Fasting, Protein and Rapamycin — What the Evidence Actually Says

I had a 68-year-old in last week. She's eating one meal a day because she read that fasting triggers cellular self-cleaning. She's lost seven pounds and she's delighted. I was not.

What were her functional numbers?

Grip strength 18 kilograms, gait speed 0.9 metres per second, couldn't hold a one-legged stance for ten seconds. Pre-sarcopenic by any threshold.

Then she's sacrificing muscle to chase a benefit that isn't well established in humans at her age. The cellular clean-up she's read about is called autophagy — cells degrading and recycling their own damaged components. The evidence that fasting meaningfully induces it in older human tissue, rather than in young mice, is weak. Mechanistic studies, not clinical endpoints.

But the pathway is real, right? mTOR, AMPK — I've seen those in the notes she brought in.

The pathway is real. mTOR is a protein kinase — an enzyme that phosphorylates other proteins — and it's the cell's main fed-state switch. High amino acids, especially leucine, activate it. Active mTOR drives protein synthesis and suppresses autophagy. AMPK runs the opposite direction when energy is low. They're mutually inhibitory, and a cell doing one isn't doing much of the other. That part is solid biochemistry.

So her logic isn't wrong, it's just the wrong patient?

Exactly. Leucine restriction to shift toward autophagy might be a reasonable hypothesis in a metabolically healthy 45-year-old with good muscle mass. In someone already pre-sarcopenic, you're withdrawing the one signal that drives muscle protein synthesis, and muscle loss at her age carries real morbidity — falls, fractures, functional dependence. That evidence is guideline-level. The autophagy benefit in humans is not.

What do I tell her practically?

Distribute protein across meals, target around 1.2 to 1.6 grams per kilogram daily — that's consistent with ESPEN sarcopenia guidelines — and pair it with resistance exercise. If she wants a fasting window, a 12 to 14 hour overnight fast doesn't gut muscle protein synthesis the way one meal a day does. That's a reasonable compromise, though the longevity endpoint remains unproven.

Separately, a colleague — 52, healthy — told me he's already sourced rapamycin. He wants my opinion.

Rapamycin is an mTOR inhibitor, approved for transplant immunosuppression and certain cancers. In multiple animal models it extends lifespan — that's reproducible. In humans, we have no randomised trial with longevity or hard clinical endpoints in healthy people. None.

What about the PEARL trial?

PEARL looked at low-dose rapamycin in healthy middle-aged adults and reported acceptable short-term safety signals and some biomarker changes. It was not powered or designed to show clinical benefit. Biomarkers are not endpoints. Known risks include impaired wound healing, dyslipidaemia, glucose intolerance, and real immunosuppression even at low doses.

Metformin gets mentioned in the same breath. Same situation?

Similar. TAME — Targeting Aging with Metformin — is ongoing and hasn't reported primary endpoints. Metformin activates AMPK and has decades of safety data in type 2 diabetes, but prescribing it to a non-diabetic for longevity is off-label with no trial evidence of benefit yet. Don't dose it, don't recommend it for that indication.

So what's the one thing you'd want me to walk away with?

The biology of growth versus maintenance is real, but the intervention that shifts the balance helpfully depends entirely on where your patient already sits. For most people over 65, muscle protection is the evidence-backed priority. Rapamycin and metformin for longevity are investigational — tell your colleague that sourcing them himself is assuming risk with no proven benefit, and leave it there.