Calorie density and weight loss: what 14 plant based experts say
Calorie density is how many calories a food carries for its weight. Vegetables and fruit are mostly water and fiber, so a full plate holds few calories; oil, nuts, flour products and sugar pack many calories into a small amount.
Most people in this guide say that filling up on low-density foods lets you eat until satisfied and still lose weight. They differ on whether calories should ever be counted, and on what else drives weight gain.
The statement we compare againstChoosing foods with fewer calories per pound is the main way to lose weight without counting calories.
- Agree 10
- Partly 3
- Disagree 1
- No statement 0
At a glance
| Expert | Position | Evidence behind it |
|---|---|---|
| Walter Kempner | Disagrees | Case series, selected for success |
| Nathan Pritikin | Agrees | Case series (uncontrolled residential programs) |
| John McDougall | Agrees | Case series (short, residential) |
| Dean Ornish | Agrees | RCT for heart outcomes; modest results in independent weight trials |
| T. Colin Campbell | Partly agrees | Ecological and animal studies |
| Neal Barnard | Agrees | Randomized controlled trials |
| Hana Kahleova | Partly agrees | Randomized controlled trials, plus cohort data |
| Alan Goldhamer | Agrees | Case series (single-arm clinic studies) |
| Doug Lisle | Agrees | No studies of his own; clinical experience |
| Jeff Novick | Agrees | No studies of his own; summarizes feeding studies |
| Chef AJ | Agrees | Personal story and program reports; no trials |
| Michael Greger | Agrees | Reviews of other people's trials |
| Rip Esselstyn | Agrees | Uncontrolled pilot and event data |
| Peter Rogers | Partly agrees | No studies of his own; population and mechanism arguments |
What each expert says
-
Walter Kempner
MD
DisagreesKempner's weight-loss program cut calories directly: 400 to 800 a day at first and always under 1,000, with daily supervision. His paper calls calorie reduction the cornerstone, and his successor Frank Neelon later said on McDougall's channel that calories do count.
Evidence: Case series, selected for success
- Kempner 1975, Archives of Internal MedicineStudy
- Frank Neelon on McDougall's channel (2022) at 58:11Someone else speaking
-
Nathan Pritikin
Engineer, no medical degree
AgreesThe center teaches eating to comfortable fullness from low-density foods, since people tend to eat about the same weight of food from day to day. On a plateau, though, it advises eating fewer calories or exercising more.
Evidence: Case series (uncontrolled residential programs)
- Pritikin center: new science on losing weightOrganization page
- Pritikin center: plateau tipsOrganization page
-
John McDougall
MD
AgreesMcDougall told people to fill up on starches and vegetables without counting or weighing anything. His explanation combined calorie dilution with his view that dietary fat turns into body fat far more easily than carbohydrate does.
Evidence: Case series (short, residential)
-
Dean Ornish
MD
AgreesOn camera Ornish says calories count, but argues that cutting fat lets you eat the same amount of food with far fewer calories: change the kind of food, not the amount. His program's written guidelines add portion control for weight loss, which he does not say himself in his videos.
Evidence: RCT for heart outcomes; modest results in independent weight trials
-
T. Colin Campbell
PhD
Partly agreesCampbell argues that people on low-fat plant diets can eat more calories and still weigh less. He credits mostly physical activity, plus a possible extra calorie burn on low-protein diets seen in rats, and calls lasting calorie restriction close to impossible for most people.
Evidence: Ecological and animal studies
-
Neal Barnard
MD
AgreesPCRM gives low energy density as one of two reasons its low-fat vegan diet works, the other being more calories burned after meals. Its trial volunteers had no calorie limits yet ate less.
Evidence: Randomized controlled trials
- PCRM weight-loss pageOrganization page
- Barnard 2005, randomized trialStudy
-
Hana Kahleova
MD, PhD
Partly agreesIn her 2020 trial, people on an unrestricted low-fat vegan diet ate about 355 fewer calories a day than controls and burned more after meals, and the authors credit both. Inside the trial the after-meal burn did not track weight change, so eating less explains most of the result.
Evidence: Randomized controlled trials, plus cohort data
-
Alan Goldhamer
DC (chiropractor, not a medical doctor)
AgreesGoldhamer teaches a ladder from salad, about 100 calories per pound, to oil, about 4,000, and says losing weight means removing the concentrated processed foods. He adds that salt, oil and sugar push people to overeat through the brain's reward system.
Evidence: Case series (single-arm clinic studies)
-
Doug Lisle
PhD
AgreesLisle says you never need to eat less than your hunger asks: make the diet less calorie-dense and the body loses weight on its own. His explanation: rich food, fat most of all, gets past the signals that normally tell the stomach it is full.
Evidence: No studies of his own; clinical experience
-
Jeff Novick
MS, RDN (registered dietitian)
AgreesCalorie density is Novick's whole method, with a number scale: eat freely below roughly 300 to 400 calories per pound, moderately up to 800, limit up to 1,800 and extremely limit above that. He rejects calorie counting outright.
Evidence: No studies of his own; summarizes feeding studies
-
Chef AJ
Chef and TV host, no medical credential
AgreesChef AJ's program opens the day with a pound or more of non-starchy vegetables, eats foods in order of rising density and stops at comfortably full. She says calories do count, but that counting them does not last.
Evidence: Personal story and program reports; no trials
-
Michael Greger
MD
AgreesGreger's weight-loss advice rests on low calorie density, fiber and whole plant foods whose cell walls are intact, with low-density foods eaten before meals. He says no counting is needed, yet notes his Daily Dozen lands at roughly 1,200 to 1,500 calories.
Evidence: Reviews of other people's trials
-
Rip Esselstyn
Former firefighter and triathlete, no medical credential
AgreesRip says calorie-counting diets do not last and that generous amounts of unprocessed plant food do the work. PLANTSTRONG's site explains weight loss as calorie density: skip liquid calories, high-fat foods and refined grains.
Evidence: Uncontrolled pilot and event data
- Engine 2 keynote (2009) at 1:13:31
- PLANTSTRONG blog: calorie densityOrganization page
-
Peter Rogers
MD
Partly agreesRogers does not count calories and has people eat vegetables first to stretch the stomach. His main explanation, though, is that dietary fat is stored as body fat, along with oils, hormone-like chemicals and stress.
Evidence: No studies of his own; population and mechanism arguments
What the evidence says
- In PCRM's 16-week trial, people on a low-fat vegan diet with no calorie limit ate roughly 490 fewer calories a day on their own and lost about 5.9 kg.
- When PCRM compared a vegan diet with portion-controlled eating over 20 weeks, the difference in weight loss was not statistically significant.
- The only independent long-term trial of an unrestricted whole-food plant based diet (BROAD, New Zealand, 2017) reported about 11.5 kg lost at 12 months, but only 23 of 33 participants were followed up.
Our reading: Calories still decide the result. Low-density food is a way to eat fewer of them without counting, for the people who stick with it.
Questions
What is calorie density?
It is the number of calories in a given weight of food, often given per pound. Raw vegetables are around 100 calories per pound, cooked starches around 500, bread around 1,500 and oil around 4,000.
Do you need to count calories on a low-calorie-density diet?
Most experts in this guide say no: low-density food fills you up before you overeat. Kempner's program counted strictly, and the Pritikin center suggests cutting calories further on a plateau.
Is calorie density proven for weight loss?
Feeding studies and PCRM's trials show people eat less on low-density diets without being told to. Long-term trials are few and small.