Why does one study call a product healthy, while another, a few years later, calls it harmful? Such contradictions have long been one of the main problems in nutritional science. Often, the reason lies not in the products themselves, but in the difficulty of separating their influence from the many other factors that determine a person's lifestyle.
Most nutritional knowledge is based on observational studies. Scientists analyze the dietary habits of thousands of people and compare them with health indicators. However, a person who regularly eats vegetables often simultaneously exercises more, smokes less, has a higher level of education, is better off financially, and undergoes preventive check-ups more frequently. Determining how much nutrition contributes versus all other lifestyle characteristics is extremely difficult. This problem is called the influence of confounding factors, or confounding.
Randomized controlled trials are considered the most reliable way to test cause-and-effect relationships. However, they face serious limitations when applied to nutrition. It is very difficult to get people to stick to the same diet for years, and chronic diseases develop over decades. Therefore, in recent years, the method of Mendelian randomization has become increasingly popular.
Its idea is based on the fact that genetic variants are randomly distributed at conception and are practically independent of a person's lifestyle or social status. Thanks to this, genes can act as a kind of natural experiment, allowing us to get closer to understanding the cause-and-effect relationships between dietary habits and diseases.
However, this approach also has limitations. Many genetic variants associated with dietary habits are simultaneously associated with educational level, income, body weight, and other characteristics that themselves affect health. As a result, it becomes difficult to understand what exactly determines the observed effect.
An international research group led by Liang-Daru Hwang from the University of Queensland and the Monell Center has proposed an original solution. Instead of searching for random genetic markers, the scientists focused on genes for taste and olfactory receptors. These directly determine the perception of taste and smell of food, and thus influence food preferences, while being significantly less often associated with a person's socio-economic characteristics.
In a study published in June 2026 in the journal BMC Medicine, data from over 160,000 participants in the UK Biobank project were analyzed. The researchers examined 1214 genetic variants in taste and olfactory receptor genes. After a multi-stage selection process, 24 high-quality genetic instruments related to the preference for 20 different products remained. A variant rs6587467 in the olfactory receptor gene OR2T6, closely linked to a love for onions, stood out particularly. This variant showed almost no correlation with income, education level, or other potential confounding factors, making it a particularly valuable tool for Mendelian randomization.
Using this genetic variant, the researchers obtained preliminary evidence that a genetic predisposition to a greater love for onions is associated with a reduction in systolic blood pressure by approximately 1.3 mmHg, diastolic by approximately 0.7 mmHg, and also with a decrease in the risk of developing type 2 diabetes by approximately 14%. No effect was found on body mass index, blood lipid levels, or the risk of coronary heart disease, suggesting a possible specific mechanism of action rather than a general improvement in health. The authors emphasize that these results require independent confirmation in other populations.
It is important to understand that the main discovery of the work is related not so much to onions themselves, but to a new way of selecting genetic instruments. Onions turned out to be a convenient model for testing the concept. The biological plausibility of the observed association also appears convincing: onions are rich in quercetin and sulfur-containing compounds, for which anti-inflammatory, antioxidant, and vasodilating properties have previously been described. However, it is precisely the new approach to selecting genetic markers that may become the research's main contribution to the development of nutritional science.
The practical significance of this work is particularly evident today, as Australia is revising its national nutrition guidelines for the first time since 2013. An updated version is expected by the end of 2026. The more accurately scientists can distinguish causal relationships from simple statistical coincidences, the more reliable the recommendations that millions of people will subsequently use will be.
Genetics will not replace clinical studies and classical epidemiology, but it can significantly enhance their capabilities. Instead of searching for random statistical coincidences, researchers gain a tool that allows them to get closer to understanding the true cause-and-effect relationships between food and health. This opens the way not only to more convincing nutritional recommendations but also to the future of personalized nutritionally, where the specifics of taste, smell, and genetics will help to more accurately tailor diets for each person.
