The Dietary Approach to Obesity: Title and subTitle BreakIs It the Diet or the Disorder?

Individuals seeking to lose weight have many diet choices, ranging from very low-fat (Ornish) to very high-fat (Atkins). However, despite great popularity and millions of dollars of sales, good-quality comparative data on the safety and effectiveness of these diets remain elusive.^{1 - 2} The study by Dansinger et al³ in this issue of JAMA could be the randomized trial many have been waiting for to help determine which diet works best. Yet, even after considering a number of important limitations, the primary outcome of weight loss at 1 year was modest and there was no difference between the low-carbohydrate approach of Atkins, the high-protein low-glycemic load approach of Sears (the Zone diet), the very low-fat approach of Ornish, and the low-calorie/portion-size approach of Weight Watchers.

The results of the study by Dansinger et al³ appear to be at odds with findings from a number of recent studies that demonstrated more weight loss, at least during the initial months, with low-carbohydrate diets.^{4 - 7} One important similarity between this trial and the other studies is a high dropout rate. In the study by Dansinger et al,³ 42% of the study participants were no longer adhering to their diets by 12 months after randomization, and those participants who did complete the trial had a weight reduction of only 5 kg or 5% from baseline weight. Adherence rates were poor for all 4 diets but were particularly poor for the Atkins and Ornish groups. In the setting of a randomized trial, in which there was at least initially substantial involvement of the study team, the results among adherent participants would have been expected to be better. After all, some studies suggest that obese patients can lose weight on their own. Data from the National Weight Control Registry indicate that some patients who successfully lose at least 10% of their weight on their own sustain the reduction for 5 years.⁸ However, Heshka et al⁹ demonstrated that the dropout rate in a self-help weight loss group was similar to that experienced in a commercial weight loss program, but the amount of weight loss was less in the self-help group. Thus, perhaps more structure and input from health care professionals does facilitate the outcome, at least over relatively short intervals.

Even 1 month after randomization in the study by Dansinger et al,³ adherence to the recommendations of the Atkins, Ornish, and Zone diets was highly variable. Participants in the Atkins diet group were ingesting a median of 68 g of carbohydrate daily, not 20 g as recommended at this stage of the diet. Participants in the Ornish diet group were consuming a median of 27 g of fat or 17% of total calories after 1 month, well above the target of 10%. By 1 year, calories from fat accounted for 32% of total intake. For participants in the Zone diet group, carbohydrate intake accounted for 45% of total calories by 1 month, identical to the target of 45%, while fat content was 34% of calories, slightly above the target of 30%.

An additional concern historically related to the dietary treatment of obesity is estimated caloric intake at baseline. The reported average caloric intake was only 2059 kcal for participants whose mean weight was 100 kg (220 lb). This is well below a very conservative estimate of 2860 calories (calculated as basal calories [weight in pounds × 10] + lifestyle activity level [weight in pounds × 3]) for an individual younger than 50 years who weighs 220 lb with a sedentary lifestyle.^{10 - 12} Even with a further 10% downward adjustment for an individual aged 50 years or older, the estimated caloric intake of 2574 is more than 25% above the average estimated by this study population (average age of 49 years).^{11 - 12} This underestimate is not surprising, given that diet records are notoriously inaccurate in obese patients.¹³

Losing weight and sustaining weight reduction are very difficult. Most health care professionals who treat obese patients would acknowledge that there is insufficient knowledge about mechanisms for recidivism in obesity. Beyond a lack of willpower is a plethora of biological explanations that challenge the weight-reduced state. These include increased appetite with a preference for energy-dense foods, reductions in energy expenditure (both basal metabolic rate and physical activity), increased insulin sensitivity, increased respiratory quotient due to more carbohydrate oxidation and less fat oxidation, and tissue-specific changes in the activity of lipoprotein lipase (more in adipose tissue and less in skeletal muscle), an enzyme that promotes fat storage vs fat oxidation.¹⁴

It seems plausible that for maintenance of reduced body mass, the right diet needs to be matched with the right patient. Ultimately, a “nutrigenomic” approach most likely will be helpful. At present, there are no data to help clinicians practicably match a diet to an individual patient’s “diet response genotype.” Even beyond this consideration, and arguably more important, once weight loss of more than several kilograms from baseline weight occurs, a substantial step-up in the amount of physical activity and conscientious monitoring as part of a more comprehensive behavior modification appear particularly important,^{15 - 16} and likely are much more relevant adaptations than the macronutrient composition of the diet.

Are there other considerations about dietary macronutrient composition and health benefits with weight reduction? In the study by Dansinger et al,³ changes in plasma lipid levels were as expected (ie, more low-density lipoprotein [LDL] and high-density lipoprotein [HDL] cholesterol lowering among participants in the Ornish diet group¹⁷ and more reduction in triglyceride levels among participants in the Atkins diet group,^{4 - 7} and presumably in the Zone diet group). Because the LDL/HDL cholesterol level ratio was similarly affected by all 4 diets and at almost identical amounts of weight loss, these differences in plasma lipids and lipoprotein levels between diets may have little clinical meaning. Moreover, there is essentially no evidence that less reduction in triglyceride levels with a higher carbohydrate diet has any related long-term adverse health consequences.

What are the next critical steps necessary to help achieve successful and sustained weight loss? It seems that more trials of diets and weight loss that examine the effects on weight reduction and the many biomarkers of the comorbidities of obesity at relatively short intervals are not needed. What is truly needed now is evidence that weight loss by diet (and exercise and behavior modification) along with risk-factor improvements can be achieved and sustained for 5 to 10 years. Given the results of the study by Dansinger et al,³ these may be difficult goals. Next, it is important to determine whether diet and other lifestyle interventions affect hard outcomes, such as death, myocardial infarction, cancer incidence, and stroke. The ongoing National Institutes of Health Look AHEAD (Action for Health in Diabetes) study,¹⁸ which will assess the long-term effects (<11.5 years) of an intensive weight loss program delivered over 4 years in overweight and obese individuals with type 2 diabetes mellitus, may be helpful. However, this study is being performed in patients with type 2 diabetes mellitus in which statistical power is more favorable with a smaller sample size (approximately 5000). Even if this study shows favorable outcomes, the results will not be transferable to obese patients without diabetes mellitus.

Arguably, the best treatment of obesity is prevention by careful dietary monitoring and lifestyle choices, along with regular physical activity. Once overweight or obesity develops, however, the best existing evidence points toward heeding the recently released joint lifestyle recommendations of 3 professional organizations: the American Cancer Society, the American Diabetes Association, and the American Heart Association, in which the recommended macronutrient mix is built on evidence that higher intake of fruits and vegetables, whole grains, and fish are associated with reduced incidences of diabetes mellitus, cancer, heart disease, and stroke.¹⁹ Although this dietary approach may lead to only modest weight changes, similar to the popular diets evaluated by Dansinger et al,³ physicians and other health care professions should teach obese patients that both quality and quantity of the diet are important, and that sustained weight loss may well be possible with the addition of physical activity and behavioral change strategies to a modest but persistent caloric restriction—the “Low Fad” approach.