VitalSTATS

Can’t Kick That Can?

by Howard Fienberg
September 2000

Recent research pointed to an insidious plot in the cola industry. Caffeine in cola “is more likely to cause addiction than improve the taste of the products,” according to the New Scientist (Aug. 19). Other newspapers offered similar treatments: “Colas add caffeine for habit, study says” (USA Today, Aug. 15); “Caffeine in soft drinks: added to be addictive?” (The Washington Post, Aug. 15). But between gulps, there were several big warning signs missed by the news media making the study a lot less newsworthy.

This study, published in the August Archives of Family Medicine, found that cola fans could not differentiate the taste of caffeinated and non-caffeinated cola. Dr. Roland R. Griffiths and his coauthors concluded that caffeine must be a nefarious additive designed to hook us into persistent cola usage.

Truth be told, this study suffers from a few key problems. A small number of subjects saves a lot of time and money, but having only 25 regular cola drinkers in your study does not inspire much confidence. Griffiths defended this small number by pointing out that other comparable studies used even fewer subjects – the equivalent of justifying your jumping from a bridge by pointing out all your friends who did it too. Media sources mostly reported the size of the study but failed to emphasize how such a small sample size damns the results.

The testing methods also appear flawed. The “secret” recipes of colas like Coke and Pepsi present a barrier to any substantial research of their contents. Griffiths and his coauthors admit that “we tested the role of caffeine by adding caffeine to a commercially available caffeine-free beverage... It is possible that other changes in the formula of this beverage render it less sensitive to the effects of caffeine.” So it is still not clear in this case how caffeine affects cola’s taste. That only two of the 25 subjects could tell the difference does not mean much in this case.

Upon finding these murky results, the researchers appear to have donned caramel-tinted glasses. They concluded that everyone should “recognize the possibility” that infatuation with cola is “more likely to reflect the mood-altering and physical dependence-producing effects of caffeine... than its subtle effects as a flavoring agent.” The most common conclusion in scientific research papers is ‘more research is required.’ We should therefore be wary when broad policy recommendations, substantiated or not, appear in the paper itself rather than in the editorial and commentary section of the journal.

In this case, the editor commented in a box at the end of the article that, “While this study was not definitive proof... I believe it is basically true.” Her belief rested in her family’s uncontrolled home experiments. News media seemed to make a similar judgment. But while snappy conclusions yield juicy headlines, they only add to a sprawling and contradictory mix of nutrition news. As observed by science writer Aries Keck, “Consumers just don’t believe anything anymore. First salt is bad, then salt is ok, then butter is bad, now butter is ok again.”

Next time, there might be more to cry about than a little spilled cola and we might find that no one wants to listen.


return to Howard Fienberg's page