You have been handed two stories about type 2 diabetes, and they cannot both be true. The one you get in the exam room is that it is chronic and progressive: you manage it, you do not beat it, and the medication list only grows. The one you get from every corner of the internet is that fasting can reverse it. The exam-room story is the one I want you to stop believing, because I have watched too many people walk it backward. The internet story is closer to the truth than your doctor will tell you, but it has a sharp edge that most of the hype leaves out, and the edge is not the fasting. It is your prescription.
That is the whole thing, so let me say it plainly before anything else. Whether a fast helps you or hurts you is decided almost entirely by what you are taking. Not by your willpower, not by your numbers, by your meds. This is not a yes-or-no article. It is a map of where you actually stand.
What Type 2 Diabetes Actually Is
To see why a fast can touch the root, you have to see what the root is, and it is not your blood sugar. In type 2 diabetes your cells have stopped listening to insulin, a state called insulin resistance. Your pancreas answers the only way it can, by making more, so you run on high insulin for years before your blood sugar ever looks wrong on a lab.
What is driving that is fat, but not the fat you can see in the mirror. It is fat piling up inside organs where it does not belong: a fatty liver that spills over into a fatty pancreas, and a fat-clogged pancreas cannot release insulin cleanly. This is Roy Taylor's twin-cycle model out of Newcastle (Taylor et al. 2018), and it explains something that trips people up: you do not have to be heavy to have this. Everyone has a personal fat threshold, and a lean person can cross theirs, carrying too much organ fat that never shows.
Can It Actually Reverse?
Often, yes, and this is real, established medicine, not wishful thinking. The honest word is remission, not cure, and it comes with conditions.
The cleanest proof is the DiRECT trial: people with type 2 diabetes ran a structured low-calorie program, and nearly half, 46 percent, were in remission at one year, off all their diabetes drugs (Lean et al. 2018). It tracked almost perfectly with fat lost: barely any remission for small losses, but 86 percent for those who dropped around 15 kilograms. When the fat left the pancreas, the insulin cells woke up. And the follow-up is honest rather than magic, with only about 13 percent still in remission at five years, because when the weight comes back, so does the diabetes (Lean et al. 2024).
I am going to be straighter with you than most articles bother to be. DiRECT was a low-calorie diet, not a fast, and certainly not a dry fast. There is no trial of dry fasting in diabetics, and for ethical reasons there never will be. What we do have is the mechanism, measured directly. In the controlled dry-fasting studies, insulin resistance does not creep down, it collapses: HOMA-IR fell from 4.8 into the normal range by day four, a 68 to 71 percent drop, with fasting insulin down about the same (Papagiannopoulos-Vatopaidinos et al. 2014). Those were healthy volunteers, not diabetics, so this is me reasoning from the engine, not from a diabetes trial. But the engine is the whole argument: create a real energy deficit, burn the organ fat, and the resistance lifts.
Why Fasting Hits the Root
Drop your food and insulin falls, and low insulin is the signal to start burning stored fat, including the fat jammed in the liver and pancreas. In DiRECT, liver fat fell by roughly four-fifths, from about 17 percent to about 3, and insulin sensitivity in the liver came back within days. That is the fire going out.
There is a mechanism here I care about more than the fat math, because it is why dry fasting in particular resets this. You spend those days pulling just enough glucose from gluconeogenesis, which means your glucose receptors get to work against a nearly empty tank. It is far easier for a cell to take sugar into an empty reservoir than a chronically flooded one, and the fast is also driving mitochondrial repair, more of the machinery that actually burns fuel. This is also why the popular advice to "just cut calories every day" so often fails: a permanent low-grade deficit keeps insulin twitching up and down and can deepen resistance over time, while a true fast empties the system and lets it recalibrate.
None of this means the drugs are useless. Metformin, the GLP-1s, the newer agents genuinely lower risk and protect hearts and kidneys, and I am not telling you to throw them away. I am telling you what most of them do not do, which is clear the organ fat or produce a drug-free remission. That gap is exactly what a fast fills.
The Real Safety Line Is Your Prescription, Not the Fast
Here is the part that actually decides your risk, and it has almost nothing to do with fasting and almost everything to do with what is in your cabinet. Read down until you find yourself.
- Diet-controlled or metformin only, the lower-risk group. Metformin does not whip insulin out of your pancreas, it mostly tells your liver to dump less sugar, so on its own it rarely causes a low (Nathan et al. 2009). This is the tier where fasting is most straightforward. The one real rule: metformin's own labeling says to hold it when you are dehydrated, and a dry fast dehydrates you on purpose, so you pause the metformin on the dry days, not the reverse.
- Insulin or a sulfonylurea, the supervised tier. These push glucose down whether or not you eat, which is the entire mechanism of a crash. Sulfonylureas are the pills whose names end in "-ide": glipizide, glimepiride, gliclazide, glyburide. This is not off-limits, but the dose has to come down as you fast or it will put you on the floor, and getting that adjustment right is a real medical job. Medical caveat: those dose changes are worked out with the prescriber who manages them, on their schedule, not on your fasting calendar.
- SGLT2 inhibitors, the "-flozin" drugs, the hidden one. Empagliflozin, dapagliflozin, canagliflozin. These can tip you into euglycemic DKA, and fasting plus a low-carb state plus dehydration are the exact triggers the FDA named (FDA 2015). A dry fast is all three at once, and a normal number on your meter is not reassurance here. These come off before a fast, full stop.
- Type 1 or insulin-dependent, the hard stop. This one is not a "be careful," it is a no. Your body makes essentially no insulin of its own, and going without can tip you into ketoacidosis even after years of stability, because it is the absence of insulin, not the fasting, that lets ketones run away (IDF-DAR 2021). Do not dry fast. If you want to explore any fasting at all, that is a conversation for the endocrinologist who runs your insulin, because it is the one situation where I will not freelance with you.
You Cannot Drink to Fix a Low, So Understand Where Lows Come From
This is the part that is specific to dry fasting, and it is the reason I steer almost anyone with diabetes to a water fast first. But most articles, including my own earlier draft of this one, get it half right, so let me go deeper, because the mechanism is the reassurance.
Start with a fact that surprises people: a fast does not crash your blood sugar. Your body defends that number ferociously. In the controlled dry-fasting work, glucose in healthy fasters did not fall off a cliff, it traced a gentle curve, dipping from about 86 to a low around 56 milligrams per deciliter near day three, then climbing back to 78 by day five with no food at all (Papagiannopoulos-Vatopaidinos et al. 2014). Read that again: it went back up while they were still fasting. Of fifty glucose readings across those days only a handful dipped below 60, and the low ones caused no real symptoms, because the body had already switched fuels. This is Cahill's old defense-of-glucose framework in action: glycogen first, then gluconeogenesis, then by around day three ketones take over, dropping the brain's glucose demand from roughly 120 grams a day toward 40 (Cahill 1970). Glucagon, growth hormone, cortisol, adrenaline, a whole fifteen-hormone counter-regulatory system, all lean on the number to hold it up.
So if the fast will not crash you, what will? A drug that forces glucose down with no food coming in. That is the only thing on this whole page that reliably causes a dangerous low, and it is exactly the insulin and sulfonylureas from the tier above. Now put that together with the "dry" part: on a dry fast, the standard rescue is gone. The textbook fix for a low is fifteen grams of fast sugar, juice, regular soda, glucose tablets, wait, recheck. On a dry fast there is no juice and no soda, and a glucose tablet chewed into a dry mouth is slow and unreliable at the exact moment speed matters. So the danger is not that fasting drops you and you cannot catch it. The danger is that a drug drops you and you have cut your own brake line. The fix is not to white-knuckle a low with no rescue. The fix is to not be on the drug that causes it before you ever dry fast, which is the whole reason the tiers matter.
The Metabolism Underneath, and Why Some People Stay Stuck
There is a layer under all of this, and it is the reason some people do everything right and still cannot move their blood sugar or their weight. It sits in the thyroid, specifically in the active hormone T3, and the connection is not loose.
The same cellular gridlock that blocks insulin blocks T3. Insulin resistance starts at the tissue level, inside the cell, long before it shows up in fasting glucose or HbA1c, and once the metabolic stress is bad enough the identical resistance that keeps insulin from working keeps T3 from getting into the cell. Your labs can read normal while the tissue runs cold. That matters here for a concrete reason: T3 does something insulin resistance cannot easily undo. It switches on the GLUT-4 glucose transporter independently of insulin (Guyenet and Schwartz 2012, and mechanistic work on thyroid and glucose uptake), opening a side door for sugar to enter the cell that bypasses the jammed insulin signaling entirely. Ketones do the same thing through their own transporters. It is why a fast, which raises ketones, and restored T3 can move blood sugar when more insulin simply cannot.
The loop runs the other way too. Cut carbs and glucose too hard for too long and you impair the liver's conversion of T4 into active T3, so blood sugar handling gets worse, which is one more reason the refeed and the metabolic layers matter as much as the fast (Ludwig 2021). If you are the person doing everything right and still stuck, this is worth ruling out before you conclude you just need to fast harder. You very likely do not. We go deep on the hidden, labs-look-normal version of this in cellular hypothyroidism and T3 therapy, and fasting and blood sugar may honestly be only half of your equation.
Breaking It Without a Spike
How you come out matters as much as the fast when blood sugar is the point. Come back on a big high-glycemic meal and you get a sharp sugar-and-insulin spike, the opposite of what you were working toward. Lead with protein, fat, and fiber, real food before fast carbs, and you blunt that spike and steady the next meal too (Jenkins et al. 1981).
Two things worth knowing. Right after a fast your insulin sensitivity is heightened, which is wonderful, but it also means that if you are still on any glucose-lowering drug, the same dose can drop you lower than expected as you eat, so that is one more reason the medication side gets handled deliberately. And the scary "refeeding syndrome" you may have read about is real but is a risk of prolonged starvation in already-malnourished people, not of a sensible short fast in a well-nourished one (Mehanna et al. 2008); if you are breaking a longer fast, the practical move is to check phosphate, potassium, and magnesium when you resume carbs. The members portal has a refeeding tool that scales the reintroduction to how long you fasted.
Where to Take This
If you are a curious self-experimenter, the mechanism under all of this is worth understanding on its own, so start with what a fast is actually doing under the hood in dry fasting, autophagy, and regeneration, and the counter-regulation that holds your sugar steady overnight is the same thing behind why you wake at 3am with your heart pounding.
But if type 2 diabetes is your actual life, not a curiosity, and especially if it comes tangled up with the fatigue, thyroid trouble, or Long Covid and ME/CFS that so often travel with metabolic disease, this stops being a general question. That is what the Scorch Protocol was built to map: the root causes, in the right order. And if you want your own numbers and your own medication list looked at against your real situation, that is what the members portal is for. Reversing this is real. The only thing standing between you and a safe run at it is sorting the two or three drugs that actually matter, and that is a solvable problem.
(This article is educational and is not medical advice. It is not a reason to start, stop, or change a diabetes medication on your own. If you take insulin or a sulfonylurea or use an SGLT2 "-flozin" drug, get the dose managed before you fast. If you have type 1 diabetes, do not dry fast.)