AINEWS
After a few drinks, your liver is steering your cravings
Drinking briefly spikes a liver hormone that once steered us toward protein — and now steers us toward the chip aisle.
Drinking briefly spikes a liver hormone that once steered us toward protein — and now steers us toward the chip aisle.
That post-drinks pull toward chips, pizza, or a late-night kebab is not a willpower failure. It is a hormonal hijack with an evolutionary backstory. Alcohol briefly raises a liver hormone called FGF21, which quiets sweet cravings and turns up the dial on savory and umami — a mechanism experimentally established in prior research and confirmed in a new population-level study from the University of Sydney Charles Perkins Centre published in Obesity Reviews (Grech et al., 2026, DOI: 10.1111/obr.70138; University of Sydney press release; Refractor summary).
FGF21, short for fibroblast growth factor 21, is a signaling molecule the liver releases during fasting, protein restriction, and alcohol metabolism. In population dietary survey data analyzed by the researchers, people ate more savory foods on drinking days than on non-drinking days, and each standard drink was associated with increased savory food intake and reduced sweet food intake. The effect was especially pronounced when the savory foods consumed were low-protein ultra-processed foods. The same surge that helps the body manage the metabolic stress of alcohol also rewires, in the moment, what the brain wants to eat next.
The wiring is old. For most of human evolution, a spike in FGF21 would have followed a stretch without protein and pushed a hunter toward the highest-protein food available: a piece of meat, a fatty organ, an egg. Today the same signal points at engineered salty snacks, fried food, and pizza — the modern stand-ins for nutrient density. The paper calls these "protein decoys": foods engineered to taste like protein without delivering it. The mismatch is not that the hormone is broken. The mismatch is that a protein-seeking signal is now landing in a pantry built from starch, oil, salt, and flavor chemistry.
That distinction matters because it reframes the late-night craving as a design problem rather than a personal one. A liver running ten-thousand-year-old software is being addressed by a food system that has spent decades learning exactly which combinations of salt, fat, and umami compounds drive repeat consumption. Ultra-processed foods are formulated to hit savory and umami receptors at intensities ancestral diets rarely matched, and they are typically low in the very protein the FGF21 signal is hunting for. The result is a craving that gets activated, satisfied briefly, and then reactivated, because the macronutrient the body wanted is missing from the food the body ate.
The study, published in Obesity Reviews, analyzed Australian national dietary survey data using nutritional geometry — a method that examines patterns across multiple dietary dimensions simultaneously. The researchers found that the correlation between umami flavor and protein, reliable in minimally processed food environments, is broken in diets rich in ultra-processed savory foods. "In this way, alcohol may contribute to overeating particularly when ultra-processed, low-protein savory foods are readily available," said senior author Professor David Raubenheimer.
A few caveats belong on the table. The study is observational and population-level rather than a controlled trial; it analyzes dietary survey data, not direct measurements of FGF21 in participants. The core hormone mechanism (FGF21 rises after alcohol, suppresses sweet, increases savory/umami) is described in the paper as building on experimentally established findings. The evolutionary mismatch framing — that the signal once pointed at protein-rich whole foods and now points at engineered snacks — is supported by the paper's "protein decoys" concept, though the evolutionary framing itself is most developed in the authors' commentary. Single-hormone stories also tend to oversimplify, since appetite is shaped by many overlapping signals, and individual responses to alcohol vary.
The takeaway is not a tip sheet. It is a recognition that a familiar craving has a specific biological author, and that the modern food environment is unusually good at answering that craving in the wrong way. Knowing the signal is coming from the liver, and that the food industry has built entire categories around mimicking the foods that signal once pointed to, is the first step in deciding whether to follow it.