The sharpest argument against water fasting is that it costs you muscle — and that concern is legitimate. It does. The answer offered, that dry fasting protects lean tissue instead, is inverted on every mechanism it names. And the real protein-sparing adaptation, the one that genuinely exists, was measured in exactly the fast being argued against.
Start by granting the premise, because it’s the strongest thing in the argument and it deserves better than a reflex. Extended water fasting does cost lean mass. That’s real, measurable, and anyone who tells you otherwise is selling something. Muscle is a chain variable — more predicts better, at essentially every level of the epidemiology — so protecting it during a fast is a completely reasonable thing to want. The question is good. What follows is an answer that gets every mechanism the wrong way round, and then recommends against the only protocol the evidence was ever collected in.
The claim is that dry fasting avoids gluconeogenesis — you don’t break down muscle for glucogenic amino acids, so lean tissue is protected. Now look at what your body is actually doing when the water runs out.
Muscle is roughly 75% water. Fat is roughly 10%. Once glycogen and its bound water are spent — that’s your first day — muscle is, by a wide margin, the best water reservoir you have left. And a dehydrated body does what any water-starved system does: it goes to the reservoir.
Two mechanisms get named, and this is where it gets instructive — because both are real, and neither does what it’s being asked to do.
Growth hormone. Fasting genuinely raises GH, and that genuinely contributes to protein sparing. Entirely true. But it rises on a water fast — that’s where it was measured. GH secretion is driven by ghrelin, hypoglycaemia, and somatostatin withdrawal. Nothing in that pathway reads hydration status. There’s no evidence dry raises it further and no proposed route by which it could.
PKA. This one isn’t subtle — it’s simply the wrong pathway. PKA sits downstream of catecholamines via cAMP, and its job in a fast is phosphorylating hormone-sensitive lipase. That is lipolysis — fat mobilisation. It is not a muscle-protection system, and it isn’t hydration-gated either. The acronym is doing rhetorical work its biology doesn’t support.
Here’s what makes this specimen worth the effort: the phenomenon he’s describing exists. Prolonged fasting really does become progressively protein-sparing. It’s classical physiology, it was characterised decades ago, and it has a name: ketoadaptation.
The mechanism is clean. Early in a fast the brain demands a lot of glucose, gluconeogenesis runs hard, and amino acids get pulled from muscle to supply it — urinary nitrogen excretion is high. As ketone bodies rise, they progressively replace glucose as the brain’s dominant fuel. Glucose demand falls, the draw on amino acids falls with it, and protein breakdown drops substantially. It takes days to develop.
The input is ketone availability. Not water status. And every measurement of it was taken in water-fasted humans.
Now a genuinely enjoyable problem. Suppose you wanted to test “dry fasting preserves lean mass.” How would you measure it?
BIA estimates lean mass from total body water — impedance changes with hydration, so a dehydrated reading is meaningless. DXA assumes lean soft tissue is a fixed ~73% water; dehydrate someone and the assumption is violated. Both standard methods route through the exact variable the intervention manipulates.
And the honest note: the artifact doesn’t even favour the claim. Both methods would report less lean mass in a dehydrated person, not more. Nobody’s being fooled in the flattering direction — the measurement is simply invalid.
The valid method is nitrogen balance — urinary urea nitrogen, measuring protein breakdown directly, exactly as the classical starvation literature did. Which requires urine. From someone who isn’t drinking.
The supporting evidence offered is a 350-pound water-fasting advocate. Set aside that it’s n=1: the causation is inverted. Water fasting didn’t make him 350 pounds. That’s an observation about a person, deployed as a verdict on a method — and it arrives attached to “I don’t recommend it to anybody,” which is a very large conclusion resting on a very small foundation.
No. It’s the only version with an evidence base at all — and that isn’t a preference, it’s an inventory. The stem-cell and regeneration work, the autophagy literature, the fasting-mimicking-diet trials, the classical therapeutic fasting research: all of it is water fasting or FMD. When a dry-fasting advocate invokes stem cells, regeneration, or protein sparing, every result being cited was produced in the protocol being dismissed.
And the muscle problem — the real one — has a real answer, which you’ve probably already worked out: shorter fasts, adequate protein on refeed days, and resistance training. A 48-hour fast with a five-day feeding window and decent protein is a coherent duty cycle that gets the fasting switch and defends the chain variable. Dehydration adds nothing to that structure except risk.
Some versions of this advice now reach five to seven days dry. That is a different category of claim from everything above. The rough physiological limit without water is around three days, and it varies with temperature, renal function, and what you did the day before — none of which you fully control. Every earlier version of this argument was at least about something survivable. Note the trajectory across sources: 12 hours → 16–18 → 24 → 48 → 72 → 5–7 days. The evidence never improved. Only the dose did.
The concern is legitimate — extended water fasting really does cost lean mass. But the proposed fix is inverted on every mechanism: muscle is ~75% water and fat is ~10%, so once glycogen’s bound water is spent, muscle is the best reservoir left — dehydration recruits it rather than sparing it. Gluconeogenesis doesn’t stop either, because ketones cover most of the brain’s fuel but not all of it. The pathways named are real and misassigned: GH does rise in fasting — on a water fast, driven by ghrelin, hypoglycaemia and somatostatin withdrawal, with no hydration input — and PKA drives lipolysis via hormone-sensitive lipase, not muscle protection. The genuine adaptation is ketoadaptation: ketones rise, glucose demand falls, urinary nitrogen drops — driven by ketone availability, measured in water-fasted humans. And the claim is unfalsifiable in practice: BIA measures body water, DXA assumes a fixed 73% hydration, and the one valid method — nitrogen balance — requires urine from someone who isn’t drinking. So no, water fasting isn’t a waste of time: it is the only version any of the cited evidence was collected in. The real muscle answer is shorter fasts, protein on refeed, and resistance training.
This article is for educational purposes only and is not medical advice, diagnosis, or treatment, and does not recommend any fasting protocol for any individual. It examines the internal logic of publicly circulated claims, not the practice of any named individual or organisation. Extended fasting of any kind is not appropriate for everyone and can be dangerous in the context of certain medical conditions, kidney disease, kidney stones, diabetes, pregnancy, older age, eating-disorder history, or many medications — including GLP-1 receptor agonists, which blunt thirst as well as appetite and carry a documented acute kidney injury signal driven by volume depletion. Extended dry fasting is potentially lethal: survival without water is typically measured in days and varies with temperature, activity, and renal function, and dehydration causes acute kidney injury, electrolyte disturbance, arrhythmia, and death without reliably signalling itself through how you feel. Physiological figures cited are representative population values that vary between individuals. Anyone considering any extended fast should do so only under the supervision of a qualified physician who can monitor renal function and electrolytes.
This lesson relates to these health systems — health works as a connected system, not isolated topics.
Prerequisite: Eight to Ten Percent: The Number That Refutes Its Own Claim