Breaking a Fast.
The risky part of a long fast is usually the meal that ends it. This checks your refeeding-syndrome risk against the NICE criteria, then gives you a graded schedule if one is appropriate.
Educational reference, not medical advice. The risk criteria are reproduced from NICE CG32, a guideline written for clinicians assessing malnourished patients, and applying it to yourself at home is an approximation of what a clinician would do with the same checklist plus bloods. It is not a substitute for that.
Frequently Asked Questions
How long should I take to break a fast?+
The rule circulating online is to refeed for half as many days as you fasted. The rule with an actual source behind it is closer to a third: the European fasting consensus and the Buchinger series both use four graded build-up days after a fast of seven to ten days. Those two rules disagree by a factor of about one and a half, and this page tells you which number came from where rather than pretending they agree.
What is refeeding syndrome, and how likely is it really?+
When you eat again after a long period without food, insulin rises and drives phosphate, potassium and magnesium out of the blood and into cells. If those stores were already depleted, blood levels can fall far enough to affect the heart and the nervous system. Honest answer on likelihood: reported incidence ranges from under 1% to around 18% depending entirely on how a study defines it, and the high figures come from hospitalised, malnourished populations. A well-nourished adult breaking a two or three day fast is at low risk. That is not the same as no risk, which is why the criteria below are worth two minutes.
Why does this ask whether my bloods were tested?+
Because one of the four criteria that alone makes someone high risk is a low potassium, phosphate or magnesium level before eating. If you have not had those measured, you cannot answer it, and a tool that treats an unticked box as a normal result is quietly asserting something it does not know. Most people who fast at home have not had those bloods, so most people will land on "not classifiable" here. That is the honest output.
What should I actually eat first?+
Something small, mostly protein and fat, low in refined carbohydrate, and eaten slowly. Carbohydrate is what drives the insulin spike that moves electrolytes into cells, so a large sugary or starchy first meal is the one to avoid. Broth, eggs, fish, yoghurt or cooked vegetables are unremarkable choices. The volume matters as much as the content: your first meal after a multi-day fast should feel too small.
What are the warning signs after I start eating?+
Swelling in the ankles or face, breathlessness, palpitations or an irregular heartbeat, confusion, and marked weakness, generally within the first 72 hours of eating again. Those warrant urgent medical attention rather than waiting to see. Mild tiredness and some digestive upset are common and are not the same thing.
Why the meal is riskier than the fast.
Going without food is metabolically quiet. Your body shifts to burning fat, insulin stays low, and phosphate, potassium and magnesium drift out of the tissues into the blood, where they look deceptively normal on a test even as total body stores fall.
Eating reverses that in minutes. Carbohydrate drives insulin up, insulin pushes glucose into cells, and those three electrolytes follow it inward. If stores were already depleted, blood levels can drop far enough to disturb heart rhythm, breathing and the nervous system. That is refeeding syndrome, and the reason it catches people out is that the fast felt fine and the meal felt like the safe part.
Worth keeping in proportion. The people it seriously harms are malnourished: long-term undernutrition, eating disorders, alcohol misuse, illness. A well-nourished adult ending a two or three day fast is at low risk. But "low" is doing real work in that sentence, and the criteria below take about two minutes to check.
The honest state of the numbers.
You will see "0.43%" quoted as the incidence of refeeding syndrome. It is not. That figure is the incidence of severe hypophosphataemia in a series of roughly 10,000 hospitalised patients, which is a related but different thing. Published incidence for refeeding syndrome itself spans about 0.4% to 18%, and the spread is not measurement noise: it is that studies use different definitions, so they are counting different events.
The refeed duration has the same problem in miniature. The rule you will read online is to take half as many days as you fasted. The rule with a citation behind it is closer to a third: the European fasting consensus and the Buchinger series both use four graded build-up days after a seven to ten day fast, with a calorie ladder of roughly 800, 1,000, 1,200 then 1,600. Those two rules disagree by about a factor of one and a half, and this calculator tells you which of its numbers came from the literature and which are our own convention rather than blending them into a single confident answer.
What the criteria can and cannot do for you.
The stratification here is reproduced from NICE CG32, a guideline written for clinicians assessing malnourished patients before feeding them. Applying it to yourself at home approximates what a clinician would do with the same checklist, minus the part where they also have your bloods.
That gap is not cosmetic. One of the four criteria that alone makes someone high risk is a low potassium, phosphate or magnesium level measured before eating. If you have never had those tested, you cannot answer it, and no honest tool can call you low risk on the strength of the questions you did answer. That is why this page has a third verdict rather than forcing everyone into a yes or a no.
- National Institute for Health and Care Excellence. (2006, updated 2017). Nutrition support for adults, CG32: criteria for identifying people at high risk of refeeding problems.
- Wilhelmi de Toledo F, et al. (2013). Fasting therapy: an expert panel update of the 2002 consensus guidelines. Forsch Komplementmed, 20(6), 434-443.
- Wilhelmi de Toledo F, et al. (2019). Safety, health improvement and well-being during a 4 to 21-day fasting period in an observational study including 1422 subjects. PLoS One, 14(1), e0209353.
- Friedli N, et al. (2017). Management and prevention of refeeding syndrome in medical inpatients: an evidence-based and consensus-supported algorithm (incidence varies by case definition).
For education, not medical advice. Results are estimates, not a diagnosis. Discuss any abnormal value or health concern with a qualified clinician.
50 High-Protein Snacks. Free.
Drop your email. We'll send the PDF straight to your downloads and add you to the weekly research drop. Unsubscribe anytime.