Eat Earlier, Sleep Better: Evening Strategies for Controlling Morning Blood Glucose

A team from Columbia University and the Salk Institute studied what would give your morning glucose a bigger boost if you have prediabetes or early type 2 diabetes: simply the duration of your overnight fast “by the clock” or the actual overnight fluctuations in sugar after your last meal. The authors introduced two concepts: chronological overnight fasting (COF), which is from the start of dinner until you wake up, and biological overnight fasting (BOF), which is from the moment your glucose has returned to its evening fasting level until you wake up.

• The conclusion is simple and practical: how sugar behaves at night is more important than “how many hours have passed since dinner.” There is a connection with morning glucose, but it is “muffled” by the composition of dinner and individual sensitivity to insulin.

Background of the study

The ideas of intermittent fasting and “extended overnight fasting” have entered the mainstream: it is often advised to simply count the hours between dinner and breakfast. But metabolically, “12 hours without food” is not the same thing for different people. Morning glycemia is influenced by both how sugar behaved after the last meal and by the nighttime regulation of glucose during sleep. At the same time, nighttime glycemia as a determinant of morning sugar has been less studied than daytime postprandial peaks.

From a chrononutritionist perspective, glucose tolerance and insulin sensitivity worsen in the evening and at night: insulin secretion and action decrease, and the contribution of hepatic glycogenolysis/gluconeogenesis increases. This explains why the same dinner in the evening produces a “heavier” curve than in the morning, and why the response to the last evening meal (LEO-PPGR) can “flow over” into the morning fasting values. However, clinical studies have traditionally looked at the composition of the dinner (e.g., the proportion of carbohydrates), rather than the actual nighttime glycemia as a predictor of morning.

The advent of affordable CGM sensors has opened up the possibility of separating the “clock” from the “biology.” Two different overnight fasting windows have been introduced: the chronological (COF) — from the start of dinner until waking up, and the biological (BOF) — from the moment glucose has returned to the evening fasting level until waking up. This operationalization allows us to separate the “tail” of the postprandial curve from the true overnight fasting and test which contributes more to morning sugar in people with prediabetes and early T2DM.

Hence the research gap: how do nighttime glucose levels and the response to the last dinner compare in strength of association with morning glycemia - and whether these associations persist after accounting for the carbohydrate load of the dinner and individual insulin sensitivity (e.g., the Matsuda index). The current work in Nutrients tests this hypothesis in a controlled 24-hour protocol with CGM and a standardized diet.

How the experiment was conducted

The study involved 33 people aged 50-75 with prediabetes or early T2DM (some on metformin, without insulin). They were given a standardized daily diet with fixed meals, and the last meal (LEO) was at 22:00. They were fitted with a "blind" continuous glucose monitoring (CGM) sensor, their sleep and meal times were tracked, and the following morning they were given an OGTT to calculate the Matsuda index (insulin sensitivity).

• COF: from the beginning of dinner until awakening.
• BOF: "clean" overnight fast only - after sugar has returned to evening fasting levels and until waking up.
• Main parameters: postprandial response to dinner (LEO-PPGR), average overnight glucose (COF/BOF) and morning fasting glucose.

What did they find?

There were many connections, but the key one was that nighttime sugar levels and the reaction to the last meal were “carried over” into the morning.

• LEO-PPGR ↔ morning sugar: the higher the average sugar/peak/area under the curve 3 hours after dinner, the higher the glucose in the morning (r≈0.53-0.71; p ≤0.001).
• Night sugar ↔ morning sugar: average glucose for COF and BOF is closely related to morning (r=0.878; p<0.001). But after taking into account dinner carbohydrates, this relationship weakens.
• The role of insulin sensitivity: adding the Matsuda index “removes” previous associations - individual insulin sensitivity explains a significant part of the morning values.

The details that matter

The authors specifically compared the “number of hours” and the “biological reality” of the night.

• COF averaged ~7 h 16 min, BOF ~4 h 48 min; in both cases, mean nighttime glucose was related to morning glucose. However, after adjusting for dinner carbohydrates and especially for Matsuda, statistical significance disappeared.
• Surprisingly, grams of dinner carbohydrate alone did not predict morning blood sugar; what mattered was the actual glycemic response to dinner (LEO-PPGR).
• From the nutrient details, it emerged that dinner fiber correlated with morning glucose (r≈0.51), but this effect also disappeared in the multivariate model.

Why is that?

Night is not just a "pause between dinner and breakfast." While you sleep, your body balances between glycogenolysis and gluconeogenesis, which is influenced by circadian rhythms, the dawn phenomenon, and individual insulin sensitivity. So "12 hours without food" is a different metabolic night for two people.

• The authors remind us that we tolerate carbohydrates worse in the evening; and for “owls” and “larks,” evening food produces different glycemic patterns.

What does this mean in practice?

If you have prediabetes/early T2D, “just extending your fasting window” isn’t always the answer. It’s more helpful to target your actual overnight blood glucose and insulin sensitivity.

• Don't look at the clock, look at your sugar: a CGM track at night (or at least glucose in the morning) will give more information than "how long the fast lasted."
• Optimize dinner: Reduce glycemic load with combination - slow carbs + protein/fat, portion size, earlier time. What matters is the actual response (PPGR), not just grams.
• Work with insulin sensitivity: Post-dinner movement, strength/aerobic training during the day, sleep and weight all change morning glucose more than the dry hours of fasting.

Things to keep in mind (limitations)

This is not a large outcome intervention, but a 24-hour controlled protocol in a subsample of 33 subjects (19 for COF/BOF), most women, some on metformin. The results are pilot but provide good direction for personalized strategies (dinner timing, composition, training, sleep).

In brief - how COF and BOF differ

• COF: from the beginning of dinner to awakening - reflects both the "tail" of the postprandial curve and the pure overnight fast.
• BOF: from the return of glucose to evening fasting until awakening - "clean" nocturnal regulation without the influence of the acute response to dinner.

Source: Diaz-Rizzolo DA et al. Biological vs. Chronological Overnight Fasting: Influence of Last Evening Meal on Morning Glucose in Dysglycemia. Nutrients. 2025;17(12):2026. https://doi.org/10.3390/nu17122026