Lies, Damn Lies, And Product Research


By Cynthia Crossen

Simon & Schuster x 272pp x $23

In the late 1980s, few environmentally concerned parents could use disposable diapers without feeling guilty. Just imagine all that plastic, paper, and chemical goo--not to mention the contents of the soiled diapers--sitting undegraded in landfills for untold years!

Then, in 1990, Procter & Gamble Co. said we could relax. According to a study funded by the maker of Pampers and Luvs, when one evaluates all the environmental costs of making, using, and discarding disposable and reusable diapers, the disposables are no worse than their cotton rivals. The same year, a study sponsored by the disposable-friendly American Paper Institute--which factored in such variables as the energy impact of exporting cotton to China to make diapers--reached the same conclusion. The two studies neatly parried previous disposable-damning research--which had been backed by cloth-diaper makers.

As Cynthia Crossen repeatedly says in Tainted Truth: The Manipulation of Fact in America, a shrewdly concocted study can prove virtually anything. The pro-disposable studies, she reports, helped kill legislation nationwide that would have imposed bans, taxes, or warning labels. Says a policy analyst she quotes: "It went from being one of the most obvious examples of a choice where one was clearly right and one was wrong to this completely muddled argument."

Dissecting studies done by and for environmentalists and the makers of drugs, breast implants, and foods, Crossen shows how often self-interest seems to dictate the outcome. The parameters of research are malleable, even within ethical limits, she notes, and buyers of research need publicize only favorable findings. A Wall Street Journal editor, Crossen also lambastes the media for uncritically reporting studies. When it comes to nutrition research, she notes, "the more the study defies common wisdom, the more likely it is to enjoy wide acclaim." Among the examples she cites:

-- Eating white bread won't make you gain weight. This was the finding of a study in which some subjects added eight slices of bread to their daily diet for eight weeks. The sponsor of this revelation? Wonder Bread.

-- Chocolate may inhibit cavities, said the Princeton Dental Resource Center, citing a study indicating tannins in cocoa inhibit plaque formation. Who funds the center? M&M/Mars Inc.

-- Almonds, walnuts, rice bran, and citrus all aid the cardiovascular system. So say separate studies touted by the Almond Board of California, the California Walnut Commission, the Rice Council, and the Florida Citrus Commission.

Crossen also explores cases in which dueling studies have weighty policy or public health implications. Consider the tobacco industry's research group, which since 1954 has doled out more than $165 million to 800 scientists who overwhelmingly tended to reassure smokers. The flood of data helped delay for years the reckoning tobacco merchants now face.

Drugmakers are another of Crossen's targets. In one case, she reports, feuding University of Pittsburgh researchers reached opposite conclusions about the treatment of children's middle-ear infections. Reviewing the same data, one pronounced the antibiotic amoxicillin effective. The other disagreed. The first scientist's lab had collected $1.6 million in grants from drugmakers, while the critic had refused such money. The opposite findings hinged on different study durations, different testing instruments, and different interpretations of results.

The study endorsing the drug's effectiveness was accepted by the prestigious New England Journal of Medicine as the version "authorized" by the university--itself a recipient of much money from drugmakers, according to Crossen. The critic's study wasn't published for nearly five years, until 1991, when the Journal of the American Medical Association ran it. As it turned out, its findings were bolstered by a broad, 3,600-child study that found higher rates of recovery in those who didn't get antibiotics.

Slicing data differently can make the truth remarkably pliable--and money makes such recasting all too alluring, Crossen persuasively warns. Onetime biotech star Centocor, she notes, saw the fortunes of its $3,500-a-dose medicine for septic shock fade when it couldn't show effectiveness 14 days after treatment. When marginal improvement seemed to show up 28 days after treatment, it asked regulators to change the study's endpoint. The Food & Drug Administration declined, and in January, 1993, a new trial vanquished the drug: More patients died using it than died on a placebo.

Some studies are more valid than others, of course, just as some researchers are more honest or more competent than others. Yet Crossen is frustratingly reluctant to choose between opposite findings. True, a detailed examination of the dozens of studies she refers to--on everything from the role of milk in juvenile diabetes to the rate of heart attacks among bald men--might be impractical and would make an already dry book downright arid. But her noncommittal stance is also a cop-out, since she asks the media to more aggressively separate the wheat from the chaff.

Still, Tainted Truth amply illustrates how easy it is to dress up a one-sided sales pitch with good-looking numbers and spin a story whichever way is convenient. No matter how impressive a new study seems, clearly caveat emptor must be the rule.

Before it's here, it's on the Bloomberg Terminal.