The consistent finding across the strongest randomized trials is equivalence. A 2020 randomized trial in JAMA Internal Medicine — the TREAT study — found time-restricted eating produced no significant weight loss versus a regular eating pattern, while a 2022 New England Journal of Medicine trial found weight loss with and without a time-restricted eating window was statistically identical at equal calories, and a 2022 alternate-day-fasting trial in Nature Medicine found fasting roughly matched continuous calorie restriction on weight and insulin sensitivity. None showed a metabolic edge from the fasting schedule itself.
This site publishes information, not medical advice. The results here belong to named trials; anyone on glucose-lowering medication, with a history of eating disorder, or pregnant should not restructure meal timing except with their own clinician, because skipped meals interact with drug dosing.
What did the most rigorous trials measure?
The 2020 TREAT study, led by Ethan Weiss at the University of California, San Francisco, randomized 116 adults to an eight-hour eating window or three structured meals daily for 12 weeks, with no calorie targets, and found no significant difference in weight, though appendicular lean mass showed a possible decrease in the fasting arm, per the report. The 2022 New England Journal of Medicine trial, by Liu and colleagues in Guangzhou, randomized 139 adults with obesity to calorie restriction with or without an eight-hour window and found weight loss of about 6 to 8 kilograms in both arms at 12 months, with no significant difference between them. A third design, alternate-day fasting, was tested in a 2022 Nature Medicine trial and in a large 2017 trial in JAMA Internal Medicine by Krista Varady's group, both finding fasting schedules performed comparably to, not better than, daily calorie restriction.
Do the trials show any advantage for fasting itself?
The honest summary is no, not on the endpoints that matter most. Trials that matched calories across arms found no difference — the 2022 New England Journal of Medicine result is the cleanest, running a full year with a primary weight endpoint. Trials where the fasting arm happened to eat less — as in some early time-restricted-eating studies by Courtney Peterson's group and in the alternate-day-fasting literature — showed weight loss attributable to the calorie deficit, per the authors' own analyses. Reviews pooling the field, including a 2022 Cochrane-network meta-analysis of intermittent fasting versus continuous energy restriction, found similar or trivially different effects on weight, blood pressure, lipids, and glucose, with the commentary literature noting adherence, not schedule, as the variable that separates winners from dropouts.
What the landmark trials found
| Trial and journal | Design | Finding | Year |
|---|---|---|---|
| TREAT, JAMA Internal Medicine | 8-hour window vs 3 meals, 12 weeks, no calorie targets | No significant weight difference | 2020 |
| Liu et al., New England Journal of Medicine | Calorie restriction with vs without 8-hour window, 12 months | Identical weight loss in both arms | 2022 |
| Varady et al., JAMA Internal Medicine | Alternate-day fasting vs daily restriction, 1 year | Comparable, with higher dropout in fasting arm | 2017 |
| Nature Medicine alternate-day trial | Alternate-day fasting vs restriction | Fasting matched restriction on weight and insulin sensitivity | 2022 |
The trials agree on a modest claim: fasting is one way to eat fewer calories, and the calories are the active ingredient.
Related stories: What studies actually show about ultra-processed foods and health outcomes · What the evidence says about the Mediterranean diet, from PREDIMED onward.
What about the early mouse and small-human studies?
The field's enthusiasm traces to rodent work — calorie restriction extended lifespan in mice across decades of studies at multiple institutions — and to small early human trials reporting metabolic improvements. Those findings are real but preliminary by the field's own standards: mice on restriction live longer, yet the human translation has no lifespan evidence and no hard-outcome trials, and early small human studies showed effects that larger, longer trials failed to replicate at matched calories. Circadian-timing research adds one nuance the equivalence trials did not test — earlier eating windows may improve glucose response per short-term studies — but that literature runs days to weeks and measures intermediate markers, not disease.
Who was never in these trials?
The exclusions matter more than in most diet research. All the major trials excluded people with type 1 or insulin-dependent diabetes, pregnant or breastfeeding people, adolescents, and histories of eating disorders — the groups for whom meal skipping carries the largest known risks. Hypoglycemia risk with sulfonylureas and insulin is documented in clinical guidance, and eating-disorder specialists have raised caution about fasting protocols as restriction cues; neither concern is testable by the trials that exist. The results therefore describe relatively healthy adults with obesity or overweight, a boundary most coverage of intermittent fasting drops.
Does any version hold an edge for adherence?
Individual, not average. The trials report wide interpersonal variation — some participants find a compressed eating window easier than counting calories, and others abandon it fast; the 2017 alternate-day trial's higher dropout in the fasting arm quantifies one direction, while other studies found the opposite preference. Meta-analyses comparing adherence rates between schedules find no consistent winner. That places intermittent fasting where most diet research ends up: a legitimate format for achieving a calorie deficit in adults whose clinician has no objection, without demonstrated superiority and without evidence of disease prevention.
When to talk to a clinician
Meal timing is a medication-timing question for anyone taking insulin, sulfonylureas, or blood-pressure drugs with food dependencies; a risk question for anyone with an eating-disorder history; and a Special Medical Question in pregnancy. Those readers need their own clinician before any fasting protocol. For adults without those conditions, the trial record supports a working conclusion: if fasting helps someone sustain a calorie deficit, the trials show it can work; if the schedule is miserable, the same trials show no biological reason to push through.