Jason FungMD · nephrologist, Scarborough Health Network, Toronto (CPSO #70179, independent practice since 29 May 2001)
Days, not hours, if you have something to reverse
What it rests onInsulin physiology, his clinic's uncontrolled experience, and two peer-reviewed case reports in which insulin was discontinued in every patient.
Source: thefastingmethod.com ↗
Satchidananda PandaPhD · circadian biologist, Salk Institute (affiliation as of 27 Jul 2026; resignation effective 13 Aug 2026)
It is a window, not a fast: 8 to 10 hours
What it rests onCircadian phase and a melatonin mechanism, plus his own randomised trials, which he then grades down as small and short.
Valter LongoPhD · biogerontologist, Professor of Gerontology, USC
12 to 14 hours daily, and five days a few times a year
What it rests onGallstone risk from reduced bile flow, centenarian eating patterns, and his own trials of the periodic protocol.
Layne NortonPhD · nutrition scientist, coach, and competitive powerlifter
Whichever one you will actually do
What it rests onControlled feeding trials in which total energy intake dominates timing effects.
These are really two different arguments. Panda, Longo and Norton are arguing about the shape of an ordinary day for an ordinary person. Fung is describing a treatment for someone with diagnosed type 2 diabetes, run under clinical supervision with medication adjusted alongside it, and his numbers start where the others stop. Longo's four-week ceiling on 16:8 sits below Fung's weekly practice, and Panda's floor for active people is 12 hours, which Fung's protocol clears before lunch on a fasting day.
So the question to ask yourself is not who is right about how long, but whether you are tuning a healthy day or treating a disease, because those have different answers and different risks.
All four are against grazing across the whole day and all four hold that a shorter, consistent eating pattern reduces intake for most people. Fung and Longo both say their longer protocols need clinical supervision and both publish exclusion criteria.
Fung and Norton converge on the practical answer despite opposite premises: Fung's "the best fast is the one that fits into your lifestyle" is Norton's adherence argument. Nobody in this set recommends fasting for children or in pregnancy.
A head-to-head trial of a multi-day-per-week fasting protocol against a 10 to 12 hour daily window, in people with type 2 diabetes, with medication discontinuation and adverse events as endpoints. It does not exist.
Longo's four-week duration cap and Fung's claim that longer is more powerful are both falsifiable and neither has been tested against the other. Failing that, an adverse-event registry for protocols above 24 hours would settle more than another mechanism argument.
You do, and your call is dated, kept, and yours to change. Follow the people you already trust and what they publish turns into a short daily practice.
Free to sign up and use. We do not ask for a card.