Why do we yearn for the explosive gustatory delight of sugar? Neural feedback loops, sensory pleasures and environmental factors like poor sleep all amplify our desire for a sugar rush. But new research suggests some of us—much more than others—may also be genetically attuned to crave such sweet sustenance.

An international team scoured the genes of more than 6,500 Danish people taking part in a large health study on heart disease. They found those who harbored one of two particular variants of the FGF21 gene were roughly 20 percent more likely to enjoy and seek out sugary substances. “This study gives us insight into the molecular basis of the sweet tooth—that’s probably the heart of it for me: Why do you have a sweet tooth at a biological level?” says Matthew Gillum, co-lead author and a metabolism researcher at the University of Copenhagen’s Novo Nordisk Foundation Center for Basic Metabolic Research. The findings were published Tuesday in Cell Metabolism.

FGF21 provides the instructions for making a hormone by the same name that is associated with food regulation in rodents and nonhuman primates. Yet this new work suggests FGF21 may modulate some appetites in humans, too. What’s more, this work also suggests the liver—the organ that secretes the FGF21 hormone and controls insulin resistance—may play a larger role in snack management than previously known as it produces this hormone and communicates with the brain.

To investigate a possible FGF21–food preference link, researchers combed through thousands of volunteers’ reports about their dietary preferences, alongside the results of their cholesterol and blood sugar tests. The team also genotyped the participants’ FGF21 genes, and found subjects who reported strong sugar preferences and consumed more of it were much more likely to have one of the two specific FGF21 variants. The same FGF21 alleles—gene versions—were also associated with increased consumption of other problematic substances: The researchers found an additional, weaker link between those gene variants and a higher prevalence of alcohol consumption and daily smoking.

The paper does not pinpoint exactly what neural pathway could be at work to explain those tantalizing relationships. But one area of the brain called the nucleus accumbens is widely considered ground zero for rewards, craving and addiction—and it is possible that cravings for sugar as well as these other substances could converge there, says David Ludwig, a professor at Boston Children’s Hospital who specializes in nutrition and obesity and was not involved with the new study. Because the Cell Metabolism work is observational, however, the FGF21–sweet tooth relationship is “more in the range of hypothesis-generating than definitive,” Ludwig says. He believes it is impossible to know if the subjects with or without these gene variants might differ in other, important ways.

Although nutrition experts have previously identified other internal factors that help control our lust for food, some elements that mediate this process and send signals about sucrose satiety have remained mysterious.

Gillum now thinks FGF21, like leptin—another hormone that regulates appetite—may suppress the neural response to rewards, both in terms of the desire to seek them and consume them“We are still working on why the liver would have evolved mechanisms to do this, but hypothesize it may be to limit excess sugar consumption—either to promote diet diversification or to prevent the [problematic] effects of excess sugar intake,” Gillum says.

There are many unanswered questions in the paper, such as why people with these gene variants and higher sugar consumption tended to have lower body mass indexes (BMIs) than their nonvariant counterparts—a surprising finding, considering the well-established relationship between higher sugar intake and obesity. “Exploring if there is some genetic determination at work in humans’ food taste preference can help guide us to understanding which populations may be especially at risk,” Ludwig says.

Next, Gillum hopes to launch further genetic studies into the effect of these variants on body weight and type 2 diabetes, among other questions. “In mice deficient in FGF21, what we’ve seen is that they eat basically twice as much sucrose as those with them,” he says. “We want to look at people who are completely deficient in FGF21 and answer, Will they be alcohol or sugar superfreaks?”