Free Meals Doubled Weight Loss at a Year. The Regain Came Anyway

The people in this trial were 202 men and women in Pittsburgh and Minneapolis-St. Paul, aged 25 to 45, healthy non-smokers on no prescription medication. Each was 14 to 32 kg (about 30 to 70 lb) overweight.

Most had already been through an organised weight-loss programme at least once, and most were White. Robert Jeffery, Rena Wing and their colleagues split them into five groups, followed them for 18 months and published the results in 1993.

Four groups got the same weight-loss programme, about 20 people to a room: weekly for 20 weeks, then monthly, each person with a calorie target and a daily food diary.

On top of that, one group was also handed a big share of its food, one got paid for losing weight, and one got both. The fifth group got no programme at all, was free to diet on its own, and was simply weighed.

At the 12-month weigh-in, the people who got the food had lost 9.1 kg (about 20 lb), against 4.5 kg (about 10 lb) for the same programme without it.

Twice as much, at a year. That date matters, because the gap was a different size at every weigh-in and the trial ran six months longer.

Are Meal Delivery Services Worth It for Weight Loss? Read the Dates

Put the three weigh-ins side by side. At six months the food groups were down 10.1 kg against 7.7 kg. At twelve, 9.1 against 4.5. At eighteen, 6.4 against 4.1.

Read it like a race and the lead does something odd. It grew for a whole year, from about a third more to twice as much, and then it shrank.

That is the half of "worth it" a trial can answer: how many kilos, and when. The other half, the price, it can't touch, because every meal in this trial was free. The blank is not special to meal deliveries: across 17 studies of trainers, meal boxes, groups, app coaches and cash, nobody measured the price.

Each of those figures is an average across two groups, and the paper doesn't show how much individual results varied.

So what exactly did "twice as much" take? Ten meals a week, and it is worth knowing which ten.

Cereal, Lean Meat and a Frozen Dinner: What the Food Groups Got

The food groups were handed prepackaged meals for five breakfasts and five dinners every week, for the whole 18 months, free.

Breakfast was mostly cereal, milk, juice and fruit. Dinner was usually lean meat, potato or rice and a vegetable. On one or two days a week it was a frozen dinner instead, such as a Weight Watchers or Lean Cuisine meal.

Every meal was made up to the person's own calorie target, 1,000 or 1,500 calories a day. A written meal plan said what to eat at which meal, recipes came with it, and lunch was left to them, with some suggestions.

One company in Pittsburgh made all of it, and the paper doesn't say where people picked it up.

Lay that week next to your own and there is still plenty nobody handed over: every lunch, two breakfasts, two dinners and whatever got eaten in between. The trial also recorded what it handed over, not whether those meals were eaten.

It would be easy to guess that the food groups simply ate what came and stopped counting.

Meal Plan or Counting Calories? In This Trial, It Was Both

Nobody in the programme got out of counting. Everyone was asked to write down what they ate every day for the first 20 weeks, then for one week a month, and to stay under their calorie target.

Here is the turn. The people handed their food filled in more of those diaries, not fewer, in every stretch of the trial. In the first 20 weeks the two food groups completed 85% and 89% of their diary days, against 67% and 81% without food. The gap was still there at the end: in the last six months the food groups filled in over a third of their diary days, the others under a quarter.

They showed up more too: in the final six months the food groups made it to 64% of the group sessions, while the programme on its own drew 26%.

People who came more often and wrote more down also tended to lose more. That is two things moving together, not proof that the diaries did it. The researchers offer their own alternative: the free food may have been a reason in itself to show up and do the homework.

Either way, the choice between a meal plan and counting never came up here. The food arrived with more of the counting done, not less.

And if showing up went with losing more, wouldn't paying people to lose weight push it further?

The $25 Cheque That Moved Nothing

The researchers tested exactly that, paying two of the groups by cheque at every weigh-in. The schedule ran $2.50 a week just for not gaining, $12.50 for being halfway to their goal, and $25 a week for reaching it and holding it.

Hit the goal and hold it for all 78 weeks and the cheques would top out at $1,950. That is a ceiling worked out from the schedule; the paper doesn't say how much was actually paid.

The cheques made no difference to weight that the analysis could find, on their own or added to the food.

The authors' verdict is dry: they suggest $25 a week isn't enough of a direct reward to keep weight-loss habits going. They add that this was a straight payment, not people putting their own money at stake, which earlier studies had found could work differently.

The food, meanwhile, came every week whether anyone lost a gram or not. Nothing about it was tied to results, and it was the side that pulled ahead.

Which brings the subscription question down to its bluntest form: meal for meal, what did the food actually buy?

Count the Meals: What Each Extra Kilo Took

Count it the way you would count a subscription. At ten meals a week, each person in the food groups had been handed roughly 520 breakfasts and dinners by the one-year mark, and about 780 by the end.

At one year the food groups were 4.6 kg ahead of the groups without food, roughly 113 meals for every extra kilo. By 18 months the lead had fallen to 2.3 kg, about 339 meals per extra kilo.

Same meals, same schedule, and each one was buying about a third as much.

These are rough sums on averaged groups. They count meals handed over, not meals eaten.

Which leaves one thing to explain: why the return shrank while the food kept coming.

Month 6 to Month 18: Deliveries on Schedule, Weight Coming Back

Because from month six, the weight started coming back, and the food didn't stop it.

Between the six-month and the 18-month weigh-ins, the food groups put back 3.7 kg and the groups without food put back 3.6 kg. Every week of that year, the meals kept arriving.

Twice the weight loss at a year. About the same regain afterwards. The food groups were still ahead at 18 months, but by about the same margin they had held at month six. A year more of food had left the lead where it stood.

Two honest footnotes. Those regain figures are averages of two groups each, and no test checked whether the two were really the same. The food groups also kept coming to more of the sessions as attendance fell everywhere.

The researchers didn't dress it up. Every group that got help slowly regained after six months, meetings, meals and all.

A result that tidy invites a harder question: which part of the food did the work?

Food, Plan or Portions: Which Part Did the Work?

The authors asked it themselves. What they handed over wasn't one thing but three bundled together: a detailed, structured diet, food that was free, and meals already packed in sensible portions.

Each piece could matter on its own. Free food, as they noted, may have been its own reason to turn up. Ready portions may have steered eating simply by being there, measured out.

This trial never pulled the three apart. For anyone about to pay for just one of them, that is the most useful gap on the page.

One trial, Wing 1996, was built to test exactly that: the written plan against the food itself. Its full text isn't available online, so this page names it and reports nothing from it.

Then there is the price: the authors point out that real life is different, especially because these people never paid for their food.

And the analysis only counts people who came back for every weigh-in. That was 90% of one food group but 65% of the plain programme, so the two sides were filtered differently before anyone compared them.

So when other researchers handed food over, did the lead last?

Fade or Last: Where the Rest of the Evidence Splits

They split, and the split is the honest picture.

A 2017 trial of 186 women handed out vouchers for single-serving main dishes, cut back from 14 a week in the first month to 4 a week from month four. The vouchers brought faster weight loss at first. That group also regained faster, so by six and twelve months it was no longer reliably different from groups given the same amount of counselling without them.[1]

A 2019 review of 23 trials of meal replacements, mostly shakes, soups and bars rather than full meals, found them still ahead at one year. Its authors say the gap appears to hold for up to four years.[2]

Those two aren't separate votes. The women's trial is one of the 23 in the review, so it is already counted there.

A 12-week trial of 112 adults in Hull, UK, found free commercial meals beat a do-it-yourself diet with the same calorie deficit: 6.6% of body weight lost against 4.3%.[3]

It was funded by Diet Chef Ltd, which the authors say had no input. It had no follow-up, and again the food was free.[3]

People stayed, too. Only 7% of the meal group dropped out, against 41% of the do-it-yourself group. Most of those who left went in the first four weeks, which the authors say may reflect disappointment at landing in the group without the food.[3]

At some point, every one of these trials found the food side ahead. Whether the lead lasts depends on which trial you read. On the hardest part of dieting, keeping it off, the 1993 paper left the question open.

Keeping It Off: The Problem This Trial Handed On

The 1993 paper's last line says it plainly: keeping weight off remains the greatest single challenge in weight-control research.

The paper's second author is Rena R. Wing. The National Weight Control Registry's own page says Rena Wing of Brown Medical School and James O. Hill set it up in 1994, the year after this trial was published. Its reason: few people were believed to succeed at long-term weight loss.[4]

The registry's page doesn't mention this trial; the link is simply the same name. What it does say is that the registry exists to study the people who kept weight off, and in 2024 it was tracking over 10,000 of them.[4]

So here is what this trial can hand you: free breakfasts and dinners on top of a group programme bought a much bigger first year, came with more logging and more showing up, and did not stop the regain.

That leaves two lines on your own ledger that no trial here filled in. Which part you would actually be paying for, when food, plan and portions came as one bundle. And what carries you once the weight starts creeping back, which in this trial began after month six, food or no food.