16:8 Intermittent Fasting: What to Know

What the 16:8 Schedule Means

16:8 means eating all of a day's food within an 8-hour window and fasting for the remaining 16 hours, most commonly by skipping breakfast or dinner. It is one of several named intermittent-fasting schedules; see how to start intermittent fasting for how it compares to 5:2 and alternate-day fasting in the broader research.

Does It Matter When Your 8-Hour Window Is (Early vs. Late)

A 12-week trial in 197 adults with overweight or obesity directly tested this question, comparing an early 8-hour window, a late 8-hour window, a self-selected 8-hour window, and usual-care education built around Mediterranean-diet counseling. The trial's own primary outcome, visceral fat measured by MRI, showed no significant difference between any of the three time-restricted-eating schedules and usual care (mean differences of -4%, -6%, and -3% relative to usual care, all p values 0.31 or higher) and none between the schedules; all three time-restricted groups did lose significantly more weight than usual care (2.9, 2.4, and 3.1 kg more for the early, late, and self-selected windows), a secondary outcome, with no significant difference in fat-free mass or lean mass across groups at 12 weeks (Dote-Montero et al., 2025) [23]. A secondary analysis of the same trial found liver fat, also measured by MRI, decreased within all three time-restricted-eating groups but not significantly more than usual care, and did not differ between early, late, or self-selected windows (Dote-Montero et al., 2026) [8]. In other words, in this trial, which 8 hours you choose did not appear to matter as much as whether you keep to a structured window at all; a significantly higher share of the time-restricted groups reached at least 5% weight loss than the usual-care group. A 12-month follow-up of 99 participants from the same trial, assessed by bioelectrical impedance a year after the 12-week intervention ended, found the early and late windows had maintained significantly greater weight loss than usual care (p=0.01 and p=0.02 respectively), a difference that held when the three time-restricted groups were pooled; fat-mass loss remained greater in the early window (p=0.04), but a reduction in fat-free mass also persisted in the late window (p=0.04) and, as a non-significant tendency, in the early window (p=0.051), and only 26% of participants reported practicing time-restricted eating at all during the follow-up year (Camacho-Cardenosa et al., 2026) [24].

What the Research Shows

Weight and Fat Mass

A 3-month trial in people with obesity and type 2 diabetes found 16:8 fasting, done 3 days a week, produced a larger weight change than control, with improved fasting glucose, HbA1c, and lipids (Sukkriang and Buranapin, 2024) [3]; that protocol ran 3 days a week, not daily, which matters if this trial is cited as evidence for a daily 16:8 schedule. A meta-analysis of 15 trials combining exercise with 16:8 found the combination reduced body weight and fat mass versus exercise alone, with smaller improvements in triglycerides and LDL cholesterol, though the authors caution the evidence base is limited by study number, sample size, and inconsistent participant characteristics (Liu et al., 2026) [7]. Set against that, a 2025 network meta-analysis of 99 randomized trials did not find time-restricted eating beating ordinary calorie restriction; only alternate-day fasting showed that advantage (Semnani-Azad et al., 2025) [1]. Two trials have tested an 8-hour window directly against a calorie-matched comparison, with different results: the longest and largest, 12 months in 139 people with obesity, found no significant weight-loss advantage over calorie restriction alone (net difference -1.8 kg, 95% CI -4.0 to 0.4, p=0.11) (Liu et al., 2022) [11]; a shorter, 14-week trial in 90 adults with obesity found an earlier 8-hour window (7am to 3pm) more effective for weight loss than a wider eating window at the same weight-loss counseling (-2.3 kg, 95% CI -3.7 to -0.9, p=0.002) and improved diastolic blood pressure, though it did not significantly affect fat loss in the main intention-to-treat analysis; a secondary analysis of 59 completers found the early window also more effective for losing body fat and trunk fat (Jamshed et al., 2022) [12]. A further secondary analysis of those 59 completers, which the authors describe as exploratory with no subgroup powered in advance, looked at effects by cardiometabolic risk factor, age, sex, and race. The early window lowered diastolic, but not systolic, blood pressure in the 41 participants with hypertension (4±2 mm Hg, p=0.04) and in a post hoc subgroup of 18 with systolic pressure of 130 mm Hg or higher (9±4 mm Hg, p=0.047), where HDL cholesterol also fell by 5.7±2.5 mg/dL (p=0.04). It lowered HbA1c by 0.3%±0.1% (p=0.04) in the 20 participants with prediabetic HbA1c, with no change in fasting glucose, insulin, or HOMA-IR (a fasting insulin-resistance index) in that group, and no significant change in anything in the 28 with impaired fasting glucose or the 37 with insulin resistance; in the 31 with high cholesterol it reduced HOMA-IR by 2.96±1.46 (p=0.04) without changing cholesterol or triglycerides. In the 25 participants with severe obesity (BMI 40 or higher), the early window raised total and LDL cholesterol by 21±9 and 17±7 mg/dL (both p=0.03), with no such change in those with less severe obesity; the authors call this unexpected, say some caution may be merited in that group, and ask for replication. There were no significant differences by age or sex. A signal that the early window improved systolic blood pressure, heart rate, and glycemic measures more in Black than in White participants rests on 13 Black participants (6 on the early window), and the authors state it should be treated as hypothesis-generating and could be a false positive (Rehman et al., 2026) [20].

Another secondary analysis of the same early-window trial looked at it from a different angle: not everyone randomized, but only the participants who stuck to their assigned schedule at least 5 days a week, every week (15 on the early window, 21 controls), a per-protocol comparison rather than the intention-to-treat result above. Among these adherent completers, the early window produced greater improvements than adherent controls in body weight (-3.7±1.2 kg, p=0.003), body fat (-2.8±1.3 kg, p=0.04), heart rate (-7±3 beats/min, p=0.02), insulin resistance (HOMA-IR -2.80±1.36, p=0.047), and fasting glucose (-9±5 mg/dL, p=0.047), with no between-group difference in fat-free mass, blood pressure, or lipids, plus greater improvements in mood, including less fatigue and anger. But the same adherers also self-reported sleeping about 30 minutes less (p=0.03), taking about 7 minutes longer to fall asleep (p=0.04), and slightly lower sleep efficiency than adherent controls, with no difference in self-rated sleep quality (Steger et al., 2023) [21]; see Safety and Cautions below for how that self-reported result compares with an objective sleep measurement from a different trial. Because this is a per-protocol analysis of people who kept to the schedule, not everyone randomized to it, its numbers run larger than the trial's intention-to-treat result (-3.7 kg versus -2.3 kg for weight); the authors note that intention-to-treat is the better guide to what prescribing the schedule achieves, and that a per-protocol analysis can be affected by selection bias.

A separate, isocaloric trial tested a wider, 10-hour window with most calories eaten before 1 p.m.: over 12 weeks in 41 adults with obesity and prediabetes or diet-controlled type 2 diabetes, it found no significant difference in weight loss between the 10-hour group and a calorie-matched usual eating pattern (average difference 0.3 kg, 95% CI -1.2 to 1.9) (Maruthur et al., 2024) [13]. A post hoc analysis of 39 participants from the same trial found no significant between-group difference in appetite hormones (leptin, ghrelin, adiponectin) or inflammatory markers (CRP, cortisol, sRAGE), though both groups improved on eating-behavior scores; the authors note the analysis was not powered to detect small differences, so the null result should not be read as proof the two patterns are equivalent (Duan et al., 2026) [14].

A pooled analysis of individual participant data from five time-restricted-eating trials lasting 3 to 6 months (three single-arm, two randomized controlled; 124 adults, each with at least one feature of the metabolic syndrome and a usual eating window of 14 hours or more) found that participants allocated to an 8-hour window, with no calorie target, lost more weight than those allocated to a 12-hour window (1.66 kg more, 95% CI 0.22-3.10, p=0.02) and more than controls, who kept their usual eating habits or received brief healthy-eating advice (2.25 kg more, 95% CI 0.84-3.66, p=0.002); the difference versus a 10-hour window was smaller and not statistically significant (1.02 kg, 95% CI -0.16 to 2.21, p=0.09) (Papageorgiou et al., 2026) [15]. Only 11 of the 124 participants were in the 8-hour group, all from a single trial, and each window length came from different trials, so this comparison cannot separate the length of the window from other differences between those trials; the authors describe the analyses as exploratory, made no adjustment for multiple comparisons, and call for larger, longer trials to confirm the findings. Blood pressure, fasting glucose, and blood lipids did not differ consistently by window length; the one exception was a larger drop in systolic blood pressure with a 10-hour than with a 12-hour window. The paper was published as an early-access manuscript ahead of final editing, so its figures may still change.

Lean Mass: The Recurring Caveat

A dedicated, purpose-built 12-week trial testing daily 16:8 against three structured meals a day found no significant weight-loss advantage and a significant reduction in lean mass in the 16:8 group (appendicular lean mass index -0.16 kg per square meter, p=0.005) (Lowe et al., 2020) [2]. This is one of several directly relevant trials for a daily 8-hour-window schedule, not a pooled estimate, and it is the only one of them to report a significant lean-mass reduction at the end of the intervention itself; the window-timing trial found no fat-free-mass difference at 12 weeks, but its 12-month follow-up reported a reduction in fat-free mass persisting in the late window versus usual care (p=0.04), measured by bioelectrical impedance (Dote-Montero et al., 2025 [23]; Camacho-Cardenosa et al., 2026 [24]). The lean-mass signal is worth taking seriously rather than treating 16:8 as cost-free for muscle, though it was not seen in the other direct trials on this page that reported lean mass at the end of their intervention.

Resistance Training and Muscle Gain

That caveat is not universal. A 12-week trial in 23 resistance-trained adults deliberately eating a calorie surplus to gain muscle found the 16:8 group had less fat gain and less body-weight gain than an unrestricted control group, while fat-free mass gains and strength gains were similar between groups (Gavanda et al., 2026) [5]. The 16:8 group ate less than its prescribed surplus, which the authors say explains the smaller fat gain, and body composition was measured by single-frequency bioimpedance, so this is not a clean test of the window alone. In this population, actively training and eating a surplus, an 8-hour window did not cost muscle or strength; the lean-mass caveat from the TREAT trial is population- and goal-dependent, not a universal finding.

Older Adults

A 2026 systematic review with a network meta-analysis of 7 randomized trials in older adults found 16:8 time-restricted eating reduced body weight (-1.92 kg, 95% CI -2.57 to -1.27) and BMI (-1.01 kg/m2, 95% CI -1.69 to -0.33) compared with usual diet, in trials lasting 6 to 12 weeks. One other protocol, a twice-weekly daytime fast combined with a daily calorie reduction, also reduced both, while a 5:2 protocol, a 12:12 window, and a healthy-living diet did not differ significantly from usual diet. The authors rate the weight evidence moderate certainty, found significant heterogeneity in the 16:8 BMI comparison, and caution that rankings in a network this small are unstable; lean mass was not pooled, though the individual 16:8 trials reported no lean-mass loss. The same review's observational studies are mixed on long daily fasts in older adults: an eating window of 10 hours or less was linked to better cognitive scores in one study and worse in another, and the longest overnight fasts were linked to higher cardiovascular mortality in one US cohort; these associations rest on self-reported meal timing, cannot show cause, and the authors call them hypothesis-generating (Couto-Alfonso et al., 2026) [16]. Older adults, particularly anyone taking regular medication, should talk to a clinician before starting; see how to start intermittent fasting for the sitewide guidance.

Liver Fat and Gut Microbiota

See "Does It Matter When Your 8-Hour Window Is" above for this trial's findings on visceral fat and liver fat (Dote-Montero et al., 2025 [23]; Dote-Montero et al., 2026 [8]): the trial's own primary measure, visceral fat by MRI, found no significant difference between any time-restricted-eating schedule and usual care, and a secondary analysis found time-restricted eating did not reduce liver fat more than usual-care education either; the authors link the liver-fat reductions seen within the time-restricted-eating groups to weight loss and baseline fatty liver rather than to window timing, and note the trial was not powered for secondary outcomes. A structured 8-hour window still helped more participants reach a meaningful weight-loss threshold than usual care.

Cognitive Performance

A small 4-week pilot in 43 healthy women (23 fasting, 20 control) found 16:8 produced a significant BMI reduction versus control, with no significant difference in working memory or attentional-control test performance, and no difference in reported adverse events (McLeod et al., 2026) [6]. This is a small, short pilot; it supports "no evidence of cognitive harm" in this trial, not proof of cognitive benefit or impairment either way.

How 16:8 Compares to Other Schedules

Where 16:8 is compared directly against alternate-day fasting, in a 3-week trial in adults with prediabetes, alternate-day fasting outperformed 16:8 on weight and BMI reduction, though both beat control (Chair et al., 2022) [4]. For the broader comparison across intermittent fasting schedules generally, see how to start intermittent fasting. For the more extreme end of the daily-fasting spectrum, a single daily meal rather than an 8-hour window, see OMAD.

Getting Started

Pay attention to hydration and electrolytes during the fasting window; see electrolytes during a fast for the mechanism and warning signs, without a home dosing recommendation. For what counts as breaking the 16-hour window, see what breaks a fast.

Who Should Talk to a Clinician First

People on insulin, a sulfonylurea, or blood-pressure medication should only start 16:8 under clinician supervision; no 16:8-specific trial in this set tested that population directly, and this is the same standing caution carried from the general intermittent-fasting guide. Pregnancy, a history of an eating disorder, being under 18, receiving cancer treatment, and any chronic condition requiring consistent food or medication timing are reasons to get clinician guidance before starting, as a matter of general clinical caution rather than a specific finding from these trials.

Common Questions

Does 16:8 beat other schedules?

Not clearly: a large 2025 pooled estimate does not show it beating calorie restriction, and a direct comparison found alternate-day fasting ahead of it on weight and BMI (Semnani-Azad et al., 2025 [1]; Chair et al., 2022 [4]).

Does it cost muscle?

That depends on the population and goal: it did in a weight-focused trial and did not in a muscle-gain-focused trial (Lowe et al., 2020 [2]; Gavanda et al., 2026 [5]).

Does the time of day matter?

Not for visceral fat, the trial's own primary measure, or for liver fat, a secondary measure, in the one trial that tested timing directly (Dote-Montero et al., 2025 [23]; Dote-Montero et al., 2026 [8]).

Does 16:8 trigger autophagy?

See our page on whether intermittent fasting triggers autophagy for what the human evidence does and does not show.

Authors and Funding of the Cited Studies

• Semnani-Azad et al., 2025 [1]: funded by the Diabetes and Nutrition Study Group of the European Association for the Study of Diabetes, which commissioned the review, and by the Canadian Institutes of Health Research. Of the 22 authors, several reported research funding, honoraria, speaker fees, or advisory-board roles with companies including Novo Nordisk, Eli Lilly, AstraZeneca, Danone, Nestle, General Mills, and multiple food-industry trade groups; all other authors declared no competing interests.
• Lowe et al., 2020 [2]: funded by the University of California, San Francisco's Cardiology Innovations Award Program, the National Institute of Diabetes and Digestive and Kidney Diseases, and the James Peter Read Foundation, with in-kind equipment gifts from iHealth Labs, MOCACARE, and Oura. Dr. Lowe reported personal fees from Keyto outside the submitted work; Dr. Vittinghoff reported personal fees from UCSF during the conduct of the study; Dr. Heymsfield reported personal fees from the Medifast Medical Advisory Board and the Tanita Medical Advisory Board outside the submitted work; Dr. Weiss reported grants from NIH and the James Peter Read Foundation and non-financial support from Mocacare and iHealth Labs during the conduct of the study, is a cofounder and equity stakeholder of Keyto, Inc., and owns stock in and was formerly on the board of Virta, Inc. No other disclosures were reported.
• Sukkriang and Buranapin, 2024 [3]: funded by the Endocrine Society of Thailand, with the article processing charge funded by Chiang Mai University. The authors reported no conflicts of interest.
• Chair et al., 2022 [4]: the authors declared no conflicts of interest, as recorded in the article's PubMed record. A funding statement could not be retrieved for this page.
• Gavanda et al., 2026 [5]: co-author C. Brinkmann (CB in the paper's statement) reported honoraria from Abbott, Lilly, and Novo Nordisk; the paper states that these companies had no role in the study design, data analysis, interpretation, or the decision to publish, and that all other authors report no conflicts of interest (as recorded in the article's PubMed record). A funding statement could not be retrieved for this page.
• McLeod et al., 2026 [6]: funded by the Health Research Council of New Zealand. The paper's competing-interest statement, as recorded in the article's PubMed record, reports that Mei Peng received financial support from the Health Research Council of New Zealand and that the other authors declare no known competing financial interests or personal relationships that could have appeared to influence the work.
• Liu, Hu, and Ma, 2026 [7]: the authors declared that financial support was not received for the work or its publication, and that the work was conducted in the absence of any commercial or financial relationships that could be construed as a conflict of interest.
• Dote-Montero et al., 2026 [8]: the authors declared no conflicts of interest and refer readers to the accompanying ICMJE disclosure forms for further details, as recorded in the article's PubMed record. The published article's own Financial Support section lists funding from the Spanish Ministry of Science, Innovation, and Universities (MCIU/AEI) and the European Regional Development Fund (ERDF, EU), the Junta de Andalucia, the University of Granada, the Government of Navarra, the Instituto de Salud Carlos III, and the EXERNET Research Network on Exercise and Health, plus named individual-researcher grants to several co-authors.
• Allaf et al., 2021 [9]: all eight review authors stated they have no conflict of interest. A funding statement for the review's "Sources of support" section could not be retrieved for this page.
• Patikorn et al., 2021 [10]: Dr. Pham reported personal fees for the Southwest Obesity Symposium outside the submitted work; Dr. Varady reported author fees from the Hachette Book Group for her book, The Every Other Day Diet. No other conflicts were reported. A separate funding statement could not be retrieved for this page.
• Liu et al., 2022 [11]: funded by the National Key Research and Development Project, the Outstanding Youths Development Scheme of Nanfang Hospital, the Key-Area Clinical Research Program of Southern Medical University, the Key-Area Research and Development Program of Guangdong Province, and the National Natural Science Foundation of China. The article states that author disclosure forms are available with the full text at NEJM.org but does not print them in the body text; itemized author disclosures were not independently accessible for this page. Participants were also provided one protein shake per day, Nutriease, made by Zhejiang Nutriease, for the first six months, which the paper describes as intended to help adherence to the permitted calorie intake, and the paper's acknowledgments thank three named nutritionists at Zhejiang Nutriease.
• Jamshed et al., 2022 [12]: funded by National Institutes of Health grants, including the National Center for Advancing Translational Sciences and the National Institute of Diabetes and Digestive and Kidney Diseases. Dr. Martin reported grants from NIH during the conduct of the study and personal fees as a scientific advisory board member for Wondr Health outside the submitted work; his institution, Pennington Biomedical Research Center/Louisiana State University, holds an intellectual-property interest in the Remote Food Photography Method and SmartIntake app, on which Dr. Martin is an inventor. Dr. Peterson reported grants from NIH during the conduct of the study. No other disclosures were reported.
• Maruthur et al., 2024 [13]: funded by American Heart Association grants and National Institutes of Health grants. The article refers readers to the authors' ICMJE disclosure forms at acponline.org; itemized disclosures were not accessible for this page.
• Duan et al., 2026 [14]: funded by an American Heart Association grant, with D. Duan separately supported by a National Institutes of Health grant and S. J. Pilla by National Institute of Diabetes and Digestive and Kidney Diseases grants and Resolve to Save Lives. D. Duan reported serving as a consultant for Rhythm Pharmaceuticals and receiving honoraria from the American Diabetes Association; S. Carnell reported receiving grant funding from Eli Lilly and Company; J. M. Clark reported participating on a Data Safety Monitoring Board for Eli Lilly and Company and serving on the advisory board of Novo Nordisk; N. M. Maruthur reported receiving royalties from UpToDate. The other authors declared no conflicts of interest.
• Papageorgiou et al., 2026 [15]: the paper's Funding statement names support for three of the five pooled trials and for two of the authors: the SeeFoodStudy was supported by the Healthy Foods Healthy Lives program and the National Institutes of Health (National Center for Advancing Translational Sciences); B. Laferrère and the TREAT pilot study (a separate New York trial, not the TREAT trial by Lowe et al., 2020 [2] cited above) were supported by the National Institute on Aging; T.-H. Collet and the SwissChronoFood study were supported by the Swiss National Science Foundation and the Swiss Society of Endocrinology and Diabetes, and T.-H. Collet's research was also supported by the Leenaards, Vontobel, SwissLife Jubiläumsstiftung, and Nutrition 2000plus foundations. Open access publication was funded by the University of Geneva. The statement names no funder for the UCSD trial or the TREMNIOS pilot study. The paper's competing-interests statement reports that L. S. Chow has an investigator-initiated grant from DEXCOM for product only, and lists no other interests.
• Couto-Alfonso et al., 2026 [16]: "This research received no external funding." Disclosure: "The authors declare no conflicts of interest."
• Chen et al., 2025 [17]: the paper's competing-interest statement, as recorded in the article's PubMed record, reads: "The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper." A funding statement from the published article could not be retrieved for this page.
• Eliopoulos et al., 2025 [18]: "The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. DS is supported by CUREPLaN, a grant from the Leducq Foundation for Cardiovascular Research (18CVD01)." Conflict of interest: "AE and KG were employed by Genosophy S.A. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The author(s) declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision."
• Clavero-Jimeno et al., 2025 [19]: Conflict of interest: "Dr Alfaro-Magallanes reported receiving funding from the MCIN/AEI and the European Union. Dr Osés reported receiving grants from the Spanish Ministry of Economy outside the submitted work. Dr de Cabo reported receiving funding from the Intramural Research Program of the National Institute on Aging. Dr Ruiz reported receiving lecture fees from Novo Nordisk and Abbott for research unrelated to this study. No other disclosures were reported." Funding: the paper's Funding/Support statement names MCIU/AEI and ERDF/EU support (Dr Ruiz); a Junta de Andalucía grant (Dr Ruiz); the University of Granada Plan Propio de Investigación Unit of Excellence on Exercise Nutrition and Health (Dr Ruiz); Spanish Ministry of Universities grants through the University of Granada (Mr Clavero-Jimeno and Dr Dote-Montero); a Government of Navarra grant (Dr Labayen); a Juan de la Cierva Formación grant funded by the Ministerio de Ciencia e Innovación and the EU NextGenerationEU/PRTR (Dr Camacho-Cardenosa); a Beatriz Galindo fellowship grant from the Spanish Ministry of Science, Innovation and Universities (Dr Migueles); and EXERNET Research Network on Exercise and Health grants (Drs Labayen and Ruiz).
• Rehman et al., 2026 [20]: Funding: "This study was supported by UL1 TR001419 from the National Center for Advancing Translational Sciences, P30 DK056336 from the National Institute of Diabetes and Digestive and Kidney Diseases of the NIH, and a start-up package to CMP. Resources and support were also provided by a Nutrition Obesity Research Center grant (P30 DK056336), a Diabetes Research Center grant (P30 DK079626), and an NIH Postdoctoral T32 Obesity Fellowship to OMA (T32 DK062710). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH." Conflicts of interest: "Starting in late spring 2026, CMP will receive consulting fees from WndrHLTH Club, Inc., for providing content knowledge on intermittent fasting, circadian rhythms, and nutrition. All other authors report no conflicts of interest." In the paper's initials, CMP is senior author C. M. Peterson and OMA is O. M. Altonji.
• Steger et al., 2023 [21]: funded by National Institutes of Health grants, including the National Center for Advancing Translational Sciences (UL1 TR001419) and the National Institute of Diabetes and Digestive and Kidney Diseases (P30 DK056336), with additional support from two Nutrition Obesity Research Center grants, a Diabetes Research Center grant, an NIH Predoctoral T32 Obesity Fellowship to C. J. Hanick, and the Louisiana Clinical and Translational Science Center. Pennington Biomedical Research Center/Louisiana State University holds an intellectual-property interest in the Remote Food Photography Method and SmartIntake app, on which co-author C. K. Martin is an inventor; all other authors declared no conflict of interest.
• Dote-Montero et al., 2025 [23]: funded by the Spanish Ministry of Science, Innovation and Universities (MCIU/AEI) and the European Regional Development Fund (ERDF, EU), the Junta de Andalucía, the University of Granada, the Government of Navarra, the Spanish Ministry of Universities, the EXERNET Research Network on Exercise and Health, and named individual-researcher grants to several co-authors; co-author R. de Cabo is funded through the Intramural Research Program of the National Institute on Aging. Conflict of interest: J. R. Ruiz reported receiving lecture fees from Novo Nordisk and Abbott for research unrelated to this study; all other authors declared no competing interests.
• Camacho-Cardenosa et al., 2026 [24]: the authors declared no conflicts of interest, as recorded in the article's PubMed record. A funding statement could not be retrieved for this page.
• Sun et al., 2024 [25]: funded by the National Key Research and Development Program of China, the Natural Science Foundation of China, and two institutional funds at Shengjing Hospital of China Medical University; the paper states the funders had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Declaration of interests, as a single blanket statement covering all 21 authors: "The authors declare no competing interests." Krista A. Varady, already disclosed above (citation 10) for book author fees, is a co-author of this paper too, but this paper's own declaration does not itemize her individually.
• Garegnani et al., 2026 [26]: the review states it had no dedicated funding. Its conflict-of-interest statement, as recorded in the review's PubMed record: eight of the ten authors declared none known; one (PA, Andrenacci) is a member of the Cochrane Hepato-Biliary Group's editorial team and was not involved in the editorial process for this review; one (MB, Boyd) serves as an Editor-in-Chief for the Nutrition Care Manual of the Academy of Nutrition and Dietetics, described in the statement as an organisation of food and nutrition professionals with a declared opinion or position on the topic.
• Disclosures are listed where the full text of the paper was available to us; a study not listed here is not a statement that its authors have no ties.

Safety and Cautions

This is general information, not medical advice. Lean or fat-free mass loss is a recurring but population-dependent signal, not a universal one: it showed up in a weight-focused trial and did not in a muscle-gain-focused trial, so state both rather than only one (Lowe et al., 2020 [2]; Gavanda et al., 2026 [5]). Most of this specific-schedule literature is short, three to twelve weeks in most of the trials reviewed here, though two trials followed participants for 12 months: a direct calorie-restriction comparison (Liu et al., 2022) [11], and a 12-month follow-up of the window-timing trial discussed above, where early and late time-restricted eating still showed significantly more weight loss maintained than usual care, though only 26% of participants reported practicing time-restricted eating at all during that follow-up year (Camacho-Cardenosa et al., 2026) [24]; shorter-trial results should not be read as durable or universal, and even this longer follow-up shows adherence dropping off.

A 2025 observational analysis of NHANES 2003-2018 data linked to the National Death Index found that, among 19,831 US adults, an average daily eating window under 8 hours was associated with a higher cardiovascular-death hazard than a 12-to-14-hour window (hazard ratio 2.35, 95% CI 1.39-3.98), but not with all-cause or cancer mortality; eating windows were averaged from two days of self-reported 24-hour dietary recall, and median follow-up was 8.1 years (Chen et al., 2025) [17]. An earlier version of this analysis was first presented as a preliminary conference abstract at an American Heart Association scientific meeting in 2024, before journal publication in 2025. This is an association from an observational cohort, not a randomized trial, so it cannot show that the short eating window caused the deaths; the study's own authors write that further research is required to understand whether the risk is due to the short eating duration itself or to residual confounding from factors that contribute to it. A 2025 perspective article discussed possible mechanisms, including lean-mass loss, circadian misalignment, and poor dietary choices linked to reward-based eating; it is a commentary rather than a new study, and its authors present these as potential effects, not established causes (Eliopoulos et al., 2025) [18].

A prespecified secondary analysis of the same 12-week trial of early, late, and self-selected 8-hour windows described above (Dote-Montero et al., 2026) [8] found no significant difference between any of the three time-restricted-eating schedules and usual care in sleep (measured by wrist accelerometry and a sleep-quality questionnaire), mood, or quality of life; for the early window versus usual care, for example, the mean difference was 0.2 hours in total sleep time (95% CI -0.2 to 0.6) and 0.2 points in depression score (95% CI -1.0 to 1.3) (Clavero-Jimeno et al., 2025) [19]. Comparing the three schedules with each other, the paper reports one statistically significant difference, in a quality-of-life bodily-pain score between the early and self-selected windows (mean difference -13.2 points, 95% CI -25.6 to -0.7); the paper's text describes this as reduced bodily pain with the early window, while its table states that higher scores on that survey reflect better quality of life, so the paper leaves the direction unclear. The authors note that the trial's sample size was calculated for its main outcome, visceral fat, so this analysis may not be well powered to detect small changes in sleep, mood, or quality of life, and that participants had only mildly impaired sleep quality at the start; a non-significant result here is not proof that time-restricted eating has no effect on these outcomes. The trial's primary report lists no serious adverse events and five mild ones, one case each of hypoglycemia and thrombosis in the early window, migraine in the late window, and headache and colitis in the self-selected window, after which those participants chose to leave the study (Dote-Montero et al., 2025) [23].

Two other secondary analyses, of two different trials, point in different directions on sleep, and neither settles the question. In the early-window trial discussed under Weight and Fat Mass, the 15 participants who adhered to the schedule at least 5 days a week self-reported sleeping about 30 minutes less (p=0.03) and taking about 7 minutes longer to fall asleep (p=0.04) than 21 adherent controls, with no difference in self-rated sleep quality, alongside the weight and metabolic improvements described there; sleep was measured by questionnaire, and the authors themselves call for objective measures in future studies (Steger et al., 2023) [21]. In the separate 10-hour-window trial also discussed under Weight and Fat Mass (Maruthur et al., 2024) [13], a post hoc analysis of 38 of its 41 participants (mean age 60, 92% women, 92% Black) using 7-day wrist actigraphy, an objective measurement rather than self-report, found the opposite: time-restricted eating increased total sleep time by 55 minutes (p=0.03) and shifted the midpoint of sleep about 44 minutes earlier (p=0.01), with sleep onset also moving earlier, from a median of 12:22 am to 11:52 pm (p=0.03), compared with the usual eating pattern, and no difference in wake time or sleep continuity (Duan et al., 2025) [22]. The two trials differ in schedule (an early 8-hour window with calorie restriction in both arms versus a calorie-matched 10-hour window), population, and how they measured sleep, self-report versus actigraphy, so the two results cannot be directly reconciled and neither should be read as overriding the other.

Talk to a clinician before starting if pregnant, under 18, managing diabetes or blood pressure with medication, with a history of an eating disorder, or receiving cancer treatment.