I recently came across the pre-print manuscript from Kevin Hall: A plant-based, low-fat diet decreases ad libitum energy intake compared to an animal-based, ketogenic diet: An inpatient randomized controlled trial.1
I wanted to write about this study as I read through it because it is truly a novel one. While previous studies have assessed plant-based, low-fat (vegan) diets and animal-based, low-carbohydrate (ketogenic) diets individually, no study has ever compared them.
What was studied?
Hall recruited a small group of young to middle-aged men and women to spend 4 weeks inside a research center where their lives could be controlled by the study investigators.
The participants didn’t have diabetes, cancer, or psychiatric conditions, but I wouldn’t say they were healthy. Most were overweight or obese, with the average body fat being 26% (range of 12–36%) for men and 41% for women (range of 32–50%).
They were randomized to consume either a vegan or keto diet for 2 weeks, then immediately switched over to the other diet for 2 weeks. Each diet was provided as 3 daily meals and a bunch of snacks, with the participants allowed to eat however much they wanted.
The total amount of food available to eat was twice each participants estimated energy requirements, an average of about 5,200 kcal, divided up as 15% protein for both groups and 75% carb, 10% fat for the vegan diet or 75% fat, 10% carb for the keto diet. Both diets included 1 kg of non-starchy vegetables and were based around whole foods.
Blood labs, glucose tolerance (continuous glucose monitor and oral glucose tolerance test), blood ketones, body composition (DXA), energy expenditure (metabolic chamber), physical activity (accelerometer), and appetite (100-point scale) were assessed throughout each of the diet periods.
What was found?
The participants ate an average of 689 kcal fewer on the vegan diet, and every single participant ate less when eating vegan compared to eating keto. The lower calorie intake was present at each meal and with snacks.
Despite this difference in calorie intake, the vegan group ate significantly more food mass (2.1 vs 1.4 kg) and the groups reported equivalent levels of hunger, satisfaction, fullness, eating capacity, pleasantness, and familiarity.
The groups were similarly active, but the vegan group had a lower total daily energy expenditure (-165 kcal; 2150 vs 2315 kcal) due to burning fewer calories while sleeping or being sedentary. This is likely owed to the calorie deficit and weight loss.
The keto dieters experienced an immediate 3.5-lb drop in fat-free mass during the first week, likely from glycogen and water losses, which stabilized for the second week. No meaningful change was seen on the vegan diet.
Comparatively, the vegan dieters showed a steady decrease in fat mass, while the keto dieters did not. On average, the rate of fat loss was 51 g/d on the vegan diet and 16 g/d on the keto diet. Neither diet affected liver fat.
Expectedly, average (94 vs 81 mg/dL) and post-meal (102 vs 80 mg/dL) blood glucose concentrations were higher with the vegan diet, but the differences were not clinically meaningful in my opinion.
Also expectedly, the keto diet quickly resulted in nutritional ketosis, with the 0.5 mM threshold being surpassed on the third day and average blood ketone concentrations being 1.8 mM.
A variety of fasted blood tests were done at the end of each diet period. Here’s my summary of some notable findings:
- No differences in HbA1c, glucose, or insulin.
- Keto group reduced their triglycerides (from 75 to 63 mg/dL) while the vegan group increased them (from 75 to 93 mg/dL).
- LDL particle counts increased on keto (from 1072 to 1224 nmol/L) while they decreased on vegan (from 1072 to 781 nmol/L).
- No difference in LDL particle size.
- Large LDL particles decreased, while small LDL particles increased on keto. Vegan didn’t affected large LDL particles but lowered small LDL particles.
- hsCRP (biomarker for systemic inflammation) was unchanged on keto and reduced on vegan (from 2.1 to 1.2 mg/L).
- Thyroid function was maintained on vegan and slightly decreased on keto (less total and free T3).
Blood glucose, insulin, and free fatty acid measurements were also taken after a standardized test meal representing each diet’s macronutrients and providing 30% of energy requirements, but I don’t think this information is valuable because the meals were liquid. It shouldn’t be surprising that liquid sugar notably increases glucose and insulin while liquid fat increases triglycerides.
Additionally, an oral glucose tolerance test was provided at the end of each diet period, and the keto diet showed higher levels of glucose intolerance, which is expected given that their body just spent 2 weeks adapting to a low carbohydrate intake.
The 2-week duration is a limitation, as things could change over the long-term. Yet, the stringent dietary control and comprehensive analyses provided a lot of much-needed insight into the metabolic, appetite, and body composition effects of ketogenic and vegan diets.
Both diets were based around whole foods and included ample fibrous vegetables, which is both a benefit and a limitation since it represents the ideal way of eating these diets but may not represent how most people would eat on these diets.
The diets were matched for protein, but many people eat high-protein ketogenic diets that would likely result in different outcomes for appetite, body composition, and some metabolic parameters.
In the short-term, a low-fat vegan diet results in lower calorie intake despite greater food intake and similar appetite ratings as a low-carb keto diet.
Metabolically, it also leads to marginally higher average blood glucose levels throughout the day, improves blood lipids other than triglycerides, and reduces inflammation compared to a ketogenic diet.
- 1.Hall KD, Guo J, Courville AB, et al. A plant-based, low-fat diet decreases ad libitum energy intake compared to an animal-based, ketogenic diet: An inpatient randomized controlled trial. Published online May 6, 2020. doi:10.31232/osf.io/rdjfb