Sniffing out smell’s effects on human behaviour
A young woman sits in a reclining chair in a laboratory, waiting to participate in what she thinks is an effort to test some electrocardiography equipment. A researcher enters the room, greets the volunteer, and shakes her hand. After a brief chat, the investigator steps back out of the lab — and in seconds, the volunteer brings her hand to her face and sniffs.
A hidden camera recorded this same behaviour in several participants over the course of a study in 20151. In some cases, a tube situated inside the nostril of the volunteers was also used to measure the strength and timing of their sniffing. Noam Sobel, a neuroscientist at the Weizmann Institute of Science in Rehovot, Israel, who led the project, thinks this behaviour is common, and analogous to how dogs inspect each other in the park to pick up cues about their emotional state, health status or availability for partnering. “We’re mostly unaware of this, but we’re constantly sampling,” he says. Although the handshake study has not been formally replicated, Sobel thinks this finding is consistent with a broader pattern of people unconsciously collecting scent data from their hands throughout the day.
Scientists have historically deemed human olfaction an underdeveloped sense compared with smell in our fellow mammals — a primitive tool that fell by the wayside over the course of evolution. Most people generally pay it little heed, with scent coming to the fore only occasionally, such as when inside a bakery or public toilet.
But olfaction researchers now think that humans can extract rich information about other people from the smells they produce, including cues that might influence our behaviour and emotional state. The COVID-19 pandemic has brought fresh urgency to this research, as many people grapple with the unexpected fallout of losing their sense of smell for weeks, or even many months. “Some are really unaffected by their olfactory loss, but most people feel hindered in their everyday life and are really affected in their social interactions,” says Camille Ferdenzi-Lemaitre, a neuroscientist at the Lyon Neuroscience Research Centre in France.
The evidence for these human olfactory signals is intriguing, and suggests the existence of a largely overlooked dimension of human social communication. But in a developing field of research, the data are largely limited to relatively modest effects in a small number of studies. Accordingly, scientists are working on more sophisticated — and realistic — study designs that could help to reveal the extent to which scents steer human behaviour.
Inferiority complex
The dismissal of human olfaction dates back to Paul Broca, the nineteenth-century French anatomist who charted the brain structures responsible for speech. Broca noted that the human olfactory bulb — the structure in the brain that processes scent signals received by receptors in the nostrils — is small compared with that in other mammalian species that are known to rely heavily on smell, such as rodents or dogs. He characterized humans as being ‘anosmatic’, meaning that smell has little to no impact on guiding our behaviour. Such misconceptions would cast a shadow over the field for more than a century. “By the time I got into human olfaction in 2001, there were very few people in the field,” says Jay Gottfried, a neuroscientist at the University of Pennsylvania in Philadelphia.
This minimization of the role of smell is due, in part, to a limited understanding of our complex odour-sensing apparatus. The relative size of the human olfactory bulb is smaller than might be expected, occupying 200 times less of the brain’s volume compared with the mouse olfactory bulb2. But both structures contain around 10 million sensory neurons. In fact, that number is remarkably consistent across a diverse range of mammalian species. And, although Gottfried notes that rodent and canine olfaction is “vastly superior” to that of humans, there is strong evidence that humans share the capacity to draw rich information from scents.
The mammalian olfactory cortex is extensively connected with the limbic system, which coordinates the emotional and behavioural response to stimuli. “In some ways, it’s the reptilian system — it’s all about detection of food and threats,” says Gottfried. He notes that humans have evolved much more sophisticated brain structures governing behaviour and decision-making, but “the same limbic circuitry has survived”. This connectivity positions our sense of smell as an important behind-the-scenes player in making decisions that steer us towards safe and desirable things, such as food and family, and away from threats. Gottfried’s lab has shown that people can orient themselves using mental maps purely on the basis of scent signals. This indicates that humans have the innate capacity to navigate using odour cues, similar to abilities found in many other animals3.
Smell might be especially crucial in our earliest days, when vision is underdeveloped. For example, scent-based navigation seems to be important in facilitating breastfeeding. Benoist Schaal, a psychobiologist at the Centre for Taste and Feeding Behaviour (CSGA) in Dijon, France, has studied suckling behaviour in rabbits, a species in which mothers typically spend just 5 minutes per day feeding their young. His team was able to uncover a single pheromone that rings the dinner bell4. “It’s a message liberated by the nipple, which attracts the pups so that they efficiently locate the nipple and suckle in the 5 minutes that they have access to the resource,” says Schaal.
Hexadecanal, a molecule secreted from the heads of newborns, mitigates aggression in men.Credit: SDI Productions/Getty Images
He and his colleagues have since found intriguing evidence that similar signals might be secreted by glands in the areolae of women who are breastfeeding. The nature of these signals — which probably include compounds present in the milk itself — remains unclear, but Schaal and colleagues have observed some striking behavioural effects. “When we took these secretions from lactating mothers and presented them to babies, all the babies reacted very strongly by mouthing and inhaling,” says Schaal. Newborns preferentially focused their attention on pads treated with these odours and exhibited feeding-associated physical behaviours, including actively moving in the direction of the maternal aroma5.
Babies also seem to use their sense of smell — which becomes functional in utero, during the third trimester of pregnancy — to recognize family members before their vision has matured. For example, Schaal and his colleagues, Diane Rekow and Arnaud Leleu at the CSGA, showed in 2021 that exposure to maternal odours sharply increased brain activity involved with face recognition in infants6. This could be an important component in helping babies to quickly establish emotional connections with their carers.
A fearful fragrance
As humans grow, sight takes over as the sense that guides us through the world. But there is evidence that olfactory cues continue to play a subtle part in navigating social situations and responding to the emotions of others. Much of the work in this field so far has centred around the reliance on olfactory sensing to steer us away from dangerous situations on the basis of odour cues produced by other people. “We can smell fear,” says Valentina Parma, a psychologist at the Monell Chemical Senses Center in Philadelphia. The scent elicits a strong physiological and psychological reaction even if people are not consciously aware of what they’re smelling, she says. “This means that we’re going to be more avoidant, or faster at scanning the environment, or we self-report that we feel more poorly.”
For example, a 2012 study at Utrecht University in the Netherlands collected armpit sweat from people watching fear-evoking scenes from horror films7. The researchers then exposed other volunteers to the odour of these samples and monitored their response. Monique Smeets, a psychologist at Utrecht University who collaborated on the study, notes that these odours triggered well-established facial and behavioural hallmarks of a fearful response, whereas sweat collected under emotionally neutral conditions did not. “That showed emotional contagion in an implicit way,” says Smeets. Sobel says he and his team have likewise observed that fear-induced sweat can provoke strong physiological responses, such as increased heart rate and higher blood pressure.
The olfactory influence of positive emotions is not as well studied, but Ferdenzi-Lemaitre thinks there are evolutionary justifications that could support the existence of scent signals encoding good feelings. “It allows people to connect with each other, to solve problems and be more creative, and it helps social cohesion,” she says. Smeets and her colleagues have conducted experiments in which the researchers collected odour samples from volunteers as they watched comedies, and then monitored the reaction of other participants exposed to those scents8. Their results showed patterns of emotional contagion that were similar to those seen in the fear experiments. But perhaps more importantly, when Smeets and her team compared their fear- and happiness-associated samples using gas chromatography and mass spectrometry, it became clear that these sweat samples contained distinct combinations of odorant chemical compounds9. “That doesn’t mean that we discovered the chemical signature of fear or happiness,” she says. “But it at least supported the notion that they are different.”
One of the most controversial areas of human olfactory research pertains to the influence of smell on selecting a partner. A few possible human pheromones have been heavily studied over several decades as having a potential influence on sexual attraction — notably, the compounds androstadienone and estratetraenol. But Tristram Wyatt, an animal-pheromone researcher at the University of Oxford, UK, is deeply sceptical of the evidence for such supposed pheromones. “Those, I’m afraid, are junk,” says Wyatt, who wrote a 2015 commentary10 in which he laid out the weaknesses in the data and defined the characteristics that a true pheromone should exhibit. These include the ability of a particular molecule or mixture to consistently elicit a defined behavioural response when present at a naturally occurring concentration. “Humans may well have pheromones, but [androstadienone and estratetraenol] haven’t met this more pragmatic definition,” says Wyatt.
Aside from pheromones, there is also another commonly proposed olfactory mechanism for mate selection in the animal kingdom. Many species are known to be guided by odorant cues associated with a highly variable set of genes involved in the immune system and that encode the major histocompatibility complex (MHC). Specifically, members of some species prefer to pair up with mates that have a scent indicating a highly genetically distinct MHC profile, which correlates with a low level of relatedness. “You can see MHC-based mating in every single vertebrate taxon: fish, amphibians, reptiles, birds and mammals,” says Craig Roberts, a psychologist at the University of Stirling, UK. In principle, this mechanism could reduce the likelihood of inbreeding and generate offspring that can elicit a more diverse immune response to infectious threats.
The source of these MHC-specific odour signals is not yet understood, but a 1995 study led by evolutionary biologist Claus Wedekind suggested that fertile women generally prefer the body odour of men who have considerably different MHC genetic sequences11. Subsequent studies have produced conflicting results, however, and a 2020 meta-analysis of the literature by Roberts and his colleagues found no clear evidence for MHC-based mate preference in people12. The authors attributed this to the limited size and scope of those studies, including a narrow focus on people of European ancestry. But the sheer complexity of human relationships also cannot be ignored. “We’re interested in personality, sense of humour, reliability as a partner, and a host of other things,” says Roberts. “Detecting a small effect is going to be difficult.”
Finding the scent
These studies highlight the challenges of teasing out the behavioural influence of olfactory signals against a cacophony of other sensory stimuli. For example, when there is strong evidence for an odour-induced behavioural response, researchers must then uncover the chemical or combination of chemicals that contributes to that effect.
In a 2021 study, Sobel’s group described one such molecule, hexadecanal13. The team showed that this compound is secreted from the scalps of infants, and seems to mitigate aggression in men and promote it in women. These behavioural effects could be beneficial for ensuring the safety and well-being of newborns, especially because maternal aggression in mammals is typically directed against threats to their offspring. But the finding was also a lucky break; Sobel credits colleagues in Germany — who had learnt about this compound’s stress-mitigating effects in mice — for bringing it to his attention.
In the absence of a clear, evolutionarily conserved starting point, researchers must systematically catalogue the chemical contents of our bodily secretions. This is challenging for many reasons, including the fact that these compounds might be present at low levels, or in quantities that vary over time. Roberts and his colleagues have begun working with a highly sensitive technology used in studies of atmospheric chemistry. The method, known as proton-transfer reaction time-of-flight mass spectrometry, can detect small changes in airborne odorant chemicals in real time. “We want to be able to record the body odour and look for that specific change,” says Roberts.
More and better-quality samples will also be crucial. Sobel is sceptical about how some groups have triggered fear in their volunteers — using scary films, for example — and has come across a more potent alternative. “We throw them out of airplanes,” he says. He is collaborating with the Israeli military to collect armpit sweat from first-time paratroopers, which has yielded thousands of fear-laden specimens for further characterization.
Some of the initial work investigating behavioural response to olfactory cues might be confounded by the factors underlying the reproducibility crisis that has rocked the psychology world over the past decade. “So much of the literature is based, like the rest of psychology, on very small samples — and the opportunity to do lots of analysis on those samples,” Wyatt says. This creates opportunities for P-hacking — fiddling with collected data to come up with a statistically feasible post-hoc explanation, rather than designing the experiment to test a predefined hypothesis. But Wyatt thinks that many researchers in the field have taken the problem to heart, and are developing improved studies that are more robustly designed.
There is also the question of how accurately these lab studies, with people sniffing scent-soaked pads or T-shirts in a controlled environment, reflect what happens in real life. “There are a lot of potential confounds”, says Gottfried, “and I’m not seeing enough consensus to convince me that this is something that is ecologically valid.” These include the impact of clothes, hygiene and chemical fragrances, as well as the nose-deadening impact of the widespread air pollution in the high-income nations where most of these studies are being conducted. This does not necessarily mean that olfactory influences on behaviour do not exist — just that more thoughtfully designed studies will be needed to understand them and the extent of their impact. This could include selecting participants from non-Western societies or from countries that are less highly industrialized, for example, or focusing on younger individuals. “In terms of mate choice, it might be very important at a younger age than we tend to test,” says Roberts. He hypothesizes that studies conducted with people who are closer to puberty — a time associated with surging hormones and changes in body odour — might reveal stronger effects.
But even if the behavioural and psychosocial impact of olfaction remains ambiguous, the COVID-19 pandemic has made it abundantly clear that anosmia, the loss of the sense of smell, can strip away important aspects of the human experience that might be impossible to replace. As part of her research, Smeets has worked with people with anosmia. “They seem to feel no longer so immersed in life,” she says. “When we talk about anosmia, it’s often about how they can’t smell or enjoy their food, and that is important, but I think there’s way more to it than we’re aware of or are measuring.”