IntroductionThe epidemics of obesity and related non-communicable diseases remains a matter of the utmost concern for public health(1). While physical exercise has been shown able to provide concrete benefits(2,3), there is good reason to believe that it cannot fully compensate for the deleterious changes produced by poor diet(4). Opinions regarding optimal diet increasingly show elements of convergence, especially regarding the association of dietary refinement and ultra-processed foods with poorer health outcomes(5). Some proposed mechanisms underlying this epidemiology of junk food include the removal from diet of health-promoting compounds due to inadequate fruit and vegetable consumption(6), and the enhancement of glycemic responses by the refinement of foods(7-9). Consistent with the latter, ad libitum low-carbohydrate diets (<50 g / day) have been shown to produce weight loss(10,11) and improve cardiovascular risk profiles(12). However, some very high carbohydrate diets have also been shown to be compatible with population-wide optimal metabolic health in non-cereal agrarian populations(13-16). Likewise, the "paleolithic style" diets used to date in small clinical trials have achieved consensus diet-beating improvements in metabolic and cardiovascular-disease risk factors(17-21) despite not restricting total carbohydrate into the <50 g / day range. Similar short-term benefits have also been shown for "the McDougall Program" a low fat, high carbohydrate diet that places a strong emphasis on food quality and minimizes processed carbohydrates(22). This wide range of carbohydrate content for ad libitum diets that produce weight loss suggests that low glycemic index and minimized insulin release are not credible explanations for their effects, especially in light of the experimental variability in the glycemic index of foods between individuals and between meal contexts(23). However, one factor that these dietary regimens do have in common is minimal consumption of processed foods, milled flours and sugar. The present article is an outline of a recently proposed mechanism(24,25) that could explain the effects of food-quality upon weight and health, independent of overall macronutrient composition.