time-restricted-eating

Time-restricted eating (TRE)

TRE confines all food intake to a narrow daily window (typically 6–12 hours) without an explicit caloric reduction target. The conceptual distinction from caloric-restriction and intermittent-fasting is that TRE is designed to be isocaloric — the therapeutic mechanism is circadian realignment of metabolism, not caloric deficit. In practice, shortened windows often produce mild spontaneous weight loss (~1–3 kg) by reducing ad libitum intake, but the metabolic benefits in some studies appear independent of weight change ¹.

TRE is currently one of the most consumer-popular dietary interventions; multiple RCTs across cardiometabolic endpoints exist, but long-term hard-endpoint (mortality, cardiovascular events, all-cause disease) data are absent.

Definitional clarifications

Protocol	Eating window	Caloric intent	Primary mechanism
TRE / TRF	6–12 h	Isocaloric (intended)	Circadian realignment
intermittent-fasting (ADF / 5:2)	24–48 h fast cycles	Net caloric reduction	Autophagy / ketosis
caloric-restriction	Unrestricted timing	20–40% sustained deficit	mTOR↓, AMPK↑, IGF-1↓
eTRF (early TRE)	6–8 h, morning-anchored (e.g., 7am–3pm)	Isocaloric	Circadian + insulin-sensitivity peak

The term “intermittent fasting” is often used loosely to include TRE, even though TRE fasting episodes (12–18 h) are considerably shorter than classic ADF/5:2 fasting (24–48 h). This conflation harms comparison across meta-analyses — many “IF” meta-analyses pool TRE and ADF protocols.

Circadian biology rationale

The foundational animal work establishing TRF (time-restricted feeding) as a circadian-metabolic intervention is the Hatori 2012 Cell Metabolism study from the Panda lab ². Male C57BL/6J mice fed a high-fat diet restricted to an 8-hour nocturnal active-phase window (ZT13–ZT21) were completely protected from obesity, hyperinsulinemia, hepatic steatosis, and inflammation compared to ad libitum-fed controls receiving identical total calories (caloric equivalence confirmed across 18 weeks; n=20–32 per group). The protective effect was attenuated when the eating window was extended toward ~12 hours, suggesting a dose-response for window length.

Mechanistic basis: metabolic organs (liver, muscle, adipose) are gated by circadian transcription factors (CLOCK/BMAL1, CRY/PER) that anticipate feeding windows and prime catabolic/anabolic programs accordingly. Disrupted or absent feeding rhythms — as in modern irregular eating patterns — decouple circadian metabolic priming from actual nutrient delivery, generating substrate mismatches that promote metabolic disease ³. no-fulltext-access — the Manoogian 2017 review PDF is behind an Elsevier paywall and could not be verified against full text; these review-sourced claims should be treated as unverified.

With age, circadian amplitude dampens across tissues, reducing the metabolic anticipatory signal. TRE may partially compensate by providing a stronger entrainment zeitgeber (a time cue) to peripheral clocks ³. This positions TRE as a potential circadian-aging intervention beyond simple cardiometabolic risk reduction.

needs-human-replication — the circadian-amplitude-rescue hypothesis for TRE in older adults has not been tested in controlled human trials.

Human evidence

Sutton 2018 (eTRF, pre-diabetic men)

The strongest early RCT evidence for metabolic-independent TRE effects ¹. Men with prediabetes (n=12 randomized; n=8 completers; crossover, 5-week periods) consumed a 6-hour early eating window (meals at ~7am, 10am, 1pm; dinner before 3pm) vs. a 12-hour control window in an isocaloric, eucaloric, weight-stable protocol — ruling out weight loss as a confound. eTRF significantly reduced:

• Insulin resistance (3-hour incremental AUC ratio, Δ=−36 ± 10 U/mg; p=0.005)
• Systolic blood pressure (Δ=−11 ± 4 mmHg; p=0.03) and diastolic blood pressure (Δ=−10 ± 4 mmHg; p=0.03)
• 8-isoprostane, an oxidative stress marker (Δ=−11 ± 5 pg/ml, ~14% reduction; p=0.05)

Note: cortisol was NOT significantly reduced (Δ=−0.1 ± 1.3 μg/dl; p=0.95). The wiki originally mis-stated “evening cortisol” as an outcome — this was incorrect. The study did not report HOMA-IR; the insulin-sensitivity endpoint was the incremental AUC ratio.

The study is limited by very small n (n=8 completers) and all-male, pre-diabetic population. needs-replication — these effect magnitudes require confirmation in larger, mixed-sex, non-diabetic samples.

Cienfuegos 2020 (4h vs 6h TRE, adults with obesity)

RCT comparing 4-hour TRE (3pm–7pm) vs 6-hour TRE (1pm–7pm) vs control over 8 weeks in 58 adults with obesity ⁴. Both TRE arms achieved significant weight loss (~3 kg), fat mass reduction, and reductions in fasting insulin vs control, with no statistically significant difference between the 4h and 6h windows. The late-afternoon eating window (ending 7pm) represents late TRE, not eTRF — window timing effects on circadian biology are not addressed.

Liu 2022 (NEJM; TRE vs CR in weight loss)

The largest and most rigorous TRE RCT to date ⁵. In 139 adults with obesity (Chinese population; n=69 TRE, n=70 CR; mean age 31.6±9.1 yr), a 16:8 TRE protocol (8am–4pm window) was compared to daily caloric restriction alone over 12 months. Crucially, both arms followed the same prescribed caloric restriction (1500–1800 kcal/day for men; 1200–1500 kcal/day for women; ~25% deficit below baseline ~2050–2075 kcal/day) — the TRE arm added only a timing constraint, it was not an ad libitum comparison. Primary finding: no significant difference in weight loss between TRE and CR (TRE: −8.0 kg, 95% CI −9.6 to −6.4; CR: −6.3 kg, 95% CI −7.8 to −4.7; net difference −1.8 kg, 95% CI −4.0 to 0.4; P=0.11). Secondary cardiometabolic outcomes (waist circumference, BMI, fat mass, lean mass, blood pressure, fasting glucose, HOMA-IR, lipids) also did not differ between groups. Caloric intake and macronutrient percentages were similar in the two groups throughout the trial.

Interpretation: TRE adds no weight-loss benefit beyond equivalent caloric restriction alone when both groups eat the same prescribed number of calories; the trial does not address whether TRE provides circadian-realignment benefits independent of caloric reduction (both arms were calorie-restricted). The Liu 2022 result is often misread as “TRE doesn’t work” — it more precisely shows “TRE ≈ CR for weight loss when both groups follow identical caloric prescriptions.” contradictory-evidence — the mechanism-specific circadian claim (independent of weight loss AND independent of caloric reduction) remains untested in large isocaloric-without-restriction RCTs.

Meta-analytic evidence (Patikorn 2021; Sun 2023)

The Patikorn 2021 umbrella review of 11 meta-analyses covering 130 RCTs of intermittent fasting protocols (including TRE) found 28 statistically significant associations out of 104 total associations tested (27%), largely for BMI, weight, fat mass, and glycemic markers ⁶. Most associations were supported by very low to low quality evidence by GRADE; only one met high-quality criteria (BMI reduction with modified alternate-day fasting); six met moderate-quality criteria. The TRE-specific signal is diluted by pooling with ADF and 5:2 protocols.

Sun 2023 meta-analyzed 8 RCTs (n=579) combining TRE with explicit caloric restriction and found significant improvements in body weight, fat mass, and waist circumference vs. CR alone — with no additional benefit on blood pressure or glycemia ⁷. Early-window TRE outperformed late-window for weight outcomes. needs-replication — TRE-as-additive-to-CR remains plausible but evidence base is small.

Sheng 2026 — TRE × organ-specific biological-age indices (NHANES, n=4,890)

Sheng et al. 2026 (npj Science of Food) cross-sectionally analyzed NHANES 2003–2018 data (n=4,890) for the relationship between TRE patterns (eating frequency + meal timing) and organ-specific biological-age indices (heart, kidney, liver, overall), frailty index, Life’s Essential 8, and cardiometabolic index, stratified by metabolic-health and obesity status ⁸. Key findings: excessively long or short fasting durations were associated with worse liver metabolic-health indices and worse cardiovascular risk markers; moderate fasting durations and eating frequencies were associated with lower biological-age indices and better health metrics across subgroups; the cardiovascular-health benefit was more pronounced in individuals who ate breakfast on time (consistent with the eTRF > late-TRE pattern). This is the largest-to-date observational signal linking TRE patterns to organ-aging biomarkers, though cross-sectional design precludes causal inference. needs-replication — RCT-grade replication with prospective biological-age clocks remains absent.

TREAD — Phase 1 TRE for Alzheimer’s MCI (planned)

Geda 2026 (J Alzheimers Dis) describe TREAD, an MCI/AD-targeted TRE trial design rationale ⁹. AD brains exhibit decreased glucose uptake while ketone utilization is preserved; ≥8–12 h fasting induces a metabolic switch to ketogenesis. Preclinical TRE in mouse models enhances cognitive function via hippocampal neurogenesis, autophagy, and reduced neuroinflammation. The TREAD trial protocol provides a theoretical model linking TRE → AD-resistance via direct (brain-energy) and indirect (cardiometabolic) pathways. No primary results yet — protocol/rationale paper. needs-replication

Mechanistic update: TRF rescues aged adipose function (Natarajan 2026)

Natarajan et al. 2026 (J Gerontol A) tested TRF (6h dark-cycle window) in 18-month-old C57BL/6 mice — the first TRF-in-aged-mice study with adipose-specific endpoints ¹⁰. TRF in aged mice improved metabolic flexibility, induced beige-adipocyte programming, and reduced fibro-inflammation in adipose tissue. The intervention has been previously well-characterized in young animals; this work extends it to aged biology and is consistent with the circadian-amplitude-rescue hypothesis. needs-human-replication — TRF effects on aged adipose tissue inflammation in older adults remain untested in RCTs.

Mechanism: how TRE differs from IF

Feature	TRE	Intermittent fasting (ADF/5:2)
Fasting duration	12–18 h/day	24–48 h/cycle
Caloric intent	Isocaloric	Net deficit
Ketosis?	Transient / absent	Significant on fast days
Autophagy depth	Mild induction	Substantial induction
Circadian entrainment	Primary mechanism	Secondary
Primary mtor pathway	Partial ↓ (nutrient-sensing)	Deeper ↓
ampk activation	Moderate	Stronger
Practical adherence	Higher	Lower

TRE’s autophagy and AMPK effects are real but likely smaller in magnitude than those triggered by full-day fasting. TRE’s unique mechanistic contribution is circadian entrainment of metabolic organ clocks — a signal IF does not consistently provide, particularly when the IF eating window is unconstrained in timing.

Eating-window timing: eTRF vs late-TRE

A consistent pattern across human and rodent data is that early eating windows (morning/midday-anchored) outperform late eating windows (afternoon/evening-anchored) for metabolic outcomes, independent of window duration ¹.

Mechanistic basis: insulin sensitivity in skeletal muscle and adipose peaks in the morning (~9am–11am) due to circadian gating of GLUT4 and insulin receptor expression. Eating early aligns nutrient delivery with peak tissue responsiveness; eating late creates post-prandial glucose/insulin loads against a background of declining tissue sensitivity — even at identical total calories and identical window duration.

Practical consequence: a noon–8pm eating window (common in consumer TRE) is chronobiologically sub-optimal compared to a 7am–3pm (eTRF) window. Most popular TRE literature does not adequately distinguish these, which may explain heterogeneity in trial outcomes. contradictory-evidence — direct head-to-head eTRF vs late-TRE trials in older adults are needed; the existing evidence is largely indirect.

Aging-specific considerations

• Circadian dampening with age — aging reduces amplitude of BMAL1/CLOCK cycling; TRE’s entraining effect may be proportionally larger in older adults, but this is untested. needs-human-replication
• Muscle preservation — unlike heavy CR, isocaloric TRE does not produce protein deficits; no consistent evidence of lean-mass loss in short-term TRE RCTs. Important for sarcopenia risk in elderly.
• Frailty and appetite — older adults may spontaneously eat less within a shortened window and risk nutritional deficiency. Supervision recommended in adults >70.
• Drug timing — time-sensitive medications (statins, blood pressure agents) require dose-timing review if eating windows shift dramatically.
• Sleep-adjacent feeding — early TRE (finish by 3pm) requires very early breakfast (~7am), which is difficult for older adults with morning stiffness or social eating patterns. Adherence is the primary practical limitation.

Practical considerations and safety

TRE has a well-established short-term safety profile across healthy adults and those with metabolic syndrome. No serious adverse events have been reported in published RCTs. The main concerns are:

• Nutritional adequacy in windows < 6 hours (micronutrient gaps possible)
• Disruption of existing eating disorder patterns
• Interaction with medications requiring food co-administration

TRE is not recommended without medical supervision for: type 1 diabetes, pregnancy or lactation, history of eating disorders, or adults with significant underweight or frailty.

Cross-organism extrapolation

Dimension	Status	Notes
Pathway conserved in humans?	yes	Circadian metabolic clocks conserved; CLOCK/BMAL1 function in humans is established
Phenotype conserved in humans?	partial	Hatori 2012 mouse findings partially replicated; magnitude smaller and weight-loss-confounded
Replicated in humans?	in-progress	Multiple RCTs exist; no hard-endpoint (mortality/CVD) trial completed

needs-human-replication — TRE’s lifespan/healthspan effects in humans are inferred from circadian biology theory + metabolic biomarker data; no human RCT has used longevity or hard cardiovascular endpoints.

Key open questions

1. Window timing vs duration — eTRF (early window) vs late-TRE head-to-head in pre-registered isocaloric RCT; DunedinPACE + HOMA-IR as primary endpoints.
2. Long-term adherence — most TRE trials are ≤12 weeks; real-world adherence beyond 1 year is unknown. long-term-unknown
3. Age-specific effects — does TRE’s circadian-entrainment benefit scale with age (where circadian dampening is greater)? No RCT in adults >65 with aging-specific endpoints.
4. Isocaloric control — very few human TRE trials succeed at isocaloric matching; the question of TRE benefit independent of caloric reduction is mostly unanswered in humans.
5. Interaction with exercise — combining resistance training with TRE for sarcopenia prevention; pilot data emerging, no large RCT.
6. Biological age clock response — DunedinPACE has been measured in caloric-restriction (Waziry 2023 slowing d=−0.25); TRE has not been tested against epigenetic clock endpoints. needs-replication

Cross-references

• caloric-restriction — overlapping but distinct; TRE is not CR; see definitional table
• intermittent-fasting — sibling intervention; see mechanism table for distinctions
• autophagy — mildly induced by TRE; more deeply induced by IF
• mtor — partially suppressed by nutrient-sensing during TRE fasted window
• ampk — moderately activated during fasted window
• sirtuin — NAD+ cycling influenced by feeding pattern
• insulin-igf1 — primary cardiometabolic target; eTRF improves insulin sensitivity
• deregulated-nutrient-sensing — primary hallmark target
• altered-intercellular-communication — circadian-mediated inter-organ signaling
• sarcopenia — potential protective effect (isocaloric; preserves lean mass); elderly supervision caveat
• dunedinpace-2022 — priority endpoint for future TRE aging trials

Sibling R23c interventions: methionine-restriction, ketogenic-diet, sleep, heat-exposure

Footnotes

Footnotes

1. doi:10.1016/j.cmet.2018.04.010 · rct (crossover) · n=12 randomized, n=8 completers · model: human males with prediabetes; eTRF 6h window (meals at ~7am/10am/1pm, dinner before 3pm) vs 12h control window, 5 weeks each, isocaloric eucaloric weight-stable · Sutton EF et al. · Cell Metabolism 2018 · eTRF reduced insulin resistance (incremental AUC ratio Δ=−36 U/mg; p=0.005), SBP (Δ=−11 mmHg; p=0.03), DBP (Δ=−10 mmHg; p=0.03), 8-isoprostane (~14%; p=0.05); cortisol NOT significantly changed (p=0.95); effects independent of weight loss · citation_percentile: 100th; 1372 citations · local PDF: failed (PMC full text verified) ↩ ↩² ↩³

2. doi:10.1016/j.cmet.2012.04.019 · in-vivo · n=20–32 per group (4 groups: NC ad lib, NC TRF, HFD ad lib, HFD TRF) · model: 8-week-old male C57BL/6J mice (Jackson Laboratory), HFD (61% fat) or NC, 8h active-phase TRF (ZT13–ZT21) vs ad libitum, >18 weeks · Hatori M et al. · Cell Metabolism 2012 · TRF prevented obesity, hyperinsulinemia, hepatic steatosis, and inflammation without caloric restriction; caloric intake equivalent between TRF and ad lib groups throughout experiment; protection attenuated as window extended toward ~12h · citation_percentile: 100th; 1904 citations · local PDF: verified ↩

3. doi:10.1016/j.arr.2016.12.006 · review · Manoogian ENC, Panda S · Ageing Research Reviews 2017 (published Dec 2016) · circadian rhythms dampen with age; feeding-fasting cycles strongly entrain peripheral clocks; animal TRF prevents metabolic disease without CR; human epidemiology: erratic eating ↑ disease risk; structured feeding-fasting may sustain circadian amplitude · 304 citations · local PDF: failed (Elsevier paywall) no-fulltext-access — claims attributed to this review could not be verified against full text; body-text claims that cite only this source should be treated as unverified ↩ ↩²

4. doi:10.1016/j.cmet.2020.06.018 · rct · n=58 randomized (4h TRF n=19, 6h TRF n=20, control n=19); n=49 completers · model: adults with obesity, late-afternoon windows (4h: 3pm–7pm; 6h: 1pm–7pm) vs control, 8 weeks · Cienfuegos S et al. · Cell Metabolism 2020 · Both TRE windows produced ~3.2% weight loss (Δ=−3.2±0.4%), reduced fasting insulin and HOMA-IR vs control; energy intake reduced ~550 kcal/day without calorie counting; 4h vs 6h not significantly different for weight or most cardiometabolic markers · citation_percentile: 100th; 536 citations · local PDF: verified ↩

5. doi:10.1056/NEJMoa2114833 · rct · n=139 randomized (TRE n=69, CR n=70); n=118 completed 12 months · model: adults with obesity (mean age 31.6 yr, mean weight 88.2 kg), Chinese population; both arms prescribed same caloric restriction (~25% deficit; 1500–1800 kcal/day men, 1200–1500 kcal/day women); TRE arm additionally constrained to 8am–4pm window; 12 months · Liu D et al. · NEJM 2022 · No significant difference in weight loss (TRE −8.0 kg [95% CI −9.6 to −6.4] vs CR −6.3 kg [95% CI −7.8 to −4.7]; net difference −1.8 kg [95% CI −4.0 to 0.4]; P=0.11); no between-group difference in waist circumference, BMI, fat mass, lean mass, blood pressure, glucose, HOMA-IR, or lipids; caloric intake similar between groups by design · citation_percentile: 100th; 450 citations · local PDF: verified ↩

6. doi:10.1001/jamanetworkopen.2021.39558 · meta-analysis (umbrella) · 11 meta-analyses, 130 RCTs of IF protocols (includes TRE + ADF + 5:2); median sample size per RCT n=38 · Patikorn C et al. · JAMA Network Open 2021 · 28/104 tested associations significant (27%); mostly BMI/weight/fat mass/glycemia/lipids; 75 associations (72%) very-low-quality GRADE; 22 (21%) low; 6 (6%) moderate; 1 (1%) high (BMI reduction with MADF); no TRE-specific high-quality associations · citation_percentile: 100th; 251 citations · local PDF: verified ↩

7. doi:10.1038/s41430-023-01311-w · meta-analysis · 8 RCTs, n=579 (overweight/obese adults) · Sun J-C et al. · European Journal of Clinical Nutrition 2023 · TRE+CR vs CR alone: significant reductions in body weight (WMD −1.40, 95% CI −1.81 to −1.00), fat mass (WMD −0.73, 95% CI −1.39 to −0.07), and waist circumference (WMD −1.87, 95% CI −3.47 to −0.26); no significant additional benefit on blood pressure, fasting glucose, insulin, HOMA-IR, or lipid profile; subgroup analysis: early TRE (eTRE) outperformed delayed TRE for weight (WMD −1.42, 95% CI −1.84 to −1.01) and fat mass (WMD −1.06, 95% CI −1.91 to −0.22) · local PDF: verified ↩

8. doi:10.1038/s41538-026-00862-z · cross-sectional · n=4,890 (NHANES 2003–2018) · Sheng L, Li E, Ji H, Cheng S, Kwan A, Bischof E, Chen Y, Pu J · npj Science of Food 2026 · TRE patterns assessed by eating frequency and meal timing; outcomes: organ-specific biological-age indices (heart, kidney, liver, overall), frailty index, Life’s Essential 8, cardiometabolic index; stratified by metabolic-health × obesity · finding: excessively long/short fasting durations associated with worse liver metabolic-health and worsened cardiovascular risk; moderate fasting durations/eating frequencies associated with lower biological-age indices and better health metrics; on-time breakfast strengthens the cardiovascular-health × healthy-metabolism association · cross-sectional limits causal inference · PMID 42091606 ↩

9. doi:10.1177/13872877261438520 · trial protocol/review · Geda YE, Krell-Roesch J, Bekele K, Gunning JA, Zaniletti I, Eagan D, Demeke M, Khan N, Chahal G, Smith T, DeCuna C, Aliskevich EL, Hettiwatte Y, Ransdell M, Racette SB · J Alzheimer’s Dis 2026 (Online Apr 6) · proposes TREAD trial — TRE in MCI/AD; theoretical model linking TRE to AD-resistance via direct (brain-energy) and indirect (cardiometabolic) pathways; AD brains exhibit decreased glucose uptake while ketone utilization preserved; ≥8–12 h fasting induces metabolic switch to ketogenesis; preclinical TRE in mouse models enhances cognitive function via hippocampal neurogenesis, autophagy, reduced neuroinflammation · no primary results yet · PMID 41940859 ↩

10. doi:10.1093/gerona/glag085 · Natarajan D, Milan M, Reyff Z, Negri S, Ekambaram S, Varshney RR, Rudolph MC, Tarantini S, Balasubramanian P · J Gerontol A Biol Sci Med Sci 2026;81(5):glag085 · in-vivo · 18-month-old C57BL/6 mice on 6h dark-cycle TRF vs ad libitum · TRF improved metabolic flexibility, induced beige-adipocyte programming, reduced fibro-inflammation in adipose tissue; first TRF study in aged mice with adipose-specific endpoints · PMID 41888918 ↩