the scent of fear

By Alison Campbell 04/09/2009

Like other animals, humans produce pheromones – chemical signals released by one individual that carry information to another. (Unlike other animals, we also spend a lot of time & money modifiying those pheromones: bathing, deodorants, perfumes…) However, while the significance of pheromones in other animal taxa has been studied fairly thoroughly, there’s been less scientific research into our own signalling chemicals, & in fact it’s sometimes said that we can’t detect them. So, some of the questions about human pheromones include: can we actually detect them? and, how do we react to these signals?

Attempts to answer these questions have often involved collecting armpit sweat & asking volunteers to sniff the samples. Sometimes this is done by getting volunteers to wear a t-shirt for a period of time & then getting others to put their noses to the test. (A while back now I had a masters student who did much the same thing, but in her case the sniffers were ferrets – she wanted to know how the mustelids reacted to the scent of human sweat. Why? If the scent was aversive, there was the possibility that trappers leaving sweat on traps could inadvertently scare the ferrets off.)

For example, in 2002 – in a paper engagingly titled The scent of fear – Kerstin Ackerl & her colleagues reported on a project examining whether it’s possible to detect fear from human body odour. They collected underarm secretions (using underarm pads) from women watching a horror movie. At the same time saliva collected before & after the movie was tested for levels of cortisol, to give a hormonal measure of the subjects’ response to fear. The control was a ‘neutral’ film, & again the same techniques were used to collect odours & hormones. Subsequently, female ‘sniffers’ were able to tell the fear & non-fear pads apart. (Interestingly, while you might expect that ‘fear’ odours would be correlated with cortisol levels, this experiment didn’t demonstrate such a relationship.) An obvious next step would be to see how the pheromones affected the sniffers’ behaviour – this is, after all, a way of telling whether one individual has communicated with another. But in humans, as in other animals, many of these behavioural changes could be quite subtle. With people, at least you can ask your subjects how they feel – but this is very subjective. What we need is a way to get inside their heads.

A new paper in PLoS One reports on a project that does just that (Mujica-Parodi et al. 2009). Again, the researchers collected underarm sweat from individuals in a state of fear (’emotional stress’) – first-time skydivers. Cortisol levels & measures of anxiety confirmed that the novice jumpers were indeed under emotional stress. A control group exercised on a treadmill, so they were physically but not emotionally stressed. The sweat samples from these two groups were then used in a series of experiments – some of which involved making fMRI scans of the sniffers’ heads, in order to see if the odours had any effect on brain function. These tests clearly showed that the sniffers’ amygdala (a part of the brain that’s involved in emotional responses) could tell stressed & non-stressed individuals apart on the basis of their sweat odours. However, it also appeared that what was happening was an emotional response, rather than discrimination between odours per se.

But what did this mean in terms of behavioural responses? After all, for this ability to exist, you could expect that it might have some adaptive significance. The research team looked into this by testing whether stress sweat [and] exercise sweat… behaviourally affected perception of subtle emotional cues in the evaluationof ambiguous faces. In other words, did exposure to stress sweat make it more likely that the sniffers could recognise individuals who might pose a threat to them, on the basis of facial expression? The answer? Yes – it did appear that stress pheromones enhanced an individual’s ability to discriminate between threatening & non-threatening facial expressions, potentially a useful survival trait.

K.Ackerl, M.Atzmueller & K.Grammer (2002) The scent of fear. Neuroendocrinology Letters 23: 79-84

L.R.Mujica-Parodi, H.H.Strey, B.Frederick, R.Savoy, D.Cox, Y.Botanov, D.Tolkunov, D.Rubin & J.Weber (2009) Chemosensory cues to conspecific emotional stress activate amygdala in humans. PLoS One 4(7): e6415 doi:10.1371/journal.pone.0006415