Just submitted our latest manuscript entitled "A plant-based, low-fat diet decreases ad libitum energy intake compared to an animal-based, ketogenic diet: An inpatient randomized controlled trial". I'm looking forward to all constructive comments!
osf.io/preprints/nutr…
osf.io/preprints/nutr…
20 adults without diabetes were admitted as inpatients to the NIH Clinical Center and randomized to consume ad libitum either a plant-based, low-fat (PBLF) diet or an animal-based, ketogenic, low-carbohydrate (ABLC) diet for 2 weeks followed immediately by the alternate diet . 
The PBLF diet was 75.2% carbohydrate, 10.3% fat, 14.5% protein and had a non-beverage energy density of 1.1 kcal/g. The ABLC diet was 75.8% fat,10.0% carbohydrate, 14.2% protein and had a non-beverage energy density of 2.2 kcal/g. Subjects consumed as much or as little as desired
The PBLF diet resulted in substantially greater postprandial glucose and insulin levels compared to the ABLC diet
The ABLC diet led to greatly increased circulating ketones compared to the PBLF diet
Interestingly, the PBLF diet resulted in a spontaneous 689±73 kcal/d decrease in ad libitum energy intake compared to the ABLC diet (p<0.0001) and this difference was consistently observed in all subjects.
The macronutrient composition of the consumed diets corresponded quite closely to the composition of the presented diets, but the ABLC diet resulted in a greater protein intake compared to the PBLF diet.
Interestingly, despite the lower energy intake during the PBLF diet, the ABLC diet led to more rapid early weight loss during the first week, but total weight loss after two weeks was not significantly different
Body composition analysis helps explain the apparent dissociation between weight loss and energy intake between the PBLF and ABLC diets. Only the PBLF diet resulted in significant loss of body fat whereas only the ABLC diet led to loss of fat-free mass.
Our study was designed to measure ad libitum energy intake when inpatient subjects were exposed to widely different food environments with meals and snacks corresponding to either an exemplary plant-based, low-fat diet versus an exemplary animal-based, ketogenic, low-carb diet.
PBLF diet had higher glycemic load and greater glucose and insulin levels compared with the ABLC diet that was higher in energy density. PBLF diet resulted in ~700 kcal/d less intake with subjects losing weight and body fat while reporting no significant differences in appetite.
These data suggest that while the ABLC diet had benefits for reducing glucose and insulin levels, the PBLF diet had benefits for appetite control. Nevertheless, it is important to emphasize that our study did not examine the effects of diet recommendations for weight loss.
Rather, subjects were told that this was not a weight loss study, were instructed not to attempt to change their weight, and were blinded to their body weight measurements. Whether our results would have been different in subjects actively trying to lose weight is unknown.
Whether long-term adaptations to the ABLC diet would eventually result in reduced appetite and energy intake compared to the PBLF diet is unknown. A previous study suggested rapid suppression of appetite with a low-carb, ketogenic diet vs a "usual diet": pubmed.ncbi.nlm.nih.gov/15767618/
What are the implications of our study? The carbohydrate-insulin model of obesity predicts that consuming a diet with high glycemic carbohydrates results in increased postprandial insulin that drives body fat accumulation thereby increasing hunger and energy intake.
Indeed, the PBLF diet contained foods with high glycemic load that significantly increased postprandial glucose and insulin levels compared to the ABLC diet. But the PBLF diet led to less energy intake which contradicts the predictions of the carbohydrate-insulin model.
Alternatively, the passive over-consumption model of obesity posits that diets high in fat & energy density drive excess calorie intake causing body fat gain. That didn't happen with the ABLC diet despite its high fat & energy density. Both obesity models fail to explain our data
