Insulin, glucose, and meal timing

What glucose and insulin mean here

Glucose is the sugar in the blood. A paper usually reports a fasting value, taken after an overnight gap. Some papers also report the rise after a test meal or a glucose drink, and some report HbA1c, a slower average of blood glucose over about two to three months. A six-week eating window is a short lever on a number that moves over months.

Insulin is a hormone the pancreas releases so glucose can move out of the blood and into muscle and other tissues. It rises after you eat, especially after a meal that contains carbohydrate, and it falls as the hours without food go on. While insulin is high, the body stores carbohydrate and releases less fat from storage. Hours after a meal, the liver lets stored carbohydrate out, then makes more glucose and burns more fat, and ketones can rise later still. That is a typical sequence from supervised studies. The next meal still needs insulin, so a lower fasting number on a morning blood test is a marker from that morning, and the aim is not to drive that number toward zero.

Many papers turn one fasting glucose and one fasting insulin into an estimate called HOMA-IR, and they use it as a rough stand-in for insulin resistance. It is a calculation from two lab numbers. It is not a diagnosis the timer can give you, and it is not the same thing as type 2 diabetes.

Meal timing here means where the eating hours sit in the day. An eight-hour window from 7 a.m. to 3 p.m. and an eight-hour window from 1 p.m. to 9 p.m. are both a sixteen-hour fast. They are different experiments, because one finishes dinner at 3 p.m. and the other finishes it at 9 p.m.

The same meal, earlier and later in the day

A mixed meal raises glucose and insulin. How high they go, and how long they stay up, depends on the size of the meal, how much carbohydrate and protein it had, what you ate earlier, how you slept, and how readily your body responds to insulin. A large late dinner is a different input from a modest lunch.

In 2020, Gloria Leung, Maxine Bonham and colleagues pooled single-meal tests in which the same person ate the same meal once in the daytime and once at night. Daytime tests fell between 7 a.m. and 4 p.m. Nighttime tests fell between 8 p.m. and 4 a.m. Across fifteen studies, ten of which could be combined, the same meal produced a smaller glucose rise in the day than at night. The insulin difference pointed the same way but was smaller, and the studies that could not be combined did not all agree. Those were lab meals, not a fasting plan, and the nighttime tests ran late into the night. A 7 p.m. dinner with the household is a different thing from a meal at midnight.

What was typically studied

Glucose and insulin are often secondary measurements, sitting beside weight, cholesterol, and blood pressure. Some papers add a glucose-drink test, a few days with a continuous glucose monitor (a sensor worn on the skin), or HbA1c. Few of them map insulin across a full day.

The usual enrollment is adults with extra weight, for weeks to a few months. A smaller set already has prediabetes or type 2 diabetes. These trials are not a sample of children, pregnancy, or type 1 diabetes, and they are not a decade of follow-up.

What trials tend to find

In 2025, Zhila Semnani-Azad, Frank Hu and colleagues reviewed 99 randomised trials (6,582 adults) of intermittent fasting, ordinary calorie cutting, and unrestricted eating. Fasting glucose and the HOMA-IR estimate were a little lower with those diet strategies than with unrestricted eating, a difference they judged small, with moderate certainty. They found no difference among intermittent fasting, ordinary calorie cutting, and unrestricted eating for HbA1c. A fasting glucose number can tick down in a short trial while the two-to-three-month average barely moves.

Inside that same review, eleven trials enrolled people with type 2 diabetes. Time-restricted eating versus unrestricted eating came with about 1.9 kilograms less body weight, a small drop in fasting glucose, and a very small change in fasting insulin and in HOMA-IR. That is a group average from those trials. It is not a diabetes treatment, and it is not a reason to keep the same medicine dose while the meals move.

In 2026, Yu-En Chen, Hui-Li Tsai and colleagues pooled 41 randomised trials (2,287 people) of daily eating windows that lasted longer than a month. The search ran through the start of 2023, so the 2026 date is when the review was published. Compared with usual eating, fasting glucose was lower by about 3.7 milligrams per decilitre (about 0.2 millimoles per litre), and fasting insulin was lower by about half a microunit per millilitre, which is the unit many lab slips use for fasting insulin. HbA1c and HOMA-IR were not clearly different. People in those windows also lost weight, on the order of about 2 kilograms versus usual eating, so the glucose change and the scale change arrived together. Semnani-Azad’s review and Chen’s review are not the identical question: one compares several fasting styles with unrestricted eating and with calorie cutting, and the other compares daily windows with usual eating.

Earlier dinner, later dinner, and weight held steady

Chen’s review split the windows by the last meal. Early meant the last meal before 5 p.m. Late meant the last meal after 7 p.m. A window with the last meal between 5 p.m. and 7 p.m. sat in the middle. Compared with the late windows, the early windows had lower fasting insulin, by about 3.3 microunits per millilitre, and lower body weight, by about 1.2 kilograms. They rated both of those differences high certainty. The insulin gap and the weight gap came as a pair. How long the window lasted, under eight hours, about eight hours, or longer than eight hours, did not line up as better or worse in a consistent way. In that review, an 18:6 day was not a higher insulin dose than a 16:8 day.

A test that held weight steady is much smaller. In 2018, Elizabeth Sutton, Courtney Peterson and colleagues fed eight men with prediabetes every meal under supervision. Each man spent five weeks on a six-hour eating period with dinner before 3 p.m., and five weeks on a twelve-hour eating period, with enough food to hold weight steady. On the early short schedule, insulin sensitivity improved, the pancreas responded more readily, and blood pressure was lower. Eight men, supervised meals, and prediabetes is a real finding and a narrow one. It compared an early six-hour schedule with a twelve-hour schedule. It did not compare an early 16:8 kept at home with a late 16:8.

A direct week of early hours versus later hours looks quieter. In 2019, Amy Hutchison, Leonie Heilbronn and colleagues had fifteen men at risk of type 2 diabetes eat from 8 a.m. to 5 p.m. for one week and from noon to 9 p.m. for another week. The glucose rise after a test meal was smaller in both weeks than in their usual-eating week, and insulin did not change. A sensor on the skin showed a lower fasting glucose only during the morning-window week compared with usual eating, and the morning window and the afternoon window were not different from each other. The test meal was served at 8 a.m. in the early week and at noon in the later week, so the clock time of the test changed too. One week and fifteen men is not a rule that dinner has to move.

Limits (read these as part of the result)

A lower fasting insulin is a change on a morning blood test. On its own, that change does not show that a disease has gone away.

Weight still explains a lot of what these papers record. Eating less because the kitchen closed in the evening is a different story from the clock changing insulin while the calories stay the same. Sutton’s five-week study is the clearest example of a change with weight held steady, and it is eight men whose meals were provided. Chen’s early-versus-late insulin gap arrived with about a kilogram more weight loss on the early side.

The people in the larger reviews mostly had extra weight. Pregnancy, being underweight, type 1 diabetes, and a complicated set of medicines are a different question. The groups and medicines that need a clinician first are in Who should be careful with fasting.

One fasting blood draw is not a week on a skin sensor. HOMA-IR is an estimate from two fasting numbers. A finger-prick meter you were not asked to use is not the measurement those trials made.

An earlier dinner can cost sleep, the family meal, or a shift that does not finish at 5 p.m. A slightly lower lab average that wrecks your sleep is a worse plan for that week. Night work is a different day, and these trials are not an instruction to eat at noon after a night shift. Sleep and a loud week are the subject of Sleep, stress, and your fasting window.

People who use insulin, or a sulfonylurea (a pill that tells the pancreas to release more insulin), can see glucose fall too far when meals move or disappear. Shaking, sweating, confusion, or feeling as if you might pass out means eat, and it means stop the fast. When to stop a fast has the rest of the stop list. The dose of those medicines belongs with the clinician who prescribed them.

What to do with the hours

Use the clock for the hours you already chose. Whether that window sits earlier or later is a practical decision: sleep, the people you eat with, hunger, shift work, and a clinician if glucose is already a medical topic for you. Start near the overnight gap you already live with. Jumping to one meal a day to chase a lab number is a bigger change than these trials support. Placing those first hours is the job of Choosing a starting window.

If glucose is already something you and a clinician watch, keep using the meter they asked you to use. This timer counts elapsed hours. It does not measure your glucose or your insulin. The stage name follows those hours, and a change of stage name means only that those hours have passed. That is not a blood test.