Analysis of Newspaper Research Report

Analysis of Newspaper Research Report

Cynthia Cohen

University of Phoenix

BSN0524 - HCS438

Rebecca Cowens-Alvarado, MPH

March 25, 2007

Analysis of Newspaper Research Report

This paper will discuss an article published in Time.com (2007, March) by Lindsey Tanner, explaining a recent research study which found that the Atkins diet beat three other diets for weight loss over a one year period. The actual research study appeared in the Journal of the American Medical Association (JAMA) on March 21, 2007. Though the findings revealed greater weight loss for the Atkins participants, this paper will discuss the reasons why the study may be flawed, and its results not appropriate for making inferences regarding the public at large.

The 12-month randomized trial consisted of 311 overweight 40 year-old pre-menopausal women. The average starting weight of the women was 189 pounds. The participants were randomly assigned to follow one of four diets: Atkins, Zone, Ornish, or LEARN {Lifestyle, Exercise, Attitudes, Relationships, and Nutrition} (Anderson, 2007). The LEARN diet is similar to United States (U.S.) dietary guidelines (Tanner, 2007). The Atkins diet is a low-carbohydrate, high-fat diet, while the remaining diets favor low-fat, higher carbohydrate intake.

Once assigned, the women read a book about their respective diet, then attended a weekly class for eight weeks, after which they were on their own. The study is reported as correlational research; women form each diet group were measured for the following variables: weight loss, HDL (high density lipoprotein) cholesterol, and blood pressure. At the end of one year, the Atkins group lost an average of 10 pounds; the Zone dieters 3.5 pounds; the Ornish dieters almost 5 pounds, and the LEARN diet almost 6 pounds (Tanner, 2007). The Time.com article reports that "the Atkins women had slightly higher HDL cholesterol, the good kind, and slightly lower blood pressure than the other three diets" (Tanner, 2007, p.2). The lead researcher believes the weight loss likely contributed to these findings.

The null hypothesis for this study is that weight loss (the primary outcome) in the Atkins group will = weight loss in the other diet groups. The research (alternate) hypothesis for the study is that the women following a low carbohydrate diet (Atkins group) will lose more weight than those following the other three diets. As per the Tanner article, the JAMA study found the weight loss in the Atkins group to be statistically significant, thereby causing researchers to reject the null hypothesis and accept the alternate hypothesis. There are however, several issues with this study that challenge its validity.

According to the study abstract in JAMA (2007, March), researchers used a significance level of p < .05 to determine that the weight loss difference between subjects was statistically significant. If the significance level is only marginally acceptable, one must challenge the true statistical significance. Since the women originally weighed an average of 189 pounds, even a mean loss of 10 pounds on