What the Research Shows About Fasting

Weight Loss and Metabolic Health

Fact 1. Pooled across 99 randomized trials in 6,582 adults, intermittent fasting and continuous calorie restriction both reduce body weight versus an unrestricted diet, with alternate-day fasting showing the largest edge among intermittent-fasting variants (Semnani-Azad et al., 2025) [1]; that alternate-day-fasting advantage over calorie restriction (-1.29 kg, moderate certainty) appeared in trials shorter than 24 weeks, while longer trials showed weight benefit only against an unrestricted diet.

Fact 2. In people with type 2 diabetes specifically, a network meta-analysis of 13 trials found no significant difference between four fasting regimens, including the fasting-mimicking diet, with twice-weekly fasting, not 16:8, ranking best on a composite score (Xiaoyu, Yuxin, and Li, 2024) [2].

Fact 3. A dedicated, year-long head-to-head randomized trial found no advantage for alternate-day fasting over daily calorie restriction on weight loss, maintenance, or adherence, and a worse LDL cholesterol signal by month 12 (Trepanowski et al., 2017) [3].

Fact 4. A 12-week randomized trial in 116 adults found that eating only between noon and 8 pm produced no more weight loss than eating three structured meals a day (between-group difference -0.26 kg, 95% CI -1.30 to 0.78), and, in the 50 participants measured in person, a significant reduction in appendicular lean mass in the time-restricted group (Lowe et al., 2020, the TREAT trial) [4].

Cardiometabolic Markers: Blood Pressure, Cholesterol, Blood Sugar

Fact 5. Across human trials of prolonged water-only fasting lasting 5 to 20 days, blood pressure consistently decreased; glycemic measures improved in people with normal blood sugar but not in people with type 1 or type 2 diabetes; and, in the few trials that examined refeeding, the metabolic improvements were no longer detectable 3 to 4 months after the fast ended, even when weight loss was maintained (Ezpeleta et al., 2024) [5]. See prolonged fasting for the full picture.

Fact 6. A meta-analysis pooling 11 fasting-mimicking-diet studies (761 participants) found reductions in HbA1c, IGF-1, and systolic and diastolic blood pressure, but no significant effect on the other cardiovascular risk factors it checked; in the authors' own sensitivity analyses, the IGF-1 reduction was no longer significant without either a breast-cancer trial or the largest single trial, and the blood-pressure reduction was no longer significant without one other trial (Mohammadzadeh et al., 2025) [6].

Body Composition: The Lean-Mass Caveat

Fact 7. Lean or fat-free mass loss alongside fat loss shows up repeatedly across this literature: in the TREAT trial above (fact 4), in a prolonged water-fasting review finding about two-thirds of weight loss was lean mass (fact 5's source), and in an umbrella review of 130 trials where intermittent fasting was found to be associated with reduced fat-free mass (Lowe et al., 2020 [4]; Ezpeleta et al., 2024 [5]; Patikorn et al., 2021 [7]).

How Certain Is This Evidence

Fact 8. Two reviews that graded the certainty of this evidence rate most of it low to very low: a 2021 Cochrane review (searches to December 2019) found the weight reduction from intermittent fasting, versus ad-libitum eating, was not clinically significant, with no study reporting on mortality, stroke, or heart attack at any follow-up (Allaf et al., 2021) [8]; a separate umbrella review of 11 meta-analyses found only 1% of 104 pooled intermittent-fasting outcome associations were backed by high-quality evidence (Patikorn et al., 2021) [7]. The largest network meta-analysis on this page rated its main alternate-day-fasting comparison moderate certainty (Semnani-Azad et al., 2025) [1]. Read every fact on this page with that certainty level in mind.

Safety Data From Large and Supervised Studies

Fact 9. A one-year observational study of 1,422 people doing a supervised, modified fast (4 to 21 days, a small daily intake of 200 to 250 kcal, not strictly zero-calorie) found significant reductions in weight, waist circumference, and blood pressure, with adverse effects reported in fewer than 1% of participants (Wilhelmi de Toledo et al., 2019) [9]. This study had no control group, and it describes a supervised, modified program; its adverse-event rate should not be generalized to unsupervised water-only fasting.

Brain, Mood, and Cognitive Performance

Fact 10. A systematic review of randomized trials assessing diet's effect on cognition found intermittent fasting "a safe intervention, and no adverse effects were found in the reviewed studies," but based this on only 9 low-quality RCTs (Senderovich, Farahneh, and Waicus, 2023) [10].

Fact 11. Two separate randomized trials examined intermittent fasting's effect on cognitive performance and mood, with different designs: an 8-week trial in women with overweight found no difference between intermittent fasting and calorie restriction on mood, sleep quality, or cognition, and performance was unchanged even during an acute 24-hour fast within the fasting group; in the same trial, intermittent fasting produced more weight and body-fat loss than calorie restriction (Teong et al., 2021) [11]. A 10-day trial found 16-hour daily fasting did not reduce cognitive performance or mood versus a waitlist control, though participants did report feeling less concentrated in the late morning before breaking the fast (Bamberg, Reichenberger, and Blechert, 2026) [12]. See hunger during intermittent fasting for more on that late-morning experience.

Fact 12. An 8-week randomized trial in 40 cognitively intact older adults with insulin resistance found 5:2 intermittent fasting produced greater weight loss than a "healthy living diet," but the two diets had comparable effects on brain-aging markers, a reduced brain-age-gap on MRI, lower brain glucose, and on most blood biomarkers; both diets improved executive function and memory, with fasting benefiting certain cognitive measures more (Kapogiannis et al., 2024) [13]. Read this as "fasting matched a comparator diet on brain markers while producing more weight loss," not as "fasting improves brain health" in a way that implies superiority the trial did not find on most of its own measures.

Autophagy: What's Actually Been Measured in Humans

Fact 13. Autophagy, the cell-recycling process most often cited as fasting's anti-aging mechanism, has been studied in people in only a small number of studies, most using indirect markers in blood cells or muscle, and their results disagree. A 6-month trial in 121 adults with obesity that measured autophagic activity (flux) in blood cells calls its own analysis "exploratory": in a post hoc analysis the intermittent-fasting group differed from a standard-care control group at 6 months but not at 2 months, the authors say this "may be driven partly by a tendency for autophagy to decrease" in the control group, and the fasting group did not change significantly from its own baseline (Bensalem et al., 2025) [14]. An 8-week intermittent-fasting trial in 50 women found muscle autophagy markers went down rather than up, which the authors link to weight loss (Chaudhary et al., 2022) [19]. See autophagy and fasting for the full set of human studies.

Gut Health and the Microbiome

Fact 14. A systematic review of 8 human studies found intermittent fasting associated with gut microbiota richness, diversity, and compositional changes, but with substantial heterogeneity: which bacteria changed varied widely study to study, and the review's own conclusion calls for more research before deciding whether the changes are beneficial (Paukkonen et al., 2024) [15]. See fasting and gut health for the full evidence set.

Fasting in Women: Reproductive Hormones

Fact 15. A systematic review of 6 studies on time-restricted eating and female reproductive hormones found randomized trials in women with general obesity consistently showed no significant change in androgens or other key hormones, while non-randomized studies in women with polycystic ovary syndrome (PCOS) reported beneficial hormone changes that may be confounded by weight loss rather than a direct hormonal effect (Jirapak et al., 2026) [16]. A single small PCOS trial in this space (18 women enrolled, 15 completers; a 1-week baseline then 5 weeks of time-restricted feeding, with no control group) found improved menstrual regularity in 11 of 15 participants alongside weight loss, testosterone, and insulin-resistance improvements, consistent with, not independent evidence against, the systematic review's confounding concern (Li et al., 2021) [17]. Read the systematic review as the more authoritative source on this question; the single small trial should not be read as a separate, stronger finding.

A Novel Schedule Worth Knowing About

Fact 16. A randomized trial of water-only 24-hour fasting performed twice weekly for 4 weeks, then once weekly for 22 weeks (the "WONDERFUL" trial, n=103) found no significant change in LDL cholesterol, its pre-specified primary outcome, but did find significant improvement in insulin resistance (HOMA-IR) and a composite metabolic syndrome score versus ad-libitum controls (Bartholomew et al., 2021) [18]. Of 103 enrolled, 71 completed; weight change, a general cognitive score, and BDNF did not differ significantly from controls. This is a useful fact precisely because it is a negative primary-outcome trial that still found secondary benefit: a clean example of why "fasting lowers cholesterol" is not a safe one-line claim.

What the Evidence Does Not Yet Show

No citation on this page establishes that any fasting schedule reduces mortality, stroke, or heart attack risk; a 2021 Cochrane review (searches to December 2019) found no study reporting on those outcomes at any follow-up (Allaf et al., 2021) [8]. No citation here shows a named schedule beating ordinary calorie restriction in a long-term head-to-head trial: the year-long alternate-day-fasting trial found no advantage (fact 3), and the one trial here that found an advantage ran 8 weeks (Teong et al., 2021) [11]. No citation here ties autophagy measurement to a proven health outcome (fact 13). No citation here resolves whether gut-microbiome changes from fasting are beneficial, harmful, or neutral (fact 14). And no citation here isolates a hormonal benefit for women with PCOS from the confounding effect of weight loss itself (fact 15).

Safety and Cautions

This is general information, not medical advice. Evidence quality across this literature is mostly low to very low certainty, and even well-known claims, such as "fasting lowers cholesterol" or "fasting improves brain health," do not hold up cleanly once a dedicated trial tests them head-to-head. Talk to a clinician before fasting of any kind or length if you are pregnant, under 18, have a history of an eating disorder, or manage diabetes or blood pressure with medication.