"The key to weight loss in a vegan diet trial wasn’t eating less, but eating more of less calorie-dense foods, particularly by drastically reducing fats and prioritizing high-volume, low-calorie options."
A recent analysis of a 16-week trial published in JAMA Network Open reveals a nuanced understanding of how plant-based diets contribute to weight loss. Contrary to assumptions that individuals in the vegan arm of the study consumed less food, data indicates they actually ate a greater total weight of food. The significant calorie reduction, leading to measurable weight loss, was achieved not by restricting overall food volume, but by strategically replacing high-calorie, dense foods with voluminous, low-calorie alternatives. This detailed examination of food group consumption highlights that the substantial impact on weight was primarily driven by a significant reduction in fat intake, particularly from animal products and added fats, rather than a simple decrease in overall food quantity.
The findings from the 16-week trial, which included 244 adults with overweight in Washington, D.C., offer a compelling look into the mechanics of dietary weight loss. Participants in the vegan group experienced a notable increase in the total weight of food consumed daily, rising from approximately 1,825 grams to 1,930 grams. This increase was deemed statistically insignificant, suggesting that participants were not actively trying to eat less. However, during the same period, their daily calorie intake saw a substantial drop of 491 calories. This stark contrast underscores a fundamental principle of weight management: the energy density of food plays a critical role.
Delving deeper into the food group data reveals the precise shifts that facilitated this calorie deficit. The consumption of meat, poultry, fish, and shellfish plummeted from around 102 grams per day to a mere 1.4 grams. Similarly, dairy intake decreased from 129 grams to 1.5 grams, and eggs dropped from 30 grams to 0.1 grams. Collectively, these animal-derived food groups, which previously contributed roughly 402 calories daily, were almost entirely eliminated. This drastic reduction in high-fat, high-protein sources formed a cornerstone of the calorie savings.
Perhaps the most surprising and under-discussed element of the dietary transformation was the significant reduction in nuts and added fats. This category, which is exceptionally calorie-dense, fell from 39 grams per day to 12.5 grams, representing a reduction of about two-thirds by weight. Given that nuts and added fats carry approximately 6.5 calories per gram, this reduction of 27 grams daily translated to a saving of about 145 calories. This highlights the disproportionate impact of reducing fats, as they pack more than double the calories of carbohydrates or proteins per gram.
To compensate for the volume lost from these calorie-dense foods, participants increased their intake of lower-density options. Vegetables saw a substantial rise of 143 grams per day, fruits increased by 80 grams, and legumes nearly tripled in consumption, going from 36 grams to 110 grams. Non-dairy alternatives also more than doubled their presence in the diet. Grains, while remaining relatively stable in the vegan group, saw a sharp decline in the control group, indicating that overall dietary patterns shifted significantly in the intervention arm. Cumulatively, these substitutions for calorie-dense foods allowed participants to consume more food by weight while significantly reducing their caloric intake.
The net effect across the trial was that the vegan group’s daily calorie intake dropped by an additional 357 calories compared to the control group, with a confidence interval ranging from 192 to 522 calories. This statistically significant difference further solidifies the impact of the dietary intervention.
A particularly counterintuitive finding emerged when examining the energy density of the diet. While the overall energy density in the vegan group decreased by approximately 30%, from about 1.1 to 0.7 calories per gram, the contribution from certain food groups actually became more calorie-dense. Legumes, for instance, saw their contribution to energy density rise from 0.7 to 1.1 calories per gram, and nuts and added fats increased from 6.5 to 7.2 calories per gram. This apparent paradox is explained by the weighted contribution of each food group to the overall diet. Legumes, despite not becoming intrinsically more calorie-dense, contributed more to the diet’s energy density simply because participants were consuming three times the quantity. The overall reduction in energy density was therefore driven by the substantial removal of extremely high-density food categories, rather than a uniform decrease across all food types.
The practical implication is that the significant drop in energy density was achieved through subtraction at the higher end of the density spectrum. Foods like vegetables, which were already low in calorie density at around 0.5 calories per gram before the trial and dropped to 0.4 calories per gram afterward, did not change dramatically. Their role was to provide bulk and nutrients without significantly increasing calorie load. The study observed that reductions in energy density correlated with greater weight loss, with an r-value of 0.28, a relationship that persisted even after adjusting for calorie intake (r=0.21). While this is an association within a specific trial population, it suggests that energy density plays a role beyond simply lowering calorie intake.
The parent trial, which randomized 244 adults with overweight, was conducted between January 2017 and February 2019. Of the initial participants, 223 completed the study. The originally published trial data indicated that 87% of participants were women, with a mean age of 54. The vegan intervention led to a significant weight advantage over the control group, with a treatment effect of 5.9 kilograms.
The examination of energy density was a post-hoc analysis, utilizing diet records collected for other purposes. The statistical analysis was performed between April and June 2026. Food intake data was gathered through three-day self-reported diaries at the beginning and end of the 16-week trial. It is important to acknowledge that self-reported dietary data has inherent limitations, including potential underreporting, which can be uneven across participants and food groups.
The control group adhered to their usual dietary habits, a design that was pre-registered on ClinicalTrials.gov. This comparison is valuable as it demonstrates the vegan dietary pattern’s efficacy against maintaining the status quo. However, it does not provide insights into how the vegan diet might compare to other dietary approaches, such as the Mediterranean diet or higher-protein diets, or the impact of simply reducing cooking oil.
Funding for the study was provided by the Physicians Committee for Responsible Medicine (PCRM), an organization that advocates for plant-based diets and employs the lead and senior authors of the research. The journal has stated that the funder played no role in the study’s design, analysis, or the decision to publish. This allows for an objective assessment of the measurements and their interpretation.
Stripping away the specific label of "vegan," the core mechanism identified in the data is a mechanical relationship governed by the caloric density of food. Fat, with its nine calories per gram, contributes significantly more energy than protein or carbohydrates, which provide four calories per gram. Water and fiber, on the other hand, add weight and volume without contributing calories. Extensive research in controlled feeding studies has consistently shown that individuals tend to consume a relatively consistent weight of food during meals. Therefore, reducing the caloric content per gram of food naturally leads to a decrease in overall calorie intake without conscious effort to eat less.
Any dietary pattern that shifts consumption towards water-rich, fibrous foods and away from concentrated fats will likely yield similar results in terms of weight management. The simple act of adding a serving of beans and reducing cooking oil by half, for example, employs the same principle. The trial’s findings strongly suggest that the reduction of added fats was a particularly potent lever in driving calorie reduction, demonstrating a greater impact per gram compared to other dietary adjustments.
It is crucial to note that these findings do not alter established clinical guidance. A secondary analysis of a trial is not intended to provide prescriptive advice. Individuals contemplating significant dietary changes, especially those who are pregnant, managing diabetes, taking medications influenced by diet, or have a history of disordered eating, should always consult with a healthcare professional or a registered dietitian. Furthermore, adherence to a vegan diet necessitates attention to essential nutrients such as vitamin B12, and potentially iron, zinc, calcium, iodine, and omega-3 fatty acids, depending on individual food choices.
Key Questions Answered
Which food change removed the most calories per gram?
Nuts and added fats. A reduction of approximately 27 grams per day from this category, which contains about 6.5 calories per gram, removed around 145 calories.
Did participants eat less food?
No. The total weight of food consumed daily increased slightly, a change deemed not statistically significant. Calorie intake decreased due to the lower energy density of the foods consumed.
Why did legumes appear to get more calorie dense?
The reported figures reflect the weighted contribution of each food group to the overall diet. As participants consumed roughly three times more legumes, their share of the diet’s energy density increased, even though the intrinsic calorie density of legumes themselves did not change.
How big was the effect against the control group?
The vegan group experienced a daily calorie reduction of 357 calories more than the control group, with a confidence interval between 192 and 522 calories.
What are the main limitations?
The study’s limitations include reliance on self-reported dietary records, the analysis being conducted post-hoc, a follow-up period of 16 weeks, a control group that made no dietary changes, and a study population composed of 87% women with a mean age of 54.
Do you have to go vegan to get this effect?
No. The underlying mechanism is compositional. Shifting to diets that incorporate more water-rich, high-fiber foods and reduce added fats will similarly lower energy density and support weight management, regardless of whether the diet is strictly vegan.
