Diets Don’t Work? Or Does Bad Science Generate Bad Data?

Dr. Sears' Blog: Diets Don’t Work? Or Does Bad Science Generate Bad Data?

One of the problems with diets is their focus on weight loss, not improved health.  A recent study in Circulation states that it doesn’t matter what diet you follow for weight loss (1).  On the other hand, we have the American Heart Association diet, the American Cancer Association diet, and the American Diabetes Association diet.  I guess these diets don’t work either since all of these chronic diseases are still the primary causes of death in the United States.

The word diet comes from the ancient Greek root meaning “way of life”.  That is the basis of the Zone Diet.  It is not a weight-loss diet but a way of life to reduce inflammation.  From that perspective, I can now analyze the latest meta-analysis study that states that popular diets don’t work.

The problem with any meta-analysis study is the choice of included studies and correspondingly the omission of germane studies that often make the final conclusions of the meta-analysis highly suspect.  The recent article in Circulation clearly illustrates this point (1).

To begin with, the authors dismissed discussing one of the four popular diets (Ornish) because they apparently didn’t wish to do any exhaustive searching of the published database under a different name as explained in their Method Section.  Nor did they attempt to seek out the wide number of published studies on low-fat, low-protein, high-carbohydrate diets that would be considered Ornish-like diets. Thus the real manuscript is the comparison of three popular diets, not four, as mentioned in the title of the article (1).

Unfortunately, the failure of the authors to do an exhaustive search of all germane diet studies that used the same macronutrient composition of the Zone Diet (40% low-glycemic-load carbohydrates, 30% protein, and 30% fat consisting of primarily monounsaturated fat) removed many important published studies (2-5) that if included, would have dramatically changed their conclusions.  Their simplistic use of a search profile using the word “Zone Diet” therefore provides only two studies on the Zone Diet (6,7).

The key factor for any diet is its ability to modify cardiovascular risk. Today it is well accepted that inflammation is the primary risk factor associated with heart disease.  Elevated insulin levels are also a consequence of increased inflammation.

In fact, one of the Zone Diet studies mentioned in the analysis demonstrated significant reduction in insulin levels that had a far greater statistical significance (p < 0.001) than in the other compared diets, even though there were no statistical differences in weight loss compared to other diets studied (6).  This fact was not mentioned in the meta-analysis (1).  Other non-included Zone Diet studies have demonstrated such statistically significant in reductions of inflammation, although the weight loss was similar compared to a control diet (2,3).

Another problem with the meta-analysis was that neither of the two mentioned “Zone Diet” studies was even close to meeting the macronutrient criteria of the Zone Diet (6,7).  A key feature of the Zone Diet is maintaining the protein level approximately equal to 30% of total calories. Neither of the included “Zone Diet” studies ever reached that protein level.  In fact, in one of the included studies the macronutrient ratios of virtually all the different diets studied at 12 months were remarkably similar, demonstrating significant regression to the mean (7).  The implication is that unless the researchers carefully control the macronutrient composition of a tested diet, it is impossible to make any meaningful statements about that diet.  Using that criteria, none of the “Zone Diet” studies should have been included for analysis because they were poorly executed.  It should be noted that many of the excluded Zone Diet studies did maintain fairly successful macronutrient maintenance (2-5).

The importance of the macronutrient control of the diet is highlighted in a 1999 study conducted by Harvard Medical School that demonstrated dramatic changes in hormonal levels after a single meal with differing macronutrient compositions in a highly controlled cross-over study (8).
Perhaps not surprisingly, the dietary recommendations of the Joslin Diabetes Research Center at Harvard Medical School to treat obesity and diabetes are remarkably similar to the Zone Diet (9).  This is because it is the reduction of inflammation that governs of true health consequences of following any defined diet, not weight loss.

This highly flawed meta-analysis only serves to further confuse physicians and the general public.  This is why many meta-analysis studies can be compared to making sausage.  This study is a particularly good example of that analogy.


1.  Atallah R et al.  “Long-term effects of 4 popular diets on weight loss and cardiovascular risk factors.”  Circulation Cardiovas Qual Outcomes doi; 10.1161/CIRCOUTCOMEs.113.000723 (2014)

  1. Pereira MA et al. “Effects of a low-glycemic load diet on resting energy expenditure and heart disease risk factors during weight loss.” JAMA 2004 292:2482-2490 (2004)
  2. Pittas AG et al. “The effects of the dietary glycemic load on type 2 diabetes risk factors during weight loss.”  Obesity 14:2200-2209 (2006)
  3. Lasker DA et al.  “Moderate-carbohydrate, moderate protein weight-loss diet reduces cardiovascular disease risk compared to high-carbohydrate, low-protein diet in obese adults: A randomized clinical trial.” Nutr Metab 5:30 (2008)
  4. Layman DK et al.  “A moderate-protein diet produces sustained weight loss and long-term changes in body composition and blood lipids in obese adults.”  J Nutr 139:514-521 (2009)
  5. Dansinger ML et al. “Comparison of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction: a randomized trial.”  JAMA 293:43-53 (2005)
  6. Gardner CD et al.  “Comparison of the Atkins, Zone, Ornish, and LEARN diets for change in weight and related risk factors among overweight premenopausal women: the A TO Z Weight Loss Study: a randomized trial.”  JAMA 297:969-977 (2007)
  7. Ludwig DS et al.  “High glycemic index foods, overeating, and obesity.”  Pediatrics 103:E26 (1999)
  8. (2007)

Tags: , , ,

About Dr. Barry Sears

Dr. Barry Sears is a leading authority on the impact of the diet on hormonal response, genetic expression, and inflammation. A former research scientist at the Boston University School of Medicine and the Massachusetts Institute of Technology, Dr. Sears has dedicated his research efforts over the past 45 years to the study of lipids. He has published 40 scientific articles and holds 14 U.S. patents in the areas of intravenous drug delivery systems and hormonal regulation for the treatment of cardiovascular disease. He has also written 14 books, including the New York Times #1 best-seller "The Zone". His books have sold more than 6 million copies in the U.S. and have been translated into 22 different languages.

Leave a Reply

Your email address will not be published. Required fields are marked *