On a very low calorie diet of around 800 calories a day, most people lose somewhere between one and two and a half kilograms a week after the first week. The first week is usually much larger, often three or four kilograms, and that number is misleading. Here is what actually happens, what the trials found, and the parts the programme brochures leave out.
The short answer
A total diet replacement programme puts you on roughly 800 calories a day, usually as three or four shakes or soups. In the supervised trials, people on this kind of programme lost around 10 to 15 kilograms over eight to twelve weeks. Week one produces the biggest number on the scale and the smallest amount of fat loss. And the uncomfortable finding from the best long term trial is that most of it comes back, whether you lose it quickly or slowly.
Week one is mostly not fat
Almost everyone has the same experience. The first week is dramatic, the second week is half as good, and by week three people start asking what they are doing wrong. Nothing. The first week includes a large amount of water.
The reason is glycogen, the body’s short term carbohydrate store in muscle and liver. Glycogen is stored bound to water. Cut carbohydrate hard and those stores run down over the first few days, and the water they were holding goes with them. That is a real physiological change, and it is not fat. It also reverses the moment you eat normally again, which is why people who come off a programme for one weekend can appear to gain two kilograms overnight and conclude they have ruined everything.
Worth being honest here: we could not find a clean trial putting a precise figure on how much of week one is water, so this page does not give you one. The mechanism is well established. The exact split is not something to invent a number for.
What the trials actually achieved
This is one of the few areas of weight management with genuinely good evidence, because total diet replacement has been tested properly in primary care.
DiRECT (PubMed 29221645) put people with type 2 diabetes on a total diet replacement of about 825 to 853 calories a day for three to five months, with structured food reintroduction and long term support. Nearly a quarter of the intervention group lost 15 kilograms or more at twelve months, and almost half had their diabetes in remission.
DROPLET (PubMed 30257983) tested the same idea in general practice without diabetes as the entry point, and found substantially greater weight loss at twelve months than usual care.
A meta-analysis of low energy formula diets (PubMed 27698345) pooled the results across studies in people with and without type 2 diabetes. These programmes work. That is not the question. The question is what happens next.
Does losing it fast make it come back faster?
Everyone has heard that slow and steady wins, and that weight lost quickly comes back quickly. An Australian randomised trial tested exactly that (PubMed 25459211) and the result is not what either camp expects.
Two hundred people were randomised to lose 15 per cent of their body weight either rapidly over 12 weeks or gradually over 36 weeks, then followed for nearly three more years. Two things came out of it.
- The fast group was far more likely to get there. 81 per cent of them reached the target, against 50 per cent of the gradual group. Slow programmes lose people.
- Neither group kept it. At the end of the follow up, both had regained around 70 to 76 per cent of what they had lost. The rate of loss made no difference to the rate of regain.
So the usual advice is wrong on the first point and the usual sales pitch is wrong on the second. Rapid loss is not the problem. Having no plan for the two years afterwards is the problem, and that is the part almost nobody sells you.
The side effect that actually turns up
Headaches, cold hands, constipation, bad breath and fatigue in the first fortnight are common and usually settle. The one worth planning around is gallstones.
In a study of 179 patients on a 605 calorie diet (PubMed 1615894), 9 per cent already had gallstones at the start and 11 per cent developed new ones during the programme. Rapid weight loss changes how the gallbladder empties and how saturated the bile is. This is well enough established that a trial tested a drug to prevent it: ursodeoxycholic acid during a very low calorie diet (PubMed 7755224).
That is not a reason to avoid these programmes. It is a reason to do one with a doctor who knows you are doing it, and to take new upper abdominal pain seriously rather than assuming it is the diet.
Broader safety reviews of very low energy diets (PubMed 36297107) and of very low calorie ketogenic diets (PubMed 31705259) reach a similar place: effective, generally tolerated, and appropriate under supervision rather than as a thing you order online and start on a Monday.
You will lose some muscle, and you can lose less
Any large energy deficit costs some lean tissue alongside fat. That matters more than it sounds, because muscle is a large part of what burns energy at rest, and losing it makes the maintenance phase harder.
A review of preserving muscle during weight loss (PubMed 28507015) points at the two levers that are actually within your control: enough protein, and resistance exercise. Formula products are built to carry a reasonable protein load for this reason, which is one of the real differences between a proper total diet replacement product and a cheap shake, and it is covered in our comparison of Optislim and Optifast.
When this needs a doctor rather than a shopping basket
Speak to a doctor before starting if any of these apply:
- You take insulin or any diabetes medication. Doses often need to come down quickly, and on 800 calories a day an unchanged dose is a hypoglycaemia risk.
- You take blood pressure medication, for the same reason in reverse.
- You take diuretics or lithium, where fluid shifts matter.
- You have a history of eating disorders. A rigid very low calorie programme is a poor fit.
- You are pregnant or breastfeeding.
- You have gallbladder, liver, kidney or heart disease.
Below roughly 800 calories a day this stops being a diet and starts being a treatment. The trials that produced the good results all had supervision built in.
How to tell whether it is working
Weigh at most once a week, at the same time of day, and expect the line to be bumpy rather than smooth. Day to day scale movement on a programme like this is mostly fluid and tells you nothing useful.
The tape measure around the waist is the better instrument from week two onwards. It ignores water shifts and it tracks the fat that actually matters for metabolic health. When the scale stalls for ten days and the tape has moved two centimetres, the tape is telling you the truth.
The part that decides the outcome
Given the regain figures above, the only question that really matters is what happens in month four and year two. The programmes that worked in the trials did not end with the shakes. DiRECT reintroduced food in a structured way and kept monthly support going for a year.
If you are considering one of these programmes, judge it on what it offers after the shakes stop. A box of sachets with no plan attached is the version that regains 76 per cent. If a programme cannot tell you what month four looks like, that is the answer to whether it will work. The same logic applies to the supplements marketed as gentler alternatives, which we looked at in natural Ozempic alternatives, and to the food and habit changes in our guide to lowering blood sugar naturally.
Frequently asked questions
How much weight will I lose in the first two weeks?
Commonly three to five kilograms across the two weeks, with most of it in the first seven days. A large share of that first week is glycogen bound water rather than fat, so do not treat it as your weekly rate.
Why has my weight loss slowed down in week three?
Because week one was not a fat loss rate. Once glycogen and its water have gone, what remains is the steadier one to two and a half kilograms a week. Nothing has gone wrong.
Is losing weight quickly worse than losing it slowly?
A randomised trial found no difference in regain between a rapid 12 week programme and a gradual 36 week one, and the rapid group was far more likely to reach its target at all. The risk of rapid loss is medical, chiefly gallstones, not a rebound penalty.
Will I put it all back on?
On the evidence, most people regain most of it within about three years unless there is a maintenance plan. In that trial both groups regained around 70 to 76 per cent. Choose a programme on its maintenance phase, not its first month.
Do I need a doctor to do this?
For a true very low calorie programme at or below 800 calories a day, yes. It is essential if you take diabetes or blood pressure medication, because doses usually need adjusting quickly.
Can a very low calorie diet put type 2 diabetes into remission?
In the DiRECT trial, nearly half of the intervention group were in remission at twelve months, and remission tracked closely with how much weight was lost and kept off. It is a real outcome, achieved with supervision and structured support rather than shakes alone.
Why do I feel cold and tired on it?
Fatigue, feeling cold, headaches, constipation and bad breath are common in the first week or two and usually settle. Tell your doctor if they do not, and treat dizziness on standing or new abdominal pain as reasons to seek advice rather than push on.
Does it cause gallstones?
It can. In one study of a 605 calorie diet, 11 per cent of patients developed new gallstones during the programme. Rapid weight loss alters gallbladder emptying and bile composition, which is why supervision matters and why new upper abdominal pain should be checked.
Medical disclaimer
This article is for information only and is not medical advice. Very low calorie diets are medical interventions and should be undertaken with a doctor, particularly if you take any prescribed medication. Do not change or stop prescribed medication because of anything you read here. Seek medical advice promptly for new upper abdominal pain, repeated dizziness, fainting, or symptoms of low blood sugar while following a programme like this.