Does Protein Keep You Full? Yes — But a Potato Beat Everything

Yes, and by a margin worth having: about 441 calories a day, eaten without anyone noticing they were eating less [1].

Then there is the awkward part, which is that the most filling food anyone has ever put on a ranked list is a boiled potato [3].

Both of those are true. Here is how they fit together.

The study that isolated protein properly

The problem with most protein research is that high-protein diets are usually also low-carb diets, so you can never tell which lever did the work.

A team at the University of Washington and Oregon Health and Science University fixed that — eleven volunteers studied in Seattle, eight in Portland. They put nineteen people through three phases and held carbohydrate constant at 50% of calories the entire time [1]. Only protein moved — from 15% of calories up to 30%, with fat coming down to make room.

And who were the nineteen? Sixteen women and three men, aged 27 to 62 and averaging 41, averaging 159 lb (72 kg), with a body mass index — the weight-to-height ratio that calls 25 to 30 overweight — from 22.5 to 30.1, averaging 26.2 [1].

The study also turned people away: a body mass index over 30, aerobic exercise more than half an hour three times a week, tobacco, more than two drinks a day, an unusually high or low carbohydrate habit to begin with, and — this one matters — any expectation of losing weight [1]. Everyone then took the same three diets in the same order, with nobody to compare them against: no control group, no randomisation, nothing hidden from anyone [1].

So the 441 is a number from mostly middle-aged women who were overweight but not obese and who did not train. It is still the test on this page that isolates protein best — three months, with carbohydrate pinned at half of calories the whole way. It is just not a measurement taken on you, and this page should have said so the first time.

Phase two is the one to watch. Protein went up, but the calories were matched to phase one exactly. Same energy, different composition.

Satiety went up markedly anyway [1].

That is the finding in its cleanest form: identical calories, and people felt fuller. Not because they ate more, because of what they ate.

Then phase three let them eat as much as they wanted for twelve weeks. They spontaneously ate 441 calories a day less, lost 10.8 lb (4.9 kg), and 8.2 lb (3.7 kg) of that was fat [1].

Nobody was on a diet. Nobody was counting. They just stopped earlier.

The detail that makes this weirder than it looks

Your appetite is partly run by two hormones. Leptin tells your brain there is fuel in reserve; ghrelin is the one that makes you want lunch.

Over those twelve weeks, leptin went down and ghrelin went up [1].

Both of those changes should have made these people hungrier. Both are exactly what happens when you lose weight and your body starts arguing about it — the well-documented reason so much lost weight comes back.

And they kept eating less anyway, for three months, with the hormones pulling the other way the whole time.

The authors do offer an explanation, and they hedge it themselves: the effect “may be mediated by increased central nervous system leptin sensitivity” [1]. In plain words — not more leptin, but a brain listening harder to the leptin already there. Note the may. What they measured was the eating. The mechanism is their proposal, and this study did not test it.

So does that mean protein wins? Not quite.

In 1995 a group in Sydney did something nobody had bothered to do: they ranked actual foods against each other for fullness.

Thirty-eight foods, each served at exactly 240 calories so the comparison was fair, with satiety measured every fifteen minutes for two hours. White bread was set at 100% and everything else scored against it [3].

The winner was boiled potatoes, at 323%. The loser was a croissant, at 47% — a seven-fold gap between the best and worst food on the list [3].

And when they checked which nutrients actually predicted the scores, protein was there — but it was nowhere near the top of the list [3]:

Water content had the strongest link of the four: a correlation of 0.64, on a scale where 1.00 would mean two things move together perfectly and 0 would mean no relationship at all. Fibre came next, at 0.46. Protein was third among the things that helped, at 0.37 — and the weakest association of any of them, because fat pulled harder in the other direction, at −0.43. More fat, less full [3].

And one more line from the same table, which is the best argument on this page: the strongest predictor of all was not a nutrient at all. It was the sheer weight of the serving, at 0.66 [3]. Bulk beat chemistry.

Which reframes the whole thing. Protein does make food more filling. It is just not the biggest lever on the board, and a food that is bulky and wet and fibrous can beat a protein-dense one that is dry and small.

A boiled potato is mostly water, has fibre, and has almost no fat. It wins on three of the four.

How strong is the evidence, honestly?

Here is a rare thing: a meta-analysis that prints a warning about itself in its own title.

The paper is called “The Effects of Increased Protein Intake on Fullness: A Meta-Analysis and Its Limitations” [2].

The result was clear enough — higher-protein preloads beat lower-protein ones on fullness, comfortably [2]. A separate count of 28 studies pointed the same way [2].

But only five studies were strict enough to pool [2]. And the authors close by warning that extending their finding beyond “tightly defined conditions” — to real eating behaviour, to body weight, to health outcomes — must be done cautiously [2].

That is researchers telling you not to oversell them.

Five studies is a thin pile, though. So how replicated is this, really? Better than that, in the short run. A 2020 review took only randomised trials in healthy adults with the calories matched between the two arms, and found 49 published reports of the short-term effect. Pooled, protein moved fullness up by 10 millimetres on the line people mark to show how full they feel, and hunger down by 7 — and the odds of a shift that big turning up if protein did nothing are under one in a thousand [4].

Then the same review turned to 19 long-term reports and found nothing to speak of: no change in fullness, no change in hunger, and no change in the gut hormones behind them bar one — GLP-1, a hormone the gut releases after a meal, which went down over the long run after going up in the short one. The review’s own summary calls the long-term trials inconclusive [4]. Note what that does and does not say, though — it is about appetite and its hormones over months, not about the pounds.

Put the two halves together and you have the calibration, which is why this verdict says Supported and not Established. The next-meal effect is one of the better-replicated findings in appetite research. The part everybody actually wants — that it keeps working for months — rests on nineteen people who all took the same diets in the same order with no control group, while the randomised long-term evidence came back undecided.

What to actually do

Put protein in the meal you most often overeat after. The fullness turns up at the next meal — and in the one long trial here it kept turning up for three months, with 10.8 lb (4.9 kg) gone by the end of them [1]. What nobody can tell you is what happens after twelve weeks, because the randomised long-term evidence is inconclusive [4].

But stop treating fullness as a protein-only problem. Water and fibre both predicted it at least as well [3], which is a roundabout argument for the least fashionable advice in nutrition: eat the potato, eat the vegetables, and be suspicious of dry, dense, low-water food that disappears in four bites.

And note what the 441-calorie finding was not. It was not a shake, a bar, or a powder. It was the protein content of ordinary meals moved from 15% to 30% [1]. If you want the separate question of how much protein you actually need, that has its own verdict. And if your question is powder rather than food, we checked whey against plant protein separately.

This is journalism about research, not medical advice. If you have been told to watch your protein for a medical reason, that decision belongs with your clinician, not with a satiety study.