Research explained: secondary analysis of a randomized crossover trial
More minimally processed animal foods went with weight gain in a vegan vs Mediterranean diet trial
A secondary analysis of a 62-person crossover trial sorted foods by plant or animal source and by processing. More minimally processed animal foods was the only link to weight gain that held up after strict statistical correction.
In a trial that compared a low-fat vegan diet with a Mediterranean diet, eating more minimally processed animal foods was linked to weight gain, and it was the only food link that survived strict statistical correction, according to a secondary analysis in Nutrition Research (published online October 1, 2026). Shifts toward minimally processed plant foods leaned the other way. The analysis has drawn news coverage in recent days because it adds a second question to the ultra-processed food debate: does the food come from a plant or an animal?
What the researchers did
The original trial randomly assigned 62 adults with overweight to a Mediterranean diet or a low-fat vegan diet for 16 weeks. After a 4-week washout, each person switched to the other diet for another 16 weeks. Body weight was the main outcome. For this new analysis, the team used 3-day food records and sorted everything people ate by source, plant or animal, and by NOVA processing level, from minimally processed to ultra-processed.
On the vegan diet, people cut minimally processed animal foods and processed foods of both kinds more, and added more minimally processed plant foods, than on the Mediterranean diet. In the first 16 weeks, rising intake of processed and ultra-processed animal foods tracked with weight gain, but these links did not pass the correction for testing many comparisons. In the second 16 weeks, more minimally processed animal foods went with weight gain (r = +0.42), and more minimally processed plant foods leaned toward lower weight (r = -0.35), a link that also did not pass correction.
How strong is the evidence?
The diets were randomized, but the food-by-food results are correlations within a small trial, so they show what was linked to weight change, not what caused it. The authors call the work exploratory. Food intake came from self-reported records. Most authors were paid by or lead the Physicians Committee for Responsible Medicine, a nonprofit that advocates plant-based diets, and one author reports research support from soy and pulse industry groups.