract patients (82 (155) μmol/l and 198 (88) μmol/l13) Both vitamin E and ethanol decrease the risk of cataract 4 5 and atherosclerosis 1 3 by about half. The residual peroxide concentrations in the aqueous humour possibly reflect those in the serum, from which the aqueous humour is formed by ultrafiltration at the ciliary body After the consumption of shaken martinis, peroxide concentrations of serum and aqueous humour could be half those found after ingestion of stirred martinis.


    Although the reason for the superior antioxidant activity of shaken martinis is not clear, is it possible that James Bond chose shaken (not stirred) martinis because of the improved antioxidant potential? This added antioxidant effect could result, of course, in a healthier beverage. There is no indication in the literature that 007 suffered from cataracts or cardiovascular disease, hence he must be considered a moderate consumer of alcoholic drinks. The authors have not examined any antioxidant contributions from olives.


    Contributors: CCT originated the idea and performed preliminary experiments on the antioxidant activity of martinis. FR and MMC performed the Student's t tests using the Excel spreadsheet program, and MMC performed the Folin phenol content determinations and statistical analysis. JW and FR, with assistance from Darin Lawrence, Adrian Lee, and Ashley MacDonald, prepared the mini-martinis and performed antioxidant assays using the luminometer. MH and JRT coordinated the study, aided in the statistical analysis, suggested appropriate tests and controls to perform in group meetings, and were mainly responsible for writing the paper. CCT and MMC suggested editorial changes to the text. JRT and MH are guarantors of the paper.


    • Funding Except for MH and JRT, all staff on the project were summer students supported by Work Study, Canada Manpower, Youth Opportunities Unlimited Ontario, and by grants from Labatt Breweries to MH and JRT Corby Distilleries provided samples of gin and vermouth.

    • Competing interests The research grants from Labatt Breweries were used for a portion of the laboratory supplies, a portion of expenses incurred by CCT, MMC, JW, and JRT in attending the conference of the Association for Research in Vision and Ophthalmology (1999), and a portion of the expenses of MH in attending the fourth international conference on toxicology in developing countries (1999).


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    View Abstract