Intermittent fasting
The diet that became a protocol
Description
By the late 2010s, intermittent fasting had become one of the dominant dietary protocols in Silicon Valley. Engineers at Google compared their feeding windows. Investors discussed their fasting schedules at conferences. Jack Dorsey, then CEO of Twitter, told a podcast he ate one meal a day and felt clearer for it. The protocol spread through tech, then into broader cultural conversation, then into mainstream nutrition writing. By 2020, surveys suggested intermittent fasting was the most popular diet in the United States, ahead of low-carb, paleo, and Mediterranean approaches. The framing had shifted from diet, with its associations of restriction and willpower, to protocol, with its associations of optimization and engineering.
Behind the cultural moment was a body of laboratory research that had accumulated through the 2000s and 2010s, much of it associated with a single neuroscientist — Mark Mattson at the National Institute on Aging. Mattson's work in mice had shown that periodic fasting could trigger autophagy, the cellular self-cleaning process that clears damaged proteins and organelles. Fasting could shift metabolism toward ketone production, modulate insulin sensitivity, and produce neuroprotective effects in animal models of neurodegeneration. The mechanistic story was elegant. The translation to humans was less so.
By the early 2020s, the gap between mouse studies and well-designed human trials of intermittent fasting had become the central tension in the field. Trials in human populations comparing intermittent fasting protocols against straightforward caloric restriction generally found similar weight loss outcomes, with no clear advantage for the timed-eating approach. The metabolic benefits that had appeared striking in rodents proved harder to replicate cleanly in humans. The protocol that had launched a thousand productivity Twitter threads turned out, on close examination, to be a moderately effective form of caloric restriction with some intriguing but not definitive ancillary benefits.
The question we're asking: what does the science actually show about intermittent fasting, and how did the gap between mechanism and outcome get covered over?
What we'll see: the religious origins, Mattson's research program, the human trial data, and what the protocol culture got right and wrong.
Table of contents
01Old practice, new framing
Periodic fasting is older than recorded medicine. Most major religious traditions — Islam during Ramadan, the Christian Lenten fasts, Jewish fasting on Yom Kippur, Hindu fasting on specific days, Buddhist monastic eating windows — incorporate some form of structured abstinence from food. The biological observation that humans can go without food for substantial periods is not a modern discovery. What is modern is the laboratory-backed framing of fasting as a metabolic intervention rather than a spiritual practice.
Clinical interest in fasting in the twentieth century focused mostly on weight loss in obese patients. Therapeutic fasts of several days, supervised in hospital settings, were used through the 1960s and 1970s for severe obesity. The approach fell out of favor as cases of fasting-related cardiac complications accumulated and as rebound weight gain after refeeding made the long-term benefit unclear. The general medical view by the 1990s was that gradual caloric reduction with sustained dietary change was the better path.
02Mattson and the mechanism
Mark Mattson, who led the Laboratory of Neurosciences at the National Institute on Aging from 2000 to 2019, is the figure most associated with the modern scientific case for intermittent fasting. His research program focused on the neurobiology of fasting, exercise, and caloric restriction, and on how these inputs affected the brain's resilience to neurodegenerative disease. The mechanistic story rested on several findings, most of them in mice.
During fasting, the body shifts from glucose-based metabolism to fat-based metabolism, producing ketone bodies — beta-hydroxybutyrate primarily that the brain can use as fuel. Ketone bodies appear to have signaling effects beyond their role as energy substrates: they activate certain transcription factors, modulate inflammation, and may stimulate the production of brain-derived neurotrophic factor, a protein involved in neuronal survival and plasticity. In mouse models of Alzheimer's, Parkinson's, and stroke, fasting protocols have produced protective effects on neuronal function and survival.
03The human trial gap
Several intermittent fasting protocols have been tested in randomized trials in humans. The most studied is time-restricted eating, typically a 16:8 or 14:10 protocol, where eating is confined to an 8-10 hour daily window. Alternate-day fasting and 5:2 fasting (five days of normal eating, two days of severe restriction) have also been studied. The trials generally last 8 to 12 weeks, with some longer follow-up studies, and compare the protocols against either no intervention or matched-calorie continuous restriction.
The most consistent finding across well-designed trials is that intermittent fasting produces weight loss similar to straightforward caloric restriction. A 2020 randomized trial in JAMA Internal Medicine, led by Ethan Weiss at UCSF, compared 16:8 time-restricted eating against unrestricted eating in 116 overweight adults over 12 weeks. The difference between groups was modest, and the fasting group also lost lean body mass at higher rates than expected. A 2022 trial in the New England Journal of Medicine, conducted in China with 139 obese adults over 12 months, found similar outcomes in time-restricted and caloric-restriction groups.
04From protocol to pattern
The Silicon Valley uptake of intermittent fasting shaped the cultural framing of the protocol more than the trial data did. The tech industry has a particular relationship with self-experimentation, biometric tracking, and the idea that personal practices can be optimized through data. Intermittent fasting fit this register cleanly. It had a precise protocol, a measurable variable, a tractable feedback loop, and a mechanistic story that mapped onto engineering intuitions. The diet became a protocol because the language of protocol was more comfortable for the tech audience than the language of dieting.
The cultural framing also smoothed over the gap between rodent research and human evidence. Mattson's mouse studies, autophagy, mTOR inhibition these became part of the popular vocabulary in ways that the more pedestrian human trial data did not. The result was a public conversation about intermittent fasting that emphasized the mechanistic upside and underemphasized the modest outcomes in human trials. People who tried the protocol and found it sustainable lost some weight and felt better, both of which are real. Whether what they were experiencing was specifically fasting biology or just sustained caloric restriction is hard to disentangle from individual experience.
05Conclusion
Intermittent fasting is a useful case study in how a research program rooted in animal models can become a mass cultural protocol on the strength of mechanism more than outcome. The animal data on caloric restriction, autophagy, and ketone signaling is real and interesting. The mouse-to-human translation in metabolism is genuinely uncertain and has produced disappointments before. The trial data on intermittent fasting in humans shows benefits comparable to other forms of caloric restriction, with no clear additional advantage from the timing alone. None of this is grounds to dismiss the practice for people who find it sustainable. It is grounds to be cautious about the more enthusiastic claims.