Why Does Intermittent Fasting Disrupt Your Sleep?

When a fasting window extends past 12-14 hours, the liver runs low on stored glycogen. To keep the brain fueled overnight, the body turns to gluconeogenesis — manufacturing glucose from amino acids and lactate — a process accompanied by rises in cortisol and growth hormone. These hormones activate the brain’s arousal centers, and can produce the characteristic 2-4am waking with sudden alertness. The effect is strongest during the first weeks of fasting before metabolic adaptation occurs.

Intermittent fasting is widely adopted for metabolic health — and a common source of sleep complaints. Many people who start time-restricted eating report disrupted sleep within the first weeks, particularly waking in the second half of the night. This reflects a measurable metabolic cascade that begins when the fasting window exceeds the liver’s glycogen storage capacity. Sleep disruption itself drives insulin resistance, undermining the metabolic benefits fasting is intended to produce. Protecting sleep during fasting is not optional — it is a prerequisite for the strategy to work.

This article covers the blood sugar mechanism: how fasting window length determines overnight glycogen availability, why that triggers gluconeogenesis and concurrent cortisol release, and which fasting schedules disrupt versus improve sleep. For the broader circadian framework, see Circadian Rhythm Disruption. For the full metabolic overview, see Metabolic Sleep Disruption.

What Happens to Blood Sugar During an Extended Overnight Fast?

The liver stores enough glycogen to supply glucose to the brain for approximately 12-14 hours after the last meal. Once that glycogen is depleted, the liver switches from releasing stored glucose to manufacturing new glucose through gluconeogenesis — a transition accompanied by rises in cortisol and growth hormone, which can activate the brain’s arousal centers.

During the first 12 hours after eating, the liver releases glucose from its glycogen reserves to keep blood sugar stable. This is glycogenolysis — the breakdown of stored glycogen into glucose — and it requires no stress hormone involvement. The brain receives steady fuel. Sleep is undisturbed.

The transition arrives when glycogen runs low. Unni et al. (2025) demonstrated this in a crossover study of 14 participants with type 2 diabetes and 15 non-diabetic controls. Under glycogen-depleted conditions — when participants consumed non-glycogen-loading meals — the percentage contribution of gluconeogenesis to total overnight glucose production rose, with participants with type 2 diabetes showing disproportionate upregulation compared to controls. The liver relied more on manufacturing new glucose from amino acids and lactate than on releasing stored glucose — a process regulated by insulin, glucagon, and glycogen availability.

When participants consumed glycogen-loading meals before sleep, their nocturnal gluconeogenic contribution decreased. The fractional gluconeogenic rates in participants with type 2 diabetes more closely resembled those of non-diabetic controls. Glycogen status was a key variable. When glycogen stores were replenished, the gluconeogenic contribution decreased.

Overnight glucose production and liver glycogen after glycogen-loading vs non-glycogen-loading meals — Endogenous glucose production (EGP) (A), gluconeogenesis (GNG) (B), and glycogenolysis (GGL) (C) at 01:00, 04:00, and 07:00 hours, and liver glycogen (D) at fed and fasted state after nonglycogen loading (NGL) and glycogen loading (GL) meals in no diabetes (ND) and type 2 diabetes (T2D). ND NGL and ND GL, T2D NGL and T2D GL compared after NGL and GL meal. *P < .05 vs GL. P < .05 vs ND based on linear mixed model analysis. Unni, U. S., Bril, F., Mugler, J. P., 3rd, Carter, R. E., Basu, A., & Basu, R. (2025). Role of Hepatic Glycogen on Nocturnal Gluconeogenesis in Type 2 Diabetes Mellitus. *The Journal of Clinical Endocrinology and Metabolism*, 110(10), 2790-2799. https://pubmed.ncbi.nlm.nih.gov/39903644/

Shapiro et al. (1991) documented what happens to blood glucose during extended fasting through 24-hour continuous blood sampling at 15-minute intervals. In 11 participants with type 2 diabetes and 7 non-diabetic controls, glucose rose during overnight fasting despite no food intake — the morning glucose maximum was 23.8% above the evening nadir. Cortisol was elevated throughout the study period compared to non-diabetic controls (P < 0.001). Growth hormone elevations correlated with the morning glucose rise (r = 0.88; P < 0.01).

The takeaway for anyone practicing intermittent fasting: the fasting window length determines whether the liver depletes glycogen before or during sleep. A 12-hour fast ending at 8pm may leave glycogen intact through the night. A 16-hour fast with the last meal at 4pm pushes the depletion point into the early morning hours — right into the 2-4am window where gluconeogenesis can activate and arousal is likely.

How Does Fasting Raise Cortisol and Fragment Sleep?

Complete fasting activates the hypothalamic-pituitary-adrenal (HPA) axis, producing cortisol elevations large enough to fragment sleep. Moderate caloric restriction does not trigger the same response. The cortisol rise is pronounced during the first weeks of fasting and attenuates over time as the body adapts.

Fasting and caloric restriction differ in their cortisol effects. Nakamura et al. (2016) conducted a meta-analysis of 13 studies totaling 357 participants and found that complete fasting produced a strong cortisol elevation effect, while VLCDs and standard low calorie diets showed no measurable cortisol increases. Total caloric absence — not moderate restriction — is the trigger for HPA activation.

This has direct implications for sleep. Cortisol is a potent arousal hormone. It suppresses melatonin, delays sleep onset, and promotes waking through its action on the brain’s reticular activating circuits. When cortisol rises in response to glycogen depletion during an overnight fast, it does what it does in any other stress context: it wakes you up.

Nakamura et al. (2016) also identified a temporal pattern through meta-regression: cortisol elevation was pronounced at caloric restriction onset and attenuated over weeks of continued caloric restriction. This means the sleep disruption risk is front-loaded. The first two to four weeks of intermittent fasting carry the highest cortisol burden and the greatest sleep fragmentation risk. People who abandon fasting because of poor sleep during this window may be quitting during the worst phase.

Chawla et al. (2021) expanded this picture in a review of 14 studies — including Ramadan fasting and non-Ramadan time-restricted eating approaches — examining cortisol and, in a subset of studies, melatonin secretion patterns. In the Ramadan fasting studies, melatonin was reduced (P < 0.05), and the normal diurnal cortisol pattern was disrupted, with flattened morning cortisol and elevated evening levels.

The dual effect matters: cortisol rises while melatonin drops. These are opposing hormones in the sleep-wake context — cortisol promotes waking, melatonin promotes sleep. When fasting pushes both in the wrong direction at the same time, the hormonal environment actively opposes sleep consolidation.

Chawla et al. (2021) also identified that which meal is skipped determines the cortisol effect. Dinner-skipping reduced evening cortisol. Breakfast-skipping reduced morning cortisol. These are opposite effects with different implications for sleep. Skipping dinner may lower the evening cortisol that delays sleep onset, while skipping breakfast may blunt the morning cortisol that supports a healthy wake time — but neither pattern addresses the overnight glycogen depletion that accompanies the 2-4am cortisol surge.

The distinction from chronic caloric restriction bears emphasis: the mechanism here is fasting duration, not cumulative energy deficit. Nakamura et al. (2016) showed this distinction with quantitative precision — moderate restriction does not fire the same HPA response. A person eating 1,600 calories in a standard three-meal pattern and a person eating 1,600 calories in a six-hour window are in different hormonal states overnight, even at identical caloric intake.

Does Your Body Adapt and Does Sleep Improve Over Time?

The cortisol elevation from fasting attenuates over weeks, and multiple studies show that time-restricted eating can improve sleep after an adaptation period — particularly with early eating windows. A 2026 meta-analysis found early time-restricted eating improved sleep quality scores by nearly double compared to mixed-timing approaches.

Nakamura et al. (2016) provides direct evidence for adaptation: cortisol elevation showed a negative association with the duration of caloric restriction, meaning the HPA response diminishes over weeks of continued fasting. The body’s stress response to the fasting state recalibrates. This aligns with the common anecdotal report that sleep disruption from intermittent fasting peaks in weeks one through three and then gradually resolves.

Bohlman et al. (2024) examined this in a review of 6 randomized controlled trials lasting at least 8 weeks with daily fasting durations of 14 hours or more. The aggregate finding: short-to-mid-term time-restricted eating does not typically worsen overall sleep. But this aggregate conceals individual variation. Within the included trials, some studies documented reduced sleep efficiency and increased sleep onset latency. Whether these effects attenuate over longer time periods remains an open question.

Average likelihood of sleep at baseline and week 12, stratified by treatment arm — The average likelihood of sleep at baseline and at week 12 of the intervention, stratified by treatment arm. Y-axis indicates a likelihood of sleep, defined as the average percentage of the recorded time classified as sleep by Cole-Kripke algorithm smoothed over 30-minute bins. The rectangular bars on the x-axis indicate assigned eating windows during the intervention (UEP 08: 00 am-12:00 am; TRE 08:00 am-06:00 pm). Shaded green areas at the top indicate the main sleep period, defined as the period between median sleep onset and offset. Duan, D., Pham, L. V., Jun, J. C., Turkson-Ocran, R. A., Pilla, S. J., Clark, J. M., & Maruthur, N. M. (2025). Effects of time-restricted eating on actigraphy-derived sleep parameters: post hoc analysis of a randomized, isocaloric feeding study. *Sleep*, 48(9), zsaf089. https://pubmed.ncbi.nlm.nih.gov/40241264/

The current quantitative synthesis comes from Jin et al. (2026), a meta-analysis of 13 studies with 638 participants. Time-restricted eating increased sleep duration by 0.13 hours (P = .03) and improved Pittsburgh Sleep Quality Index (PSQI) scores by -0.47 points (P < .01), where lower scores indicate better sleep. Early time-restricted eating -- eating windows placed toward the morning -- produced the larger benefit: PSQI improved by -0.77 points (P < .01), nearly double the effect of mixed-timing approaches.

An important caveat from Jin et al. (2026): when the analysis was restricted to controlled trials only, excluding uncontrolled pre-post designs, no detectable differences in sleep duration or quality remained. Study design contributes to the observed effect sizes. The positive results should be interpreted with appropriate weight.

Duan et al. (2025) provides direct mechanistic evidence. In a randomized isocaloric feeding study — total calories held constant between groups — time-restricted eating increased total sleep time by 55 minutes. Within the TRE group, sleep onset moved approximately 30 minutes earlier, though between-group differences in sleep onset were not statistically significant (P = .73). Because calories were identical, this isolates meal timing as the active variable. The eating window itself, independent of energy restriction, changed when participants fell asleep and how long they slept.

The picture that emerges: the first 2-4 weeks of fasting carry the highest sleep disruption risk due to acute HPA activation. Those who persist through adaptation — and particularly those who place their eating window earlier in the day — may see net sleep improvements. Adaptation varies, and individual variability is high, but the trajectory bends toward improvement after the initial metabolic adjustment.

Which Fasting Schedules Disrupt Sleep and Which Ones Protect It?

Four variables determine whether a fasting schedule disrupts or improves sleep: fasting window length (longer than 14 hours depletes glycogen in many people), window placement (earlier eating windows produce better sleep outcomes), last meal composition (carbohydrate-containing meals replenish glycogen), and adaptation period (the first 2-4 weeks carry the highest risk).

Chawla et al. (2021) identified the placement of the eating window relative to the light-dark cycle as an underexplored but relevant variable in time-restricted eating research. Where the window falls in the day determines which hormones are affected and in which direction. An eating window of 8am-4pm has a different cortisol and melatonin impact than an eating window of 12pm-8pm, even at the same duration.

The data on early versus late eating windows is converging. Jin et al. (2026) found that early time-restricted eating produced the largest sleep quality improvements — a PSQI improvement of -0.77 compared to -0.47 for mixed-timing approaches. Duan et al. (2025) found that earlier eating windows increased total sleep time by 55 minutes, with within-group sleep onset moving approximately 30 minutes earlier. Eating earlier aligns the feeding window with the body’s natural cortisol rhythm, with insulin sensitivity (which peaks in the morning), and with the light-dark cycle.

Last meal composition is the other modifiable variable with direct evidence. Unni et al. (2025) demonstrated that glycogen-loading meals before the fasting window reduced nocturnal gluconeogenesis — a process that coincides with overnight cortisol and growth hormone release. A last meal containing adequate carbohydrate replenishes the glycogen stores that fuel the brain overnight. A last meal that is low in carbohydrate — common in people combining intermittent fasting with carbohydrate restriction — empties glycogen stores earlier and pushes the gluconeogenesis trigger point into the sleep window.

The variables that modulate sleep disruption risk from fasting, based on the available research:

Fasting window length. Windows longer than 14 hours deplete hepatic glycogen in many individuals, initiating the gluconeogenesis cascade during sleep. The 16:8 schedule — 16 hours fasting, 8 hours eating — puts many people past the glycogen threshold by 2-4am.

Window placement. Late eating windows (e.g., 12pm-8pm) place the end of digestion close to bedtime, while early eating windows (e.g., 8am-4pm) create a longer pre-sleep fasting gap but align with circadian insulin sensitivity. The evidence favors early windows for sleep quality, despite the longer pre-sleep fast.

Last meal composition. Carbohydrate-containing meals in the final hours of the eating window replenish liver glycogen. This extends the period before gluconeogenesis activates. A dinner low in carbohydrate shortens that buffer.

Adaptation period. Cortisol responses attenuate over weeks of sustained fasting. The first 2-4 weeks carry the highest disruption risk. Individuals who experience poor sleep during early adoption are experiencing the acute HPA response, not a permanent incompatibility with fasting.

These are the variables the research identifies as relevant to sleep outcomes during intermittent fasting. Individual responses vary based on metabolic status, insulin sensitivity, liver glycogen capacity, and baseline sleep architecture.

Many people waking at 3am have more than one factor at work. Blood sugar instability from fasting might be compounding a cortisol rhythm problem, a circadian misalignment, or an underlying metabolic issue. The interaction between these causes determines the waking pattern — and addressing only one may not resolve it.

Find out which causes might be driving your 3am wakeups

Frequently Asked Questions

Can a Calorie Deficit Cause Insomnia?

Yes. Fasting elevates cortisol, which delays the overnight cortisol nadir and increases nocturnal awakenings. But the mechanism differs depending on the source of the deficit. Fasting — total caloric absence for extended periods — produces a consistent HPA axis activation with strong cortisol elevations. Moderate, steady caloric restriction does not produce the same response.

Nakamura et al. (2016) quantified this distinction: complete fasting produced a strong cortisol elevation effect across 13 studies, while moderate caloric restriction (low calorie diets and standard low calorie diets) did not reach the threshold for measurable cortisol elevation. The type of deficit matters as much as the magnitude. A 500-calorie daily deficit spread across meals and a 500-calorie deficit achieved by eliminating an entire meal are metabolically different — the latter triggers the acute fasting response.

For a detailed breakdown of chronic caloric restriction and sleep, see Can a Calorie Deficit Cause Insomnia?.

Why Does Not Eating Enough Affect Sleep?

The body responds to insufficient fuel as an energy shortage. When energy intake drops below what the brain requires overnight, the HPA axis activates cortisol and growth hormone to drive gluconeogenesis — manufacturing glucose from stored protein and fat. These counter-regulatory hormones are the same ones that trigger waking. The brain does not distinguish between starvation and a fasting schedule.

The threshold for this response varies between individuals. Metabolic status, liver glycogen capacity, and insulin sensitivity all modulate when the stress response fires. Unni et al. (2025) showed that glycogen status is a key factor — when glycogen stores were replenished, the nocturnal gluconeogenic contribution decreased even in participants with type 2 diabetes. Shapiro et al. (1991) documented the downstream consequence: glucose rising 23.8% above the evening nadir during overnight fasting, driven by concurrent cortisol and growth hormone elevation.

This connection to sleep is direct. The hormones that rescue blood glucose are the same hormones that activate waking.

How Do You Address Sleep Problems From Calorie Restriction?

The research points to three variables rather than a single solution: eating window placement (earlier windows produce better sleep outcomes), last meal composition (carbohydrate-containing meals replenish the glycogen stores that fuel the brain overnight), and adaptation period (cortisol responses attenuate over weeks, so the worst sleep disruption is temporary).

Jin et al. (2026) found that early time-restricted eating produced sleep quality improvements nearly double those of mixed-timing approaches (PSQI -0.77 vs. -0.47). Unni et al. (2025) demonstrated that glycogen-loading meals reduce nocturnal gluconeogenesis. These are modifiable variables. Placing the eating window earlier and including adequate carbohydrate in the last meal address two of the three primary drivers of fasting-related sleep disruption.

The third variable — adaptation — requires patience. The cortisol response diminishes over weeks of sustained fasting. Persistent sleep disruption beyond 4-6 weeks may indicate the fasting schedule is not appropriate for your current metabolic status.

Can Intermittent Fasting Improve Sleep?

In some cases, yes — particularly with early eating windows and after the initial adaptation period. Jin et al. (2026) found that early time-restricted eating improved sleep quality scores by -0.77 PSQI points (P < .01). Duan et al. (2025) found that time-restricted eating increased total sleep time by 55 minutes in an isocaloric design, isolating meal timing -- not caloric restriction -- as the active variable.

After adaptation, the aggregate data leans positive. But the evidence has an important limitation: when Jin et al. (2026) restricted the analysis to controlled trials only, no detectable differences in sleep duration or quality remained. The headline numbers are driven partly by uncontrolled pre-post designs, which are more susceptible to placebo effects and regression to the mean.

Individual variability is high. Some people experience lasting sleep improvement from time-restricted eating, particularly when the eating window aligns with the first half of the day. Others experience persistent disruption, particularly with late eating windows, fasting durations exceeding 16 hours, or underlying insulin resistance that amplifies the nocturnal gluconeogenesis response. The research identifies the variables — window timing, window length, meal composition, adaptation status — but does not predict individual outcomes with certainty.

References

1. Shapiro, E. T., Polonsky, K. S., Copinschi, G., Bosson, D., Tillil, H., Blackman, J., Lewis, G., & Van Cauter, E. (1991). Nocturnal elevation of glucose levels during fasting in noninsulin-dependent diabetes. The Journal of Clinical Endocrinology and Metabolism, 72(2), 444-454. https://pubmed.ncbi.nlm.nih.gov/1991813/

2. Nakamura, Y., Walker, B. R., & Ikuta, T. (2016). Systematic review and meta-analysis reveals acutely elevated plasma cortisol following fasting but not less severe calorie restriction. Stress, 19(2), 151-157. https://pubmed.ncbi.nlm.nih.gov/26586092/

3. Chawla, S., Beretoulis, S., Deere, A., & Radenkovic, D. (2021). The Window Matters: A Systematic Review of Time Restricted Eating Strategies in Relation to Cortisol and Melatonin Secretion. Nutrients, 13(8), 2525. https://pubmed.ncbi.nlm.nih.gov/34444685/

4. Bohlman, C., McLaren, C., Ezzati, A., Vial, P., Ibrahim, D., & Anton, S. D. (2024). The effects of time-restricted eating on sleep in adults: a systematic review of randomized controlled trials. Frontiers in Nutrition, 11, 1419811. https://pubmed.ncbi.nlm.nih.gov/39144285/

5. Unni, U. S., Bril, F., Mugler, J. P., 3rd, Carter, R. E., Basu, A., & Basu, R. (2025). Role of Hepatic Glycogen on Nocturnal Gluconeogenesis in Type 2 Diabetes Mellitus. The Journal of Clinical Endocrinology and Metabolism, 110(10), 2790-2799. https://pubmed.ncbi.nlm.nih.gov/39903644/

6. Duan, D., Pham, L. V., Jun, J. C., Turkson-Ocran, R. A., Pilla, S. J., Clark, J. M., & Maruthur, N. M. (2025). Effects of time-restricted eating on actigraphy-derived sleep parameters: post hoc analysis of a randomized, isocaloric feeding study. Sleep, 48(9), zsaf089. https://pubmed.ncbi.nlm.nih.gov/40241264/

7. Jin, X., Li, T., Xu, X., & Rong, S. (2026). The Effect of Time-Restricted Eating on Sleep: A Systematic Review and Meta-analysis. Nutrition Reviews, 84(3), 514-526. https://pubmed.ncbi.nlm.nih.gov/40498475/

Written by Kat Fu, M.S., M.S. — Last reviewed: May 2026 — 7 references cited