Fruit and vegetables for weight loss: what adding more changes

More produce does not guarantee weight loss. Trials show why adding fruit and vegetables differs from replacing other foods in your meals.

日本語で読む（Japanese version）

A salad sits beside dinner, and fruit follows the meal. The usual main course and the usual sweets are still there. You have added produce, but no other food has decreased. Replacing a higher-fat food with vegetables or choosing whole fruit instead of a sweet changes the meal differently. In the reviews, reducing other energy sources alongside increased produce intake matters to the weight-loss findings.[2][4]

The trial results give reason to examine the whole meal. Reviews of randomized trials encouraging people to eat more produce have reached different answers: one found a small difference in weight change, while another found no clear weight-loss effect.[1][2] The useful question is what changed across the whole meal, rather than how much healthier an individual plate looks.

This article examines the amount of fruit and vegetables people eat and its relationship with weight. The separate feature on meal sequence considers eating vegetables first and changes in blood glucose. Meal order and adding another dish to dinner are different interventions. Here, the starting point is what ends up on the table, what it replaces, and what the studies actually measured.

Adding more produced a small difference in weight

What happens when people increase fruit and vegetables without being specifically told to reduce other foods? Mytton and colleagues addressed that question in a systematic review and meta-analysis. They selected randomized trials of increased produce intake and statistically combined their results. In these trials, participants entered their comparison groups by chance, allowing the review to examine the effects of the interventions rather than only existing eating habits.[1]

There were 8 trials with 1,026 participants in total. Their average duration was 14.7 weeks, ranging from 4 to 52 weeks. The average difference in fruit and vegetable consumption between groups was 133 g. Weight change in the higher-intake groups was, on average, 0.54 kg lower than in the lower-intake groups. The reported effect was −0.54 kg, with a 95% confidence interval of −1.05 to −0.04 kg.[1]

For Mytton’s weight estimate, the confidence interval shows the range of uncertainty around the pooled difference. Reading the average alongside that range gives a more useful picture than treating the average as a precise promise. Here, the estimated difference was small, and the interval extended from a larger difference to one close to zero. That is a measured comparison between groups, rather than a forecast for everyone who starts eating salad.[1]

In particular, the result does not mean that every participant lost 0.54 kg. It combines differences in weight change between people encouraged to eat more produce and their comparison groups. Losing weight from your own starting point and having a more favorable change than another group are different outcomes. If the comparison group gained weight while the higher-intake group maintained it, a difference could still emerge. That possibility helps explain why the authors discussed weight maintenance as well as loss.[1]

The results also varied substantially across studies. The measure of heterogeneity, or variation between study results, was 73%.[1] “Eat more” covered different amounts and different lengths of intervention. Combining those studies gives an overall estimate, but it does not make their conditions identical. A bag of vegetables comes without a matching weight-loss guarantee.

Energy intake complicates the picture further. In the 6 trials measuring changes in energy consumed, the estimate pointed toward higher daily intake in the higher-produce groups: +368 kJ, with a 95% confidence interval of −27 to 762 kJ. The between-group difference was not statistically significant.[1] Kilojoules are a unit of energy. This result does not show that increasing produce automatically reduced the energy in the whole diet.

The authors concluded that encouraging fruit and vegetables without explicit advice to reduce other foods was unlikely to cause short-term weight gain and might have a role in maintenance or weight loss. They also called for more research on longer-term effects.[1] Those conclusions support a measured expectation. Planning substantial weight loss from the addition alone would go beyond what these trials established.

Another review found no clear weight-loss effect

Kaiser and colleagues published a different meta-analysis in 2014. They used stricter selection criteria: at least 15 participants per group, interventions lasting at least 8 weeks, body weight as an outcome, and a stated aim of weight loss or prevention of weight gain. The produce had to include a variety of minimally processed fruit and vegetables.[2]

Only 2 trials met every criterion. Their combined standardized mean difference was −0.16, with a 95% confidence interval of −0.78 to 0.46. There was no clear weight-loss effect. When the authors included studies meeting all or nearly all the criteria, the analysis covered 7 trials and produced a standardized mean difference of 0.04, with a 95% confidence interval of −0.10 to 0.17.[2]

A standardized mean difference places results on a common scale. It is not measured in kilograms. Reading 0.04 as a weight change would therefore misread the study. Equally, putting this value beside −0.54 kg and comparing their numerical sizes directly would confuse different ways of expressing results. The two estimates use different scales, so their numerical sizes cannot be compared directly.[1][2]

The authors concluded that the trials available at the time did not support the claim that recommending more produce, delivering it to homes, or providing it would itself cause weight loss. They argued that advice to increase intake should explicitly accompany efforts to reduce other sources of energy when the goal is treating or preventing obesity.[2]

How can this coexist with Mytton’s small favorable difference? The reviews selected trials differently. One focused on interventions without specific encouragement to remove other foods; the other applied requirements concerning weight-related aims, duration, group size, and the processing of foods. Similar titles do not mean that two analyses repeatedly tested the same participants eating the same meals.[1][2]

“No clear effect” also does not establish that produce is useless for every person. The confidence intervals leave uncertainty, just as a small statistically significant finding still requires attention to its setting. Read together, these reviews provide weak grounds for presenting an increase in produce as a powerful, self-contained weight-loss method.[1][2]

Consider a set lunch in Tokyo. Adding another vegetable side while keeping everything else gives you more vegetables. Choosing whole fruit in place of the usual sweet also increases fruit intake, but another food has decreased. Both choices can be described as eating more produce. Only the second explicitly changes the other energy source. That is the distinction Kaiser and colleagues wanted weight-loss advice to address.[2]

Larger losses belong to particular people and interventions

Some reviews report more favorable results. Arnotti and colleagues’ 2020 meta-analysis examined interventions in people with overweight or obesity. It combined 16 studies, providing 17 comparisons across 3,719 participants, and reported an overall effect of −2.81 kg.[11] This is larger than the −0.54 kg difference in Mytton’s analysis and differs in impression from Kaiser’s finding of no clear effect.[1][2][11]

The tempting headline is that more vegetables can produce a loss of about 3 kg. The study conditions make that too broad. Participants had a body mass index above 25. Body mass index is a measure calculated from weight and height. The investigators also examined whether results differed with participant, intervention, methodological, and study characteristics.[11] The pooled value cannot become an expected loss for all adults, irrespective of body size or eating pattern.

Reported effects included −4.99 kg in lower socioeconomic groups, −3.7 kg among people recruited through healthcare or programs, and −5.2 kg among those recruited in workplace or academic settings. Results also differed with starting body size.[11] These results concern particular study categories and starting conditions; they cannot predict the loss for adults whose circumstances differ.

No confidence interval for the overall pooled effect or detail on how far other foods were reduced in each study was reported.[11] That limits how confidently the result can be attributed to simply adding produce. We can acknowledge the favorable result without treating every intervention in the review as the same isolated change. A larger estimate deserves careful reading just as much as a small or uncertain one.

Dreher and colleagues took a different approach in a review focused on women. They examined prospective observational studies and randomized trials and judged the evidence for increased fruit and vegetable intake contributing to weight loss positively. They also reported that the effect was enhanced when higher-energy-density or high-fat foods were restricted at the same time.[4] A pooled kilogram estimate and confidence interval were not reported.[4]

Energy density means the amount of energy in a given quantity of food. The review discussed possible pathways through which whole fruit and vegetables might influence weight, including eating rate, satisfaction, low energy density, and dietary fiber.[4] These are proposed connections between the properties of foods and weight-related outcomes. They help explain the reasoning behind substitution without giving each serving a predictable effect on the scale.

For a reader, the useful point is both whom the review concerns and what else changed. A vegetable side can become more prominent while a higher-fat food becomes less prominent. Alternatively, both can grow. Fruit can take the place of a sweet, or be eaten after that sweet. Looking at those differences makes a meal easier to interpret in light of the research.[2][4]

Numbers: fruit, vegetables, and body weight

8 trials, 1,026 participants: the review examining increased intake without advice to reduce other foods.[1]

−0.54 kg: the average between-group difference in weight change; 95% confidence interval −1.05 to −0.04 kg.[1]

7 trials, standardized mean difference 0.04: a separate analysis found no clear weight-loss effect; 95% confidence interval −0.10 to 0.17.[2]

−2.81 kg: the pooled effect in a meta-analysis of people with overweight or obesity; the abstract does not report its confidence interval.[11]

16 weeks: the duration of a trial providing fruit and vegetables to adults with overweight or obesity.[7]