The Baby Grow Project: Part 1

Show Notes

The First Eight Months

This podcast documents an extraordinary experiment revealing how the way infants wriggle, from birth, helps prepare their brains and bodies to navigate a complex physical and social world. This podcast is for anyone interested in how social and communication skills emerge over the first years of life and explores if the quality of a baby’s first movements can act as an early detection system for neurodevelopmental conditions like autism.

Recommended Sources:

BBC Reels:

https://www.bbc.co.uk/reel/video/p0ggy0fn/how-i-rewired-my-brain-in-six-weeks

University of Sussex: https://www.youtube.com/watch?v=dDnPhP-gPuY

Newsweek: https://www.newsweek.com/baby-movements-predict-neurodivergence-1920020

Me Human website: www.mehuman.io

Prof Gilly Forrester media: www.gillianforrester.com

Prof Gilly Forrester, University of Sussex: https://profiles.sussex.ac.uk/p170236-gillian-forrester/about

Publications about this project:

• Sotoodeh, S., Ossmy, O., Hall, J., Rowan, H, Donati, G., Forrester, G.S. (in press). Automatic Pose Estimation in Newborn Infants: Lessons from the Baby Grow Study, Behavior Research Methods
• Ossmy, O., Rowan, H., Sotoodeh, S., Hall, J., Donati, G., Forrester, G.S. (2025). Frequent, Scalable and Global Use of “Intelligent Onesies” to Quantify Newborns’ Spontaneous Movements in Natural Settings, 54, IEEE International Conference on Development and Learning (ICDL), Prague.

https://sussex.figshare.com/articles/conference_contribution/Frequent_scalable_and_global_use_of_intelligentonesies_to_quantify_newborns_spontaneous_movements_in_natural_settings/31035037?file=60899593

• Ossmy O, Donati G, Kaur A, Sotoodeh MS, Forrester GS (2025) Towards automatic assessment of atypical early motor development? Brain Research Bulletin,

https://doi.org/10.1016/j.brainresbull.2025.111311

Acknowledgments

This podcast episode was written and presented by Gilly Forrester, produced by Will Hood of The Academic Podcast Agency, funded by the Simons Foundation and with support from the University of Sussex.

What is Me, Human?

Do you identify as an upright, walking, talking tool using Great Ape? Well, like it or not, you are. My name is Gilly Forrester. I'm a professor of comparative cognition, and this podcast channel will be featuring some of the many research projects I'm involved in.

I'm proud of my ape identity because it's a unique window into who we are as human beings. And I really want to understand how our bodies and our brains work together to allow us to think, feel, act the way that we do. A big part of understanding who we are requires that we appreciate a shared evolutionary history with other apes. This has led to the inheritance of much of our modern skills and cultural traits.

Me Human challenges everyone to consider humans within this evolutionary framework, from infancy to adulthood, and asks these fundamental questions about who we are as human apes.

speaker-0: ⁓ The Academic Agency

speaker-4: Have you ever wondered what your first movements were like? Anybody who's spent time around young infants will know that they are incessantly wriggling, these involuntary movements happening all the time.

speaker-5: But what we now know is actually what they're doing is practicing every single muscle movement, every joint rotation, breaking the X, Y, and Z planes so that their bodies are preparing the muscles and all of the coordination to eventually make different postures which will allow them to navigate the social and the physical world.

speaker-4: single New research is exploring how these movements from birth are different for different babies and might hold the key to early autism screenings. They might also open the door to a host of new therapies and interventions.

speaker-0: I'm Jilly Forrester. I'm a professor of comparative cognition at the University of Sussex. And you're listening to the Me Human Podcast. Episode 1, The Baby Grow Project.

speaker-5: So what we've done with our study is try to capture in a very objective fashion what complexity and variability means.

speaker-6: ⁓

speaker-0: It's a novel attempt to better understand how babies' movements from the first days of life can support and predict their later development. We have over 100 participants all over the world, and they do the study at home in their own natural and stress-free environment. Our parents are asked to take weekly videos of their infants from birth wearing a custom onesie and placed on a Haffi yoga mat. where they clip their smartphone into a grip and take a couple of minutes of video for us. And in addition, they're answering questions from a survey called The Vineland, which tracks the baby's social and communication development over time. So there are two primary aims of this study. The first is that we want to understand ⁓ with greater clarity the relationship between the quality of early motor movements and later social and communication ability. Secondly, but just as importantly, we want to be able to detect earlier than we currently do markers that might be indicators of higher likelihood. for developing neurodiverse conditions like autism.

speaker-3: I'm Dr Hannah Rowan, I'm the research assistant on the Baby Grow project.

speaker-0: To successfully conduct a project as ambitious as the Baby Grow project, you really need a dedicated and talented team.

speaker-3: I saw a job advertisement and thought that sounds like something that I would be suited to, very much so.

speaker-0: Hannah knows these families inside and out. She builds incredibly intimate relationships with them from the first days of their babies' lives.

speaker-3: My background is in infant nutrition and you're often listening more to someone's emotional issues rather than what they're doing with their food.

speaker-0: She is the one responsible for bringing in the masses of the data that this project produces.

speaker-3: Basically from birth I've been calling parents every couple of weeks and asking them a series of questions on how their baby's developing. The violin is separated into sections. For example we have understanding and listening, expressing and talking, interpersonal relationships, have play and leisure, gross motor and fine motor skills. So those are the six sections we've been looking at. And as babies grow, the number of questions that I'm asking them expands. So at the start, it might be something as basic as, your baby looking at you? Are they making eye contact? Are they telling you when they're not happy about something by crying? Are they able to sit supported for a minute? Are they able to reach out for a toy? And you can build up a picture of when they're actually attaining these motor and interpersonal skills. And we're comparing the ages that babies reach these milestones between our high likelihood and low likelihood groups. So this morning we to a school called the Sons, they're two years old and they've been involved in the study.

speaker-2: driving out to meet one of the participants, Flo, her coming up

speaker-3: So for the first seven days, Flo would have been sending in videos of...

speaker-2: James weekly. At the same time we do phone calls.

speaker-3: with participants every couple of weeks where I go through survey. After a year we switch to an online form every couple of weeks and then parents switch doing the every month from 18 months to 30 months. He'll age out at the of then he'll

speaker-2: through the questions on the vine. to D**k Island. the study when he turns two and a half which will be next summer. ⁓

speaker-3: full set of data from James. ⁓ do an assessment at the end of the session with some questionnaires and a 30-month and then that's it for our participants. But at this stage they've been taken to the which is ⁓

speaker-2: study behavioural observation. taking part for two years. It's quite a long time to be involved in a...

speaker-3: study with your child.

speaker-5: project has really been founded on an understanding that there is a connection between movement of our bodies and the way our minds develop, so our cognitive development. It's important to explain that the children who go into our high and low likelihoods are based on whether or not there's a previous incidence of autism or other neurodevelopmental conditions in the family. We're particularly interested in individuals who... will go on to receive a diagnosis of autism because many people who gain a diagnosis will also have some sort of motor condition. So we've been talking with clinicians and training with clinicians on what they know about this already and what the clinicians see is that those babies don't have as much complexity or variability in the way they move their bodies. We're going to look for those differences and those patterns across those two baby groups. And the data will tell us, and hopefully this is one of the big reveals of the project, what is it exactly about the features of movement in those early days that are different in those two groups, what's different in the complexity and the variability.

speaker-3: me park. I

speaker-2: That's alright.

speaker-3: take our shoes.

speaker-1: So my name is Florence and I joined the study in the autumn of 2023 and I found the study online I think I saw an advert for it pasted and I just thought it looked really interesting at the time I was pregnant with my son James and I just thought it'd be a really interesting thing to do and a really nice way to contribute so I signed up and I think you contacted me Hannah saying Do you still want to be part of it? Because I think you didn't hear from me for some time because James was two and half weeks late so he didn't turn up until December even though I think I put my due date down as sort of mid-November so there was a bit of a delay there but as soon as he was born I sort of got in touch again and you sent me out the kit and we got started.

speaker-2: And how would you describe your involvement in the Baby Grow project to people that ask you about it?

speaker-1: I'd just say it was a study that we found that was involved in looking at the really early indicators of possible developmental disorders like autism and it was a way of studying James's movement when he was really small and still developing to see whether were any kind of traits that might in the future be a useful way of diagnosing developmental disorders. But it certainly wasn't anything I knew about at all so it was a, you know, it's been a learning curve for me as well. Because I was constantly thinking about the survey questions and watching James through the videos, it made me really think about how he was changing and developing on a pretty much a daily basis in some cases. I remember to start with, for example, he only moved one side of his body quite a lot and not the other side. And I remember thinking, ⁓ that's really interesting. And then a couple of weeks later, he started moving the other side as well. ⁓ And that's the kind of thing I just wouldn't have noticed, I think, if I hadn't been part of the study.

speaker-2: So you started in early January 2024. Was it very stressful to have these extra things to do whilst being the mother to a newborn?

speaker-1: Yeah, I think it was early January. I don't think so particularly. mean, being a mum to a newborn, it's really overwhelming to start with, but also you've got to remember their wake windows are quite short. So there's quite a lot of time when James was either sleeping or not really doing anything. So that was quite a good opportunity to kind of like plan my day and think, okay, great. Next time he has a little wake window, then we'll pop him in the seat. We'll do the video. ⁓ And it just meant that it gave us something in a way to structure our week by. I know that sounds really odd, but Sometimes it's quite helpful to be like, every week we know we need to do this video because the maternity leave can be a big kind of haze of goodness knows what you're going to do with your time. And it was nice to have that as a sort of focal point.

speaker-2: That's some feedback that I've had previously is that especially also doing the Vineland's every couple of weeks. Yes. It gives you something to focus on and to actually think about how your baby's developing. Because the questions are the same sort of every couple of weeks more or less they get they get more complex.

speaker-1: Yeah, no, It's quite encouraging as well to start with, know, when I think when you do the first survey it's sort of no, no, no, never, never, anything. And then after a couple of weeks they do start doing new things and start moving differently or being more aware or following you with their eyes and all the kind of things. And it was lovely to kind of feel like you were able to notice that because you were actually specifically being asked about it. So I really like that.

speaker-2: And then the communication was mainly over the phone, I think, for the first year or so. That's right.

speaker-1: And emails as well, also emailing the videos over, dropboxing them, getting replies and occasionally having to notify you if James has had a meltdown and we haven't been able to get them. Which occasionally happens.

speaker-2: Which happens. Yeah, of course. And how did you find that? Chatting over the phone every couple of weeks.

speaker-1: It was really easy, really nice. And the beauty of doing it on the phone, it means you don't have to be sort of in one place in one time. So I think we had some conversations and I was taking the dog out of the fields or, you know, I might have been between driving between places to take James somewhere. So it was a really nice and flexible way of doing it. And ⁓ like super convenient for me, super easy. And I think occasionally a couple of them I missed. So sorry about that. And you but you're always very kind about letting me reschedule and it was just

speaker-2: Yeah. and convenient for you.

speaker-1: It was a very stress-free experience, which is always good. One of the questions before we started the survey was whether either me or my husband Ian had any kind of signs or symptoms of developmental differences. Was it an issue that we were both barely standing?

speaker-2: No, we were looking for people with autism and ADHD in the family and those without as we're looking to compare two different groups and how their babies might show different signs. So our aim is to let all the participants know about any results that we have and hopefully in the next few months we should be having some more results that are published and so we're always happy to let our participants know what's been going on. So is this something you'd do again?

speaker-1: Yeah, definitely. that's good. tonight. It's study now. I'll do it again.

speaker-2: That's lovely.

speaker-0: It's important to say that we're not clinicians and therefore we don't provide feedback to individual families about their children because all of the methods that we're using, we're using as research tools, not as clinical diagnostic tools. However, all of our families are kept abreast of the findings of the wider project.

speaker-5: Hannah, we've had a really good retention rate of our families who have joined, yet a good retention rate ⁓ is still kind of what, 50, 60 percent of the people who first signed up. So that means actually quite a lot of families did drop out, regardless. Can you tell us a little bit about some of the challenges that parents faced when, in the early days of the project, why some people might have had to leave the project?

speaker-3: I think from a practical point of view, some babies just do not want to be put down. And part of the project is filming babies when they're lying on their back for five to ten minutes. And some babies just won't do this. They want to be held all the time. And so for those families we had to say, okay, this isn't working, it's stressful for you and the baby, absolutely, that's fine, no problem at all. For other mothers, they were struggling with post-natal depression. and were going through a tough time. again, for those parents, actually a couple of them have carried on with the project, but for some it was the best thing not to carry on at all. It was just adding to the burden of stress.

speaker-5: Were there any specific instances where you felt like you were able to retain a parent in the project because you were able to speak to them beyond your capacity as a research assistant?

speaker-3: Yeah, absolutely. think being a parent myself, I was able to empathise and sympathise with the struggles they were going through, feeling tired and just acknowledging the situation that they were in, which if you weren't a parent yourself would have been difficult to do. Funny, funny instances like, you know, a having a poon army in the middle of a filming session and having to stop and start again. And, you know, you can laugh about these things because you've been there yourself.

speaker-5: And I think it's important to say too that there's no shame in dropping out of studies when you sign up for them of this nature. They are a big commitment and we understand that. Our priority is always the family's well-being and the infant's well-being and that goes without saying that they need to do what's right for them. The project is secondary to their well-being. I think this is a really good example of why research is never totally straightforward. It always has obstacles and the data you collect is never fully complete in the way one would hope. Yet we've had enough babies join the project for a ⁓ long enough period of time that we've got an excellent and novel data set that nobody else has.

speaker-6: I'm Saber, I have a PhD in Movement Science and I'm a Data Scientist.

speaker-7: ⁓ in the Baby-Girl project.

speaker-6: A movement scientist is a person who analyzes usually human movements to understand the link between brain

speaker-7: behaviors ⁓ and by behaviors I mean movement because we know that there is bidirectional

speaker-6: Tractional link between brain and behavior.

speaker-0: Another hugely important member of our BabyCrow team is Saber. Now, Saber is a postdoctoral research fellow and data analyst, and he deals with the monumental massive task of dealing with all the video data that comes in.

speaker-6: We've been receiving two different sets. One of them.

speaker-7: data. was videos but the

speaker-6: and the other one was a questionnaire, a survey, which has been called Vineland.

speaker-0: He looks at similarities and differences in the way babies move based on which group they're in, or low risk.

speaker-6: So for the videos, let's imagine like 1,000 videos.

speaker-7: on average seven minutes. It means like 7,000 minutes of videos, on average.

speaker-6: When it comes to data analysis, we are going to look at the frame by

speaker-7: the frame it means data sets we have to multiply it

speaker-6: billions of then in 17 points, which each point x and

speaker-7: has, why, and can you imagine how many data points would be? Like billions of data points, I would say.

speaker-6: Machine learning techniques helped a lot when we worked with huge data sets. It is not easy to code all those movements manually. When it comes to thousands of hours of videos, who can code these huge amounts?

speaker-7: you know. data. It's like

speaker-6: impossible.

speaker-7: safe.

speaker-6: Machines can work 24

speaker-7: work for seven they don't get tired

speaker-6: and in some cases they are more precise

speaker-7: they are more accurate than human eyes.

speaker-6: specifically looking at the tying we can easily describe someone's movement as fast but we can't say how was the jerk of

speaker-7: Any movement. slow. movement, how was the acceleration of movement. Computers do this kind of analysis for us.

speaker-6: can do.

speaker-0: Now once Saber makes sense of all of the movement data that comes out of the infant videos, he's also tasked with seeing how it relates to the Vineland questionnaire.

speaker-6: So Vineland has different subscales. It provides us information about the development of movement and social communication and daily life. Looking at the literature, we know that there is a link between motor development, social communication,

speaker-7: Babies. mmm skills right

speaker-6: So later on, we are going to find if there is a link between those early movements, the Vineland questioners, and later social communication.

speaker-7: skills in our babies.

speaker-0: Jazz is a PhD student on the Baby Grow project.

speaker-8: So my name is Jasmine and I'm in the third year of my PhD.

speaker-0: Now her job is to code all of the joint movements in the videos of our babies.

speaker-8: had a vision of looking at the relationship between very early motor behaviour and social communication skills. Because we know that behavioural clinical diagnosis for conditions such as autism are pretty much based on the severity of social communication skills, which is usually detected around three years of age. But often to get a diagnosis can be really challenging and it's quite often that it's just more severe cases that gets flagged.

speaker-0: Jazz's coding provides ground truth. Now that means that she's actually being the human eye to see how these joints are moving.

speaker-8: So I am trained in the general movement assessment and so I was able to take a five minute video and narrow it down to the two most optimal minutes of the video. Sometimes in one video it would be a thousand movement points which sometimes honestly took days and it was so labour intensive, was a lot but it's not been done before, it's not been done yet. No one's been able to say, Here's an AI model of a general movement assessment. No one's been able to do that, but I think what they're missing is the human eye to detect it, to create these markers that will feed into an AI machine. Every elbow flexion extension, every wrist, radial ulna, all the different parts, the way the bodies move, I was creating a data point for that.

speaker-0: These data provide Saber's machine learning with the information it needs to make sure that it's trained effectively.

speaker-8: He's able to take those data points and create a model around that. Every parent watches their baby's motor milestones. When are they sitting? When are they rolling? When are they walking or grasping? But when it comes down to these general motor patterns, we don't know that much about them as parents. So don't think it's very publicly known. A lot of the GPs even that I've spoken to, they don't know what they are either. which just really surprised me.

speaker-6: So we have been interacting with people in the society and in many cases I have been surprised seeing parents who are not aware of their kids differences. And sometimes they blame them because of and in some cases they might

speaker-7: being different. label their kids, which affects.

speaker-6: obviously could their mental health. We believe the earlier we can

speaker-7: have a diagnosis for the neurodiversity kind of conditions like autism and ADHD, the better.

speaker-6: any kind of conditions. the sooner we can understand differences in

speaker-7: the different babies, we can inform their parents, we can increase the awareness in the people of the babies.

speaker-6: around

speaker-3: It's important so that these children who might be at higher risk of these conditions, so that their families are supported. If we can look at babies who have a higher chance of developing these conditions and identify them earlier, it might lead to earlier interventions, which is really important in terms of language ability, motor ability. If we can implement support for these children and their families, then that's a good thing,

speaker-5: We've got a tendency to think cognition, our problem-solving skills, our ability to speak and solve maths problems, that's where our special qualities are. That's what makes us humans and when there's a breakdown in that, that's when we offer a screening and or a diagnosis. But that stuff doesn't just emerge out of nowhere, it's built upon much more rudimentary brain systems and motor sensory behavior. So if you don't have that stuff as the building blocks, you actually don't get the cognitive bits. And that's been ignored not just in development, it's also been ignored in evolution. So it's a gap.

speaker-6: Mmm.

speaker-0: I think the Baby Grow project has many objectives that are important. One is that we better understand how early motor movements from babies associate with their social and communication development.

speaker-5: Two

speaker-0: is that we understand and work out whether or not we can see earlier markers of children at high or low likelihood of different neurodiverse conditions. But three is that it really forces us to recognize that our brains are not operating in isolation of our bodies. We are embodied humans. Our experience of interacting with the physical world, with our senses and our Motor movements are integral and form a foundation for cognition to emerge through development, just as it did through evolution. So we really need to pay attention to what might look like more rudimentary or simplistic forms of behavior as actually key components and building blocks to all of our sophisticated behaviors.

speaker-4: In the next

speaker-0: of the Baby Grow podcast. We'll be looking at

speaker-5: Yeah

speaker-4: the results from the finished study. We'll examining what this means, both for our participants and their families, to clinicians to see what it might mean for the future of diagnoses and interventions.

speaker-0: We'll be. but also talking. Thank you so much to all my team for working so hard on this study. Thank you to the Simons Foundation for funding our public engagement. And thank you for taking your time to listen to this podcast. The Baby Grow project is due to conclude in 2027. But if you'd like to read some of the papers that have been published around this work, check the links in the show notes. You can also find a link to the Me Human website which will keep you updated on developments of the Baby Grow project but also other research projects that I'm involved with. If you want to get in contact you can find my contact information on the Me Human website or follow me on social media, look for Julie Forrester or Me Human. The Me Human podcast was produced by the Academic Podcast Agency and was presented by me, Jilly Forrester.