Fasting Is Three Interventions
Nine practices, three mechanisms. One changes when you eat, one changes what the body burns, one changes whether you drink. All three work by moving pathways — the same pathways the engine already maps, listed in full below. Tap any of them to see its graded evidence.
I. The clock — when you eat
Food concentrated into part of the day or week. Across randomized trials the high-certainty gains are specific: waist circumference and fat mass fall, triacylglycerols and fasting insulin fall, HDL cholesterol rises — and fat-free mass increases, which is the result ordinary restriction does not deliver.
A daily eating window, the fast carried overnight. Placed early in the day it improved insulin sensitivity, blood pressure and oxidative stress with no weight lost at all — the clearest demonstration anywhere that timing is an active variable in its own right.
The window narrowed to a single daily meal. Studied in lean adults, where metabolic markers and physical performance each respond to the hour the meal is taken — placement is part of the intervention, not a detail of it.
A dawn-to-dusk fast from food and fluid across a lunar month — the largest natural experiment in fasting that exists. Pooled studies show the endocrine panel holds steady: thyroid hormones, testosterone, LH, FSH, prolactin and parathyroid hormone all unchanged, while cortisol follows the shifted day. The hormones are stable; the clock is what moves.
II. The substrate — what the body burns
The unmodified form. As the fast lengthens the body crosses to fat-derived fuel, and the crossing runs through the engine’s own machinery: the low-fuel sensor rising, the growth complex standing down, stored fat released and oxidized.
The one practice here that reaches a fasting fuel state while eating. Carbohydrate is restricted far enough that the liver produces ketone bodies, and the body runs on the substrate an extended fast would have made for it.
A periodic, low-energy, plant-based cycle that produces the fasting response while food continues. It carries the cleanest randomized data on this page: across three monthly cycles in 100 adults, the principal growth factor, systolic pressure, body weight and trunk fat all fell.
III. The water — whether you drink
Abstention from fluid as well as food, studied chiefly in its dawn-to-dusk form. In adults with metabolic syndrome it shifted the proteome of circulating immune cells toward an anti-inflammatory and anti-atherosclerotic profile, with a companion study reporting lower circulating inflammatory cytokines at raised body mass.
The partial form — energy withheld, fluid reduced rather than stopped. Its evidence is read across from the dawn-to-dusk studies, whose shape it shares, and which an umbrella review has since pooled for cardiometabolic and obesity outcomes.
Every pathway fasting moves
44 pathways, each one already carried by the engine and graded on its own evidence. Tap any of them.
What it corrects
The conditions these pathways sit under. Each opens its own pathway map, with the blends built for it.
The growth switch, measured in people
Autophagy is the part everyone names. What switches it on is the part almost nobody does. Nutrient and growth signaling run through mTOR, and the chief upstream input to mTOR is insulin-like growth factor 1. Fasting lowers both ends of that chain, and the human figures are real.
In healthy lean men across a 72-hour fast: total IGF-1 fell by more than 50%, and the free fraction — the part that actually signals — fell by about 75%. Insulin fell by over 70%. IGF-binding protein 1, which sequesters IGF-1 and lowers the free fraction further, rose significantly. Two things the popular account gets backwards, measured in the same study: growth hormone rose rather than fell, and IGFBP-3 did not change at all.
And the limits on that figure
Those numbers come from a 72-hour fast in six evaluable men, within-subject and uncontrolled on the fast-versus-fed contrast. They are not what a daily eating window does. The nearest figure for a shorter, repeatable practice is a five-day fasting-mimicking trial in 100 participants where IGF-1 fell about 24% and was still roughly 15% below baseline a week after normal eating resumed — read in a review rather than in the primary trial, and labeled here as such.
And the step most accounts take next — lower mTOR, therefore longer life — is established in animals, not demonstrated as a human lifespan outcome. No trial of that length exists. We name the mechanism and stop where the evidence stops.
One thing we will not claim
Fasting is popularly credited with switching on cellular self-digestion at a named hour. There is no established method to measure autophagic flux in humans — the markers used as proxies rise both when degradation speeds up and when it is blocked. We name the pathway, because it is real. We do not put an hour on it.
Educational only. Not evaluated by the FDA. Nothing here is a protocol or a recommendation to begin any fast. Anyone pregnant or nursing, managing a medical condition, or taking prescription medication should speak with their own physician first.