How Long Should You Fast to Lower Blood Sugar?

Fasting for roughly 12 to 16 hours on a regular basis appears to lower fasting blood sugar modestly in most people, with randomized trials in people with type 2 diabetes showing fasting glucose drops of about 28 to 31 mg/dL over several months of daily time-restricted eating. But the picture is more nuanced than picking a single number of hours: what your body does with blood sugar during a fast changes dramatically depending on how long you go without food, what time of day you eat, what medications you take, and whether you have diabetes. Shorter daily fasts and longer multi-day fasts involve entirely different metabolic machinery, and the blood sugar response is not always a straight downward line.

What Happens to Blood Sugar During a Fast

Your liver is the main organ keeping blood sugar stable when you stop eating. It does this through two processes: breaking down stored glycogen and manufacturing new glucose from scratch. After an overnight fast of about 10 to 12 hours, roughly two-thirds of the glucose entering your bloodstream comes from glycogen breakdown, while about a third comes from new glucose production.

As the fast stretches beyond 12 hours, those proportions shift. Glycogen stores start running low, and new glucose production picks up. By about 60 hours without food, new glucose production accounts for roughly 78% of the glucose your liver releases, and the contribution from glycogen is almost negligible.

¹ The liver does not just run out of fuel and let blood sugar crash; it ramps up alternative production pathways, even breaking down amino acids and recycling them into glucose to keep the brain supplied.²

This is why blood sugar does not simply plummet the longer you fast. In healthy people, it drops gently over the first several hours and then more or less plateaus. A large population study found that each additional hour of fasting lowered blood glucose by only about 0.4 mg/dL in men and 0.2 mg/dL in women, and that most of the measurable decline happened within the first three hours after eating. After about eight hours, blood sugar levels barely budged further.³ Your body is designed to keep glucose within a tight range, and it has powerful hormonal tools to do so.

Daily Time-Restricted Eating and Blood Sugar

The fasting approach with the most clinical data for blood sugar is time-restricted eating, where you compress all your meals into a window of 8 to 10 hours and fast for the remaining 14 to 16 hours each day. This is the pattern most people mean when they talk about intermittent fasting.

In a randomized trial of people with obesity and type 2 diabetes, both a 16:8 protocol (eating within 8 hours, fasting 16) and a 14:10 protocol (eating within 10 hours, fasting 14) produced significant drops in fasting blood sugar compared to a control group receiving standard dietary advice. The 16:8 group saw fasting glucose fall by about 31 mg/dL on average, while the 14:10 group dropped about 28 mg/dL. Both groups also lowered their HbA1c, a measure of average blood sugar over the previous two to three months, by about half a percentage point. There was no meaningful difference between the two fasting durations.⁴

A separate trial comparing alternate-day fasting and 16:8 time-restricted eating found a similar pattern: both approaches lowered fasting blood sugar compared to controls, and neither was clearly superior to the other for glucose control.⁵ The practical takeaway from these trials is that you probably do not need to push past 14 hours of daily fasting to get meaningful blood sugar benefits. The jump from 14 to 16 hours did not consistently produce better glucose numbers.

The 5:2 Approach

Another popular pattern is the 5:2 diet, where you eat normally five days a week and restrict calories sharply (typically to about 500 to 600 calories) on two non-consecutive days. This is less about continuous fasting hours and more about periodic severe calorie restriction.

A striking randomized trial in people with newly diagnosed type 2 diabetes compared a 5:2 meal-replacement plan against starting metformin or empagliflozin, two common diabetes drugs. Over 16 weeks, the 5:2 group reduced HbA1c by about 1.9 percentage points, which was meaningfully larger than the drop seen with metformin (1.6 points) or empagliflozin (1.5 points).⁶ That is a noteworthy result, though it involved structured meal replacements on fasting days rather than simply skipping meals. And when researchers compared 5:2 fasting against standard daily calorie restriction for liver fat and metabolic markers, the two approaches produced equivalent improvements in blood sugar measures.⁷ The mechanism likely has as much to do with overall calorie reduction and weight loss as with fasting itself.

Does the Time of Day You Eat Matter?

If you are going to eat within a restricted window, it seems to matter whether that window falls in the morning or the evening, at least for insulin sensitivity. Your body handles glucose better earlier in the day, partly because insulin sensitivity follows a circadian rhythm that peaks in the morning and declines at night.

A meta-analysis comparing early time-restricted eating (finishing meals by mid-afternoon) against later eating windows found that early eating produced a modest but statistically significant drop in fasting glucose of about 2.2 mg/dL compared to unrestricted eating, while later eating windows did not significantly lower fasting glucose at all. The bigger difference showed up in insulin resistance: early eaters improved their insulin resistance scores roughly twice as much as late eaters.⁸

However, the evidence is not unanimous. A more recent systematic review of 16:8 time-restricted eating studies found no clear differences in glucose outcomes between early, late, and self-selected eating windows, suggesting that simply restricting the eating window to eight hours may matter more than precisely when it falls.⁹ If you have flexibility, eating earlier and finishing dinner early is probably the better bet for blood sugar, but the difference is small enough that adherence matters more than timing perfection.

Why Extended Fasting Can Backfire for Blood Sugar

A common misconception is that fasting longer always means lower blood sugar. Past about 24 to 36 hours, the metabolic landscape changes in ways that can push blood sugar in unexpected directions.

When healthy people fasted for 36 hours and then took a glucose tolerance test, their two-hour blood sugar readings were significantly higher than after a normal overnight fast: about 110 mg/dL versus 89 mg/dL. Their insulin secretion was also markedly lower in both its early and late phases.¹⁰ In other words, after a day and a half without food, these people temporarily handled an incoming sugar load worse than they would have after an ordinary overnight fast. The pancreas had dialed down insulin production because it was not needed during the fast, and it took time to ramp back up.

Around 18 to 36 hours of fasting, the body also begins producing significant amounts of ketone bodies as it shifts toward burning fat for fuel. Ketone levels jump by roughly threefold between 18 and 36 hours of fasting and continue rising beyond that.¹¹ This shift is normal in healthy people, but it introduces a risk for people with diabetes, especially type 1 diabetes, where unchecked ketone production can become dangerous.

The dawn phenomenon adds another layer of complexity. Even during an overnight fast, the body increases insulin production in the early morning hours to counter a natural rise in insulin resistance. In healthy individuals, insulin levels climb by roughly a third between 3 a.m. and 8 a.m. even though blood sugar stays flat.¹² People with diabetes often cannot mount this compensatory insulin surge, which is why morning fasting readings are frequently their highest of the day, even after not eating all night. Extending the fast further into the morning does not necessarily fix this.

What Continuous Glucose Monitoring Reveals

Continuous glucose monitors give a more detailed picture of what happens during real-world fasting than spot checks. Studies using these monitors during Ramadan, when observant Muslims fast from dawn to sunset (typically 12 to 16 hours), have revealed a characteristic pattern in people with diabetes: blood sugar slowly drifts downward during fasting hours, then spikes sharply after the sunset meal. The magnitude of that post-meal spike was largest in people taking insulin and smallest in those on diet control alone.¹³

The key insight is that fasting-induced blood sugar control can be partially undone by how and what you eat when the fast ends. A large, carbohydrate-heavy meal after a prolonged fast tends to produce glucose excursions that can be worse than what you would see after a smaller meal on a normal eating schedule. The overall average glucose did not change much during Ramadan fasting in the study above, even though the shape of the glucose curve changed dramatically. This suggests the lowering effect of the fast was offset by the post-fast spike.

Safety for People on Diabetes Medications

Fasting is not a universally safe blood sugar strategy, and the people most motivated to try it are often the ones at highest risk. If you take insulin or sulfonylureas, fasting meaningfully increases your risk of hypoglycemia. A randomized trial of intermittent fasting in people with type 2 diabetes found that fasting roughly doubled the rate of hypoglycemic episodes compared to a non-fasting control group, even though researchers had reduced participants’ medication doses ahead of time.¹⁴ Whether the fasting days were consecutive or non-consecutive did not change this risk.

Insulin, sulfonylureas, and food insecurity are all recognized risk factors for hypoglycemia in type 2 diabetes, and fasting sits alongside them.¹⁵ This does not mean people on these medications cannot fast, but it means they should not start an intermittent fasting protocol without working with their prescriber to adjust doses. People on metformin alone, or those managing blood sugar through diet, face substantially less risk.

Muscle Mass and Blood Sugar Disposal

Skeletal muscle is the largest consumer of glucose in your body, absorbing the bulk of blood sugar after a meal. People with more muscle mass tend to have lower fasting blood sugar, and research has found an inverse relationship between skeletal muscle mass and blood glucose levels.¹⁶ Low muscle mass reduces the body’s capacity to dispose of glucose, and population studies have linked declining muscle mass over time to increased diabetes risk.¹⁷

This matters for fasting because extended or very frequent fasting can lead to muscle loss if protein intake is inadequate or if resistance training is absent. The blood sugar benefit you gain from fasting can be partially offset by losing the muscle that helps clear glucose from the bloodstream. For someone whose primary goal is blood sugar management, combining intermittent fasting with resistance exercise and adequate protein is likely more effective than fasting alone.

Exercising in a Fasted State

Working out during a fast changes how your body handles fuel. In a training study comparing six weeks of endurance exercise done either fasted or after a carbohydrate meal, the fasted exercise group developed greater fat-burning capacity and showed higher activity of key fat-oxidation enzymes. The fasted group also avoided the drop in blood sugar that typically occurs during prolonged exercise.¹⁸ The body essentially becomes better at using fat for fuel when it trains without incoming carbohydrates, which helps stabilize blood sugar during and after exercise.

That said, this was studied in healthy young men, not in people with diabetes on medication. For someone taking insulin or sulfonylureas, exercising in a fasted state amplifies hypoglycemia risk considerably. Even for people not on medication, fasted high-intensity exercise can sometimes trigger a stress-hormone response that temporarily raises blood sugar. Moderate-intensity activity like brisk walking or cycling in a fasted state is generally well tolerated and appears to produce the best metabolic adaptations.

Sex Differences in the Fasting Response

Men and women do not respond identically to fasting. After about 38 hours without food, women tend to have lower blood sugar than men, along with higher levels of free fatty acids in their blood. Despite those higher fatty acid levels, women appear to be relatively protected from the insulin resistance that fatty acids normally cause, possibly because their muscles are better at preventing the accumulation of certain fat-derived molecules that interfere with insulin signaling.¹⁹

During shorter fasts, however, the differences in glucose metabolism between men and women are small. One study that carefully measured glucose and fat metabolism during both basal conditions and short-term fasting found clear sex differences in fat metabolism but not in glucose handling.²⁰ The practical implication is that for the daily 14-to-16-hour fasting windows most people use, men and women can expect broadly similar blood sugar effects. The divergence becomes more pronounced only with extended fasts of a day or longer.

What Fasting Does to Insulin-Producing Cells

Beyond the immediate glucose-lowering effect, intermittent fasting may help preserve the pancreatic beta cells that produce insulin. In animal models of obesity-induced diabetes, intermittent fasting restored a cellular cleanup process called autophagy in the insulin-producing islands of the pancreas. This improved the cells’ ability to secrete insulin in response to glucose, enhanced beta cell survival, and stimulated markers associated with pancreatic regeneration.²¹ The animals continued eating a high-fat diet on non-fasting days, yet their glucose tolerance still improved.

This research is in mice, not humans, and translating animal findings to clinical practice is always uncertain. But the idea is compelling: type 2 diabetes progressively destroys beta cells over time, and if fasting can slow that destruction or even stimulate some regeneration, that would represent a benefit beyond simple calorie reduction. Whether this mechanism operates in human beta cells at the fasting durations people actually practice remains an open question that clinical researchers are actively exploring.

When HbA1c Does Not Tell the Whole Story

If you are using HbA1c to track the effects of fasting on your blood sugar, be aware that the test has blind spots. HbA1c reflects average blood sugar over the previous two to three months by measuring how much glucose has attached to hemoglobin in red blood cells. But if your red blood cells live shorter or longer than average, the number can be misleading. People with glucose-6-phosphate dehydrogenase deficiency or thalassemia, for example, have shorter-lived red blood cells, and their HbA1c readings systematically underestimate their true average glucose.²² Iron-deficiency anemia, recent blood loss, and pregnancy can also skew results.

For people tracking fasting’s effects closely, a combination of fasting glucose, post-meal glucose checks, and HbA1c gives a fuller picture than any single measure. Continuous glucose monitoring, where available, adds even more detail by showing the swings that averages hide. A person might have an HbA1c that looks well-controlled while still experiencing large glucose spikes after breaking a fast, or they might have an apparently high HbA1c that reflects a red-blood-cell issue rather than poor glucose control.