In 2014, a team led by Chia-Wei Cheng at the University of Southern California reported something most readers found hard to believe at first. After three or more cycles of two to four day water fasts, mice showed a measurable rebuild of their blood and immune systems, driven by stem cells in the bone marrow waking up and producing fresh white blood cells.1 The same paper included a small pilot in cancer patients who fasted for 72 hours before chemotherapy, and the early markers pointed in the same direction.
The headline is real. The fine print is bigger than the headline. Most of the work was done in mice, the human arm was tiny, and the researchers were upfront that the regeneration showed up only after several rounds of fasting, not a single skipped weekend.
What the 2014 study actually found
The Cheng paper, published in Cell Stem Cell, focused on a question that had nagged longevity researchers for years. Long fasts were known to lower a growth signal called insulin-like growth factor 1, or IGF-1, and to quiet down an enzyme called protein kinase A, or PKA. What no one had shown before was what happened to the body’s blood-forming stem cells when those two signals dropped at the same time.
In mice, the answer was striking. Cycles of 48 to 96 hour fasts, separated by a week of normal eating, pushed hematopoietic stem cells out of their resting state and into active production. After several rounds, older mice and mice whose immune systems had been damaged by chemotherapy showed regenerated white blood cell counts, with new lymphoid and myeloid cells appearing in roughly normal ratios.1 The authors described it as a “regenerative switch” rather than a tune up.
One detail that often gets dropped from social media versions is that the effect was specifically tied to PKA activity falling inside the stem cells themselves. Drug interventions that lowered PKA in the same way reproduced part of the effect, which made the team confident the fasting wasn’t just a coincidence sitting next to the regeneration.
Why low IGF-1 seems to flip a switch
IGF-1 is the body’s growth telegram. When food, especially protein, is plentiful, the liver releases more of it, and tissues respond by building. Pull food away for long enough and IGF-1 drops, sometimes by 30 to 40 percent during a multi day fast.1 A separate 2014 paper from the same USC group, working with Italian and other US collaborators, found that adults under 65 who ate a high-protein diet had four times the cancer mortality risk of low-protein eaters, and IGF-1 was the suspected mechanism connecting the two.5
Stem cells in the bone marrow read low IGF-1 as a signal that the body is in repair mode rather than growth mode. Combined with low PKA activity, the cells appear to clear out damaged or worn out progenitor cells (a process the Cheng team showed in mice using fluorescent labeling) and then push the surviving stem cells to divide and replenish the population.
Whether human stem cells follow the same playbook is still partly an open question. The 2014 pilot in cancer patients suggested they do, but the sample was small, the fasts were closely supervised, and the patients were already in an unusual physiological state because of their disease.
What happened in actual humans
The cancer patient arm of the Cheng study involved fewer than a dozen volunteers fasting for 72 hours before a round of chemotherapy. Their white blood cell counts dropped during the fast, which is what you would expect, and then rebounded after refeeding. Some markers of immune cell stress fell. The authors were careful to call this “preliminary.”
A larger and cleaner human window came in 2016, when Sebastian Di Biase and colleagues at USC tested a fasting-mimicking diet in cancer patients and tumor-bearing mice and reported that fasting cycles increased the activity of cytotoxic T cells, the white blood cells that kill tumor cells.2 A 2017 randomized trial of the same fasting-mimicking diet in 100 healthy adults, led by Min Wei, found that three monthly cycles of a five-day low-calorie regimen lowered IGF-1, fasting glucose, and several aging biomarkers without harming the participants.4
None of these studies measured a freshly rebuilt immune system in humans the way the mouse work did. They measured the upstream signals that drive the rebuild, plus some functional changes in immune cell behavior. That is suggestive, not conclusive, and the researchers say so in plain language.
Mice are not small humans
Six months in a mouse is roughly a quarter of its lifespan. Six months in a human is half a year. That difference matters more than people realize when they read about fasting cycles done “for six months.”
Mouse metabolism also runs hotter and faster than human metabolism. A 48 hour fast in a mouse produces deeper drops in IGF-1, glucose, and ketone-related signaling than the same hourly fast in a person.3 A 2019 review in the New England Journal of Medicine, written by Rafael de Cabo and Mark Mattson, walked through this carefully and concluded that human evidence for cellular regeneration from fasting was promising in narrow windows (cancer therapy support, metabolic syndrome, type 2 diabetes management) but thin everywhere else.3
The de Cabo and Mattson piece is the closest thing to a sober consensus document the field has. It explicitly warns that translating mouse fasting protocols directly to humans, especially older or medically fragile humans, is not safe without supervision.
Who should not fast for several days
Prolonged fasting carries real risk for some people. Pregnant or breastfeeding women, anyone with type 1 diabetes, people on insulin or sulfonylurea medication, people with a history of eating disorders, anyone underweight, children, and people taking medications that require food are routinely excluded from fasting research for good reason.
Even healthy adults can run into low blood pressure, electrolyte issues, or rebound binge eating after a 72 hour fast. The Cheng team’s human volunteers were monitored daily; the Wei trial used a structured fasting-mimicking diet with around 800 calories on day one and 250 to 400 on the next four days, not zero calories.4 That distinction tends to get lost on social media.
If a person is curious about prolonged fasting, the smallest reasonable first step is a conversation with their primary care doctor, blood work that includes glucose and electrolytes, and a plan for breaking the fast that does not involve a large meal. Refeeding syndrome is rare but real, and the people most at risk for it are precisely the people most attracted to long fasts, which is to say those who have already been undereating for weeks before the fast even starts.
Two practical points come up often. First, hydration matters more than people expect. Plain water, salt, and sometimes potassium are the bare minimum during a multi day fast. Second, sleep tends to get worse around night two or three for many fasters, then settles. None of this shows up in a single sentence on Facebook, and none of it should be treated as optional.
It is not just one study
One frequent worry about the Cheng paper is that the findings rest on a single research group. They do not. Several independent lines of work, in different labs and on different protocols, point in compatible directions.
The Wei et al. trial of the fasting-mimicking diet, published in Science Translational Medicine, replicated the IGF-1 drop in 100 healthy adults and showed that the marker stayed low between cycles in many participants.4 The de Cabo and Mattson review pulled together more than a decade of intermittent fasting work and reported consistent reductions in inflammatory markers and improvements in insulin sensitivity across small to mid sized human trials.3 Di Biase and colleagues, working separately, showed that fasting cycles changed the behavior of T cells in tumor environments, which is a different question but a related mechanism.2
The pattern is not “fasting cures everything.” It is more like a tentative arrow. Periodic stretches without food appear to shift several biological levers (IGF-1, insulin, autophagy related signals, possibly PKA) in directions that look beneficial for repair, and the immune system seems to be one of the systems that benefits.
How long does the effect last?
Honest answer: nobody knows yet for humans. In the Cheng mouse work, the regenerated stem cell pool persisted for weeks after the final fasting cycle, but mice were sacrificed at fixed time points, so long term durability was not measured.1 In the Wei human trial, IGF-1 reductions and other metabolic improvements were strongest right after each five day cycle and partly faded between rounds, which is why the protocol was monthly.4
For someone who fasts once and then resumes normal eating, the cellular changes from a single round are likely small and short lived. The repeated cycle structure was central to the original finding.
Common questions about fasting and the immune system
Does skipping breakfast count as fasting in the Cheng sense?
No. The 2014 paper used continuous fasts of 48 to 96 hours. Skipping one or two meals does not produce the same drop in IGF-1 or PKA activity, although shorter daily fasts do have other documented benefits for blood sugar and weight management.
Will fasting wipe out my existing immunity, like vaccines?
There is no evidence that periodic fasting erases established immune memory. The Cheng work focused on stem cell driven replenishment of the cell pool, not on memory B cells or T cells that hold long term protection. Researchers have not reported loss of vaccine response after fasting cycles.
Is fasting the same as the fasting-mimicking diet?
Not quite. The fasting-mimicking diet, developed by Valter Longo’s lab, uses a five day plant-based, low-calorie, low-protein meal plan that produces many of the same hormonal changes as a water fast while keeping people fed. The Wei trial tested this version, not water fasting, and most clinicians who recommend the protocol use it instead of multi day water fasts because the safety profile is better.4
How often do researchers actually do these cycles?
In published human protocols, monthly is the most common cadence, with some studies running quarterly or every six months. The Cheng team’s mouse work used cycles roughly every two weeks because mice live shorter lives and metabolize faster.
Could short term fasting hurt my immune system?
White blood cell counts drop during a long fast and recover during refeeding, which is normal. In otherwise healthy adults, this temporary dip has not been linked to higher infection rates in supervised trials. In medically fragile people, the risk picture is different, which is why supervision matters.
Where this leaves a curious reader
“Fasting rebuilds your immune system” is a fair short summary of what one well designed mouse study and a small human pilot reported in 2014, plus several follow up trials that point in compatible directions. It is not a fair summary of settled medicine. The honest version sounds less viral. Repeated multi day fasting, or a fasting-mimicking diet done in cycles, appears to lower growth signals and prompt some cellular cleanup and renewal in ways that may benefit the immune system over time, especially in people whose systems are already worn down.
Anyone considering this on their own deserves to hear the cautious version too. Talk to a clinician, especially if any medication or chronic condition is in the picture. Track how a first cycle feels before scheduling a second one, write down the numbers that matter (weight, blood pressure, energy, sleep) on day one and day five so the next round is not a guess. And remember the original paper’s punchline: the regeneration came from cycles, not a one off skipped weekend.
Sources
- Cheng CW et al. Prolonged fasting reduces IGF-1/PKA to promote hematopoietic-stem-cell-based regeneration and reverse immunosuppression. Cell Stem Cell, 2014. PubMed: 24905167
- Di Biase S et al. Fasting-Mimicking Diet Reduces HO-1 to Promote T Cell-Mediated Tumor Cytotoxicity. Cancer Cell, 2016. PubMed: 27411588
- de Cabo R, Mattson MP. Effects of Intermittent Fasting on Health, Aging, and Disease. New England Journal of Medicine, 2019. PubMed: 31881139
- Wei M et al. Fasting-mimicking diet and markers/risk factors for aging, diabetes, cancer, and cardiovascular disease. Science Translational Medicine, 2017. PubMed: 28202779
- Levine ME et al. Low protein intake is associated with a major reduction in IGF-1, cancer, and overall mortality in the 65 and younger but not older population. Cell Metabolism, 2014. PubMed: 24606898
