Fasting as Detox: Autophagy Instead of Slag
Your body has no slag depot. But it does have a waste service inside every single cell. And fasting appears to give it more work.
The way you feel after three days of fasting is real. The explanation you are sold for it is not. It is not slag leaving your body. It is something far more interesting: your cells begin to take themselves apart.
I bet you know the moment. Day two or three without solid food. Your head becomes clear in a strange way. Your belly is quiet. A feeling of lightness, almost of being clean. And then somebody tells you: see, now the slag is coming out.
I have heard that sentence hundreds of times. In consultations, in fasting groups, in podcasts. And I understand why it works so well. It gives a strong physical experience a simple story. A picture of a blocked pipe that finally runs free.
Except that this pipe does not exist. There is no lab value for slag. No imaging method shows it. No pathologist finds it in tissue. The term comes from metal smelting, not from physiology.
And still, something about the feeling is right. During fasting a kind of cleanup really does happen in the body. It simply takes place somewhere completely different from what most people assume. Not in the gut. Not in connective tissue. But inside every single one of your cells.
The 2016 Nobel Prize in Physiology or Medicine was awarded for this. The process is called autophagy. By current understanding it is the best scientific core behind the detox feeling during fasting. It is also considerably less spectacular than the advertising promises. That is exactly why it is worth understanding properly.
What this article covers
- Why the term slag does not exist in medicine
- How liver and kidneys actually process foreign substances
- What autophagy is, explained in simple images
- The nutrient switch inside your cell: mTOR and AMPK
- What has really been measured in humans, and what has not
- Why losing fat can release pollutants in the short term
- When fasting is not a good idea
- Three levers you can put into practice from tomorrow
Why fasting feels like a cleanup
Before we talk about cell biology, let us take the experience seriously. Because it is real and it has several reasons, all of which can be measured.
After roughly twelve to sixteen hours without food, the glycogen stores in the liver are largely empty. The body switches over. It pulls fatty acids out of adipose tissue and the liver builds ketone bodies from them. George Cahill studied this switch in humans over decades and described how D beta hydroxybutyrate in particular allows the brain to get through longer periods of starvation.
Who: Cahill, Harvard Medical School, summarised his life's work on fuel metabolism in the starvation state in the Annual Review of Nutrition in 2006.
What he observed: During longer fasting, energy supply shifts from glucose to free fatty acids and ketone bodies. D beta hydroxybutyrate is used by the brain and, according to Cahill, is what makes survival of longer periods of starvation possible in the first place.
What this means for you: The famous mental clarity on the second day of fasting is not an esoteric phenomenon. It coincides in time with a real change of brain fuel. That is not causal proof of the subjective experience, but it is a very plausible contributor.
Cahill GF. Fuel metabolism in starvation. Annu Rev Nutr. 2006;26:1-22. DOI: 10.1146/annurev.nutr.26.061505.111258Then there is digestive rest. The gut gets no work. Bloating, fullness and reflux disappear in many people within a day. Blood sugar settles at a low, stable level. The rollercoaster of sugar peak and crash falls away.
And finally something comes into play that no lab measures: you made a decision and you are sticking with it. That alone changes how your body feels.
The cleansing feeling during fasting does not need a slag explanation in order to be real. Ketosis, digestive rest, stable blood sugar and a piece of self efficacy are entirely sufficient.
The interesting question is not whether something changes. The interesting question is what actually happens at the level of the cell. And that is where it gets really good.
And now you know why the experience feels so convincing: it is real. Only the usual explanation for it does not hold.
Slag: a term without a finding
May I ask you an uncomfortable question? When somebody tells you your body is full of slag, have you ever asked how that is measured?
I ask without mockery. I used the term myself for years before I started asking questions. It sounds medical. It is not.
There is no parameter in medicine called slag. There is no reference range, no cutoff, no method of detection. What does exist are concrete substances with concrete names: urea from protein breakdown, uric acid from purine breakdown, bilirubin from the breakdown of red blood cells, creatinine from muscle metabolism. All measurable. All with their own route of excretion.
Who: Klein and Kiat, Macquarie University Sydney and Cardiac Health Institute Sydney, examined the evidence behind commercial detox diets in the Journal of Human Nutrition and Dietetics in 2015.
What they observed: By the authors' own account, no randomised controlled trials had been conducted up to that point testing the efficacy of commercial detox diets in humans. They describe the few existing clinical investigations as methodologically flawed and too small in sample size.
What this means for you: When a programme advertises that it clears toxins out of you, there is almost never a measurement behind it. That does not automatically mean a week of fasting achieves nothing. It means the reasoning has to be something other than slag.
Klein AV, Kiat H. Detox diets for toxin elimination and weight management: a critical review of the evidence. J Hum Nutr Diet. 2015;28(6):675-686. DOI: 10.1111/jhn.12286What detox advertising says
- Slag deposits itself in connective tissue
- It makes you tired, heavy and ill
- Fasting flushes it out
- A tea or a juice speeds that up
- Afterwards the body is cleansed
What can actually be described
- Metabolic end products with a name and a lab value
- Fat soluble foreign substances in adipose tissue
- Liver and kidneys work on this continuously
- Fasting may stimulate cellular cleanup
- An endpoint called cleansed does not exist
I am not devaluing the fasting tradition here. Therapeutic fasting has a long clinical history in Europe and is properly supervised in specialised clinics. Many colleagues who use the term slag mean something sensible by it. They are simply describing with an old image something for which we now have a better one.
My point is not who is right. My point is that you get an explanation that still holds up when you question it.
And now you know why I replace the term in conversation rather than simply striking it out.
What your body actually does with foreign substances
Imagine your liver as a chemical plant running around the clock. Substances come in that your body never intended to build: medication residues, alcohol breakdown products, pesticides, plasticisers, components of smoke. In technical language they are called xenobiotics, meaning substances foreign to the body.
The problem with most of them: they are fat soluble. Fat soluble molecules cannot simply be flushed out via urine. They would migrate back into the blood from the kidney tubules. The body therefore has to make them water soluble first.
Phase I: creating the docking point
Enzymes of the cytochrome P450 family oxidise the molecule. They attach a reactive group to it, rather like screwing a handle onto a smooth package. Intermediate products can briefly be more reactive than the original substance. That is why phase I alone is not enough.
Phase II: attaching the water soluble tag
Transferases couple a water soluble partner to that handle: glucuronic acid, sulfate, glutathione, an acetyl or a methyl group. Jancova and colleagues describe phase II reactions as the detoxifying step of drug metabolism. Now the molecule can be transported.
Phase III: the transport out
Transport proteins move the finished conjugate out of the liver cell. From there it goes either via bile into the gut and out with the stool, or via the blood to the kidney and out with the urine.
Who: Jancova, Anzenbacher and Anzenbacherova, Palacky University Olomouc, published an overview of the phase II enzymes of drug metabolism in 2010.
What they describe: Phase II reactions, also called conjugation reactions, generally serve as the detoxifying step in drug metabolism. A reduced capacity of these enzymes can, according to the authors, lead to toxic effects of clinically used medications.
What this means for you: In your body, detoxification is not an event but a continuous operation. It is running while you read this. When something falters, it is usually at the level of enzyme capacity, nutrient supply or a genetic variant, not because of a slag depot.
Jancova P, Anzenbacher P, Anzenbacherova E. Phase II drug metabolizing enzymes. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2010;154(2):103-116. DOI: 10.5507/bp.2010.017The liver needs building blocks, not motivation
Glutathione, glycine, sulfate, glucuronic acid and methyl groups are real substrates. They come from protein, from sulfur compounds and from B vitamins. A programme without protein gives the system less material, not more.
The kidney filters, it does not flush
The kidneys filter around 180 litres of primary urine per day and take most of it back. What they excrete is selected on purpose. Drinking more can make sense, but it does not raise detoxification capacity as such.
The gut is an exit, not a storage room
Conjugates released via bile can be partly split again in the gut and reabsorbed. Fibre binds such substances and can take them out of that circuit with the stool.
Adipose tissue is the actual depot
Fat soluble environmental substances accumulate in adipose tissue over years. That is the one place where the depot idea does touch physiology. Only this depot behaves differently during weight loss than people assume. More on that further down.
And now you know why, when someone asks about detoxification, I talk first about liver building blocks and gut passage and only afterwards about fasting windows.
Autophagy: the waste service inside your cell
Now to the actual core. And I will start with an image.
Imagine a cell as a small workshop. Machines work inside it: mitochondria for energy, ribosomes for building proteins, membranes as partitions. These machines get damaged in operation. Proteins fold into the wrong tangles. Mitochondria become leaky and then produce more free radicals than energy.
A workshop that simply leaves broken machines standing around eventually clogs up. So it needs a clearing system. That is precisely what autophagy is. The word comes from Greek and literally means self eating.
Wrapping
A double membrane grows around the defective component, like a rubbish bag that wraps itself around the scrap. The finished package is called an autophagosome.
Delivering
The package travels to the lysosome. The lysosome is the cell's recycling plant: an acidic vesicle full of digestive enzymes. Both membranes fuse.
Breaking down
The enzymes split the contents into amino acids, fatty acids and sugars. The old engine becomes individual screws again.
Reusing
These building blocks go back into the cell interior. The cell builds new proteins from them or burns them for energy. That is exactly why autophagy is so valuable during hunger: it is clearing up and resupply in a single process.
Who: Takeshige, Baba, Tsuboi, Noda and Ohsumi published the work that laid the foundation in the Journal of Cell Biology in 1992. They observed yeast cells under nutrient deprivation.
What they observed: After one hour in nutrient free medium, a few spherical bodies appeared in the vacuole. After three hours they filled the vacuole almost completely. Their diameter ranged between 400 and 900 nanometres, and their contents were indistinguishable from cytoplasm. The researchers called them autophagic bodies.
What this means for you: This is the moment when nutrient deprivation and self breakdown were first seen together. Yeast is not a human being. But the machinery Ohsumi decoded back then exists in human cells in the same way.
Takeshige K, Baba M, Tsuboi S, Noda T, Ohsumi Y. Autophagy in yeast demonstrated with proteinase-deficient mutants and conditions for its induction. J Cell Biol. 1992;119(2):301-311. DOI: 10.1083/jcb.119.2.301On 3 October 2016 Yoshinori Ohsumi received the Nobel Prize in Physiology or Medicine, explicitly for his discoveries of the mechanisms of autophagy. Rubinsztein and Frake of the University of Cambridge placed this award in context in a commentary and described autophagy as an intracellular degradation pathway that maintains balance in the cytoplasm.
Autophagy is not excretion. It is reuse. Your cell throws nothing away. It takes things apart and uses the pieces again.
Mizushima and Komatsu put it precisely in the journal Cell in 2011: the purpose of autophagy is not the simple disposal of material but a dynamic recycling system that supplies new building blocks and energy for cellular renewal and homeostasis.
If you have been thinking that fasting rinses something out of you, turn the picture around once: during fasting nothing comes out. Things are rebuilt inside.
That is the difference between flushing a drain and renovating a room. And the renovation is by far the more interesting story.
And now you know why I no longer speak of clearing things out when the topic is fasting.
The switch: mTOR, AMPK and why eating pauses the waste service
But why does a lack of food of all things turn this clearing system on? Because your cell has two opposing sensors. One for abundance and one for scarcity.
The growth sensor
mTOR stands for mechanistic target of rapamycin. It registers amino acids, insulin and growth factors. When food is there, mTOR gives the signal: build, grow, make proteins. Clearing up would be a waste right now.
The energy sensor
AMPK measures the ratio of spent to charged energy carrier in the cell. When energy drops, AMPK becomes active. Its message is the opposite: save, reuse, live off what is there.
The shared switch
Both sensors meet at the same starting enzyme of autophagy. It is called ULK1. Which of the two touches it first decides whether the waste service drives out or stays in the garage.
The decision
If mTOR is active, it blocks ULK1. If AMPK is active, it releases ULK1. Fasting lowers mTOR and raises AMPK. That clears the path for autophagy.
Who: Kim, Kundu, Viollet and Guan showed in Nature Cell Biology in 2011 how this switch works at the molecular level.
What they observed: Under glucose deprivation, AMPK activates the starting enzyme ULK1 directly, through phosphorylation at serine 317 and serine 777. When nutrient supply is sufficient, high mTOR activity prevents this activation by phosphorylating ULK1 at serine 757 and disrupting the connection between ULK1 and AMPK.
What this means for you: The sentence eating slows autophagy is not a metaphor. It is three concrete docking points on one enzyme. What remains important: this was shown in cells and in the mouse model, not in fasting humans.
Kim J, Kundu M, Viollet B, Guan KL. AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1. Nat Cell Biol. 2011;13(2):132-141. DOI: 10.1038/ncb2152Carosi, Fourrier, Bensalem and Sargeant of the South Australian Health and Medical Research Institute described this axis as a central control point of ageing in FEBS Open Bio in 2021. Their conclusion: the autophagy lysosome pathway is an excellent target for approaches that aim to slow biological ageing and extend healthspan.
Does a splash of milk in your coffee break autophagy?
This question comes up in almost every discussion about intermittent fasting. The honest answer is: in humans it has not been properly investigated.
The thought is plausible enough. Amino acids can activate mTOR, and active mTOR blocks ULK1. A protein drink supplies plenty of such amino acids, black coffee practically none. How large the effect of small amounts actually is has not been measured in humans so far.
That is exactly why the Break Fast study in Adelaide was set up in the first place. There, autophagic flux in blood is to be determined before and after a protein rich drink. Until results are available, every hour limit and every milk quantity rule remains an assumption.
What gets lost in all this in practice: the big levers are sleep, movement, eating rhythm and adequate protein supply. Debating millilitres of oat milk while those four are unresolved shifts attention to the smallest adjusting screw.
And now you know why the switch exists and why I still do not recommend turning it into a religion.
What has really been measured in humans
This is where it gets uncomfortable. And that is exactly why it matters to me.
The cell biology of autophagy is excellently documented. The question of what happens in which tissue after how many hours in a fasting human is not. The reason is banal: measuring autophagy was long possible only in tissue samples or cell cultures. You would have to take liver or muscle tissue from a person, repeatedly, over time.
Human observation, cohort or review in humans
In vivo animal study
In vitro cell culture, mechanism or basic research
Who: Alirezaei and colleagues at the Scripps Research Institute investigated in the journal Autophagy in 2010 whether short term fasting also stimulates autophagy in the brain. Until then the brain was considered metabolically privileged and therefore exempt.
What they observed: In mice, short term fasting led to what the authors call a dramatic upregulation of neuronal autophagy in cortical neurons and Purkinje cells. In parallel, mTOR activity in the neurons fell. Electron microscopy confirmed the increased autophagosomes.
What this means for you: This is one of the most cited papers behind the claim that fasting cleans the brain. It was carried out in mice. The authors themselves phrase the conclusion as a speculation, not a promise: occasional fasting might be a simple, safe and inexpensive way to promote this neuronal response.
Alirezaei M, Kemball CC, Flynn CT, Wood MR, Whitton JL, Kiosses WB. Short-term fasting induces profound neuronal autophagy. Autophagy. 2010;6(6):702-710. DOI: 10.4161/auto.6.6.12376Who: Dethlefsen and colleagues at the University of Copenhagen had untrained and trained participants fast for 36 hours and took muscle biopsies after 2, 12, 24 and 36 hours.
What they observed: Fasting lowered the content of LC3I, LC3II and p62 in muscle, and did so exclusively in the untrained participants. The authors summarise that autophagy in human skeletal muscle was only moderately affected by 36 hours of fasting and that training state co-determines the effect.
What this means for you: In humans, autophagy is not simply a dial that fasting turns up. It is tissue dependent, state dependent and hard to interpret. Anyone who gives you a simple number of hours is simplifying heavily.
Dethlefsen MM, Bertholdt L, Gudiksen A, Stankiewicz T, Bangsbo J, van Hall G, Plomgaard P, Pilegaard H. Training state and skeletal muscle autophagy in response to 36 h of fasting. J Appl Physiol (1985). 2018;125(5):1609-1619. DOI: 10.1152/japplphysiol.01146.2017Who: Chaudhary and colleagues from Adelaide examined in the journal Nutrition in 2022 whether intermittent fasting changes autophagy markers in liver and muscle. Fifty women aged 51 ± 2 years were randomised to one of two intermittent fasting protocols, with 24 hour fasting days on three non-consecutive days per week over eight weeks.
What they observed: In mice, autophagy markers rose in the liver but not in muscle. In humans, a 24 hour fast raised SQSTM1. After a twelve hour overnight fast, BECLIN1, SQSTM1 and LAMP2 were even lowered in the group with reduced energy intake, in the authors' assessment probably as a consequence of weight loss.
What this means for you: The result does not fit the simple narrative. It shows mixed and partly opposing findings. Studies like this one are exactly what is missing from most fasting podcasts.
Chaudhary R, Liu B, Bensalem J, Sargeant TJ, Page AJ, Wittert GA, Hutchison AT, Heilbronn LK. Intermittent fasting activates markers of autophagy in mouse liver, but not muscle from mouse or humans. Nutrition. 2022;101:111662. DOI: 10.1016/j.nut.2022.111662Who: Jamshed and colleagues investigated in Nutrients in 2019 what influence an early eating window has on gene expression, hormones and blood sugar. Eleven adults with overweight ate, in a randomised crossover over four days, either between 8 am and 2 pm or between 8 am and 8 pm.
What they observed: The early eating window lowered mean 24 hour glucose by 4 ± 1 mg/dl and glucose excursions by 12 ± 3 mg/dl. In the morning before breakfast, ketones and cholesterol rose, as did the expression of an ageing and stress response gene and of the autophagy gene LC3A. In the evening, by contrast, the expression of a central nutrient sensitive growth regulator rose.
What this means for you: This is one of the few direct indications that an autophagy related gene can be moved in humans through meal timing. What was measured was gene expression in blood, not the actual breakdown process inside the cell. An indication, not a proof.
Jamshed H, Beyl RA, Della Manna DL, Yang ES, Ravussin E, Peterson CM. Early Time-Restricted Feeding Improves 24-Hour Glucose Levels and Affects Markers of the Circadian Clock, Aging, and Autophagy in Humans. Nutrients. 2019;11(6):1234. DOI: 10.3390/nu11061234The measurement problem is currently being tackled. Bensalem and colleagues published a study protocol in BMC Nutrition in 2022 in which autophagic flux in humans is to be determined with a newly developed blood test, before and after a protein rich drink. In their rationale they write explicitly that human data are limited because autophagic flux has been difficult to measure in humans so far.
"The cell biology behind fasting is strong. The human evidence is considerably thinner than the hype. Saying both at once is not a weakness. It is honesty."
Shukri Jarmoukli, ViveCura BerlinIf somebody tells you that autophagy starts at exactly 16 hours, that sounds like precision. It is not. For humans, that number does not exist in the literature.
Autophagy runs at a baseline level anyway. It is not a switch you flip at four o'clock in the afternoon. It is more like a dial that moves at different speeds in different tissues.
And now you know why I become sceptical about hour counts and still stay interested in the subject.
The uncomfortable point: losing fat can release pollutants
There is one aspect that never appears in detox advertising, although it is the only place where the depot image really does touch something.
Fat soluble environmental substances deposit themselves in adipose tissue over years. There they are relatively immobilised. When you break down fat, that fat goes into the circulation. And the substances dissolved in it go along with it.
Who: Malarvannan and colleagues at the Toxicological Centre of the University of Antwerp examined 94 adolescents with obesity, 34 boys and 60 girls aged 11 to 19, before and after five months of weight reduction treatment.
What they observed: After weight loss of 4 to 42 kilograms, serum concentrations of most persistent organic pollutants rose significantly. For most substance groups the increase was 1 to 3.5 percent per kilogram of weight lost. The authors write that fat soluble pollutants were released from adipose tissue into the blood and redistributed within the body. Whether this rise in internal burden is harmful to health remains, in their words, to be clarified.
What this means for you: Losing weight can raise pollutant concentrations in blood in the short term. That is not an argument against weight loss, quite the opposite. It is an argument for thinking about the route of excretion as well, instead of only accelerating release.
Malarvannan G, Van Hoorenbeeck K, Deguchtenaere A, Verhulst SL, Dirinck E, Van Gaal L, Jorens PG, Covaci A. Dynamics of persistent organic pollutants in obese adolescents during weight loss. Environ Int. 2018;110:80-87. DOI: 10.1016/j.envint.2017.10.009The popular detox logic goes: mobilise, mobilise, mobilise. The physiological logic reads differently. What you mobilise, you also have to be able to move out.
Rapid fat loss without sufficient phase II capacity and without binding in the gut is like a move where you stack all the boxes in the hallway and then realise that no van has turned up.
What follows from this in practice
- Do not run protein supply down to zero. Glutathione and glycine need amino acids as raw material.
- Keep fibre in view. It can bind in the gut what has been released via bile.
- Choose the pace deliberately. Very rapid weight loss means very rapid release.
- With a known pollutant burden, seek medical supervision rather than forcing things on your own.
And now you know why, on the topic of detoxification, I almost never start with fasting but with the excretion side.
When fasting is not a good idea
Fasting has a friendly image. That leads to it being seen as harmless. It is not. It is a strong intervention in metabolism, and strong interventions have limits.
The good news first: under supervision, tolerability is well documented.
Who: Wilhelmi de Toledo, Grundler, Bergouignan, Drinda and Michalsen prospectively documented 1422 people in PLoS One in 2019 who fasted between 4 and 21 days in a specialised clinic according to Buchinger guidelines, with a daily intake of 200 to 250 kilocalories.
What they observed: Weight, waist circumference and blood pressure fell across the whole group. Blood glucose dropped into the lower normal range, ketone bodies rose. Of 404 people with pre-existing complaints, 341, meaning 84.4 percent, reported improvement. Absence of hunger was reported by 93.2 percent. Adverse effects occurred in fewer than 1 percent of participants.
What this means for you: Fasting from 4 to 21 days was safe and well tolerated in this setting. The setting is what matters: a specialised clinic, daily monitoring, an accompanying exercise programme. That is something different from a week alone at home.
Wilhelmi de Toledo F, Grundler F, Bergouignan A, Drinda S, Michalsen A. Safety, health improvement and well-being during a 4 to 21-day fasting period in an observational study including 1422 subjects. PLoS One. 2019;14(1):e0209353. DOI: 10.1371/journal.pone.0209353Situations in which fasting belongs in medical hands
- Being underweight or unintended weight loss in recent months
- Eating disorder in your history or currently, including in a milder form
- Pregnancy and breastfeeding
- Diabetes with glucose lowering medication, in particular insulin or sulfonylureas, because of the risk of hypoglycaemia
- Advanced kidney or liver disease
- Cardiac arrhythmia or known electrolyte problems
- Recovery after surgery, infection or injury
- Long term medication whose blood levels may shift during fasting
- Children and adolescents who are still growing
- High performance pressure at work or in sport without any window for recovery
And the question of whether intermittent fasting is superior to calorie restriction can by now be answered just as soberly.
Who: Wang and colleagues at Peking Union Medical College Hospital compared intermittent fasting with continuous calorie restriction in people with type 2 diabetes or metabolic syndrome in Diabetes Research and Clinical Practice in 2021. Of 84 screened studies, five met the criteria, and four with a total of 355 participants entered the meta-analysis.
What they observed: For HbA1c the difference was minus 0.06 with a 95 percent confidence interval of minus 0.27 to 0.16, meaning no relevant difference. For weight there was an advantage for intermittent fasting of minus 1.70 kilograms, confidence interval minus 3.28 to minus 0.11. Hypoglycaemia occurred at a similar frequency in both groups.
What this means for you: Intermittent fasting can be a safe and workable option. On this data it is not a magically superior principle. The best method remains the one you can keep up.
Wang X, Li Q, Liu Y, Jiang H, Chen W. Intermittent fasting versus continuous energy-restricted diet for patients with type 2 diabetes mellitus and metabolic syndrome for glycemic control: A systematic review and meta-analysis of randomized controlled trials. Diabetes Res Clin Pract. 2021;179:109003. DOI: 10.1016/j.diabres.2021.109003And now you know why, with fasting, I ask about history and medication first and only afterwards talk about protocols.
Three levers I consider worthwhile
If you have read this far, you probably want to know what to do with it. I am deliberately giving you three directions and no recipe. An individual approach belongs in a consultation, not in a blog article.
Lever 1: A real overnight eating break
Most people spread their eating across a window of 14 to 16 hours. The last bite often falls late. If you shorten that window by stopping earlier in the evening rather than starting later in the morning, you are also working with your daily rhythm. That was exactly the setup in the investigation by Jamshed and colleagues with the early eating window.
Start small. A break you keep five days a week can do more than a perfect protocol that collapses after ten days.
Lever 2: Giving the liver building blocks instead of irritants
Phase II reactions need material: sulfur containing amino acids, glycine, methyl groups from B vitamins. Cruciferous vegetables such as broccoli, Brussels sprouts and rocket supply sulfur compounds. Bitter substances can stimulate bile flow. Sufficient protein supplies the amino acids from which glutathione is built. Fibre binds in the gut.
What gives the liver work, by contrast, is well known and unspectacular: alcohol, a lot of fructose from drinks, constant small meals, lack of sleep.
Lever 3: Movement, the second route to AMPK
Fasting is not the only way to activate AMPK. Muscle work also lowers the energy charge in the cell. Both act on the same axis that Kim and colleagues described. A brisk walk at the end of an eating break is therefore not an add on but a second route to the same switch.
This is not about a clean blood panel as an end in itself. It is about you feeling able to act within your own body again.
That is exactly the difference between a programme and a rhythm. A programme makes you dependent on the next programme. A rhythm gives you back something that belongs to you.
And now you know why I would rather talk about rhythm than about cleansing.
Frequently asked questions about fasting, detoxification and autophagy
Does fasting really detoxify?
Not in the sense that detox advertising suggests. There is no evidence that fasting flushes any kind of slag out of your tissue.
What can be measured is something else: a lack of nutrients lowers mTOR activity and activates AMPK. That releases autophagy, the cell's own recycling of defective proteins and damaged mitochondria. It is a kind of inner cleanup, but it happens inside the cell, not in the gut or in connective tissue.
The actual processing of foreign substances continues to run through the liver and the kidneys, around the clock, including while you eat.
What is autophagy, explained simply?
Autophagy literally means self eating. The cell wraps broken components in a double membrane, transports the package to the lysosome and breaks it down there into its building blocks. Those building blocks are reused.
Yoshinori Ohsumi worked out the mechanism in yeast cells and received the 2016 Nobel Prize in Physiology or Medicine for it.
Imagine a workshop that receives no new parts and therefore takes old machines apart to build something new out of the screws. That is why Mizushima and Komatsu describe autophagy not as disposal but as a dynamic recycling system.
Does slag exist in the body?
As a medical term, slag does not exist. There is no lab value for it, no imaging that shows it and no accepted definition.
A critical review by Klein and Kiat found not a single randomised controlled trial testing the efficacy of commercial detox programmes in humans.
What does exist are metabolic end products such as urea, uric acid, creatinine and bilirubin, plus fat soluble foreign substances that can accumulate in adipose tissue. For all of that the body has its own well described pathways.
At what point does autophagy start during fasting?
The honest answer: for humans there is no reliable number of hours. The figures of 16 or 18 hours circulating online come mostly from animal models or from analogy.
Autophagy runs continuously at a baseline level anyway. Fasting appears to turn it up, but how strongly and from when is tissue dependent in humans and has not been measured cleanly. In a Copenhagen investigation, autophagy markers in muscle changed only moderately after 36 hours of fasting, and only in untrained participants.
Anyone who gives you an exact clock time is going beyond the data.
Has autophagy actually been demonstrated in humans?
Autophagy as a process, yes. The fasting response in humans, only in part.
The problem is measurability: autophagic flux could long be determined only in tissue samples or cell cultures. In a study of untrained and trained men fasting for 36 hours, individual markers in muscle changed only moderately. In a study of 50 women doing intermittent fasting, autophagy markers rose in mouse liver but not in human muscle.
New blood tests for autophagic flux are in development, among others at the South Australian Health and Medical Research Institute. The cell biology is solid, the human evidence is thinner than the hype.
What is the difference between autophagy and detoxification?
Autophagy is the recycling of the body's own material inside the cell: old proteins, broken mitochondria, membrane remnants.
Detoxification in the pharmacological sense means the conversion and excretion of substances foreign to the body. That happens mainly in the liver via phase I, meaning oxidation by cytochrome P450 enzymes, and phase II, meaning conjugation with glucuronic acid, sulfate or glutathione. Then comes phase III, transport into bile or urine.
Two different systems that are constantly confused in everyday language. Fasting mainly addresses the first. Detox advertising promises the second.
Do coffee, lemon water or collagen break autophagy?
For humans this has not been cleanly settled.
It is mechanistically plausible that amino acids activate mTOR and can therefore slow autophagy. A protein drink or collagen powder delivers exactly those amino acids. Black coffee without milk delivers practically none.
Robust human studies measuring autophagic flux before and after are only just starting. Anyone telling you that a squeeze of lemon destroys your autophagy is arguing from assumption, not from data. And fear at the table is rarely a good adviser.
How long do I have to fast for it to be worthwhile?
That depends on the goal.
For blood sugar and metabolism, effects showed up even with short eating windows: in a crossover study with eleven adults with overweight, an eating window from 8 am to 2 pm lowered mean 24 hour glucose by 4 ± 1 mg/dl and glucose excursions by 12 ± 3 mg/dl.
For longer fasting periods there are large observational data, for example from a cohort of 1422 people. For autophagy itself, no proven minimum duration exists in humans. More useful than counting hours is the question of which pattern you can sustain over months.
Who should not fast?
Fasting is not one of the harmless self experiments.
Being underweight, a current or past eating disorder, pregnancy and breastfeeding all argue against it. With diabetes on glucose lowering medication, in particular insulin or sulfonylureas, there is a risk of hypoglycaemia. Advanced kidney or liver disease, cardiac arrhythmia and recovery after surgery also call for medical supervision. Children and adolescents who are still growing should not fast.
If you take medication, its effect may change during fasting. Talk to your doctor beforehand. And bear in mind that rebuilding after longer fasting also needs supervision, keyword refeeding.
Is intermittent fasting better than a normal diet?
For blood sugar control, apparently not by much.
A meta-analysis of four randomised trials with 355 participants who had type 2 diabetes or metabolic syndrome found no relevant difference in HbA1c between intermittent fasting and continuous calorie restriction. For weight there was a small advantage for intermittent fasting of 1.70 kilograms.
Translated, that means intermittent fasting can be a workable option if time windows come easier to you than daily counting. On this data it is not a fundamentally superior principle.
What does ketosis have to do with autophagy?
Both are answers to the same situation, namely the absence of supply from outside. They are not the same thing, though.
Ketosis describes a switch in energy supply: the liver builds ketone bodies from fatty acids, which the brain can use. George Cahill described this in humans over decades. Autophagy, by contrast, describes the breakdown and reuse of defective cell components.
You can be in ketosis without a particular rate of autophagy following from it. The widespread equation of ketosis with autophagy is a simplification the data does not support.
Can I stimulate autophagy without fasting?
Mechanistically there is a case for it. The shared switch for autophagy is the enzyme ULK1, on which mTOR and AMPK act in opposite directions.
AMPK becomes active not only through a lack of food but also through muscle work, because the energy charge inside the cell drops. Sleep plays a role too, because many repair processes run at night.
In humans this has barely been quantified cleanly so far. So I would not say that exercise replaces fasting. I would say it plausibly acts at the same point, and it is easier for most people to fit into daily life.
More in the Detoxification Guide
- Detoxifying the liver: what counts and what is only marketing
- Phase 1 and phase 2 of liver detoxification, simply explained
- Glutathione: the master antioxidant and its role in detoxification
- Detox programmes: what juice cures and detox teas actually deliver
- Fasting as detox: autophagy instead of slag (you are here)
Related reading from other areas
Intermittent fasting 16:8
What happens in the body hour by hour
Intermittent fasting for women over 40
Why the cycle and the hormones have a say
Fasting and the cycle
When fasting can be unfavourable for women
Gut reset
The exit through which a great deal leaves the body
Leaky gut and zonulin
When the gut barrier becomes more permeable
Detoxifying the liver
The plant behind every real detoxification
Sources
- Takeshige K, Baba M, Tsuboi S, Noda T, Ohsumi Y. Autophagy in yeast demonstrated with proteinase-deficient mutants and conditions for its induction. J Cell Biol. 1992;119(2):301-311. DOI: 10.1083/jcb.119.2.301 [In vitro, yeast cells]
- Rubinsztein DC, Frake RA. Yoshinori Ohsumi's Nobel Prize for mechanisms of autophagy: from basic yeast biology to therapeutic potential. J R Coll Physicians Edinb. 2016;46(4):228-233. DOI: 10.4997/jrcpe.2016.403 [Review]
- Mizushima N, Komatsu M. Autophagy: renovation of cells and tissues. Cell. 2011;147(4):728-741. DOI: 10.1016/j.cell.2011.10.026 [Mechanism review]
- Kim J, Kundu M, Viollet B, Guan KL. AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1. Nat Cell Biol. 2011;13(2):132-141. DOI: 10.1038/ncb2152 [In vitro, human cell lines and mouse]
- Carosi JM, Fourrier C, Bensalem J, Sargeant TJ. The mTOR-lysosome axis at the centre of ageing. FEBS Open Bio. 2022;12(4):739-757. DOI: 10.1002/2211-5463.13347 [Mechanism review]
- Alirezaei M, Kemball CC, Flynn CT, Wood MR, Whitton JL, Kiosses WB. Short-term fasting induces profound neuronal autophagy. Autophagy. 2010;6(6):702-710. DOI: 10.4161/auto.6.6.12376 [In vivo, mouse]
- Dethlefsen MM, Bertholdt L, Gudiksen A, Stankiewicz T, Bangsbo J, van Hall G, Plomgaard P, Pilegaard H. Training state and skeletal muscle autophagy in response to 36 h of fasting. J Appl Physiol (1985). 2018;125(5):1609-1619. DOI: 10.1152/japplphysiol.01146.2017 [Human study, serial muscle biopsies]
- Chaudhary R, Liu B, Bensalem J, Sargeant TJ, Page AJ, Wittert GA, Hutchison AT, Heilbronn LK. Intermittent fasting activates markers of autophagy in mouse liver, but not muscle from mouse or humans. Nutrition. 2022;101:111662. DOI: 10.1016/j.nut.2022.111662 [RCT, n=50 women, plus mouse model]
- Jamshed H, Beyl RA, Della Manna DL, Yang ES, Ravussin E, Peterson CM. Early Time-Restricted Feeding Improves 24-Hour Glucose Levels and Affects Markers of the Circadian Clock, Aging, and Autophagy in Humans. Nutrients. 2019;11(6):1234. DOI: 10.3390/nu11061234 [RCT, crossover, n=11]
- Bensalem J, Heilbronn LK, Gore JR, Hutchison AT, Sargeant TJ, Fourrier C. The Break-Fast study protocol: a single arm pre-post study to measure the effect of a protein-rich breakfast on autophagic flux in fasting healthy individuals. BMC Nutr. 2022;8(1):120. DOI: 10.1186/s40795-022-00617-5 [Study protocol, human]
- Cahill GF. Fuel metabolism in starvation. Annu Rev Nutr. 2006;26:1-22. DOI: 10.1146/annurev.nutr.26.061505.111258 [Review, human]
- de Cabo R, Mattson MP. Effects of Intermittent Fasting on Health, Aging, and Disease. N Engl J Med. 2019;381(26):2541-2551. DOI: 10.1056/NEJMra1905136 [Review]
- Wilhelmi de Toledo F, Grundler F, Bergouignan A, Drinda S, Michalsen A. Safety, health improvement and well-being during a 4 to 21-day fasting period in an observational study including 1422 subjects. PLoS One. 2019;14(1):e0209353. DOI: 10.1371/journal.pone.0209353 [Observational study, n=1422]
- Klein AV, Kiat H. Detox diets for toxin elimination and weight management: a critical review of the evidence. J Hum Nutr Diet. 2015;28(6):675-686. DOI: 10.1111/jhn.12286 [Systematic review]
- Jancova P, Anzenbacher P, Anzenbacherova E. Phase II drug metabolizing enzymes. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2010;154(2):103-116. DOI: 10.5507/bp.2010.017 [Mechanism review]
- Malarvannan G, Van Hoorenbeeck K, Deguchtenaere A, Verhulst SL, Dirinck E, Van Gaal L, Jorens PG, Covaci A. Dynamics of persistent organic pollutants in obese adolescents during weight loss. Environ Int. 2018;110:80-87. DOI: 10.1016/j.envint.2017.10.009 [Prospective cohort, n=94]
- Wang X, Li Q, Liu Y, Jiang H, Chen W. Intermittent fasting versus continuous energy-restricted diet for patients with type 2 diabetes mellitus and metabolic syndrome for glycemic control: A systematic review and meta-analysis of randomized controlled trials. Diabetes Res Clin Pract. 2021;179:109003. DOI: 10.1016/j.diabres.2021.109003 [Meta-analysis, k=4, n=355]