122: Can AI help us treat eating disorders? How to change your social media habits to help your recovery - with lead researcher Dr Gemma Sharp

Speaker 1: 0:39

you've done some very interesting research during COVID-19 times eating disorder rates and body image concerns. Can you talk to me about that?

Speaker 2: 0:46

so there was just a lot going on.

Speaker 1: 0:48

That was unfortunately a bit of a perfect storm kids don't want to come to school anymore, and it's multiple teachers who reported this. You're looking into whether or not AI can assist people with eating disorder treatments?

Speaker 2: 1:00

and the results have been really excellent so far, which has been great. I will always ask my clients, patients, people with eating disorders what is it that you like to do before you had an eating disorder? That part of you is still there.

Speaker 1: 1:20

Hi everyone and welcome to no Wellness Wankery. I'm dietitian and nutritionist, Lindy Cohen, and this is a place where we get science-backed information so you actually know how to look after your lovely, lovely body. And my guest today. Well, it's a little bit tricky to sum up just everything that she has done in her very short career, because she has achieved so much.

Speaker 1: 1:42

Gemma Sharp is an associate professor of research and leads body image and eating disorder research at Monash University in Melbourne, which is in Australia.

Speaker 1: 1:49

If you don't know, she has degrees in molecular biology, microbiology, oncology, Japanese and psychology, culminating in a PhD in clinical psychology. That is a mouthful and it is very impressive. Her research spans across multiple countries Australia, Japan, the UK and she's very much focused on body image, on eating disorders and, I guess, innovative therapies and interventions, things like digital technology, like AI, and that's something we're going to talk about in today's podcast. She's also, in case you needed more things, she is also a registered clinical psychologist who's been recognized for her significant contributions to national and state eating disorder strategies and clinical guidelines for cosmetic procedures. She's won countless awards so many I found it really hard to count them all. She's been named one of ABC Radio's national top five under 40 scientists and was invited to deliver a TED Talk, so I'm very excited to have Associate Professor Gemma Sharp here to discuss breaking down beauty standards and identifying disordered eating in the world of AI. Associate Professor Gemma Sharp, I'm very, very, very happy to have you on the show.

Speaker 2: 2:58

Thank you so much for having me. It's great to be with you.

Speaker 1: 3:01

I'm delighted and I really want to hear from the get-go what got you interested in body image and eating disorder research.

Speaker 2: 3:08

My very first foray into research was actually cancer research. I was particularly speaking with a lot of breast cancer patients and they were talking to me about body image concerns after having their breasts removed and things like that body image concerns after having their breasts removed and things like that and so my cancer interest was still there. But the body image research was just so appealing to me so I kind of came on a bit of a tangent, I suppose. And then the eating disorder element. I was very lucky to be chosen to do one of my clinical placements in an eating disorder clinic and I had this wonderful mentor and she was just super inspiring and she said that obviously eating disorders needed a lot more attention and advocacy. As your listeners will understand and believe me, I love rooting for an underdog and I was like this is my cause, so I suppose maybe it's not what people think it would be, but yeah, so sort of on a tangent and rooting for an underdog, really.

Speaker 1: 4:04

No, and I do love it. I appreciate you rooting for all the people with eating disorders, with disordered eating and one of those things where disordered eating I feel like it's this huge number of people who we're statistically not really capturing right, like we have an idea of those who are diagnosable, but then there's this whole group of people who are like listen, we know you're borderline or you've got some behaviors that we don't quite consider healthy and I certainly talk to people all the time and I think that is a very much an underestimated group and I think some of your research touches on that.

Speaker 2: 4:36

Absolutely so. As you said, we were sort of quite clear on those diagnostic criteria, so it's a little bit easier to categorize people, but people will often go in and out of diagnostic criteria as well as between eating disorder diagnosis and then sub-threshold disordered eating, and I mean it really is the gateway to eating disorders, isn't it? So I think there's so much more we need to do in that prevention space around disordered eating, and unfortunately we're seeing it with really young people as well. I'm seeing lots of things about children not eating their lunches at school or being lunchbox shamed, and so it can start really early from sources that are supposedly meant to protect our young people, and I and I think people probably everyone, it would be with rare exception would have experienced some disordered eating at some stage in their life.

Speaker 1: 5:31

And do? Speaking about young people, a lot of your research has focused on the impact of social media and body image. I assume most of our listeners are going to be with the understanding that if you're using social media, what we see is not often correct and following people who share certain images are going to make us feel a certain way. What did you learn from your many pieces of research that you did on this topic?

Speaker 2: 5:54

Thank you for saying many pieces as well. I appreciate that I'm a bit of a publication junkie if I can be, I like to get the good word out there as much as possible. So I think I suppose some really fascinating research that I've been a part of or have read just showed that if we actually showed people, if the images had been edited so we had like labels like you might find on cigarette packets, for example, it actually didn't change their dissatisfaction with their bodies. So I think there's an awareness that we all have nowadays that everything is edited, everything's quite unrealistic, but it doesn't stop us from wanting that ideal, that wonderful lifestyle that we're seeing, and I think it shows how aspirational humans are. I think that they want the best for themselves, and what's wrong with that?

Speaker 2: 6:46

But I think it's really hard to get around sometimes because I think if those warning labels had have been effective, we would all be using them, but sadly not. I think it's just. I think potentially what we need to be doing more of is just a lot more diversity of representations, and I love those influencers who show us behind the scenes and or sort of, I suppose, have a bit of a joke with some of the stylized photos. I think people really do love that and that does tend to make people feel a bit better.

Speaker 1: 7:20

And I think that's the misconception those influencers who are trying to appear perfect have is that people will like them more if they turn up as their shiny selves or their filtered selves.

Speaker 2: 7:30

Exactly, there's research to show that we actually like leaders who have a bit of imperfection. They seem a bit more human. Surely that's why my team love me here at Monash. So, yes, there is research to back that up, that having some imperfections is actually very likeable. Tell your influencers.

Speaker 1: 7:50

that's what we like Exactly. You've done some very interesting research during COVID-19 times and eating disorder rates and body image concerns. Can you talk?

Speaker 2: 8:02

to me about that, absolutely. So what an interesting time of everyone's life, hey, getting to live through a global pandemic and tick that off the bucket list. Hopefully we don't have to do that again. Yes, I think what we saw was quite a lot of things, so a lot of impacts. We were spending a lot more time on screens, like everyone was doing online study, online schooling, online work, online socializing so we were looking at images of ourselves a lot more and I'm not sure if the term has persisted, but we did sort of talk about Zoom dysmorphia there for a while and people were actually getting more facial cosmetic procedures done because they didn't like what they were seeing in the Zoom camera and I think, in terms more broadly so, just not being able to go to school, not being able to do all the things we enjoyed, the control that we have in our lives was taken away from us very quickly, and that was a huge trigger for a lot of people's eating disorders.

Speaker 2: 9:02

So either they were already experiencing one and it exacerbated or it triggered off new ones, unfortunately for people at risk. So there was just a lot going on that was unfortunately a bit of a perfect storm for a body image concerns and eating disorders, sadly. So I think probably we could have predicted it, but obviously none of us could predict the pandemic. So none of us could predict the pandemic, so none of us thought it would all come at once.

Speaker 1: 9:26

Yeah, and I think it affected people in so many ways. I've been speaking to some teachers recently who are talking about how kids don't want to come to school anymore, and it's multiple teachers who've reported this. So there are long lasting impacts as well that we're feeling from, I guess, covid-19.

Speaker 2: 9:43

Yeah, I mean we're going to be doing these longitudinal studies. I think the people who are young people in COVID-19 are probably going to be the most studied people ever moving forward, so I hope they enjoy participating in research, including ours. But yes, I think it has shaped their social development, and I think so. We obviously had students who had only done, for example, online university and are now coming to us for their, like, phd and higher level studies and they've never actually spent a day in a lab before and I'm like, oh well, welcome. So it is pretty curious.

Speaker 1: 10:57

It is curious. Can you talk to me about some other curious work you're doing at the moment around AI? Oh, yes, I think it's so interesting what's unfolding. What are you doing here? And I guess what I can see is that you're looking into whether or not AI can assist people with eating disorder treatments. Can it be something that can actually be helpful and useful? Because what we know with eating disorder support is it requires a large amount of time for us to get recovery, requires multiple interventions, reminders. So can you talk to me about KIT, the Positive Body Image Chatbot?

Speaker 2: 11:29

Yeah, Well, that's. I mean we actually have a whole family of them now, Lindy, so I'll introduce them all quickly like our chatbot family, because they all have like avatars that are similar, so we call them our family. You know, people who don't have enough friends, like myself, will obviously adopt these kinds of tools. So Kit is now known as Jem and Jem lives on Monash's website. If anyone can access it, feel free to contact me if you can't find it.

Speaker 2: 12:01

But Gem started out during the COVID-19 pandemic. The idea was formulated beforehand but the pandemic accelerated its development and we really just wanted a tool that could help. It was Butterfly Foundation's helpline at the time because it was so oversubscribed, everyone needing help at once. So Kit slash Gem was there to help shoulder some of that load, give those in-the-moment coping skills. We actually saw quite a very high usage like post-dinner, which shows when people were experiencing distress and had maybe a bit of time on their hands. So our micro skills were very helpful there. So that was GEM, and GEM is actually going to be launched in North America later this year. So we're doing a sort of North American GEM right now, which is super exciting and going to help them with their helpline load.

Speaker 2: 12:50

And then we also have EDSE, which stands for Eating Disorder Electronic Single Session Intervention. That's why we call it EDSE we don't say that name all the time and EDSE is currently in trials around Australia for the many people on wait lists for eating disorder treatment. So GEM was kind of based in the community, whereas EDSE is very much that first point of contact in treatment and the results have been really excellent so far, which has been great. So people are like, oh great, I get this free chatbot session while I'm waiting, because we know if we leave people on wait lists, unfortunately their symptoms get worse. I mean, what a surprise. But I think people are quite appreciating that they can get that early intervention and we've seen that they do better in treatment when they do get to start. But I should say still in trial. We're looking forward to publishing those results and hopefully EDSE will be more of a mainstream tool in Australia in 2025.

Speaker 1: 13:49

I think this is such interesting research. Can I ask you, with Jem, is it a kind of chatbot where you'd log in, I chat to you, I get the support that I needed. I log off. If I re-log in, would Gem remember me so?

Speaker 2: 14:03

I think we did talk a lot about that. So, edse, you can, because it's kind of like your treatment, whereas Gem, I suppose, is a little bit more like the bank chatbot and the Optus chatbot. That it won't, but I suppose it's a feature we can look into. It's just we wanted that anonymous usage with Gem and that's why we didn't have the memory feature.

Speaker 1: 14:31

Absolutely. Now, if anyone isn't up to date with this whole kind of AI development, there is I don't even know what it's called, but a program where people have got an AI that they're forming a relationship with and some people come and they have very intimate relationships with this chatbot who essentially would remember hey listen, your mother passed away. How have you been feeling and can handle quite complex things. In fact, people can attach a huge amount of, they get a large amount of attachments to this bot and I guess I was always thinking about how would it be if you had a long-term relationship with a chatbot and someone who would remember hey listen, I know that last week you found a particularly triggering with this and this how have you been finding it? Kind of like how you'd have that clinician one-on-one support, and I guess this is kind of like you are in the beginning stages of looking at how this could be applied.

Speaker 2: 15:25

We're actually building that right now. So you're right on the money. I suppose what they've been able to do with those ones that are available now they're kind of generally trained, so we know what they're giving is good for general advice, but we can't say that there's a strong sort of evidence base from a clinical perspective for them. So I'm really glad that people are using them because they're wonderful tools, but I think in terms of actually getting them into our health services and having them recognised as clinical tools, they need to go through a certain pathway that we're going through. But that doesn't mean that other people won't find these freely accessible tools to be useful. I think we just.

Speaker 2: 16:08

I suppose it's a really tricky time in AI. So I think the health services are a bit behind technology. What a surprise. So at the moment health services are using AI, but it's more like for administrative tasks, it's less about clinical service provision. So we're moving ahead in terms of our clinical service provision, but step by step really. But we love the idea of people being able to come back and chat with, like a sort of a friend, which we've called Gem AI. I guess all how narcissistic it's all my name, but we'll change the name. Obviously I like Gem. I suppose something else we've been doing is just targeted intervention, so your phone knows when you might be feeling down or you might be about to engage in an eating disorder behaviour. So that's another part that we've been working on too, that predictive power of people's phones. So it's not just you choosing to chat, it's that your phone knows when you might need to chat as well.

Speaker 1: 17:09

And this is incredibly interesting and exciting, I think, because when you think about eating disorders, we know this is a very secretive kind of thing we tend to hide, especially in younger people. Sometimes you might not feel willing to speak out loud to a clinician. We have COVID-19 babies coming through who perhaps have social skills that need a little bit of extra support, but it means that they are perhaps more comfortable with sending a message to someone and typing and texting to someone.

Speaker 2: 17:38

Who makes phone calls nowadays. I mean, I feel like I'm an old millennial and I make phone calls and I'm slow at texting. I don't know what I am. Yes, I agree, I think. I mean, I think we need to cater for all different types of communication and if people are happier with texting, well you know, that's why we have that form of communication and we can absolutely leverage it.

Speaker 1: 18:06

What are you most excited for in the future of eating disorder research?

Speaker 2: 18:11

Oh God, that's what a big question. Thank you. I mean gosh, so many things. I mean. I think something our group does well that I absolutely love is that we're coming at it from all angles. So that genetics, biological side so you know we have our work that we do at the lab bench obviously this technology side as well. We're looking at it from a psychology and social perspective, just as we were talking before impacts of social media, using that as a channel for good. So I think what I'm most interested in is the intersection of all of those coming together to really tackle eating disorders, because eating disorders are so multifaceted, so you need that multifaceted solution. So I'm just glad we are finally exploring all the avenues we need to instead of, I suppose, just psychological support. Obviously, that's very important, but I feel there's lots of ways we can help people.

Speaker 1: 19:08

Gee, I find that really exciting. And what about? The thing that you find as a researcher is most challenging still?

Speaker 2: 19:16

Sorry for my laughter. Everything Gosh. I wow Sometimes me personally, I think because I'm a younger woman. I know I sound old to your listeners Sorry, listeners, I swear.

Speaker 2: 19:34

In terms of research I'm not super old and I think sometimes I might not get taken as seriously as some of my more senior colleagues who might be another gender. So that think sometimes I might not get taken as seriously as some of my more senior colleagues who might be another gender. So that's sometimes a bit of a challenge. But obviously I think we're making good progress there and I think sometimes eating disorder research does not receive the funding it deserves, particularly given the prevalence of it and the scope of the issue that it impacts people's lives so dramatically and their loved ones. So sometimes it can be hard to fund the research that you want to do. But I wouldn't change my job for anything. I have so much fun. I get to think of new ideas all the time and explore most of them, and we get to launch cool things and get to chat on podcasts. So I think my job is a fantastic one and I don't have much cause to complain at all.

Speaker 1: 20:29

Yes, and I'm very grateful to have a young female like yourself in the work that you do. I think it's very, very important. I always found that one of the challenges in research and perhaps this is me from my media perspective is how the research can often get confused and you get these clickbaity headlines. So by the time it actually reaches the masses if it reaches the masses at all it can get butchered and the authenticity or the meaning or the validity of it is totally missed, and what we end up with is just a clickbait headline and nothing else. So that's something that I would love to see shifted as well.

Speaker 2: 21:04

Absolutely yeah, and I think, well, you're doing a fantastic job by, I suppose, allowing the researchers to tell their own stories, and I think I've generally had very good experiences with the media because I think I've I've been allowed to have my own quotes, my own recordings, et cetera in my own voice and I think, as well as lived experience, folks who are part of our research too. They're almost always part of our stories, they're certainly part of our research, because I think that really brings research to life. It takes it away from those statistical numbers or that one percentage or whatever it just it needs to be brought to life, as you said, and not, I suppose, iterated such that it no longer resembles what it started out as.

Speaker 1: 21:50

Indeed. Can I finish with one last question, of course, for the audience listening, what is something that they can do on social media based on your learnings, your research that could help to improve their body image?

Speaker 2: 22:04

Cat videos. No, that's what I do.

Speaker 2: 22:10

That's the small silly cat on Twitter, slash X. I'm a huge fan of that one. I think it is. I think it is. It is finding what interests you and finding what you find meaning in on social media, because I think there's so much content out there and if we, I suppose, keep clicking on the same kinds of stories, that's what we get fed back because of the way social media algorithms tend to work. So I would always set an intention when I go on social media that I'm going to look at something meaningful. Maybe it's a quote, maybe it's a beautiful picture, maybe it's something to do with gardening. I just think there are just there's so much content there that we're not accessing. That would actually make us feel better about ourselves in general not just our bodies, but our minds and souls as well.

Speaker 1: 23:01

Spot on, I mean. I think sometimes, when you're going through an eating disorder, it can occupy your entire headspace and it doesn't leave much room for hobbies and things that make you feel something greater. And I think finding what it is that you even like you know, yeah, just ask that question.

Speaker 2: 23:19

Like I will always ask my clients, patients, people with eating disorders what is it that you like to do before you had an eating disorder? They'll say, I really like horses or I like roller skating or I like painting. Like there's always something and I think that part of you is still there, maybe just dormant, but it's still there.

Speaker 1: 23:40

And I think living that full life, or at least leaning into your interests more, can help to give more of your headspace to those things that make your life bigger, as you're saying.

Speaker 2: 23:50

Absolutely yeah, even if it's some Netflix series I mean, isn't everyone watching baby reindeer at the moment? Or if it's cat videos. Whatever you know, you're fine. I know. You See I'm showing some insights into my own viewing habits here. Stalkers and cats Anyway, stalker cat, I should make a show about that.

Speaker 1: 24:14

Thank you so much for your time on the podcast and please, if you do have more research, which no doubt you will have very shortly we'd love to hear an update.

Speaker 2: 24:22

Thank you so much for having me. It's been an absolute pleasure.