Can smells help you give up smoking?

Now, smell is a mysterious sense in many ways, but one thing that struck researchers is most odours don’t wake you up. There are exceptions: researchers in Japan are developing a wasabi scented fire alarm for deaf people, but most odours can really reek, and you’ll still sleep. So this gave scientists an opportunity to look at a hot topic - whether or not we can learn in our sleep. Audio books that supposedly teach you another language while you doze are still being sold, but there is no evidence that they actually work. But Anat Arzi, at the University of Cambridge, wanted to investigate whether smells might be different, and whether we can learn to associate odours in our sleep. This might, they think, provide a way to help people quit smoking! Georgia Mills spoke to Anat about how they did it...

Anat - We used a unique measure that we have in our FACTION and this is the ‘sniff response.’ I can tell if you like or dislike an odour only if I measure your respiration. For example, if you go by your favourite bakery and they’ve just baked fresh bread, you will take a deep inhale. However, if you go by a public toilet you might take a smaller inhale. And this change in nasal airflow in accordance with the properties of the odour is the sniff response. We are having a sniff response every day, on a daily basis automatically without being aware of it. What we discovered is that we also have it during sleep, so while you’re asleep you can perceive information from the environment and general an adequate behavioural response: a large sniff for a pleasant odour and a small sniff for an unpleasant odour.

Georgia - So when people are sleeping, if you shoved a croissant in front of their nose they might go “Mmm”, still not waking up but you know that means that they like the smell? And I suppose that makes sense because if there is a nasty smell it’s probably not a good idea for us to inhale a whole load of it?

Anal - Exactly. The sniff is a smart mechanism for us. And then, after discovering that we have this super cool implicit measure for processing in sleep we said okay, let’s see if we can test with this if we can learn during sleep.

Georgia - To test this, Anat and her team looked at one of the simplest forms of learning: conditioning, where you learn to associate two things together. So they took some people, while they asleep, and played them a note and then they played that note repeatedly with a bad smell, like rotten fish or eggs. This resulted in a small sniff response. And then they also gave some people a note and paired it with a nice smell, like shampoo, and this would have resulted in a larger sniff response. But then later that night, they looked at what happened when they played only the tones without the smell.

Anat - What we discovered is that when we presented the conditioning during sleep people then, during the same night of sleep, took a different sniff to the tone depending on which odour was presented. So if I presented a tone with no odour whatsoever during sleep, then the person took a smaller sniff if the tone was associated with an unpleasant odour. This means sleeping humans can learn a new association during sleep and implement them within the same night.

Then to test whether they can actual retrieve this information in the morning, we presented the same tones upon awake and we discovered that also, when they wake up they change their sniff to the tone even when there is no odour there. This means that they can learn a new association in sleep and retrieve them again upon awake.

Georgia - Did they have any awareness; were they aware of why they were doing this smell or were they like “oh, that’s odd”?

Anat - Excellent question. Absolutely not. We asked them if they smelt anything during the night or if they heard anything, and they said no, they had no clue.

Georgia - So does this kind of learning last for longer than the next morning? To find out more, Anat and her team, decided to look at smoking…

Anat - What we did, we invited to our lab people who are smokers who wanted to quit smoking and then, while they were sleeping, we presented a cigarette odour that was paired with a profoundly unpleasant odour. We made sure they really disliked this odour before they went to bed.

And then we asked them to fill in a smoking diary: they were filling in seven days before the experiment how many cigarettes they were smoking every day, and then seven days after the night in the lab. We presented conditioning either during non-REM sleep or during REM sleep - rapid eye movement sleep. What we discovered is that if we conditioned the cigarette odour with profoundly unpleasant odour during non-REM sleep, people reduced smoking about 30% in the week after the conditioning in the lab. If they were conditioned during REM sleep, they reduced smoking only in about 10%. What was fascinating was when the conditioning was presented during wake and they knew what is happening, they didn’t change their smoking habits so, only when the condition was implicit, it reduced smoking. It doesn’t mean that we found an alternative treatment for smoking - not at all. But it proves the concept that learning during sleep can moderate behaviour during wake.

Georgia - My question is then, were you a little bit worried doing this study that by pairing the nasty smell - a poopy smell - with the smell of cigarettes, were you a bit worried it might have gone the other way and they started to really get a bit addicted to the smell of poo?

Anat - I think the unpleasant odour are so unpleasant, and the cigarette odour even for a smoker is not that nice that it wasn’t a big concern.

Georgia - Now you’ve got this proof of concept that we can learn these basic things in our sleep, what’s really interesting is only seems to work when we’re asleep. What does this tell us about the brain and what next for the research?

Anat - There are several different lines of research we can continue from here and many open questions. One of them is to understand what is possible to learn during sleep; is it something that is unique to the sense of smell or can we learn different basic form of learning during sleep as well? And we still have a lot of  work to do in order to understand where is the line between what we can learn in sleep and what we cannot.