Yes. Fasting lowers glucose levels in most people, typically by 5 to 15 mg/dL after a 12 to 16 hour overnight fast, with larger drops of 20 to 40 mg/dL during extended fasts of 24 to 72 hours. The size of the drop depends on your starting glucose, your glycogen reserves, and how insulin-sensitive your tissues are.
Glucose is not a single number that drifts passively. It is the output of a balance between glucose entering the blood and glucose leaving it. Fasting changes both sides of that balance, and the body defends a floor rather than driving glucose toward zero. That defense is why fasting can lower glucose substantially without causing hypoglycemia in healthy people, and why the same fast can be dangerous in someone taking insulin or a sulfonylurea.
This explainer walks through the mechanism, the numbers, the edge cases, the common errors, and the practical thresholds you can actually act on.
How fasting lowers blood glucose
After a meal, dietary carbohydrate is broken into glucose, and the pancreas releases insulin to move that glucose into muscle and fat and to signal the liver to stop making its own. Within roughly 4 to 8 hours, that incoming glucose is cleared. Fasting physiology then takes over.
The first lever is glycogenolysis. The liver stores glycogen, roughly 80 to 120 grams in a typical adult, and breaks it down to release glucose into the blood. Glycogenolysis supplies most of the body's glucose needs for the first 12 to 24 hours of a fast in a fed, glycogen-replete person. As glycogen falls, output falls with it, and blood glucose trends down.
The second lever is insulin. Fasting insulin falls within hours of the last meal. Lower insulin means less suppression of fat breakdown, so adipose tissue releases free fatty acids. Muscle shifts toward fatty-acid oxidation and reduces its own glucose use, sparing glucose for the brain. This shift is a major reason fasting glucose does not simply keep dropping.
The third lever is gluconeogenesis, the liver's manufacture of new glucose from lactate, glycerol, and amino acids. This process ramps up as glycogen depletes, typically after 12 to 24 hours. It sets a floor. In healthy adults, the liver and kidneys together produce enough glucose to keep fasting levels in a defended range even during multi-day fasts.
In plain terms: fasting lowers glucose because the liver's stored output runs down and insulin falls, but the liver then builds a replacement supply that puts a brake on the decline.
The time course, with numbers
Glucose after a fast follows a curve rather than a straight line. The table below describes a representative healthy adult with normal glucose regulation; individual values vary widely.
| Time since last meal | Dominant glucose source | Typical blood glucose (mg/dL) | What is happening |
|---|---|---|---|
| 0 to 2 hours | Dietary glucose | 110 to 140 peak, then falling | Post-meal rise, insulin rising |
| 4 to 8 hours | Dietary absorption tapering, liver beginning glycogenolysis | 85 to 110 | Insulin falling, glucose stabilizing |
| 8 to 12 hours | Liver glycogenolysis | 80 to 100 | Fasting baseline established |
| 12 to 18 hours | Glycogenolysis plus rising gluconeogenesis | 75 to 95 | Fat oxidation increasing |
| 18 to 24 hours | Glycogen partly depleted, gluconeogenesis leading | 65 to 90 | Ketones beginning to rise |
| 24 to 48 hours | Gluconeogenesis dominant | 60 to 85 | Ketones partly replace brain glucose |
| 48 to 72 hours and beyond | Gluconeogenesis, ketones | 55 to 80 | Defended floor, ketosis established |
The counterintuitive detail: for many people with insulin resistance, fasting glucose measured first thing in the morning can be higher than expected, because the early-morning rise in cortisol and growth hormone drives a surge of liver glucose. This is the dawn phenomenon. A single morning reading after an overnight fast may overstate the average. Continuous glucose monitoring usually shows a dip overnight followed by a rise in the pre-dawn hours.
Quantified context: how much does it drop?
The size of the effect depends on where you start. Someone with a fasting glucose of 95 mg/dL has little room to fall before the body's counter-regulatory hormones push back. Someone with a fasting glucose of 130 mg/dL and insulin resistance has far more room.
For people with type 2 diabetes, studies of intermittent fasting and time-restricted eating consistently report fasting glucose reductions in the range of 10 to 30 mg/dL over weeks, alongside HbA1c decreases of roughly 0.3 to 1.0 percentage points. That is a range, not a promise. Response varies with baseline control, medication, diet quality, and adherence.
Some context for the thresholds. The American Diabetes Association defines normal fasting glucose as below 100 mg/dL, prediabetes as 100 to 125 mg/dL, and diabetes as 126 mg/dL or higher on two occasions. A fasting protocol that moves someone from 118 to 102 mg/dL has moved them out of the prediabetes band on that measurement, but one reading does not reclassify anyone; repeat testing is required.
Edge cases and red flags
Fasting does not lower glucose uniformly, and in some situations it can raise it or create danger.
Type 1 diabetes, and type 2 diabetes treated with insulin or a sulfonylurea, carry a real risk of hypoglycemia during fasting. Do not fast on these medications without a clinician's plan. Pregnancy, a history of eating disorders, and being underweight make fasting risky for different reasons. Acute illness, surgery, or infection raise counter-regulatory hormones and can push glucose up, not down. Very low-calorie intake and prolonged fasts can produce the dawn phenomenon and elevated morning glucose even as overall control improves. Ketosis in type 1 diabetes can progress to diabetic ketoacidosis, a medical emergency.
The professional caveat: a lower fasting glucose number on a meter is not the same as better glucose control. Average glucose, post-meal excursions, and HbA1c tell a fuller story than a single fasting reading.
Common misconceptions
A frequent error is treating fasting as a way to drive glucose as low as possible. The body actively defends a floor, and chasing very low readings is not a goal.
Another is assuming that an elevated morning fasting glucose means fasting is not working. If average glucose and post-meal numbers improve while the fasting morning number stays high, the dawn phenomenon is often the explanation.
A third is that fasting and calorie restriction are the same thing. Fasting lowers insulin and depletes glycogen on a schedule; the same total calories eaten across many small meals produce a different hormonal pattern even if weight loss is similar.
A fourth is the assumption that fasting replaces medication management. It does not, and combining fasting with glucose-lowering drugs without adjustment is where serious harm occurs. The table below compares fasting with common alternatives.
| Approach | Typical effect on fasting glucose | Main mechanism | Key caveat |
|---|---|---|---|
| Overnight fast (12 to 16 h) | 5 to 15 mg/dL lower | Glycogen depletion, lower insulin | Modest effect, easy to sustain |
| Extended fast (24 to 72 h) | 20 to 40 mg/dL lower | Gluconeogenesis, ketosis | Hypoglycemia risk on some drugs |
| Time-restricted eating (8 to 10 h window) | 10 to 20 mg/dL lower over weeks | Reduced late eating, lower insulin load | Adherence and sleep timing matter |
| Calorie restriction alone | 10 to 25 mg/dL lower over weeks | Weight loss, improved insulin sensitivity | Slower than fasting, steady |
| Low-carbohydrate diet | 10 to 20 mg/dL lower over weeks | Reduced dietary glucose load | Requires sustained adherence |
| Metformin (for comparison) | 15 to 25 mg/dL lower | Reduced liver glucose output | Prescription, GI side effects |
Practical actions and thresholds
Use these concrete steps rather than guessing:
Measure fasting glucose after at least 8 hours without calories, at the same time each morning, and average at least 3 readings before drawing conclusions. If your fasting glucose is 100 to 125 mg/dL, a 12 to 14 hour overnight fast with a consistent eating window is a reasonable starting point, with rechecking after 4 to 6 weeks. If you take insulin or a sulfonylurea, ask your clinician for a medication adjustment plan before any fast longer than 12 hours. Stop a fast and treat with fast-acting carbohydrate if glucose falls below 70 mg/dL, or below 80 mg/dL if you have symptoms or take glucose-lowering medication. If fasting glucose is consistently above 126 mg/dL, get formal testing rather than self-managing with fasting alone. Track average glucose or HbA1c alongside fasting readings, because the two can move in different directions.
These thresholds are the ones to reconcile: 70 mg/dL is the standard hypoglycemia cutoff for most adults; 100 and 125 mg/dL define normal and prediabetes; 126 mg/dL is the diabetes threshold on repeat testing; and 180 mg/dL is a common upper target for post-meal glucose. A fasting protocol should be judged against the whole set, not one number.
What the evidence cannot settle
Fasting research has real limits. Many studies are small, short, and use different fasting definitions, from 12 hours to alternate-day protocols. Adherence is hard to measure, and weight loss often confounds the glucose effect. Long-term outcomes, meaning cardiovascular events and mortality rather than glucose numbers, remain understudied for most fasting patterns.
The mechanisms are stable and well characterized. The magnitude of benefit in any one person is not, and no fasting schedule is a substitute for diagnosis, medication review, or ongoing monitoring. Fasting lowers glucose levels through glycogen depletion, lower insulin, and a gluconeogenic floor, and that same physiology is why the effect is real, bounded, and occasionally hazardous.