Plant-based diets and weight loss depend on what trials compare

Plant-based diets can reduce weight, but trial conditions matter. Read the findings on vegan diets, calorie intake, and reviews showing little difference.

日本語で読む（Japanese version）

You are shopping for dinner and put the usual pack of meat back on the shelf, reaching for tofu instead. The number on the scale has been bothering you, and eating more plant foods seems like a manageable change. Back home, a search turns up advice ranging from giving up every animal product to adding another vegetable dish. The same phrase, “plant-based,” points toward very different dinners.

Research has a similar problem. Many trials of plant-based eating use a low-fat vegan diet. They change more than the origin of the ingredients: fat intake and energy intake can change too. Calling a diet plant-based does not, by itself, identify the mechanism behind any weight loss.[7][8]

Some studies report lower weight, while a review finds little or no weight difference compared with other diets.[1][10] Before changing your shopping basket, it helps to ask who took part, what they ate, what the comparison group did, and how much changed. Those details make the evidence useful without turning a study average into a promise about your own body.

Plant-based does not describe a single diet

An actual comparison trial helps make sense of the labels. In adults with overweight, researchers compared vegan, vegetarian, pesco-vegetarian, semi-vegetarian, and omnivorous diets. The groups differed in how far they moved toward plant foods.[2] Here, vegan means excluding animal foods. Vegetarian is a broader category, and when a diet includes eggs and dairy, that condition needs to stay visible.[1][2] A person replacing meat with beans may be making a different change from someone removing dairy as well.

The dietary changes tested in trials often include reducing fat alongside excluding animal foods. A systematic review found that most of its studies compared low-fat vegan diets with omnivorous diets in people with overweight, type 2 diabetes, or cardiovascular disease. Restrictions on fat intake, followed by lower energy intake, were important to interpreting the results.[7] The label tells you something about the foods, but it does not tell you everything about the intervention.

The reviews considered here collect dietary trials under their own eligibility rules, and the meta-analyses combine those trials to estimate the average weight difference between the tested diets. Reviews with similar titles may include different participants, dietary rules, and comparison diets, so they may not be answering the same question.[1][6][8] “What happens when people start a vegetarian diet?” and “Does it outperform another diet designed with health in mind?” need separate answers.

That distinction also matters in the kitchen. Write down what you actually plan to replace. Will a bean dish take the place of a meat dish, will vegetables be added to the existing meal, or will dairy products disappear as well? Then consider how the fat content and the whole meal will change. Choosing what goes into the refrigerator is more concrete than choosing a dietary identity. It also gives you a clearer description of the change you are trying to sustain.

The related feature on the Mediterranean diet considers another dietary pattern. The features on whole grains and energy density focus more closely on properties of foods and meals. The central question concerns what sits inside the plant-based label when researchers test the diet as a complete intervention. Keeping that scope clear prevents a result about a whole eating pattern from becoming a claim about every food carrying a plant-based label.

An average weight difference is not everyone’s target

A frequently cited meta-analysis combined 12 randomized controlled trials involving 1,151 participants. In a randomized trial, people are assigned to groups by chance so researchers can compare the effects of the diets. The median intervention lasted 18 weeks. Participants assigned to vegetarian diets lost an average of 2.02 kg more than those assigned to nonvegetarian diets, with a 95% confidence interval of −2.80 to −1.23 kg.[1]

The words “more than” do substantial work here. The −2.02 kg figure is an average difference between the groups’ weight changes. If the comparison group also lost weight, that loss is taken into account in the difference. It is neither the total weight loss of every vegetarian participant nor a personal target. The researchers also analyzed results by diet type, energy restriction, study population, and follow-up period.[1]

The vegan subgroup illustrates how uncertainty differs between dietary comparisons: the difference was −2.52 kg, with a 95% confidence interval of −3.02 to −1.98 kg. For lacto-ovo-vegetarian diets, which include eggs and dairy, the estimate was −1.48 kg, with an interval of −3.43 to 0.47 kg. That interval crosses zero, the value representing no difference between groups.[1]

The reported interval for the egg-and-dairy subgroup does not establish a clear between-group difference.[1] It is also premature to turn the subgroup results into a rule that vegan eating always produces more weight loss. Pooling different trials within separate subgroups is different from directly comparing everyone under the same conditions.

Another meta-analysis focused on people with overweight or obesity. It included 9 trials and 1,628 participants, and estimated a weight difference of −3.60 kg, with a 95% confidence interval of −4.75 to −2.46 kg. However, the certainty of evidence was rated very low for most outcomes. There were no clear differences in waist circumference, body mass index, or some other measures.[6] An encouraging weight estimate can coexist with uncertainty about how dependable the broader evidence is.

A further meta-analysis covered vegan diets lasting at least 12 weeks in people with overweight or type 2 diabetes. Across 11 trials and 796 participants, the weight difference was −4.1 kg, with a 95% confidence interval of −5.9 to −2.4 kg. The authors noted that differences in nutrient composition and energy intake might account for part of the effect.[8] Selecting only the largest number would leave behind the population and dietary conditions that give it meaning.

Give reviews showing little difference equal attention

Reading only studies that favor weight loss can leave the impression that plant-based diets win regardless of the alternative. A review published in 2026 reached a different assessment. It included 10 randomized controlled trials in people with overweight or obesity and found little or no difference in body weight between plant-based and omnivorous diets. Its assessment of body fat mass was similar, and the authors described the available evidence as limited and uncertain.[10]

The reviews differ in their assessment of a weight-loss advantage.[10] The cause of that disagreement cannot be established from the reported findings.

Another systematic review included 22 publications from 19 studies. Although every study reported weight reduction, the review described both 7 publications with significant intervention-versus-control differences and 4 without significant differences.[7] Losing weight within a group and losing more weight than another group are separate findings. Recognizing that a plant-based diet may be useful while remaining uncertain about its superiority is a coherent reading of the evidence.

The 2026 review considered plant-based diets a possible dietary option for people with overweight or obesity, with uncertainty about their advantage.[10] It searched for research through January 3, 2024, identified 2,664 records, and included 10 randomized controlled trials. A further 6 ongoing studies met the eligibility criteria. Publication in 2026 does not mean that the review contains every study available in that year; the search date defines the evidence it could gather.[10]

The comparison group deserves attention too. In the trial comparing low-fat vegan eating with other dietary advice, both sides received some form of support.[2] The difference between those groups is not the same as the change from living without dietary advice. When reading a result, put the comparison group’s instructions beside the intervention’s name. This makes it easier to see what the study actually tested and what it leaves unanswered.

The 2026 review also examined outcomes beyond weight. It found little or no difference in systolic and diastolic blood pressure, serum glucose and insulin, insulin sensitivity, total cholesterol, triglycerides, and body fat mass. It judged that LDL cholesterol, often called “bad” cholesterol, might decrease slightly.[10] These findings do not all move together, and a dietary choice should not be assessed as if every measurement were interchangeable.

Body mass index, a measure relating weight to height, and glycated hemoglobin, a blood glucose measure, might also decrease. However, certainty for these outcomes was very low. Interventions lasting 14 weeks or more showed greater improvements in body mass index, LDL cholesterol, and glycated hemoglobin.[10] Even a review finding little weight advantage can contain possible changes in selected outcomes. The useful question is which measurement changed, and how certain the evidence for that change was.

Numbers: plant-based diets and body weight

12 trials and 1,151 participants: the mean weight difference between vegetarian and nonvegetarian diets was −2.02 kg; the median intervention lasted 18 weeks.[1]

−3.60 kg: the average difference in 9 trials involving people with overweight or obesity, with very low certainty for most outcomes.[6]

−4.1 kg: the average difference across 11 trials of vegan diets lasting at least 12 weeks in people with overweight or type 2 diabetes.[8]

−7.5% and −3.1%: weight changes after 6 months in the vegan and omnivorous groups of a small dietary comparison trial.[2]

10 trials: the number of randomized controlled trials in the 2026 review, which assessed body weight as showing little or no difference.[10]