In Brief

A new study of more than 450 adolescent girls finds that those who took extreme dieting measures, were depressed or who reported having an obese parent were more likely to become obese four years later. Notably, the Journal of Consulting and Clinical Psychology (Vol. 73, No. 2) study found that the girls' self-reported binge eating, high-fat food consumption and exercise didn't predict obesity.

The researchers, lead by University of Texas at Austin psychologist Eric Stice, PhD, say the findings don't controvert experts' assertions that exercise and healthy eating prevent obesity. Rather, the team says its findings call into question the accuracy of the self-report measures researchers commonly use to measure people's eating and exercise habits.

The study followed 496 adolescent girls over four years. Each year, the researchers measured the teens' height, weight and depressive symptoms and administered questionnaires about their exercise, dieting and eating habits. The girls also reported if one or both of their parents was obese.

Girls who became obese were far more likely to report having an obese parent, dieting and employing radical weight-loss measures such as purging and using laxatives. The researchers were also more likely to identify them as being depressed using structured interviews and an adapted version of the Schedule for Affective Disorders and Schizophrenia for School-Age Children.

Some experts have taken similar research findings to mean that dieting behavior may increase the risk of becoming obese. But Stice disagrees. Instead, he argues that the self-report dieting and exercise measures are identifying people struggling with their weight. Why? Research has shown that people consistently underreport how much they eat: What many people consider to be "dieting" actually isn't intense enough to reduce their weight.

"My suspicion is that the people who are engaging in effective weight-control behaviors don't think of themselves as dieters," says Stice. So, the self-report dieting measures miss them, he says, and instead identify overeaters unsuccessfully trying to battle the bulge.

That's still an important group to identify, he notes, since his study's results do indicate that people the dieting measures identify--those who repeatedly diet, binge or take more extreme weight-loss measures--may especially benefit from tailored obesity-prevention programs.

As a result, Stice and his colleagues are studying whether intervention trials that use psychosocial principles reduce both weight and poor dieting strategies. A study by Stice and his colleagues in press in Health Psychology finds that such a program does in fact reduce bulimic behavior, Stice says.

Overall, the results suggest that researchers should use more objective measure of healthy eating and exercise than self-reports. For example, Stice has conducted studies that weigh participants' food before and after they eat; other researchers have used "doubly labeled water"--water laced with an isotope that indirectly measures participants' caloric intake.

--D. SMITH BAILEY