I’ve been doing 16:8 intermittent fasting for years and recently started 48-hour fasts — dropping about three pounds each fast, gaining one or two back, and trending steadily downward. I wanted to understand what the research actually says about what I’m doing to myself, so I worked with Claude (Anthropic’s AI) to produce this series. I set the structure, chose the topics, pushed back on claims that felt hand-wavy, and guided the editorial tone. Claude did the writing and research synthesis. My curiosity driving Claude’s research and prose.
Autophagy
Research brief — what autophagy is, what the evidence actually shows, and where the common claims outrun the science.
What Autophagy Is
Autophagy — from Greek auto (“self”) and phagein (“to eat”) — is the process by which cells degrade and recycle damaged organelles, misfolded proteins, and dysfunctional mitochondria. Think of it as cellular housekeeping: damaged parts get broken down and the raw materials get repurposed for repair and new construction.
Yoshinori Ohsumi won the 2016 Nobel Prize in Physiology or Medicine for discovering the genetic mechanisms of autophagy. Working in baker’s yeast in the early 1990s, he identified the genes essential for the process. The mechanisms are highly conserved across species including humans. (1)
The molecular trigger is well-understood: nutrient deprivation suppresses insulin and mTOR signaling and activates AMPK — all of which are upstream regulators of autophagy. The logical chain from “fasting → reduced insulin/mTOR → autophagy activation” is mechanistically solid. Disrupted autophagy has been linked to Parkinson’s disease, type 2 diabetes, Alzheimer’s, certain cancers, and many infections. (2)
What We Don’t Know in Humans
This is where the common claims outrun the evidence. Most of what people confidently state about fasting and autophagy in humans is extrapolated from animal models.
The measurement problem: You can’t biopsy someone’s liver or brain during a fast to count autophagosomes. The available approaches in living humans:
- Peripheral blood mononuclear cells (PBMCs): Researchers measure the LC3B-II/LC3B-I ratio in blood cells treated with chloroquine ex vivo. This is an indirect proxy — it tells you something is happening in immune cells, not necessarily what’s happening in liver, muscle, or brain tissue.
- LC3 and p62/SQSTM1 in tissue biopsies: More direct but rarely done in fasting studies on healthy humans because it requires invasive sampling.
- Interpretation challenge: Elevated LC3 levels can reflect either enhanced autophagy or impaired autophagic flux (a blocked process, not an active one). The marker alone doesn’t tell you which.
A registered clinical trial (NCT04842864) titled “Time Course for Fasting-induced Autophagy in Humans” exists on ClinicalTrials.gov — the fact that this is still an active research question tells you something about how unsettled the timing is. (3)
A 2022 study by Chaudhary et al. found that intermittent fasting activated markers of autophagy in mouse liver but not in muscle from either mice or humans. Tissue-specific responses mean blanket claims about “autophagy ramping up everywhere at 24 hours” are oversimplified. (4)
What Can Be Said Honestly
The commonly cited “autophagy kicks in at 24–48 hours” is a reasonable estimate based on animal data and indirect human markers, not a precisely measured human threshold. The molecular machinery is real. The upstream triggers (low insulin, low mTOR, active AMPK) are clearly activated during a 48-hour fast. A 48-hour fast almost certainly activates autophagy in at least some human tissues. But “how much” and “exactly when” are not precisely known in humans, and the response varies by tissue type.
The metabolic switch framework (Part 1) places autophagy activation as one of several adaptive cellular stress responses triggered by the shift from glucose to ketone metabolism. (5) It doesn’t happen in isolation — it’s part of a broader cascade that includes DNA repair, protein quality control, and mitochondrial biogenesis.
Evidence strength: Moderate. Mechanism is solid. Direct human measurement is thin. The claim that autophagy is active during a 48-hour fast is well-supported mechanistically but not precisely quantified in humans.
Sources
- Nobel Prize press release (Ohsumi 2016): https://www.nobelprize.org/prizes/medicine/2016/press-release/
- Nobel Prize advanced information — autophagy mechanisms and disease links: https://www.nobelprize.org/prizes/medicine/2016/advanced-information/
- ClinicalTrials.gov — “Time Course for Fasting-induced Autophagy in Humans”: https://clinicaltrials.gov/study/NCT04842864
- Chaudhary et al. 2022, Nutrition — IF activates autophagy in mouse liver but not human/mouse muscle: https://pubmed.ncbi.nlm.nih.gov/35660501/
- de Cabo & Mattson 2019, NEJM — metabolic switching framework: https://pubmed.ncbi.nlm.nih.gov/31881139/