2.3. Experimental Design
After the pre-testing visit, subjects attended the laboratory for an oral glucose tolerance test (OGTT) and then continued their habitual food intake for six days. On the seventh day, subjects consumed the experimental diet that was provided to them. The experimental diet was designed to be high in fat (68% total energy) and provide an energy excess (+78% kJ). High-fat foods were specifically chosen because individuals tend to overconsume more readily with high-fat foods due to the greater palatability and the higher energy density of this macronutrient, and because individuals who binge eat are known to selectively increase their intake of such foods . Individual diet plans were designed using NetWISP nutrition software (Tinuviel Software Ltd., Llanfechell, Anglesey, UK). All foods were purchased and prepared by the research team. Subjects were instructed to consume all food provided, and to avoid consuming additional food or nutritive beverages. Food intake followed a normal daily feeding pattern (i.e., breakfast, lunch, dinner, and snacks) and water intake was allowed ad libitum throughout the dietary intervention. An example diet plan for one subject can be viewed in the supplementary material online. No subjects reported any issues with dietary adherence. The next day (Day 8), subjects returned to the laboratory for a second OGTT.
I checked ref #7 for obvious problems. Found 'em. It's animal models. Binge eating is induced by prior restriction. The selective aspect is innate, not arbitrary, yet manifested from arbitrary choice by experimenter.
In this experiment, it's humans, not induced by prior restriction, arbitrary choice by experimenter a priori.
A couple significant points.
200g of carbs per day for the test meals/day. This shuts down ketogenesis. It's probably what allowed the binge. Otherwise, ketones flow, hunger is suppressed, binge becomes much harder if not outright impossible. I'm not sure by which mechanism ketones suppress hunger, but apparently they do, so I'm going with that. But then again subjects' normal diet probably shut down ketogenesis anyways so the test meals/day is just more of the same.
In an animal model, dietary fat activates PPAR-alpha to maintain homeostatis of various energy substrate, i.e. glucose, lipids, etc. https://www.ncbi.nlm.nih.gov/pubmed/16054078
If this animal model translates to humans, it should show up in this experiment. It does not. On the contrary, the apparent excess dietary fat does not improve insulin sensivity which is an otherwise expected effect of activation of PPAR-alpha in the liver by dietary fat.
Feltham did something similar. His results were opposite. Maybe Sam lied? Dunno, let's say he was truthful. What was done differently? Sam didn't eat 200g of carbs per day for his high-fat experiment. No, he ate something like 10-15g of carbs per day, barely enough to show up on blood tests. (did he do an OGTT?) Physical effects of insulin resistance did not show up with Sam. On the contrary, his gut shrank, in spite of gaining a few lbs otherwise.
So, this experiment didn't replicate Sam's, so the results will not be replicated either. Rather, we can now see clearly why people say "low-carb doesn't work". It's cuz it ain't low-carb.
The word "ketone(s)" appears exactly none in this paper. Any serious researcher will at least mention in discussion, if not measure in experiment, all pertinent aspects, not just those that support their hypothesis and/or conclusions. For example, "we did not measure ketones in this experiment because it would confound the results".