A diet by any other name is still about energy.

The obesity epidemic has fostered increasing interest among manypeople to seekeffective treatment strategies.Relatively few research studies have compared the effectiveness of various diets with different macronutrient composition, and even fewer studies have comparednamedor branded (trade-marked)weight lossprogramsusingcomparable,wellcontrolledassessmentmethodsandoutcomemeasures.There is significant interest among the public as well as health care professionals regarding theefficacy, safety, and long-termfeasibility of adhering to these branded diets for the purposes of guiding decisions regarding the best choices. In this issueofJAMA, Johnstonandcolleagues1 report findings fromanetworkmeta-analysisdesigned toassess thecomparative effectiveness of branded competing diets that have theavailabilityofpublisheddata fromrandomizedclinical trials (RCTs). Networkmeta-analysis offers amethod of estimating the relative effectiveness of available diets in the absence of direct comparisonsbyusing indirect comparisonsof 2diets vs common comparators. This is helpful for standardizing these analyses in a manner that is unlikely to be conducted by the commercial diet vendors and also offers results that can help manage expectationswhen attempting to choose one dietary approach over another. This carefully conductedmeta-analysis describes eligibility criteria for inclusion,durationand interventiondetails, and specific outcomes regardingweight loss at 6 and 12months of follow-up.1 The search strategy, study selection process, and data extraction system further lend credibility to this review and analysis. The choice of comparators includingwait-listed controls and no previously assigned diet or competing dietary program is sound. The use of the Lifestyle, Exercise, Attitudes, Relationships, and Nutrition (LEARN) dietary programfirstdescribedbyBrownell2 reflectsa reasonablestandard for comparison. The approaches to data analyses and detailed discussion of the confidence estimates and assessment of publication bias further attest to the strong study design. A total of 48 RCTswere identified,met eligibility criteria, and were evaluated (N = 7286 individuals). Of these studies, 43 reported weight loss at 6 months and 25 trials reported weight loss at 12 months. Ultimately, Johnston et al1 concluded that both the low-carbohydrate and low-fat dietswere associated with average weight losses of approximately 8 kg at 6 months and 6 to 7 kg at 12 months compared with no diet.1 The LEARN and moderate macronutrient distribution diets (including Biggest Loser, Jenny Craig, Nutrisystem, Volumetrics, andWeightWatchers) were associatedwith loss of body weight of about 2 kg less overall in these comparisons. Further considerations regarding the inclusion of exercise and behavioral support are important. Theauthorsarecareful topointout thatwhile thestrengths of these analyses include rigorous application of criteria used todetermine the eligibility of these studies and ranking of the overall quality in the review, the limitations remain important. There aremore low-carbohydratediet (Atkins, Zone, and South Beach) studies reported than any other study type including low fat. Thenumerousmoderatemacronutrient studies, which are based on diet composition, aremixed-branded diets that involve use of food replacement products, different types of behavioral approaches, and encompass a variety of other differences that make it difficult to differentiate and compare with other diets. Also, no sensitivity analyses were provided to help readers compare the level of adherence to these different diets andwhether the diet composition or the level of adherence (representing successful caloric restriction) were the underlying causes of less weight lost. Aswell doneas these comparative analyses are andas reasonable as the conclusions by Johnston et al1 appear to be (eg, thatboth low-carbohydrateand low-fatdiets appear toachieve similar weight loss results), several questions and additional details would facilitate overall interpretation of these findings. For example, even though the results are based solely on weight loss, it would be helpful to know more about the differences innutrient quality, long-term levels of dietary adherence, and energy intake associatedwith these diets involving very different diet compositions. Protein intakes of 30% of kilocalories, ordoublewhat theotherdietsprovide, raisequestions about possible long-term influences on kidney function, calcium losses, and other questions that should be explored. The adverse effects reported are certainly benign, ranging from headache to halitosis during the experimental dietingperiod, butwhat, if any, further changesoccurwithadditionalmonths or years of this high protein intake? Aremeal replacement diets affordable and tolerable long term and are there economic, social isolation, andother limitations thatput these diets out of reach for medium to low income subgroups? Are there compromises in the intake of sodium, solid fat, or sugar, or are there other adverse factors that may enhance adherence to these meal replacements but limit nutrient quality? The Preventing Overweight Using Novel Dietary Strategies (POUNDS LOST) trial3,4 was an RCT that tested differences in weight loss achieved by 4 different approaches to diet composition: (1) a low-fat, average protein diet (20% fat, 15% protein, and 65% carbohydrate), (2) a low-fat, highprotein diet (20% fat, 25% protein, and 55% carbohydrate), (3) a high-fat, average protein diet (40% fat, 15% protein, and Related article page 923 Opinion