‘It’s an intriguing world that is opening up'

Can you take me back to when you first became interested in how bodily signals can influence emotion and memory?
I was working with people with post-traumatic stress disorder. I was just looking at neural signals and the mechanisms underlying persistent fear memories. But I was also taking skin conductance as a measure of fear, and that was the first time I had used physiology. I was struck by the way that some people had very large bodily responses to fear, and others were just flat. I spoke to other researchers and they said 'Oh, we throw out non-responders' – they would just discard their data. But their bodies were doing something so different. That intrigued me. What might that mean for the brain?

My research was putting people in 'dangerous' and 'safe' rooms, and showing that people with PTSD couldn't use external context to guide memory recall. So I was thinking, is there something about their internal state which says 'fear fear fear'. What might that look like, what might that be that keeps them constantly primed?

And you settled on the heart, which makes sense because we intuitively feel things in our heart. It's not just chugging away in the background like a metronome.
The beautiful thing about the heart is that it's not regular, and I love that. It's such an honest signal. With neuroimaging you can analyse your data in different ways with different models, but with the heart you can just see the variability… it's there, it's apparent, and it's real. A simple signal that can be so informative.

And different conditions are associated with different variability. The fact that schizophrenia has such a regularly beating heart is intriguing. But my work deals more with the heart–brain connection, and how sensitive individuals are to detecting bodily sensations. That's interoception.

Explain that heart–brain connection. Fundamentally, your brain reacts each time your heart beats?
Isn't that beautiful? You do get a suppression effect with most people, so when your heart is beating your brain gives you a smooth perception of the world. Your brain edits out these signals that could potentially interfere. It's trying to downplay events that happen in sync with your heart. If I gave you a pain stimulus when your heart is beating, you'd perceive it as less painful. There are effects on memory too… if we see stimuli on the heartbeat we are more likely to forget them. So that's back to the inhibitory effect. That paper, 'what the heart forgets', showed that if a word coincides with a heartbeat you're more likely to forget it.   

The one exception to that is how fear is processed when the heart is beating. Maybe we'll find more things, but this was a paradigm shift to find that fear is selectively enhanced with these heartbeats. And your amygdala fires more. So you have subjective effects and neural effects associated with the heightened processing of fear with each heartbeat. We think that channel may be more active in individuals prone to anxiety… It's an intriguing world that is opening up.

So take me back to this idea of interoception…   
Interoceptive accuracy is an individual difference in how good we are at perceiving internal sensations, and that's something you can train. So when you sit here at rest, do you feel your heart beating?

And there's a difference between how good you are at that, and how good you think you are at that?
Yes. My dad thought he couldn't do it, but I tested him and he was almost perfect. He didn't have insight into it. So you get these interesting dissociations. I'm not the best at it, but I'm all right, and I have decent insight.

What are the implications of being really good at it and knowing it?
A BBC Radio 4 programme got me to work with a hostage negotiator. He claimed he was really sensitive to other people's emotions. Being sensitive to your own heart potentially tells you how you're feeling, but could also give you information about how other people are feeling as well. This guy was 100 per cent perfect. So he's certainly sensitive at reading himself, the next question is, does his body change in response to the emotional cues of others?

You'd think in some ways that wouldn't be an advantage to a hostage negotiator. Wouldn't it get overwhelming?
My favourite quote in Middlemarch is about how overwhelming empathy can be… 'If we had a keen vision and feeling of all ordinary human life, it would be like hearing the grass grow and the squirrel's heart beat, and we should die of that roar which lies on the other side of silence.' I think that in order to be a good hostage negotiator you need to be aware of what's going on but also be able to regulate it so that it doesn't overwhelm you. Use it and be in control of it.

People with autism spectrum conditions are interesting in this respect… a lot of the literature says they have deficits in empathy, but their bodies are actually very sensitive to other people's pain.

A study by Xiaosi Gu and colleagues demonstrated that autistic individuals had a heightened skin conductance response to other people's pain. They have often have hypersensitivity to extroceptive signals – touch, noises in the environment etc. – yet not be able to read their own bodily signals. So their heart may be racing in response to the pain of others but they're not able to read that in the same way.

There's a big overlap with anxiety there. You're suggesting we should help people with anxiety tune into their own bodies?
To get more accurate at it. An internal focus can be a sign of anxiety, so what you actually want is bodily precision. People with anxiety might think they're good without actually being good. Our new work shows that if you train people to have better accuracy, they have reduced anxiety.

How quickly could you make me better at that?
We initially gave people eight sessions. We did a baseline, a mid-point and a final assessment, and by the mid-point everyone was pretty good. We didn't need the last four sessions! You'd probably see a significant improvement after just 40 minutes of me giving you some feedback.

You've taken this work into other conditions.
I've been based in psychiatry for a decade. My PhD was getting students drunk and looking at their memory, which was an awful lot of fun, and I found interesting effects, but I felt a bit of an existential crisis, I wanted to help the world. So I did a fellowship in neuroscience and psychiatry, building on the memory work. I was working with people with PTSD in the States, and then I moved to work with Hugo Critchley, who is one of the world leaders in that body–brain relationship. I got tenure a year ago and now run my own lab in the medical school.

My aim is always to look at basic mechanisms and then look at clinical disturbance, with the potential for opening out new avenues of treatment. It's a mechanistic approach that touches on what the general public instinctively think might be right anyway… take mindfulness, lots of people are subscribing to that but nobody is quite sure how or why it works. Having these more controlled studies looking at bodily precision and interoception could give some idea. Mindfulness won't necessarily increase your accuracy in the way that training will, but the two together… But also the 'just noticing' part of mindfulness could be useful: it's good to have interoceptive precision, but then just notice and not worry if you feel a heartbeat.

You've worked in the area of dissociation.
We have a paper in the pipeline on individuals who were first episode psychosis, who are more likely to have dissociative symptoms – to not feel like their body is integrated in the world, and to feel like their memories didn't really happen to them. It turns out they are people that have deficits in interoception, specifically in the metacognitive domain. So they don't have insight into what their body is doing, in terms of whether they know if tones that we present to them are in sync with their heart. They might get it correct, but not know they are correct. That metacognitive deficit in interoceptive was associated with higher dissociative symptoms.

Your journey so far has involved a lot of neat findings and cool titles… have there been dead ends along the way, or is the heart just opening up in the way you are hypothesising?
It's never like that. There are always things in file drawers! The key for me is not to be disillusioned, and also I like working in teams. If one thing doesn't work, then another thing will, and people feel ownership of both. I run a collaborative lab and that seems to work well. Working with Hugo Critchley is a joy, and all my mentees and colleagues are great.

You're finding things that are statistically significant, but how clinically significant are they as targets for interventions?
I think that we spent so long as scientists focusing on the brain… we as the general public know that our brain feels better if we exercise, we sleep, we eat well. There's such an amazing dynamic interaction between body and brain, yet looking at and understanding the two hasn't been investigated that much. Partly that has been due to the equipment. All specialties have been developed in isolation and you're limited by the technology. Going back not that far, people were talking about imaging the brain and heart together by finding really small people! It's relatively recently that we've been able to do things like look at blood pressure and the brain together. Looking at these dynamic relationships between peripheral bodily signatures concurrently with brain is a relatively new thing, and I think it has the potential to open up new insights and make therapeutic progressions we hadn't thought of.

It's exciting to see where it may lead.

- Sarah Garfinkel is Professor in Psychiatry (Neuroscience and Imaging) at the University of Sussex. Email [email protected] or find her on Twitter @DrSFink. She was recently named by Nature as one of their 'Rising stars': see tinyurl.com/natrisingstar