Have you ever had the nose completely stuck due to a cold and discover that your favorite dish has become tastelessalmost like sawdust? This illusion of “loss of taste” reveals to us that what we call flavor is a symphony complex, where the taste perceived on the tongue via the taste buds and the perceived aroma retronasally they come together in a common experience. A recent one scientific researchpublished on Nature Communications in September 2025, goes in search of the place where themeeting between taste and smelldiscovering that it happens much earlier than we previously thought: in fact, smells, both sweet and salty, activate the same areas of the brain that light up when we are actually eating that dish.
Smell and taste: how they work together
Usually, the brain handles sensory information through dedicated departmentsand only subsequently the various elements are combined (in what are called “association cortices“) to build a single, unitary and coherent representation of the reality before us. For example, as regards the tastethe signals sent by the tongue through the taste buds reach theinsulaalso known as primary gustatory cortex: let’s imagine it as the center of taste control. Instead, the aromas they odorsespecially those that come up from the nose while we chew (the famous retronasal sense of smell), are initially processed in the piriform cortex.
Now, what was believed until before this study was that the meeting point between taste and smell was in the orbitofrontal cortexand that the two signals joined in this area only after passing on parallel tracks that did not meet before. But, after the investigation carried out in the Karolinska Institutet in Stockholm, it was seen that theinsula it is activated both by the sensations of “sweet” and “salty” coming from the tongue and by that coming from the retronasal sense of smell.
The experiment with vanilla
An aroma, like that of vanillacan cause a sweet sensation even in the absence of taste receptors. If theintegration occurs late, in the orbitofrontal cortex, as explained by theimmediacy of the sweet sensation as soon as we smell the vanilla aroma? To answer this question, researchers created a experiment ingenious using the fMRIa tool that allows you to see brain activations in real time, on healthy subjects who were subjected to congruent tastes and odors (for example, a sweet taste with a sweet smell; an acidic taste with an acidic odor, and so on). After this “familiarization” phase, inside the fMRI the participants received the same sensory components, but this time separately. The experiment, thus constructed, allowed the brain to be directly interrogated to see if the smell, alone, was able to activate the same neural circuits as taste with which it had previously been associated.
The discovery of the taste-olfactory code of the insula
The results have surprised all (otherwise we would not have written this article): the brain activation patterns evoked by retronasal odors they were indistinguishable by the patterns evoked by tastes associates. In particular a sub-section of the insula, the anterior agranularwas activated not only for taste (unlike the granular subsection of the insula, which instead is activated exclusively for taste), but also for the associated odorcrossing the two data to provide a first and quick “holistic flavor identikit“.
The reason why we had not been able to see this peculiar activation until now lies in analysis model. The classic modelcalled mass-univariate general linear modelis based on the simple average amplitude of the BOLD signal in a region, i.e. on the amount of oxygen average in a brain area: if an increase in oxygen is observed in an area, it is assumed that that area is active more than normal, given the increased flow of blood signaled by rising oxygen levels. The type of analysis implemented in this experiment, however, is called multivariate pattern analysis: it is more sensitive to subtle variations and a greater number of details, because it is not based on the average activity of an area (as the previous technique does), but analyzes how the activity is distributed in space within the area. It’s kind of based on the behavior of individual neurons in an area, and not just on average behavior.
This study therefore has the merit of having “promoted” the role of the insula, elevating it from a simple receiver of primary gustatory signals to a crucial and preliminary center for flavor integration, and revealed to us a deeper and more complex integration between these two senses usually considered “minor”.
