Read: Don’t Believe

In early February, a startling headline jumped off a page in the Boston Globe: "Study Finds No Major Benefits of Low-Fat Diet."

Hold on. Haven't we been told for years to eat less fat to keep our hearts healthy and our waistlines somewhat less than equatorial? On closer inspection, this sort of sensationalist headline was indeed misleading. The multiyear study, designed specifically to track the ability of a low-fat diet to prevent breast cancer, colorectal cancer, and heart disease, monitored the effects of diet only on postmenopausal women ages fifty to seventy-nine. What's more, it didn't differentiate between the various kinds of fats these women consumed -- olive oil and cheesecake counted just the same.

Hence, this rule of thumb regarding health studies and their headlines: "If it seems too good to be true," warns Dr. Jeannie Moloo, a registered dietitian and a spokesperson for the American Dietetic Association (ADA), "chances are it's not."

Sad news, but sound advice. Science is complex and speculative, and it's ever changing. Gina Kolata, science reporter for the New York Times, contends that most of what we read in the media about science and health matters consists of oversimplified syntheses of hypothesis-generating studies, not the final answer about any disease, diet, or food. But when coffee is touted as a preventive measure against alcohol-induced cirrhosis of the liver, it's awfully tempting to down a mocha Frappuccino and head straight for the pub.

What should you do instead?

"Start to think like a detective," advises Roberta Anding, a dietitian at the Texas Children's Hospital Sports Medicine Clinic and spokesperson for the ADA. Get out your magnifying glass and ask yourself these questions:

* Who were the research subjects? "If it's an animal-based study, wait," Kolata cautions. (Rats, after all, aren't people.) Likewise, if it's a comparative study between two very large and diverse populations, be wary -- it's impossible to rule out the effects of every variable. How did the researchers get their information? Results for observational studies (ones that track participants' health without controlling their behavior) are likely to be subjective or skewed. The same goes for surveys in which participants report on their own habits and health. Trust the stats only if it is a randomized, double-blind trial -- that is, one in which participants are randomly assigned to receive the drug or diet in question or a placebo (a neutral substitute), and where neither the researchers nor the subjects know who's in which group. How many people participated? A measly 200 subjects, or a meaty 200,000? More is better.

* How long did the study take? Even the low-fat-diet study, which spanned eight years, may not have been long enough to assess a diet's long-term effects.

* Who paid for it? Drug trials conducted by the pharmaceutical company making the drug, for example, may be suspect. (Case in point: The harmful effects of a Merck-designed arthritis drug called Vioxx were found in May 2006 to have been underrepresented in an earlier, manufacturer-conducted trial.)

* What is not mentioned? If any of the above is left out (and it often is), run up the red flag.

Ultimately, professionals suggest that consumers be patient. When it comes to health, only overall trends that have emerged over time -- such as the need to reduce weight and increase exercise -- are worth making lifestyle changes for.

Read the articles with interest, but resist jumping to conclusions. "The basic thing that people need to understand," says Dr. Catherine DeAngelis, editor in chief of the Journal of the American Medical Association, "is that it is rare -- extremely rare -- that any one study has all the answers." At best, she adds, a landmark study is "perhaps one small piece of a very, very large puzzle."

Get Started

The best way to get correct information is to search for it. Often, the Web sites of medical journals, hospitals, medical schools, nonprofit research organizations, and national institutes are great resources regarding a specific health study or health concern.

• The American Cancer Society
• The American Diabetes Association
• The American Dietetic Association
• The American Heart Association
• Johns Hopkins Medicine
• The Journal of the American Medical Association
• The Mayo Clinic
• The New England Journal of Medicine
• The U.S. Centers for Disease Control and Prevention
• The U.S. Food and Drug Administration
• The University of California San Francisco School of Medicine

Sara Bernard is a former staff writer and multimedia producer for Edutopia.