The hardest part comes on day three. Hunger has ebbed and returned in waves, the fridge has become a study in restraint, and the body is no longer improvising. It is reorganizing.
A team of researchers from Queen Mary University of London’s Precision Healthcare University Research Institute and the Norwegian School of Sport Sciences set out to watch that reorganization unfold in real time. In a tightly supervised experiment, they guided 12 healthy adults through a water-only fast that lasted a full week. Blood was drawn before the fast, every day during it, and again after eating resumed. The samples were then sifted through high-resolution proteomics, a method that tracks thousands of circulating proteins at once, converting a week without calories into a day-by-day map of the body’s priorities.
The resulting paper, published in Nature Metabolism, is less a how-to guide than a field atlas. It captures when systems turn up or down, which signals move together, and how shifts in fuel use are matched by changes far beyond metabolism. As the authors write,
“Surviving long periods without food has shaped human evolution.”
The question they posed was what that survival machinery looks like when viewed with modern tools.
What changes, and when
The study enrolled five women and seven men. Over seven days of complete caloric restriction, participants lost an average of 5.7 kilograms, or about 12.5 pounds. Scale readings, though, were only the outline. The details came from temporal proteomics that quantified roughly 3,000 proteins each day.
Early changes were modest and scattered. The inflection point arrived around day three, when the blood proteome reorganized in a coordinated way that then persisted through the week. To make sense of the complexity, the team grouped proteins by how they rose or fell over time. Nine distinct patterns emerged. More than 1,000 proteins changed significantly, revealing not only a shift in energy supply but a wider recalibration of maintenance and defense.
One surprise was the strength of signals tied to the extracellular matrix, the meshwork that supports cells and orchestrates communication among them. The fasting signature was strongly enriched for matrix proteins from multiple tissues. Among the most striking movers was tenascin R, typically discussed in the context of brain structure. Its swing in the circulation during fasting does not answer what it means for neural function, but it flags a new set of questions about how nutrient scarcity touches the nervous system’s scaffolding.
Hormones follow the fuel
Our fuels shifted along a familiar path. In the first 24 to 48 hours, participants relied mainly on stored carbohydrate. As the fast continued, fat became the dominant source and ketone bodies took center stage. The proteomic timeline lined up with that transition. Leptin, the hormone produced by fat cells that signals energy sufficiency, declined as reserves were tapped. In contrast, leptin receptor levels in blood increased, a combination consistent with heightened sensitivity as the leptin signal weakens.
Other hormone-like proteins moved in directions that fit the changing fuel mix. FGF21 rose, in keeping with a state that favors fat and ketone use. Follistatin increased. Adiponectin tended to decrease. None of these movements by itself is a verdict on benefit or harm. Together, though, they sketch a system that is reorganizing to run on internal stores while protecting key functions.
Beyond weight loss
Body composition scans added texture. Dual-energy X-ray absorptiometry captured changes in both fat mass and lean tissue, a more nuanced picture than body weight alone. Urine measurements showed that nitrogen excretion declined over the week, a sign that the body was adjusting how it handled amino acids and conserving protein as fasting went on.
To gauge potential health implications, the researchers paired their fasting data with large-scale genetic studies of proteins and disease. Using proteogenomic techniques, they examined 212 proteins that changed during fasting against roughly 500 clinical outcomes. The analysis flagged candidates that may be favorable if nudged in the same direction as fasting, as well as proteins whose shifts might carry trade-offs. The paper points, for example, to links between SWAP70 and rheumatoid arthritis, and between HYOU1 and heart disease. These are not prescriptions. They are leads for testing whether specific molecular changes seen during fasting might be mimicked safely without asking people to stop eating for a week.
The clock inside the fast
What stands out is timing. The body does not flip into a single fasting mode on day one. It eases in. Around midweek, a coordinated wave appears in the blood, as multiple networks adjust in concert. Some proteins climb steadily. Others drop and stay low. A few spike at specific points, then drift back toward baseline even while the fast continues. That choreography hints at priorities: secure essential fuel, conserve protein, maintain structure and signaling, then fine-tune as scarcity persists.
This time-resolved map matters because it replaces an either-or picture with a sequence. It suggests that interventions which aim to borrow the upsides of fasting could be timed or targeted, rather than simply prolonged. If matrix remodeling in the brain is part of the fasting response, for example, should future studies test whether short, controlled windows can capture useful plasticity without extended deprivation? If leptin signaling becomes more sensitive as levels fall, can that adaptation be triggered in safer ways in people with metabolic disease?
A careful reading, not a playbook
Seven days without calories is extreme, and in this study it was done under strict medical supervision. Twelve people is a small sample. The participants were healthy to start. Proteins in the blood tell a story about coordinated change, but they are not the whole story. The authors are clear about these limits and about the work ahead. They also note that the multi-organ response appears remarkably consistent across volunteers, at least within the bounds of this experiment.
That mix of caution and clarity is the point. The paper delivers a reference map of what happens when humans stop eating for a week: when switches flip, which systems speak up, how the conversation changes day by day. It documents a fuel handoff from carbohydrate to fat and ketones. It shows immune and structural signals moving with that shift. It outlines where protein conservation begins to take hold. It identifies candidate molecules that might be worth modulating in future trials.
None of this argues for unsupervised water-only fasts. Many people should not attempt them at all, including those with underlying health conditions or on medication. The promise here is not self-denial as a lifestyle. It is the possibility that the body’s fasting playbook can be read, then excerpted safely. With the proteomic timeline on the table, researchers now have a clearer sense of where to look.