This is not medical advice. General information for self-tracking only. If you have diabetes, heart disease or another condition that affects what you eat, talk to a doctor or a registered dietitian before you start counting calories to lose weight. The same goes if you are under 18, pregnant or breastfeeding, or have ever had an eating disorder or a difficult relationship with food. The numbers behind MyPlate's own targets are on the health sources page.

What volume eating means: calories per gram

Every food has a number that never appears on a calorie target: how many calories it carries per gram. Researchers call it energy density. Volume eating is the idea of building meals from foods where that number is low, so the same calories come as more food on the plate.

The range is wider than most people expect. These are the USDA's standard figures for a few everyday foods, with the weight of each that comes to 100 calories.1

Food (USDA entry)Calories per gramGrams for 100 calories
Cucumber, with peel, raw0.15667
Spinach, raw0.23435
Broccoli, raw0.34294
White rice, long-grain, cooked1.3077
Chicken breast, meat only, roasted1.6561
Cheddar cheese, sharp4.1024
Potato chips, plain, salted5.3219
Almonds5.7917
Olive oil8.8411

Two things set the number. Water adds weight and no calories, which is why the vegetables sit at the top. Fat adds the most calories per gram of anything on the list, which is why oil sits at the bottom, with 11 grams of it worth as much as two thirds of a kilo of cucumber.

None of this says one food is better than another. Chicken breast is denser than cooked rice and is still where a lot of people get their protein. The question this post asks is narrower: does eating lower down the table change how much you eat, and does that help if you are already counting calories to lose weight?

People eat the same weight of food, not the same calories

The finding that started this line of research is simple. In a 1998 study at Penn State, 18 women of normal weight had their meals provided for two days at a time, three times over. The main dish at lunch, dinner and an evening snack was made in a low, medium or high energy density version, matched for taste. The women ate about the same weight of food each time. So they took in 1,800 calories in the high-density condition against 1,376 in the low-density one, and they reported no difference in hunger or fullness.2

A 2004 study from the same group put energy density and portion size together. Thirty-nine women were served a lunch entrée at 1.25 or 1.75 calories per gram, in portions of 500, 700 or 900 grams. Both changes raised intake, and they added up: the largest portion of the denser version led to 56% more energy at lunch than the smallest portion of the lighter one, about 221 calories. The women did not eat less at the next meal to make up for it, and their hunger ratings did not track what they had eaten.3

That last part is the reason this matters to anyone keeping a log. If fullness follows the weight of food more than its calories, the calories in a dense meal can climb without the meal feeling any bigger.

Salad first, soup first: what a low-density starter does

The same idea can be tested at the start of a meal. In a 2004 study, 42 women ate lunch in a laboratory once a week for seven weeks. Some weeks they had to finish a first-course salad before a pasta main course they could eat freely. The salads came in 150 or 300 gram portions at 0.33, 0.67 or 1.33 calories per gram, and the difference between them was the amount and type of dressing and cheese.4

Compared with no salad at all, the low-density salads cut the whole meal's energy by 7% for the small portion and 12% for the large one. The high-density salads raised it, by 8% and 17%. When two salads had the same calories, the bigger, lighter one led to the smaller meal.4 So it was the dressing and the cheese that decided which way the meal went.

Soup does something similar. Sixty men and women of normal weight came in for lunch once a week for five weeks. On four of those weeks they had a vegetable soup first, served chunky, puréed or as broth and vegetables separately. Counting the soup itself, the meals came out 20% lower in energy than lunch with no soup, an average of 134 calories, and the form of the soup made no difference.5

Water seems to work only when it is part of the food. In a 1999 study, 24 lean women had the same 270-calorie starter before lunch on different days: a chicken and rice casserole, the casserole with a glass of water, or the same ingredients made into soup. After the soup they ate about 289 calories at lunch. After the casserole, with or without the glass of water, they ate about 396 and 392, and they did not make up the difference at dinner.6 The water was the same and only its place changed. What the evidence says about a glass of water before meals is covered in liquid calories.

A quarter off the density, a quarter off the calories

A starter is one meal. In 2006 the Penn State group changed every food for two days at a time. Twenty-four young women were served the same menus over four sessions, with portion size and energy density each set at the standard level or cut to 75%. Cutting the portions by a quarter reduced what they ate by 10%, 231 calories a day. Cutting the energy density by a quarter reduced it by 24%, 575 calories a day. The two effects added together, held from meal to meal, and came with no difference in hunger or fullness ratings.7

A 2013 study asked whether the way you lower the density matters. Fifty-nine adults ate three meals a day in the laboratory once a week for four weeks. In three of the weeks the main dishes were made 20% less dense, by cutting fat, by adding fruit and vegetables, or by adding water. Against a standard day of about 2,667 calories, people ate 396 fewer calories a day when the fat was cut, 308 fewer with more fruit and vegetables, and 230 fewer with added water.8 All three worked, and less fat worked best.

These were lab studies of a day or two in small groups, mostly of young women. They show what happens to intake when someone else makes the food lighter. They do not show what happens over months when you are the one cooking.

What happened over a year

The longest direct test is a 2007 trial of 97 women with obesity. One group was counselled to eat less fat. The other was counselled to eat less fat and more water-rich foods, particularly fruit and vegetables. Neither group was given a calorie goal or a fat goal, and both were told to eat as much as they wanted within the rules of their diet.9

After a year, the 71 women who finished had lost 7.9 kg on the fruit and vegetable diet and 6.4 kg on the low-fat diet alone. The difference also showed when everyone who started was counted. Both groups cut their fat by about the same amount, but the fruit and vegetable group ate a greater weight of food at a lower energy density, and reported less hunger.9 A 1.5 kg difference after a year is real but modest, and a quarter of the women did not finish.

A second line of evidence comes from the PREMIER trial, which tested lifestyle advice for 658 people with high or borderline-high blood pressure. When researchers grouped participants by how much the energy density of their diet fell over six months, the third with the largest fall lost 5.9 kg, the middle third 4.0 kg and the smallest 2.4 kg. The two thirds whose density fell most ate 300 and 80 grams more food a day while eating 500 and 250 fewer calories.10 People were not randomised to change their energy density, though. They were grouped by what they reported in two 24-hour recalls, so this shows an association, not a cause.

Volume eating when counting calories to lose weight

Most volume eating advice skips a detail of these studies. In nearly every one of them, people ate as much as they liked, and lower density worked by making them eat fewer calories without noticing. If you count calories, you have already fixed the number. A lighter plate cannot lower a target you are hitting anyway.

One trial tested it with a calorie target in place. In 2005, dietitians put 200 adults with overweight or obesity on an energy-restricted diet. On top of the diet, people were randomised to eat one or two servings a day of a low-density soup, two servings of a high-density snack with the same calories, or nothing extra. After a year the two-soup group had lost 7.2 kg and the two-snack group 4.8 kg, a difference the authors describe as 50% more weight loss on the same calories.11

The group with no extra food, though, lost the most: 8.1 kg. The soup did not beat simply following the diet. The snacks did worse than both. Weight loss tracked the drop in energy density in the first two months but not at six or twelve.11 My reading is that, with a calorie target, low-density food did not add weight loss so much as dense food in the same calories took some away. Why that happened, the trial cannot say. Under-counting the snacks is one possibility, and I could not find a trial that settles it.

So for someone already counting, volume eating changes what the number buys you rather than the number itself. A 1,700-calorie day built low on the table is a lot more food than the same day built from the bottom, and the lab studies suggest it is likely to feel like more food too. That may be the difference between a target you hold and one you give up on in week three. The trials show that much for intake and hunger ratings. None that I found measured whether it keeps people counting.

What the reviews add up to

A 2022 meta-analysis pooled 31 experiments, mostly in adults, that changed the energy density of the food people were served and measured what they ate over at least a day. Lower density led to a large drop in daily intake, with little sign that people made it up at other meals, whether or not the macronutrients were changed to get there. Only five studies measured weight, and in those the lower-density food came with 0.7 kg more weight loss, a difference that was not statistically significant.12 The lead author has had funding from the American Beverage Association and Unilever for unrelated work, and the paper says so.12

A 2016 meta-analysis of 13 studies covering 3,628 adults with obesity found low energy density foods went with 0.53 kg more weight loss. The studies varied greatly in who took part, how they were designed and what diets they used, and they mixed trials with observational studies.13

The reviews disagree on how strong the case is. The committee behind the 2010 US Dietary Guidelines reviewed 17 studies in adults and called the evidence that lower-density diets improve weight loss and maintenance strong and consistent.14 Two years later, a review from the USDA nutrition research centre at Tufts called the short-term evidence convincing and found insufficient data to judge long-term weight loss, partly because studies do not agree on how to define energy density, including whether drinks count.15

Does it help keep the weight off?

Possibly, but the trials point both ways. In an observational follow-up of 74 people about two years after a programme built on low-density eating, the ones who had kept their weight off reported eating about the same weight of food as those who had regained, at a lower energy density. Their reported intake was lower too, 1,608 against 1,989 calories a day, though after adjustment that difference was not statistically significant.16

Two randomised trials tested it directly. In a 2008 trial, 103 women who had just lost weight were randomised to one of three 14-week maintenance programmes, one of which taught them to lower the energy density of their diet. That group did lower it, but kept no more weight off than the others, and all three groups regained between 6 and 18 months after the programme.17 A 2014 trial in primary care patients found that the group taught reduced energy density eating, without meal replacements, regained significantly less than the control group two years later. The authors called that result notable and said methodological issues and the lack of an explanation for it make replication essential.18 Its lead author sits on a scientific advisory committee for Weight Watchers, which the paper discloses.18

And in 2017, the Penn State group's own one-year trial of 186 women found that teaching people to choose portions by energy density, with food scales, did no better than standard advice to eat less. Everyone lost about 6% of their starting weight, and use of the portion strategies faded over time.19

What the evidence does not settle

Whether it works when you count. Almost all of the intake studies let people eat freely, and the one year-long trial with a calorie target found that low-density soup did not beat the diet alone.11 The case for volume eating on a fixed target rests on hunger, and hunger was measured in the lab, not over months at home.

Whether lab results survive a real kitchen. The studies that showed the biggest effects provided the food, for one meal or for a day or two at a time.78 The trials where people chose their own food found smaller differences, or none.1719

How much rests on one laboratory. Most of the experiments in this post came from one group at Penn State. That is not a reason to doubt them, and the 2022 meta-analysis found the effect across 31 studies,12 but the longer trials are few, and a Tufts review still called the long-term evidence insufficient.15

What counts as energy density. Researchers calculate it differently, with or without drinks, and a 2014 review named the lack of a standard definition as one of the main barriers to progress.15

What a calorie counter app can see on a low-density plate, and what it cannot

Energy density is just calories divided by grams, so a tracker that shows both shows it. When MyPlate scans a meal, it breaks the plate into ingredients and gives each one a weight in grams and its calories, along with the dish's total weight. Put 300 grams of salad beside 60 grams of cheese and you can see which one is carrying the meal. The app does not work out calories per gram for you or rank foods by it. You divide.

The salad study above is a good picture of where a photo struggles. Its light and heavy salads differed in dressing and cheese,4 and dressing on a bowl of leaves, or oil in a pan, is the densest thing on the plate and the easiest to miss. A dish screen can show you that guess instead of hiding it. Under "What I went with", MyPlate lists what the scan assumed, such as how much oil went in, and a row the camera could not see, like the oil, is marked "Added, often hidden". Tap the guess and you can swap it, say for no oil, with the calorie change shown before you pick. When a lot was left to guess, the scan says "Rough estimate", so you know the number is loose before you trust it.

What it cannot see is the bit that makes volume eating work, which is how full you felt. The app also has no food database and no barcode scanner, and a scan is an estimate rather than a measurement, so a single plate can be off in either direction.

How to use volume eating with calorie tracking

Keep your target where it is. Volume eating works inside it. The app sets that target from your sex, age, height, weight, activity and goal, and never below 1,200 calories a day for women or 1,500 for men, the top of the low-calorie ranges in the NHLBI guidelines.20 A plate of vegetables is not a reason to go lower.

Then change one meal and look at what happens. Here is the same lunch two ways, worked out from the USDA figures in the table above.1

Lunch ALunch B
Cooked white rice200 g, 260 kcal120 g, 156 kcal
Roasted chicken breast120 g, 198 kcal120 g, 198 kcal
Olive oil10 g, 88 kcal5 g, 44 kcal
Raw spinach and cucumbernone100 g each, 38 kcal
Total330 g, 546 kcal445 g, 436 kcal
Calories per gram1.650.98

Lunch B is 115 grams more food for 110 fewer calories. The chicken did not change, so the protein did not either, and protein matters for its own reasons when you are in a deficit (does protein make you eat less goes through the evidence). Half of the saving came from five grams of oil, which is the cheapest place to look first.

You can spend that saving two ways. Bank it, or keep the 546 calories and eat a bigger lunch B. Either is fine.

If you want a starter, the studies support a low-density one: a broth-based vegetable soup, or a salad with a light dressing.54 Log the starter too. A heavy dressing turned the salad in that study from something that shrank the meal into something that grew it.4

Don't strip the fat out entirely to make the numbers look good. Fat is not the enemy (macros explained covers why you need some), and a plate that leaves you hungry an hour later defeats the purpose. Ultra-processed food is a related question with its own evidence on eating speed and density, covered in what a calorie and nutrient tracker cannot see.

And if weighing your plate by the gram has started to make meals feel like a test, stop and talk to a doctor or a registered dietitian. What the research says about when tracking stops helping is in can calorie tracking cause an eating disorder.

Common questions

What is volume eating?

Eating mostly foods that are low in calories per gram, so the same calories come as more food. Raw vegetables such as cucumber, spinach and broccoli are at the low end, at well under half a calorie per gram, and oil is at the top, at almost nine.1

Does volume eating help you lose weight?

Somewhat, in the trials. In a year-long trial, women told to eat more fruit and vegetables as well as less fat lost 7.9 kg against 6.4 kg on less fat alone, with no calorie goal. A 2022 meta-analysis found lower-density food cut daily intake a lot in short studies, but its five weight studies showed a 0.7 kg difference that was not statistically significant.912

Is volume eating worth it if you already count calories?

Probably, for hunger rather than for the number. In a one-year trial on a calorie-restricted diet, people given low-density soup lost 7.2 kg and people given the same calories as dense snacks lost 4.8 kg, but people given neither lost the most, 8.1 kg. Lab studies suggest a lighter plate feels like more food for the same calories, which may make a target easier to hold.117

How do you count calories in food for volume eating?

Divide the calories by the grams. Raw cucumber is about 0.15 calories per gram and olive oil about 8.84, on USDA figures, so 11 grams of oil has as many calories as 667 grams of cucumber. A tracker that shows each ingredient's weight and calories lets you see which part of a meal is dense.1

What should a food tracker for weight loss show if you eat a lot of vegetables?

The weight and calories of each part of the meal, and the oil or dressing as its own line. That is where most of the calories in a vegetable-heavy plate hide. No study has compared trackers on this. MyPlate shows grams and calories per ingredient and marks oil it could not see as a guess you can change, but a scan is still an estimate.

Fill the plate, keep the number

People tend to eat a similar weight of food from meal to meal, so lighter food means fewer calories when nobody is counting, and more food when you are. The trials support that for intake and hunger over days, and modestly for weight over a year. They do not show that it beats simply sticking to a target. Keep your target, look at the grams beside the calories, start with the oil, and use the space you free up for whatever keeps you going in week six.

Sources

  1. US Department of Agriculture, Agricultural Research Service (2018). FoodData Central, Standard Reference Legacy Release, April 2018. USDA National Agricultural Library. Read on FoodData Central →
  2. Bell EA, Castellanos VH, Pelkman CL, Thorwart ML, Rolls BJ (1998). Energy density of foods affects energy intake in normal-weight women. American Journal of Clinical Nutrition, 67(3):412-420. Read on PubMed →
  3. Kral TV, Roe LS, Rolls BJ (2004). Combined effects of energy density and portion size on energy intake in women. American Journal of Clinical Nutrition, 79(6):962-968. Read on PubMed →
  4. Rolls BJ, Roe LS, Meengs JS (2004). Salad and satiety: energy density and portion size of a first-course salad affect energy intake at lunch. Journal of the American Dietetic Association, 104(10):1570-1576. Read on PubMed →
  5. Flood JE, Rolls BJ (2007). Soup preloads in a variety of forms reduce meal energy intake. Appetite, 49(3):626-634. Read on PubMed →
  6. Rolls BJ, Bell EA, Thorwart ML (1999). Water incorporated into a food but not served with a food decreases energy intake in lean women. American Journal of Clinical Nutrition, 70(4):448-455. Read on PubMed →
  7. Rolls BJ, Roe LS, Meengs JS (2006). Reductions in portion size and energy density of foods are additive and lead to sustained decreases in energy intake. American Journal of Clinical Nutrition, 83(1):11-17. Read on PubMed →
  8. Williams RA, Roe LS, Rolls BJ (2013). Comparison of three methods to reduce energy density. Effects on daily energy intake. Appetite, 66:75-83. Read on PubMed →
  9. Ello-Martin JA, Roe LS, Ledikwe JH, Beach AM, Rolls BJ (2007). Dietary energy density in the treatment of obesity: a year-long trial comparing 2 weight-loss diets. American Journal of Clinical Nutrition, 85(6):1465-1477. Read on PubMed →
  10. Ledikwe JH, Rolls BJ, Smiciklas-Wright H, et al. (2007). Reductions in dietary energy density are associated with weight loss in overweight and obese participants in the PREMIER trial. American Journal of Clinical Nutrition, 85(5):1212-1221. Read on PubMed →
  11. Rolls BJ, Roe LS, Beach AM, Kris-Etherton PM (2005). Provision of foods differing in energy density affects long-term weight loss. Obesity Research, 13(6):1052-1060. Read on PubMed →
  12. Robinson E, Khuttan M, McFarland-Lesser I, Patel Z, Jones A (2022). Calorie reformulation: a systematic review and meta-analysis examining the effect of manipulating food energy density on daily energy intake. International Journal of Behavioral Nutrition and Physical Activity, 19(1):48. Read on PubMed →
  13. Stelmach-Mardas M, Rodacki T, Dobrowolska-Iwanek J, et al. (2016). Link between food energy density and body weight changes in obese adults. Nutrients, 8(4):229. Read on PubMed →
  14. Pérez-Escamilla R, Obbagy JE, Altman JM, et al. (2012). Dietary energy density and body weight in adults and children: a systematic review. Journal of the Academy of Nutrition and Dietetics, 112(5):671-684. Read on PubMed →
  15. Karl JP, Roberts SB (2014). Energy density, energy intake, and body weight regulation in adults. Advances in Nutrition, 5(6):835-850. Read on PubMed →
  16. Greene LF, Malpede CZ, Henson CS, et al. (2006). Weight maintenance 2 years after participation in a weight loss program promoting low-energy density foods. Obesity, 14(10):1795-1801. Read on PubMed →
  17. Lowe MR, Tappe KA, Annunziato RA, et al. (2008). The effect of training in reduced energy density eating and food self-monitoring accuracy on weight loss maintenance. Obesity, 16(9):2016-2023. Read on PubMed →
  18. Lowe MR, Butryn ML, Thomas JG, Coletta M (2014). Meal replacements, reduced energy density eating, and weight loss maintenance in primary care patients: a randomized controlled trial. Obesity, 22(1):94-100. Read on PubMed →
  19. Rolls BJ, Roe LS, James BL, Sanchez CE (2017). Does the incorporation of portion-control strategies in a behavioral program improve weight loss in a 1-year randomized controlled trial? International Journal of Obesity, 41(3):434-442. Read on PubMed →
  20. National Heart, Lung, and Blood Institute, NIH (1998). Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. NIH Publication 98-4083. Read the guidelines →

The formulas and limits behind MyPlate's own numbers, with their citations, are on the health sources page.