[@hubermanlab] Essentials: How Smells Influence Our Hormones, Health & Behavior | Dr. Noam Sobel

Link: https://youtu.be/arMMF0ISRRM

Duration: 34 min

Transcript: Download plain text

Short Summary

Huberman Lab Essentials host Andrew Huberman speaks with Stanford neuroscientist Noam Sobel about chemosensation, olfaction, pheromones, and hidden social perception. Sobel, a professor of neurobiology and ophthalmology at Stanford School of Medicine, explains how the nose and brain encode smell, how body odor drives friendship and aggression, and how his lab is digitizing smell. The episode covers handshakes and sniffing, "click friendships," hexadecanal's sex-dependent effects, emotional tears lowering testosterone, and transmitting violet scent over the internet.

Key Quotes

1. "And there was a remarkable increase in the hand going to the nose after handshake. It's unreal." (00:07:46)
2. "All of this was predicted by the electronic nose. So people who smell more similar to each other think that the other person is more likely to be their friend, is more likely to be a nice person, etc. etc. So we could actually predict friendship using the electronic nose." (00:13:37)
3. "So, you're winning. So, the bottom line of all this is that hexadecanal is the most abundant semi-volatile in baby heads. It's tons of it coming out of baby heads. Well, babies are conducting chemical warfare, right? They're reducing aggression in their fathers or males around them and increasing aggression in their mothers or females around them, and both of those things are good for them." (00:20:13)
4. "And you have uh about a 14% drop in free testosterone. And people who study testosterone, which is not me, but they tell me this is a really strong effect. Like it's it's hard to even pharmacologically get an effect like that that fast." (00:23:06)
5. "we generated olfactory metamers. So, we measured mixtures completely non-overlapping in their molecular structure but they smell exactly the same." (00:27:48)

Detailed Summary

Overview

Andrew Huberman hosts this Huberman Lab Essentials conversation with Dr. Noam Sobel, a Stanford School of Medicine professor of neurobiology and ophthalmology. The episode connects the neural machinery of smell with human chemosignals that shape friendship, aggression, fear, and emotional behavior, then examines efforts to encode, reconstruct, and transmit odors digitally.

• Humans do not simply use smell to identify substances; body odor also supplies hidden social information that can influence how people perceive and respond to one another.
• Sobel’s laboratory treats chemosensation as a bridge between molecular biology and social behavior, asking how odor molecules become neural signals and how those signals affect interpersonal judgments.
• The episode moves from the architecture of the olfactory system to human experiments, animal research, clinical implications, and early work on creating a digital platform for smell.

How Smell Works: From Molecules to Brain

Primary olfaction begins with orthonasal sniffing, which carries airborne molecules into the nose and allows them to reach the olfactory epithelium. Once an odorant is detected, chemical information is translated into neural activity that reaches the olfactory bulb and multiple brain regions involved in perception, emotion, and bodily responses.

• In orthonasal olfaction, airborne odor molecules travel approximately 6 or 7 centimeters up the nose before reaching the olfactory epithelium, where sensory neurons detect them.
• Humans possess roughly 7 million olfactory receptors drawn from about 350 receptor subtypes, making olfactory receptor genes a meaningful part of the human genome.
• Odorant binding is converted into a neural signal carried as action potentials, rather than gradient potentials, so olfactory receptor neurons communicate in the conventional spike-based format used by much of the nervous system.
• The olfactory nerve carries those signals through openings in the cribriform plate, the thin, perforated portion of the skull located above the nasal cavity.
• The olfactory bulb is the first major target of olfactory input and contains the mammalian nervous system’s most extreme convergence: receptors of the same subtype converge onto a small number of glomeruli.
• The textbook description of one glomerulus per receptor subtype is a slight simplification, because each side of the olfactory system has its own corresponding glomerulus.
• Each receptor subtype responds to a limited set of molecular shapes called odoritopes, so an odor is represented by the combined pattern of receptor subtypes it activates.
• Coffee illustrates this combinatorial code: the scent activates olfactory receptor types 1, 3, and 7, producing a characteristic pattern of glomerular activity.
• The primary olfactory cortex includes the piriform and entorhinal cortex on the ventral surface of the lower temporal lobe, while olfactory signals also project directly to the amygdala, probably to the hypothalamus, and possibly to the cerebellum.
• The receptor-identity map visible in the olfactory bulb probably does not recur elsewhere in the brain, and how olfactory information is represented beyond the bulb remains poorly understood.

Humans as Smelly Social Animals

Humans possess a remarkable sense of smell even though they usually conceal it more explicitly than many other animals. Sobel places human smelling within a broader social context rather than treating it solely as a way to detect food or danger.

• Dogs visibly sniff one another during social interactions, whereas people generally avoid sniffing strangers; people are instead socially expected to smell babies and readily smell loved ones.
• The contrast suggests that humans regulate and conceal olfactory sampling in many adult interactions, even though the underlying sensory capacity remains strong.
• Sobel disputes the common explanation, repeated on Wikipedia, that handshaking evolved to demonstrate that a person was not holding a weapon.
• He argues that there is no good evidence for that account and that the evolutionary origin of handshaking remains poorly understood.

The Handshake and Unconscious Smelling

A study led by Idan Frumin tested whether a brief physical interaction changes how people sample scent from another person. The results showed a pronounced increase in hand-to-nose behavior after a handshake, supported by an airflow control showing that participants actually sniffed their hands.

• Participants were secretly filmed during a three-minute baseline in which hand-to-face touches—including touches to the nose, forehead, and chin—were recorded.
• Three minutes later, a new experimenter entered with a scripted cover story about still setting up equipment in another room and shook hands with half the participants.
• Handshaking produced a marked increase in hand-to-nose behavior, an effect Sobel described as remarkable, even though baseline self-touching was not unusually low.
• The researchers unobtrusively measured nasal airflow to test whether participants merely touched their faces or actually sniffed their hands.
• Nasal airflow confirmed that participants sniffed after the handshake, indicating that the behavior was olfactory rather than simply tactile.
• Participants also smelled the hand that had not been shaken, which the researchers interpreted as self-sampling that allowed participants to compare their own odor with the stranger’s.

“Click Friendships” and Body-Odor Similarity

Research led by Inbal Raviv examined whether body odor similarity contributes to the immediate rapport known as a “click friendship.” The results indicate that odor similarity can be an input into friendship formation rather than merely a consequence of an existing relationship.

• Raviv recruited same-sex, non-romantic friend dyads from across Israel through social-media posts, focusing on pairs who reported an instant connection when they first met.
• Participants used odorless shampoo and soap for three weeks before collection, then slept alone for two nights in a T-shirt from which their body odor could be extracted.
• An electronic nose equipped with MOX, or metal-oxide-covered, sensors found that click friends’ body odors were significantly more similar to each other than the odors of randomly paired strangers.
• Human raters independently judged the odors of click friends to be more similar, showing that the electronic-nose result was not an artifact of the instrument alone.
• To test causality, researchers measured the body odors of complete strangers before placing them in close proximity during a nonverbal mirror game.
• The strangers played in a round-robin design in which every participant played every other participant, allowing odor similarity to be compared with subsequent social judgments.
• One prediction failed: strangers whose body odors were more similar did not necessarily mirror each other’s movements better.
• Post-interaction ratings nevertheless showed that electronic-nose measurements predicted how much participants wanted to be friends with their partner, how nice they thought the partner was, and how affectionate the partner seemed.
• Together, these findings support a causal role for body-odor similarity in the early formation of friendships.

Chemosignals and the Smell of Fear

Sobel’s lab deliberately uses “chemosignal” rather than “pheromone” in its publications, while arguing that humans unquestionably emit chemical signals that alter other humans’ behavior. Fear-related odor provides one of the clearest examples of a human emotional state producing a socially consequential smell.

• The lab avoids the term “pheromone” because it considers that label too restrictive for the range of human chemical communication being studied.
• The “smell of fear” became a major focus of chemosensory research during the decade preceding the conversation.
• Odor collected from frightened people can increase another person’s autonomic arousal and sympathetic state, providing a possible mechanism for contagious fear.
• The ability of dogs to detect human fear had long been assumed but was only demonstrated experimentally about a year and a half before the recording.
• Research from the Denise Chen and Haviland Jones laboratories, together with Sobel’s lab, showed that people can distinguish fear odor from the same individuals’ non-fear odor.
• Those studies also demonstrated that fear odor can influence the behavior of people exposed to it, rather than merely making fear detectable.
• Charles Zucker, a Columbia University professor who has worked on vision, olfaction, and thirst sensing, has also investigated the smell of the back of a grandchild’s neck.

Hexadecanal and Sex-Dependent Aggression

Hexadecanal is a particularly striking chemosignal because it produces opposite behavioral effects in men and women. Research conducted by graduate student Eva Mishor found that exposure reduced male aggression but increased female aggression, a difference Sobel connects to the evolutionary risks faced by offspring.

• Heinz Breer first studied hexadecanal as a mouse chemosignal associated with social buffering and identified it in mouse feces.
• Breer and colleague Stortzman identified the molecule’s receptor and found that it is highly conserved across mammalian evolution, leading them to hypothesize that it may be a universal mammalian signal.
• Mishor’s earlier effort to collect body odor from an MMA club had stalled, so Breer mailed hexadecanal to the laboratory by FedEx, allowing the aggression research to proceed.
• The experiments used the Taylor Aggression Paradigm, a standard, well-validated protocol used worldwide to quantify aggressive behavior.
• Hexadecanal exposure consistently reduced aggression in men, with an effect size described as unusually large by standards in the aggression field.
• In the same experiments, hexadecanal produced an equally significant increase in aggression in women.
• The sex-dependent result replicated when the experiment was rerun inside an fMRI magnet while the researchers tracked brain activity.
• Sobel’s offspring-survival explanation is that paternal aggression—including infanticide and male aggression toward human children—is harmful, whereas maternal aggression is protective.
• A molecule that calms males while increasing female aggression would therefore favor the survival of the mother and baby.
• Hexadecanal is also the most abundant semi-volatile compound found in baby heads, making it a plausible candidate for this evolutionary role.

Emotional Tears and Testosterone

Emotional tears appear to carry a social signal even though people cannot consciously detect them as odors. The best-established effect is a rapid reduction in free testosterone in men who smell women’s emotional tears.

• In The Expression of the Emotions in Man and Animals, Darwin devoted the sixth chapter to tears and concluded that they were an epiphenomenon, or byproduct, because he could identify no preceding function for them.
• Researchers collected emotional tears and exposed participants to them, finding that the tears themselves were completely odorless and undetectable by conscious perception.
• Men who sniffed women’s tears experienced an approximately 14% decrease in free testosterone within about 20 minutes.
• A hormone researcher who studies testosterone described a change of that magnitude and speed as tremendous and difficult to achieve pharmacologically.
• About two years after the original result, an independent South Korean group led by Oh and colleagues replicated the testosterone effect with essentially the same numbers.
• The original team had worried that no other laboratory would successfully replicate the finding, making the South Korean result especially important.
• fMRI scans showed pronounced dampening of brain activity during an arousing state in both the hypothalamus and fusiform gyrus, although researchers did not know why those regions were affected.
• A Nature paper published two years later showed that mouse pup tears reduce aggression from adult male mice in a smell-dependent manner and identified the relevant tear chemosignal component.
• Across these findings, tears appear to act like a chemical blanket that covers the person crying and protects that person from aggression.

Toward a Digital Smell Platform

Sobel’s laboratory is using algorithms to represent the perceived similarity of molecular mixtures and to design new mixtures that evoke specified scents. The work could eventually support digital odor reconstruction, scent design, and medical applications.

• Google released an April Fools spoof video about “Google Smell,” but the company is also genuinely attempting to digitize smell alongside other companies pursuing the same goal.
• A Nature paper led by Aaron Raviv, with Kobe Snitz as a major contributor, presented an algorithmic framework that predicts the perceptual similarity of any two molecular mixtures with very high accuracy.
• The framework can also design mixtures intended to have a known perceptual relationship to an existing odor.
• This capability produced olfactory metamers: mixtures with completely non-overlapping molecular structures that nevertheless smell exactly the same.
• Classical perfumers had considered such substitution impossible because they focus on the ingredients in a fragrance rather than on the perceptual result alone.
• The team generated a metamer for Chanel No. 5 containing none of Chanel No. 5’s own components, and the result is reproducible by anyone.
• The laboratory maintains a map of 4,000 molecules whose perceived smells are known.
• A new mixture can be projected onto that map and compared with existing odors through pairwise distance; a mixture that lands on the lemon region smells like lemon, while one that lands on tomato smells like tomato.
• Researchers are now developing a set of components they call “odor primaries” that could be combined to reproduce any odor a person can perceive.

Transmitting Smell Over the Internet

The digital-smell program has already produced a proof of concept for transmitting odor information through an IP connection. The violet demonstration succeeded, but failures with rose and anise show that the system remains far from a general-purpose odor transmitter.

• The work is conducted with Jonathan Williams’s laboratory at Max Planck in Munich; Williams is an atmospheric chemist experienced in using GC-MS instruments to measure odorant molecules.
• Williams measured odorants in Germany and transmitted the resulting information to Sobel’s team over IP.
• The transmitted data were entered into the perceptual-similarity algorithm and then used to create a physical odor with a device that mixes the odor primaries.
• In a four-odor test, the system failed for rose, which most participants perceived as bubblegum, and for anise, which most described as cherry.
• The gasoline reconstruction was slightly but significantly better than chance.
• The team succeeded with violets, and 15 of 16 participants correctly identified the transmitted scent as violet.
• The researchers describe the violet result as the first odor ever transmitted over IP, while emphasizing that it is an early proof of concept rather than a consumer-ready system.
• Their technological ambition is to move from the olfactory equivalent of circa-1956 black-and-white television toward “4K” odor transmission with much greater resolution.
• A near-term goal is to generate something vaguely recognizable as wine, including distinguishing grapes such as cabernet from merlot.
• The longer-term medical prediction follows from the fact that diseases are specific metabolic processes whose metabolites can have characteristic smells.
• If digital olfaction becomes sufficiently precise, the team envisions an electronic nose in every bathroom or toilet that could perform continuous medical diagnostics.

Closing

Huberman closes by thanking Sobel and invoking the late Ben Barres’s advice to keep going. He also emphasizes the persistence and spirit of science despite the political pressures surrounding it.