STATS ARTICLES 2003
Fighting Alcohol Abuse with Bad Statistics?
February 25, 2003
America has a big drinking problem and big business is to blame, according to a new study that appears in the Journal of the American Medical Association (JAMA).
A new study from Columbia University's National Center on Addiction and Substance Abuse (CASA) claims that adults who drink excessively and youths who drink illegally account for over half of the alcohol consumed in the United States, and that the alcoholic beverage industry makes too much money from these groups to ever voluntarily address the problem.
What the country needs, the authors say, is "an aggressive public health campaign similar to those that address smoking and illegal drug use." That means the federal government should raise taxes on alcohol and restrict advertising and marketing of alcoholic beverages.
There is only one problem with launching a public health campaign based on these claims: they don't add up. "Excessive drinking" sounds like it refers to people that have a serious alcohol problem. But the study significantly lowers the bar on "excessive," by defining it as any more than two drinks per day. That means that one glass of wine or beer at lunch and dinner and a brandy at bedtime makes you an "excessive" drinker.
The researchers take their baseline from government dietary guidelines on moderate drinking, which are defined as two drinks per day for a man and one for a woman. Then they simply assume that anything more can be described as excessive. But an actual scientific definition of "excessive" drinking depends on factors such as a person's body weight and metabolism.
In research on alcohol abuse that has appeared before in JAMA, seven drinks per day has been used as a baseline for heavy consumption. Moreover, recent scientific studies on the benefits of alcohol consumption have shown that there's considerably more latitude in the definition of moderate than the researchers would like to make their case.
In one study, researchers combined the results from 51 existing studies and concluded that the protective role of alcohol extended to three drinks per day. Another major study of almost a half million men and women found that even those who consumed as many as five drinks a day had lower mortality rates from all causes than those who did not drink at all.
Alcohol's biggest impact is on coronary heart disease, preventing the narrowing of coronary arteries through the buildup of plaques. In a study of over 80,000 American women, moderate drinkers had half the risk of heart attack than those who were teetotalers, even when exercise, smoking and weight were taken into account. For many years, this benefit has been described as the "French Paradox" - due to the low rate of heart disease in France despite a diet high in both fat and alcohol. But studies have shown that there is a German Paradox, an Italian Paradox and a Japanese Paradox as well.
One might say that there is a CASA paradox too, and that it involves over-estimating the impact of underage drinking on alcohol consumption despite evidence to the contrary. Their claim that underage drinkers account for 20 percent of all alcohol consumption not only rings false, it has the ring of deja vu. Almost exactly a year ago, CASA president Joseph Califano made headlines by claiming that a quarter of all alcohol was consumed by teenagers.
The next day, embarrassed news organizations ranging from The New York Times to CNN were forced to run retractions, after critics pointed out that CASA's math was wrong, and that the accepted figure among researchers in the field was about 11 percent. Now CASA are back with another figure, this time the result of combining data from two very different surveys so as to produce a higher number.
This is just the latest in a series of highly publicized studies warning the nation of the perils of drink, especially among the young. A study last September by the Center on Alcohol Marketing and Youth charged that the industry was "targeting" underage drinkers by advertising in teen-oriented magazines. However, they defined "teen-oriented" in such a broad way that it caught such unlikely offenders as Popular Science, Popular Mechanics, and People magazine.
Then in May 2002, the National Institute on Alcohol Abuse and Alcoholism released a report blaming alcohol for the deaths of 1,400 college students, and a staggering 600,000 assaults on campus each year. On closer inspection, these shocking numbers turned out to reflect some shockingly dubious assumptions.
Although the NIAAA referred to "on campus" problems, they defined "student" in a way that extended to anyone who enrolls in a college course after work for one night a week. And though most of the alcohol-related fatalities came from traffic accidents, the researchers included any accident in which the victim or anyone else had consumed any measurable amount of alcohol. In other words, a pedestrian who drank a glass of wine hours before being struck by a teetotal driver could be classified as an alcohol-related fatality.
Finally, they reached their massive count of 600,000 assaults, not from crime reports, but from a survey that asked students whether they had been "assaulted, pushed, or hit" sometime during the school year "because of other students' drinking." Only in the fight against demon drink can you cite a push from a student, who may have actually been sober, as an alcohol-related assault.
You might think that the health risks and social problems of alcohol abuse are obvious enough to require no exaggeration. But increasingly public health advocates are pursuing their agendas by first finding a substance unsafe and then suing the industry for making an unsafe product.
This is reflected in recent campaigns against alcohol, fast foods, and even guns and bullets. CASA's new study is simply another example of agenda-driven research crafted to scandalize the public and spur reform. The trick is to catch the attention of the media by identifying a legitimate problem, and then spin the numbers to exaggerate its gravity.
But why should we care if a little exaggeration leads people to act on genuine problems they might otherwise ignore? Because good intentions never justify bad science, and false or unproven claims have real social and economic costs - not least of which is betraying the public's faith in science as a guide to solving our most challenging problems.