How to Start Intermittent Fasting

Choose a Schedule That Fits Your Life

Several named approaches show up across the research: a daily 16:8 or 14:10 eating window, alternate-day fasting, a 5:2 pattern of two reduced-calorie days a week, one-meal-a-day (OMAD) eating, and a complete 24-hour fast once or twice a week. None of these schedules has been shown to be clearly superior to the others for most people. A 2026 Bayesian network meta-analysis of 167 randomized trials in 11,998 adults with overweight, obesity, or metabolic abnormalities (searches to December 2022) found that most intermittent-fasting regimens produced weight loss comparable to continuous calorie restriction with a similar absolute energy restriction, and concluded that weight loss depended mainly on how much energy intake was restricted, regardless of meal timing. In its comparisons matched for energy restriction, the authors report mean weight changes of 11.50 kg (95% CI 10.07 to 12.93, moderate-certainty evidence) for severe continuous calorie restriction, 6.09 kg (95% CI 5.26 to 6.93, moderate certainty) for moderate continuous calorie restriction, and 5.07 kg (95% CI 3.44 to 6.72, high certainty) for alternate-day fasting; the weight-loss effects of the intermittent-fasting regimens, but not of continuous calorie restriction, rebounded after 12 weeks (Wu et al., 2026) [13]. In people with type 2 diabetes specifically, a network meta-analysis of 13 trials found no significant difference in effect between four fasting regimens, with twice-weekly fasting ranking best, not the more commonly recommended 16:8 (Xiaoyu, Yuxin, and Li, 2024) [11]. The practical starting point is a schedule sustainable enough to stick with, begun gradually rather than jumping straight to the longest window.

Comparing Fasting Methods: What the Trials Found

The section above covers schedules within intermittent fasting. This one is about choosing between fasting methods themselves: no single approach, daily time-restricted eating, alternate-day fasting, one meal a day, water-only fasting, prolonged fasting, or the fasting-mimicking diet, has been shown to clearly outperform the others for most people. Sustainability and each method's own safety cautions are a better starting point than a claim that one method wins.

The methods compared here, each with its own page: 16:8 intermittent fasting, alternate-day fasting, one meal a day (OMAD), water fasting, prolonged fasting, and the fasting-mimicking diet.

What Head-to-Head Trials Actually Compared

Among intermittent-fasting variants specifically, two trials that gave a fasting arm and a daily-calorie-restriction arm the same calorie target found no weight-loss advantage for fasting: alternate-day fasting against daily calorie restriction at one year (Trepanowski et al., 2017) [7], and time-restricted eating plus calorie restriction against calorie restriction alone at 12 months (Liu et al., 2022) [12]. A 12-week trial with no calorie targets found time-restricted eating produced no more weight loss than three structured meals a day (Lowe et al., 2020) [8]. Pooled analyses find a small edge for alternate-day fasting over daily calorie restriction, mainly in shorter trials (Semnani-Azad et al., 2025) [1], so for alternate-day fasting specifically the evidence disagrees.

Separately, in a different population and on a different outcome, a network meta-analysis of 13 trials in people with type 2 diabetes compared four fasting regimens, including the fasting-mimicking diet, on glucose control and found no significant difference between the regimens (Xiaoyu, Yuxin, and Li, 2024) [11]. That is a glycemic-control finding in people with type 2 diabetes, not a weight-loss finding, and should be read on its own rather than combined with the trials above into one "all methods are equal" claim.

No study cited on this site compares water-only fasting or prolonged fasting head-to-head against daily intermittent fasting or the fasting-mimicking diet. The calorie-matched results described above apply to the trials that tested it; it does not extend to water or prolonged fasting, which do not have that head-to-head evidence.

Rather than a single best method, the practical differences between these approaches are duration and intensity: a daily eating window, a repeating pattern like alternate-day fasting, a single fast of a day or more like water fasting, or a structured 5-day fasting-mimicking diet program. Each method's own page, linked above, covers who should talk to a clinician before starting that specific method.

What the Research Shows About Weight and Metabolic Health

A 2025 systematic review and network meta-analysis of 99 randomized trials in 6,582 adults found that both intermittent fasting and continuous calorie restriction reduce body weight compared with an unrestricted diet, with alternate-day fasting showing the largest edge among the intermittent-fasting variants studied (Semnani-Azad et al., 2025) [1]. A larger 2026 network meta-analysis of 167 trials, with searches ending earlier (December 2022), found most intermittent-fasting regimens produced weight loss comparable to calorie restriction at a similar energy deficit (Wu et al., 2026) [13]. A systematic review of 27 smaller, mostly short trials (2 to 26 weeks, two running a year) found weight loss of 0.8% to 13.0% of baseline body weight, no serious adverse events, and results broadly similar to calorie restriction; five of those trials included people with type 2 diabetes and all five reported improved glycemic control (Welton et al., 2020) [2]. A separate review of alternate-day fasting, 5:2, and time-restricted eating more broadly reports mild to moderate weight loss (roughly 1% to 8% of baseline) and a consistent reduction in energy intake, and notes intermittent fasting may also improve blood pressure, insulin resistance, and oxidative stress markers, with no evidence in the trials it reviewed of increased disordered eating behavior (Varady et al., 2021) [3].

In people with obesity and type 2 diabetes, a three-month randomized trial of 16:8 fasting three days a week produced a larger weight change than a control group, with improved fasting blood sugar, HbA1c, and lipids (Sukkriang and Buranapin, 2024) [4]. In people with insulin-treated type 2 diabetes, a 12-week supervised trial with dietary counseling and continuous glucose monitoring improved HbA1c and reduced insulin dose without severe hypoglycemia (Obermayer et al., 2023) [5], and a three-week trial in adults with prediabetes found both alternate-day fasting and 16:8 reduced weight, BMI, waist circumference, and blood glucose, with alternate-day fasting outperforming 16:8 on weight and BMI specifically; triglycerides fell only in the 16:8 group, and the authors report no triglyceride or LDL effect for alternate-day fasting (Chair et al., 2022) [6].

For adults 60 and older specifically, a 2026 systematic review with a network meta-analysis of 7 randomized trials, each lasting 6 to 12 weeks, found two protocols reduced body weight and BMI significantly compared with usual diet: 16:8 time-restricted eating (-1.92 kg, 95% CI -2.57 to -1.27; BMI -1.01 kg/m2, 95% CI -1.69 to -0.33) and a twice-weekly daytime fast combined with a 300 to 500 kcal daily calorie reduction, tested in two trials in men aged 50 to 70 (-2.36 kg, 95% CI -2.93 to -1.79; BMI -0.81 kg/m2, 95% CI -1.43 to -0.19). 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 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 (Couto-Alfonso et al., 2026) [14].

That is the supportive side of the evidence. It is not the whole picture: see "Where the studies disagree" in the evidence box below for the findings that complicate a simple "intermittent fasting works" summary, including a one-year trial that found no advantage for alternate-day fasting over calorie restriction, two randomized trials that found no weight-loss benefit from time-restricted eating over calorie restriction specifically, and three separate evidence-quality reviews rating most of this literature as low to very low certainty.

How Long Before You See a Change

Most of the trials behind these findings ran from a few weeks to several months, in adults who were already overweight or managing an existing metabolic condition, so "how long" is better read as "how long in these specific trials" than as a universal timeline. Short trials of three to twelve weeks report measurable changes in weight and glycemic markers (Chair et al., 2022 [6]; Obermayer et al., 2023 [5]), while the longer, one-year head-to-head trial found weight loss plateaued at similar levels for alternate-day fasting and daily calorie restriction by six months (Trepanowski et al., 2017) [7]. A lean- or fat-free-mass change alongside fat loss shows up in more than one trial, though the direction is not consistent across the larger pooled evidence (Lowe et al., 2020 [8]; Patikorn et al., 2021 [10]; Sun et al., 2024 [16]; see Safety and Cautions below for the fuller picture), so "how long before you see a change" is not only a question about the scale. Over a longer timeline, a 12-month trial of time-restricted eating with calorie restriction found no significant difference in weight loss or metabolic risk factors against calorie restriction alone by the end of that year (Liu et al., 2022) [12].

Preparing for Your First Fast

Prepare for a first fast by choosing a modest starting schedule, such as a 12-hour or 16-hour eating window, eating normally rather than overeating the day before, and planning ahead for water, electrolytes, and the meal that will break the fast. Anyone on glucose-lowering medication should set up a fasting plan with a clinician first.

Eat a normal meal the day before rather than treating it as a last chance to eat; this is general practical guidance, not a research finding, and nothing on this page suggests that eating more than usual beforehand offers any advantage.

Hydration and electrolyte needs during the fasting window follow the same principles whether this is a first fast or a fiftieth; see water fasting for what the evidence shows, particularly as fasts run longer.

Hunger, especially in the first several days of a new schedule, is worth expecting rather than being caught off guard by; see hunger during intermittent fasting for what the research says about how it changes over time.

Decide in advance what the first meal back will be, rather than leaving it until the fast ends; see the refeeding guide for how to do that safely. The point here is having a plan ready before starting, not a new how-to.

If you take insulin, a sulfonylurea, or another medication with food-timing requirements, talk to a clinician before you start, not after. This repeats the caution below on purpose, at the point a first-timer needs it most.

Getting Started, Step by Step

Start with a shorter eating window, such as 12:12, and extend it gradually toward a schedule like 16:8 rather than beginning at the longest window attempted. Pay attention to hydration and electrolytes during the fasting window; full detail on that lives on the water fasting page, since the same principles apply at longer durations. Plan the first meal back deliberately rather than as an afterthought; see the refeeding guide for how to do that safely, particularly after a longer fast.

What Breaks a Fast While You're Doing This

Coffee, gum, electrolytes, and medication timing all raise the same practical question: does this end the fasted state? See what breaks a fast for a detailed, cited answer to each.

Who Should Talk to a Clinician First

People on insulin or sulfonylureas should only fast under clinician supervision; the favorable safety result in the insulin-treated trial above came from a protocol with monitoring built in, which is evidence that intermittent fasting can be done safely on insulin, not evidence that it is safe to do unsupervised (Obermayer et al., 2023) [5]. A two-year follow-up of that same trial, in 17 fasting-group and 22 control participants who returned at two years (of 22 and 24 randomized), found the intermittent-fasting group had kept off more weight than the control group (-4.2 vs -0.4 kg from baseline, P = 0.021), with no severe hypoglycemia, defined as needing third-party help or hospital care, reported over the full follow-up; but almost 40% of the fasting group had stopped fasting after the 12-week trial, adherence was not tracked and relied on participants' own reports, and the follow-up's authors state plainly that intermittent fasting in this population "requires constant medical supervision, vigilant self-monitoring and treatment adjustments" (Azhar et al., 2025) [15]. For older adults, the same review's observational studies point in both directions on long daily fasts: 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 and cannot show cause, and the authors call them hypothesis-generating; the review also could not assess interactions with medication, which many older adults take, so older adults should talk to a clinician before starting any schedule (Couto-Alfonso et al., 2026) [14]. 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

Most of the named schedules (16:8, 5:2, alternate-day fasting, OMAD) have not been shown to clearly outperform one another in head-to-head evidence; pick one that fits your routine. This holds even in a specific head-to-head comparison: a network meta-analysis in people with type 2 diabetes found no significant difference between four fasting regimens, including the fasting-mimicking diet (Xiaoyu, Yuxin, and Li, 2024) [11]. The evidence quality behind most intermittent-fasting health claims is rated low to very low certainty by the three separate reviews with the most rigorous methodology in this set (Allaf et al., 2021 [9]; Patikorn et al., 2021 [10]; Garegnani et al., 2026 [17]), so claims of dramatic or guaranteed results should be treated with skepticism.

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 (including the Almond Board of California, the National Honey Board, and the International Nut and Dried Fruit Council); all other authors declared no competing interests. The paper states that the study group's Clinical Practice Guidelines Committee had input on all aspects of the work and that no other funders had a role in the design or conduct of the study.
• Welton et al., 2020 [2]: no funding or conflict-of-interest statement could be retrieved for this page after checking four sources.
• Varady et al., 2021 [3]: funded in part by the National Institute of Diabetes and Digestive and Kidney Diseases. First author Krista A. Varady reported author fees from the Hachette Book Group for her book, The Every Other Day Diet; the other authors stated they were not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of the review.
• Sukkriang and Buranapin, 2024 [4]: funded by the Endocrine Society of Thailand, with the article processing charge funded by Chiang Mai University. The authors reported no conflicts of interest.
• Obermayer et al., 2023 [5]: funded by the Austrian Science Fund, the Austrian Research Promotion Agency, the Integrative Metabolism Research Center Graz, the Styrian Government, the City of Graz, and BioTechMed-Graz. No potential conflicts of interest relevant to the article were reported.
• Chair et al., 2022 [6]: the authors declare no conflicts of interest. The article prints no funding statement.
• Trepanowski et al., 2017 [7]: funded by the National Institutes of Health/National Heart, Lung, and Blood Institute and the National Institute of Diabetes and Digestive and Kidney Diseases. Senior author Krista A. Varady reported receiving an advance for her book, The Every-Other-Day Diet, published by Hachette Book Group. No other disclosures were reported.
• Lowe et al., 2020 [8]: 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.
• Allaf et al., 2021 [9]: all eight review authors stated they have no conflict of interest. The review's "Sources of support" section lists no internal sources of support and reports external support from the NIHR (UK) via Cochrane Infrastructure funding to the Heart Group.
• 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. The article text available to us contains no separate funding statement.
• Xiaoyu, Yuxin, and Li, 2024 [11]: the authors stated the research was conducted without financial support and declared no competing interests.
• Liu et al., 2022 [12]: 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.
• Couto-Alfonso et al., 2026 [14]: "This research received no external funding." Disclosure: "The authors declare no conflicts of interest."
• Azhar et al., 2025 [15]: funded by the Austrian Science Fund (grants KLI851-B and KLI-1076, to senior author Harald Sourij), the same funder already disclosed above for Obermayer et al., 2023 [5]. Disclosure, as printed in the paper: "All authors declare no conflict of interest relevant to this article."
• Sun et al., 2024 [16]: 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 [17]: 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. Some loss of lean or fat-free mass alongside fat loss shows up in this literature, but it is not a one-direction, settled finding. An umbrella review of 11 meta-analyses found reduced fat-free mass a recurring pattern (Patikorn et al., 2021) [10], while a larger 2024 umbrella review of 23 meta-analyses disagrees with itself on the same question: among the pooled results it rated significant at moderate or high certainty, all pooled against people eating their own usual, unrestricted diet, the largest and highest-certainty one (13 studies) found the intermittent-fasting group ended up with higher fat-free mass than the comparison group (+0.98 kg, 95% CI 0.18 to 1.78, high certainty), while three smaller ones (3, 5, and 8 studies, two with substantial heterogeneity) found the opposite, lower fat-free or lean mass than the comparison group by 0.63 to 0.69 kg, moderate certainty (Sun et al., 2024) [16]. A single trial on this page also found lean mass fell in a time-restricted-eating group relative to control (Lowe et al., 2020) [8]. Read together, lean- or fat-free-mass change is mixed and depends on which trials are pooled and against what comparison, not a settled loss. Most of the supportive evidence above comes from short trials in adults who were already overweight or managing a metabolic condition; it should not be read as durable, universal results for everyone. Talk to a clinician before fasting of any length if you are pregnant, under 18, manage diabetes or blood pressure with medication, have a history of an eating disorder, or are receiving cancer treatment.