Autophagy and fasting.

What the evidence shows

Mizushima and Komatsu's 2011 review (Cell) describes autophagy as the major intracellular degradation system, by which cytoplasmic material is delivered to and broken down in the lysosome, and frames it as central to the renovation of cells and tissues — including the clearance of damaged organelles and protein aggregates. Levine and Kroemer's 2019 review (Cell) catalogues how autophagy genes function across human physiology and disease, with particular attention to neurodegenerative, inflammatory, and cancer contexts. These are the core references establishing that autophagy matters for cellular health.

On the link to fasting specifically, a 2018 review (Bagherniya et al., Ageing Research Reviews) surveyed the literature and concluded that the evidence "overwhelmingly" suggests both fasting and calorie restriction induce autophagy across a wide variety of tissues and organs in response to food deprivation. The de Cabo and Mattson 2019 NEJM review likewise lists increased autophagy among the adaptive responses associated with the fasting metabolic switch.

What's uncertain

The human evidence is largely indirect. Autophagy is difficult to quantify in living people — there is no simple, validated blood test for autophagic flux — so much of the fasting-and-autophagy literature rests on animal models, cell culture, and surrogate markers rather than direct measurement of the process in human tissue.

Fasting is not the only trigger, and more is not obviously better. The cell-biology reviews make clear that autophagy is constitutively active and tightly regulated; exercise and certain drugs also modulate it, and both insufficient and excessive or dysregulated autophagy are implicated in disease. That regulatory complexity argues against treating "more autophagy" as a simple good to be maximised by longer fasts.

Specific clinical benefits remain unproven. The claim that fasting-induced autophagy prevents or reverses particular human diseases is not established by the reviews cited here; the disease links they describe are largely mechanistic or drawn from preclinical work.