Portion size and weight loss: change the serving before the plate

Smaller servings can reduce calorie intake. Smaller plates alone are less reliable. Discover what meal experiments show and what they cannot predict.

日本語で読む（Japanese version）

While washing the dinner dishes, you notice that everyone finished their meal again. What is harder to remember is how much food went onto those plates before anyone started eating. A more generous scoop of rice, or the last of the main dish added to an already full plate, can leave exactly the same evidence afterward: an empty dish.

The advice to use a smaller plate sounds appealing because it asks so little of your food preferences. Yet experiments that reduce the amount of food served and experiments that shrink the plate have produced different results. Smaller servings have reduced energy intake; smaller plates have not shown a consistent effect.[1][5][8] The diameter of the dish and the quantity of dinner are separate things to change.

That distinction shifts attention away from judging your willpower after a meal. It gives you a more concrete question to ask beforehand: what will you serve, and how much? The research offers useful evidence about that decision, provided we keep the amount served, the amount eaten, and any later change in body weight separate.

The amount served can change the amount eaten

Here, portion size means the quantity of food offered at a meal. Some papers define it in terms of the energy content of an eating occasion, so it matters whether an experiment changes food weight or the amount of energy offered.[1][3] A portion measured in grams describes how much a food weighs. A portion described in kcal describes its energy content. Those measurements answer related questions, but they are not interchangeable descriptions of a meal.

Simply comparing people who choose large meals with people who choose small meals would leave many questions unresolved. The groups might differ in hunger or daily routines before ordering anything. Experiments that deliberately change the amount offered help examine the effect of that serving decision. They also allow researchers to measure what remains, rather than assuming that everyone eats everything placed in front of them.[1][7]

A systematic review and meta-analysis examined experiments measuring energy intake for at least a full day. The review selected experiments that changed the quantity served and followed intake for a full day or longer, then pooled 85 effects from 14 eligible studies. Participants offered smaller portions consumed approximately 235 kcal less per day than those offered larger portions.[1] This was a difference in daily intake, not a measurement of weight lost.

The standardized mean difference was −0.709, with a 95% confidence interval of −0.956 to −0.461.[1] A standardized mean difference puts results from different studies onto a common scale. For this portion-size comparison, the interval indicates uncertainty around the estimated intake reduction on that common scale. The approximately 235 kcal difference and the standardized effect describe the same finding in different ways.

Another meta-analysis of randomized controlled trials studied healthy adults aged 19–65. Across 24 portion-size comparisons, larger portions were associated with 295 kcal more daily intake, with a 95% confidence interval of 202–388 kcal.[3] Randomization means assigning study conditions by chance, which helps reduce bias arising from differences between participants. However, the results varied substantially between studies, and the review found insufficient information to determine chronic effects on body weight.[3]

The approximately 235 kcal and 295 kcal estimates come from different reviews with different study selections and ways of combining comparisons.[1][3] Their disagreement in size does not erase their shared direction: changing the amount offered can change total energy intake.

Nor does this appear to be a finding that applies only to people who describe themselves as poor at resisting food. A pooled analysis used data from 3 randomized controlled trials involving 111 adults. Lunch portions were compared at a large-normal amount, about 75% of that amount, and about 50%. The analysis examined characteristics including emotional eating, plate-clearing tendencies, self-control, sex, and body weight. None reliably changed the effect of reducing portion size.[6]

Participants included 55 men and 56 women, with body mass index values ranging from 19 to 42. Body mass index is an indicator of body size calculated from weight and height. Most participants ate less when portions were reduced; consistently reducing intake was more common than consistently eating a similar amount despite the smaller portions.[6] The finding supports treating the serving as part of the eating situation, rather than interpreting a response to a large meal as evidence of a character flaw.

Serving yourself does not remove the portion effect

A classic lunch experiment involved 51 men and women of normal weight or overweight. They attended lunch once a week for 4 weeks. The main dish was macaroni and cheese, offered in amounts of 500, 625, 750, or 1000 g. Participants consumed 30% more energy when offered the largest portion than when offered the smallest.[7] The serving amounts doubled between those extremes, but intake increased by 30%. What was offered and what was eaten were measured separately.

This experiment also tested who put the food onto the individual plate. One group received a plated portion. Another received the food in a serving dish and could take as much as they wanted. The effect of portion size did not depend on who determined the amount on the plate.[7] Having control of the serving spoon did not make the quantity available irrelevant.

That is useful when thinking about a family dinner with shared dishes. The first helping on your plate is only part of the serving arrangement. The amount sitting in the dish from which everyone takes food is also something you can consider. Keeping some of the cooked food away from the table initially, with more available if needed, is a practical application of the finding about self-serving.[7] The amount needed will also depend on the number of people and their food preferences.

The experiment also found that responses to serving size did not differ according to sex, body mass index, dietary restraint, or disinhibition.[7] Dietary restraint describes trying to restrict eating; disinhibition describes a tendency for that restraint to give way. Together with the pooled analysis, this makes the evidence difficult to reduce to a story about one particular kind of eater.[6][7] The amount available deserves attention even if you usually serve yourself and think carefully about food.

A separate randomized crossover trial extended the question across consecutive days. In a crossover design, the same participants experience different conditions, allowing comparisons within each person. Here, 32 adults from a university community received food and beverage portions at 100%, 150%, or 200% of baseline amounts. Increasing portions by 50% increased daily energy intake by 16%; doubling portions increased intake by 26%. The increase persisted across 2 consecutive days.[9]

With the 50% larger portions, women consumed 335 kcal more per day and men 504 kcal more. With doubled portions, the differences were 530 kcal for women and 812 kcal for men.[9] Participants ate their main meals in a controlled setting and were given snacks for the periods between meals. The same menus were used each week, while the quantities of foods and beverages changed. This helped distinguish the amount offered from a change in what foods were available.

These findings did not support the expectation that a larger meal would necessarily be balanced out by eating less later during the observed period.[9] They also did not establish a long-term change in body weight. Remembering the 2-day duration alongside the intake figures prevents a short feeding experiment from becoming a prediction about the next several months. Its contribution is more specific: a serving-size effect can continue beyond the first meal.

A smaller plate does not necessarily mean less food

A review of experiments manipulating dishware size included 9 experiments from 8 publications. Most found no significant difference in food intake between smaller and larger dishware. The overall standardized mean difference was −0.18, with a 95% confidence interval of −0.35 to 0.00. The effect was small, and findings varied substantially across studies.[8] A nonsignificant result means the data did not establish a clear difference; it does not mean that every participant behaved identically.

The review concluded that dishware size had not been shown to affect food intake consistently.[8] That is weak support for promising that a smaller plate will make everyone eat less. More importantly, it addresses a different intervention from reducing the food offered. If the amount remains available and people can refill their plates, changing the container does not necessarily change the total meal.

A broader review examined portion-control tools, analyzing 36 publications and including 28 in a meta-analysis. These tools included smaller bowls and spoons as well as calibrated or divided tableware. Across tools, the effect size for food intake was −0.22. For smaller plates alone, it was −0.03, with a 95% confidence interval of −0.12 to 0.06, which did not establish a clear effect.[5]

The category of “portion-control tool” therefore matters. A serving utensil can change how much is scooped. A marked dish can provide a reference for how much is placed on it. A smaller plate changes the surface available for a helping. These items may sit together in a store, but the review did not find that they performed equally well. Taking the favorable finding for the whole category and calling it the effect of small plates would misrepresent the comparison.[5]

In the tools review, combinations of smaller bowls and spoons contributed to reductions in serving sizes and intake. No effect was detected on awareness of appropriate portion sizes, although few studies measured that outcome.[5] This also separates a behavioral result from an educational one: eating less in a condition does not automatically demonstrate that someone has learned to identify an appropriate serving. Choosing a tool starts with deciding what you want to change in the actual serving process.

Three crossover experiments in adults likewise found no significant reduction in energy intake from using smaller plates. They tested serving from a large dish, presenting an equal amount of food on different plates, and choosing foods from a buffet. At the buffet, participants given the smallest plate made more trips to collect food.[12] The plate had less space, but the route back to the food remained open.

For a home meal, this makes refills part of the question rather than an afterthought. Refills after using a smaller plate can leave the total quantity eaten unchanged. Conversely, you can change the amount initially served while keeping your familiar dinnerware. The useful observation is whether the first serving and the additions together have changed, not whether the dish itself looks like a portion-control product.[5][12]

Numbers: portion size and food intake

Across 14 studies, smaller servings meant approximately 235 kcal less daily intake. This was an intake difference, not weight loss.[1]

Across 24 adult trial comparisons, larger portions meant 295 kcal more per day; the 95% confidence interval was 202–388 kcal.[3]

In a lunch experiment with 51 adults, the largest serving produced 30% more energy intake than the smallest. The comparison used a single main dish.[7]

For smaller plates alone, the effect size was −0.03, with a 95% confidence interval of −0.12 to 0.06. A clear effect was not established.[5]

In 24 young women, a 25% reduction in portions reduced energy intake by 10% across 2 consecutive days.[19]