Intermittent fasting triggers autophagy

In animals, going without food reliably switches on autophagy, the process cells use to break down and recycle their own worn or unneeded parts. In people, the evidence is a handful of small studies, and their results are mixed. The largest, a six-month randomized trial in 121 adults with obesity, found no significant rise in autophagy within the fasting group itself; autophagy in blood immune cells was higher in the fasting group than in a standard-care group at six months only in an after-the-fact comparison, which the authors say may be partly because autophagy tended to fall in the standard-care group. Studies of Ramadan fasting and of shorter daily eating windows mostly report higher activity of autophagy-related genes or proteins in blood, but not every marker moved, and the one Ramadan study with a non-fasting comparison group found some markers higher and others lower.

Blood markers are not the same as autophagy taking place in organs. The only human tissue studied here is thigh muscle, and it did not show a clear rise: after eight weeks of 24-hour fasts three days a week, the authors of a trial in 50 women describe autophagy markers in muscle as reduced, which they attribute to weight loss, and after a single 36-hour fast, autophagy markers fell in untrained men and did not change in trained men, and after 72 hours the markers pointed in opposite directions. No human study here measured liver or brain, or any effect on health or aging. For people taking glipizide, a sulfonylurea diabetes drug, the US prescribing information states that low blood sugar is more likely when calorie intake is deficient.

The six-month trial (Bensalem 2025) was run within a larger Australian trial in adults at increased risk of type 2 diabetes. The fasting group ate 30 percent of their energy needs between 8am and noon and then fasted for about 20 hours, on three non-consecutive days a week, eating freely on other days. The comparison groups cut daily intake to 70 percent of energy needs or received a weight-loss booklet. Autophagy was measured as flux, the rate of the process, using the autophagy protein LC3B-II in blood immune cells. No group differed at two months, daily calorie cutting did not differ from standard care at six months, and the difference between fasting and standard care came from a post hoc test on an outcome the trial registry lists outside its primary and secondary outcomes.

The four-day study (Jamshed 2019) was a randomized crossover in which the same 11 adults with overweight followed each eating schedule in turn. Before breakfast, the early schedule raised expression of the autophagy gene LC3A in blood cells, along with ketones and expression of the gene SIRT1; in the evening it also raised expression of MTOR, a nutrient-sensing gene that regulates cell growth. Autophagy genes were not among the outcomes the study registered in advance. An eight-week randomized trial in 48 adults with overweight or obesity and fatty liver disease (Ozlu Karahan 2025) is the only study here that used the common 16:8 schedule. Both groups ate the same calorie-reduced diet, and one group also ate only within an eight-hour window each day. Blood levels of the autophagy protein ATG5 rose in the 16:8 group and not in the diet-only group, while a second autophagy protein, Beclin-1, did not change in either group; the paper reports no direct comparison between the groups on these proteins, and the trial was registered after it ended.

Three studies followed Ramadan fasting, from dawn to sunset without food or drink for about a month. In 51 adults with overweight or obesity (Bou Malhab 2025), blood expression of the autophagy genes LAMP2, LC3B and ATG5 was higher at the end of Ramadan than before it, with no non-fasting comparison group. In 25 healthy young men in Germany who fasted 17 to 19 hours a day through Ramadan 2015, the first report (Erlangga 2023) found expression of the autophagy genes ATG5, ULK1 and BECN1, the gene for Beclin-1, raised two weeks in and lower again a week after the fast ended. A 2026 re-analysis of the same men (Erlangga 2026) found ULK1 and ATG5 higher after about 30 days than a week before Ramadan, partly returning toward baseline a week after it ended; the authors describe it as exploratory and not prespecified. The third study (Dastghaib 2025) sampled 24 adults who had fasted and 26 who had not, once, the day after Ramadan: blood levels of the autophagy proteins Beclin-1 and LC3 did not differ, and in blood immune cells Beclin-1 gene activity was about 2.3 times higher in the fasting group while LC3 and p62 gene activity were about 15 percent lower.

The tissue evidence comes from thigh muscle biopsies and from mice. In a trial of 50 women with overweight or obesity (Chaudhary 2022), all fasted for 24 hours on three non-consecutive days a week for eight weeks and were randomly assigned to eat 70 or 100 percent of their energy needs overall. After an overnight fast at eight weeks, activity of the genes for Beclin-1, p62 and LAMP2 in thigh muscle had fallen in the 70 percent group, and a 24-hour fast raised activity of the p62 gene; the authors attribute the lower markers to weight loss. In mice on the same kind of schedule, fasting raised autophagy markers in the liver but not in muscle. In 13 young men who fasted for 36 hours (Dethlefsen 2018), the autophagy markers LC3 and p62 fell in the seven untrained men and did not change in the six trained men, and the authors found no sign that autophagy had been switched on. In eight young men who fasted for 72 hours (Vendelbo 2014), the autophagosome marker LC3B-II rose by about 30 percent, but p62, which usually falls when autophagy speeds up, rose by about 10 percent, so the authors could not say whether the process had increased. In mice engineered so that autophagosomes glow under a microscope, 24 hours without food produced a clear rise in autophagosomes in skeletal muscle and a slight rise in the liver, where levels were almost back to baseline by 48 hours, and the response differed from organ to organ (Mizushima 2004). No source here reports how long a fast has to last to raise autophagy in any human tissue, so the mouse timings are not a human figure.

All the human studies enrolled adults. The four-day study excluded people with diabetes, people using diabetes medicines, and anyone pregnant or breastfeeding; the 16:8 trial excluded anyone pregnant or breastfeeding and people using insulin, because of the risk of low blood sugar; the 51-person Ramadan study excluded people with diabetes; and the single-fast muscle studies enrolled only healthy young men, while the eight-week muscle trial enrolled women with overweight or obesity, average age 51, whose exclusions could not be read. For a reader on glucose-lowering medication, the relevant information here comes from the glipizide labeling rather than from any autophagy study.