Does Age of Smartphone Ownership Affect Child Health? With Ran Barzilay, MD, PhD

How do online environments shape children’s health and well-being? Which features have the biggest impact, and how can we make sense of the very different ways individual children experience the same digital world? On this episode of Screen Deep, host Kris Perry discusses new research into this “digital exposome” with Dr. Ran Barzilay, Assistant Professor of Psychiatry at the University of Pennsylvania. Dr. Barzilay explains how researchers are beginning to measure and quantify online environmental factors, cutting-edge findings from his study on adolescent smartphone ownership and health risks, and why making recommendations can be challenging when conclusive evidence is lacking. He also provides concrete suggestions for parents, including signs to look for when deciding whether it might be appropriate to introduce a smartphone. 

In this episode, you will learn:

• How researchers are using new data techniques to quantify parts of the digital environment that may affect child health and why that matters for families 
• What new, cutting-edge research suggests about the age of first smartphone ownership and negative health outcomes
• What emerging evidence is showing about other digital use patterns and child health - and what questions scientists are still trying to answer 
• How one leading researcher is  thinking about digital parenting decisions for his own family based on the new evidence

For more resources and research on this topic visit the Learn and Explore section of the Children and Screens website (https://www.childrenandscreens.org)

--------------

Follow Children and Screens on: 

Facebook: Children and Screens: Institute of Digital Media and Child Development
Instagram: @childrenandscreens
LinkedIn: Children and Screens: Institute of Digital Media and Child Development
X: @childrenscreens
Bluesky: @childrenandscreens.bsky.social

---------------

Music: 'Life in Silico' by Scott Buckley - released under CC-BY 4.0. www.scottbuckley.com.au

Transcript

[Kris Perry]: Welcome to Screen Deep, where we decode young brains and behavior in a digital world. I'm your host, Kris Perry, Executive Director of Children and Screens. Today's conversation examines how multiple stressors in a child's life interact, accumulate, and shape mental health over time, including stress from digital media. 

Our guest today is Dr. Ran Barzilay, a child and adolescent psychiatrist and translational neuroscientist. Dr. Barzilay is an Assistant Professor of Psychiatry at the Perelman School of Medicine at the University of Pennsylvania; a Clinician at the Children's Hospital of Philadelphia's Youth Suicide Prevention, Intervention, and Research Center; and the Principal Investigator of Barzilay Lab at the Lifespan Brain Institute. Dr. Barzilay studies how environmental exposures from family stress and adversity to social media use and cyberbullying interact with biology to influence mental health risk and resilience in children and adolescents. He's been a leader in advancing the concept of the exposome and, more recently, the digital exposome, which offers new insights into how cumulative digital stress may affect youth mental health and suicide risk. In this episode, we'll discuss what the exposome is, how digital media fits into a child's broader developmental environment, and what his research tells us for families navigating decisions like when to introduce a smartphone. 

Let's get started. Ran, you're doing important work in linking up how exposure to various stressors, including digital ones, are impacting children's health and development. What feels different about the research you're doing is that you're aiming to find ways to come up with a more complete picture of experiences and environmental exposure to help us understand how these experiences affect health. So, first off, in your work, you reference something called an “exposome.” Can you explain this concept of the exposome and how it relates to child health and development?

[Dr. Ran Barzilay]: Sure, Kris. The exposome is a conceptual framework that tries to quantify environment and its contribution to downstream biology and health. The term was coined just more than 20 years ago in 2005 by Christopher Wild, who is a cancer epidemiologist. And basically what he said is that, we know that some environmental factors, toxins, chemical, air pollution can cause cancer, but we also know that they tend to come together, so it makes very little sense to study them in isolation. So, he said we should study environment for its interconnectedness, for its network structure, and he coined the term “exposome,” which brings omics into exposures. So, like, we have genome to study genomics, which is the entire structure of the genome, not just one single gene at a time. There is the concept of the exposome, which talks about studying environment beyond the lens of one isolated exposure.

[Kris Perry]: How does the concept of the exposome add to or replace the old dichotomy nature versus nurture debate? And how is that related to wellness?

[Dr. Ran Barzilay]: So I wouldn't say that it replaces it. I think that it broadens the understanding in which we can try to learn how genes and environment shape development of health. It's never – I mean, we know now that it's not just nature and not just nurture, it's a combo of them. It's not just the genes, it's not just environment. The challenge is that if we want to combine both biology and environment – genes and environment, nature and nurture – we need to come up with methods that will allow us quantification of environment in a way that can be one, more reliable, more reproducible. So, if different people around the world study environment, they know that they study a similar construct. And this is, for example, when you think of biology of genetics, when we study genetics here or another place in the other side of the Atlantic, we use the same genetic code. But when we talk about environment, it has so many facets to it, it's so many contexts with cultural context. So, the exposome framework tries to solve this problem by giving us a framework where we can be at a better place to advance the science of environment to the level where it can be on the same stage like genetics, like what you call nature, biology, in one holistic framework because it's not just environment and not just genes.

[Kris Perry]: Did something happen that drove your interest in this, and this framework that you're describing?

[Dr. Ran Barzilay]:  Yes, it wasn't – I would say it was not one thing, it was more my own developmental trajectory as a neuroscientist and child analysis psychiatrist later. It was more, I would say, a sequence of events in my training. Some of them were clinical encounters with patients. 

I can give a couple of examples. I remember, for example, as a very young and junior resident doing a night shift in an inpatient psychiatric hospital, I was walking with the attending who came at the middle of the night to go over some patients in the ward. And we discussed a patient that was in very, very acute distress. And I was very junior, new, not ashamed to ask questions of what can explain this very acute, extreme presentation. And I asked – where I came from, which is more the biological side of training, I came from a lab where I studied stem cells – whether there is something genetic, whether there is something that is, like, a very unique presentation. And then this mentor of mine, later she became a very experienced attending, she said, “Well, it can be that, but it can also be very severe trauma earlier in lifetime.” 

Now, it was then I learned from the perspective of a very experienced clinician how to think of a very strong environmental adversity as something that can shape behavior long term because the patient was an adult. And then later in the clinic, when I started seeing young patients and I became more experienced, I worked a lot with kids on the autism spectrum. Now, autism is an example of a neurodevelopmental condition that has a very well established, strong genetic basis. No one argues that autism is environmental, this is a highly genetic disorder. But even so, in my population of kids that I had the privilege of following them as they developed, I consistently saw that the environment in which they developed meant family context, educational context, had critical implications to their developmental trajectory. And sometimes I said, “this is even more than so-called the biology they were born into, the environment plays a great role in shaping it.” So this was a clinical influence that I got to try and focus more on environment. Also, I think there's something optimistic about environment. I know that oftentimes people think that it's not the case, but for me, environment is most – or a lot of environment is modifiable. So, these clinical encounters led me to focus more on environment. 

And then from my research side, I learned more and more that there's emerging evidence originally in the context of trauma and ACEs – what's called “adverse childhood events and experiences” – with growing literature over the last, I would say now 30, 40 years that trauma and adversity, especially in earlier timepoints in life, early in the lifespan have long lasting health implications. 

So, taking this together, my clinical experience, my research knowledge, I became more interested, “Okay, how can we now take this concept of environment and bring it to the level that we can integrate it into translational neuroscience, into biomedical research, as something that can really help us one, understand, and hopefully will lead to more impact on interventions?” So, that's the story.

[Kris Perry]: You mentioned that the exposome attempts to represent, essentially, a quantification of environment. Other fields have done similar things, such as adverse child experiences, which you mentioned, also called ACEs. Tell us what an exposome score is and how is it calculated, and how does it relate to other scores, such as an ACEs score across child well-being research?

[Dr. Ran Barzilay]: So, going back to the exposome framework and to Christopher Wilde, basically the idea is that exposome is an umbrella term for everything that is non-genetic. So, everything that you experience in your environment – it can be the immediate psychosocial environment, it can be the more structural air you breathe, climate you live in, political climate. It can also be physical exposures, like chemicals. Everything can be conceptualized under the exposome. Whereas ACEs, or trauma, are more limited to a certain set of – in ACEs it's very, I would say, hardcore adverse experiences that were just, in a single study, now a few years ago, were measured, counted, quantified, and have shown the relationship to health outcomes. The exposome is more a framework. You look at the data and you ask yourself, “What is it in this data that can be conceptualized as part of the exposome?” And then, it can be a data set that may have, I don't know, 5,000 measures of environment ranging from in-utero exposures, physical exposures, psychosocial exposures, chemical exposures, structural exposures, like the geo-coded address of where you lived as a kid that is connected to many external data sets. 

So, you take all of these measures and you run through data-driven, in most times, sophisticated methodologies. The overall idea is that it can either be done just based on the correlations of the network of environmental exposures or it can be done in a specific relationship to a specific outcome. So, for example, in one of the recent studies that my lab has published, we calculated an exposome score for risk of adolescent suicide attempts. And to calculate the score, what we did, we screened hundreds – and, in one scenario, even thousands – of environmental and lifestyle factors and saw how they relate to adolescent suicide attempt. And after we do all this data-driven screening, what we end up with is a bunch of measures or exposures that may be associated with suicide attempt. Then what we do, we take the ones that were associated with suicide attempt and we lump them up in a score. But the good thing about this score is that it's not just that it includes many different exposures that are relevant, it also can assign a weight to what is the significance of this exposure to suicide attempt. Because for example, as you can imagine, an ACE would get – which is an adverse childhood experience – would most likely get much larger weight on the overall score than something that is less salient. Let's say, for example, nutritional habits. Let's say that we find something there. So, the exposome score takes everything together. 

And the concept of the exposome, by the way, another advantage that it also allows you to identify protective environmental and lifestyle exposures. And it's very, very important in biomedical research when we think of environment that we don't just paint environment in dark pictures. Everything is traumatic, everything is adverse, and these are the only things that contribute. We know it is not the case. There are several environmental components that are fostering, enabling, enhancing resilience, promoting resilience. 

So, the exposome framework, by taking everything we have, really gives us a chance to identify both risk and resilience factors.

[Kris Perry]: I really appreciate that you brought up protective factors, not just risk or adverse experiences. So many of those play out every single day, right? Parents are trying all the time, providers to create a protective environment for kids. So, we know that plays a part in their future success and well-being. 

You're one of the leading researchers on how the digital exposome, specifically, is impacting youth and adolescents. Why are you focusing on the digital exposome?

[Dr. Ran Barzilay]: Well, I think this is more observations from life, first as a parent to three kids across the developmental spectrum. I have my oldest is 18, she's a senior in high school, and my middle son just started high school is 14, and my youngest son is nine, third grader. And looking at them, they also represent some shift in the digital environment because as I mentioned, my oldest is 18. And it is very clear just for me as a parent that when I think of environment these days, I don't know if it's gonna be too much to say that most of the environment of the younger generation these days is online. If I want to be less provocative, maybe I will say at least half is happening online. I mean, in terms of what influences how kids develop, and when you think specifically about adolescence when it's about identity formation and it's about figuring out our social place, making connections, learning to live with feedback from our peers, experiencing rejection – at least half of it is done online. 

Now, if you add to that the COVID pandemic, where we were forced to move many interactions online because of necessity, you come to a point that me as a researcher who studies the exposome, and did not start the research on the digital exposome, I was thinking to myself, “Okay, if you want to measure environment, but you, let's say you focus on ACES, you entirely missing huge chunk of what is environment these days to kids.” So then we came up with this concept of the digital exposome and how it relates to the offline exposome and how we as scientists try to understand how environment shapes health in teens growing up in the digital era, I would say we must consider this a big chunk of the environment, if we really want to capture the complexity and the entirety of the environment, if we try to live by the motto of the exposome framework that is everything is environment that is not genetic.

[Kris Perry]: What aspects of the digital exposome seem to have the most impact on children and adolescents?

[Dr. Ran Barzilay]: So, here I will be more careful in my responses because we still do not know and this is a field that is now I think just in its first steps, baby steps. But we do know some things. So I want to separate my answer to this question to at least two different domains. One is what we know about patterns of use that are more associated with worst health outcomes. And the other would be more content of, what do you do on the digital environment and what apps you interact with. 

So, let's start with the first one, patterns of use. So we know, for example, that if the pattern of use becomes problematic – which is a very vague term and I use it intentionally and I don't use the word “addictive” because it's very, very difficult to quantify what means addiction, so, I want to stick with problematic – so, if we think of use that becomes just too much, it takes control over one's life, it interferes with day-to-day functions. A kid – by the way, it's true also for adults, but as a child psychiatrist, child adolescent, I prefer talking about youth – a kid is staying all night long, goes to sleep 4 a.m. on a weekday because they are on their phone, that's a problem. I mean, you don't need me to say that as so-called expert, correct? Because we realize something is off. A kid that has no offline experiences, almost at all – everything is done on a screen. Again, I don't think we need me to say something is off here because the kid loses on interactions that are important, that are done in the offline. A kid that cannot be one minute without their smartphone even when the family wants to get to a dinner, or where they need to do something and engage in an activity that requires not having a phone around them, and it becomes, he or she become very agitated about it. So, these are the type of behaviors that relate to use of smartphones or screens that we can say they are problematic. And when I talk about the digital exposome, you can see it's a very broad, soft concept. So, I also refer to the way of use, not just to the exposure themselves. I would say, we can say comfortably now that we know, and we have sufficient evidence to say, that these levels of problematic use are not ideal and have bad associations with health outcomes. I think this is now in consensus. At the same time, I want to emphasize for most people and most kids, they do not meet the criteria I just described. And I say to myself as a clinician, we should be especially or particularly mindful for that. Because as clinicians, we see the kids who come with problems most of the time. But when you think broadly on society, for most kids, they're not addicted to their phone. So, we need to remember that. So, this is one part of my answer to your question about what is it about the digital exposome. 

Now for the second part, more about content and what is actually happening online, we have far less concrete evidence there. One thing that we know is that cyberbullying, which is basically bullying done in the online environment, is bad for mental health, and there's a lot of literature on that. And also from our lab, we had a paper a couple of years ago showing that cyberbullying, over and above offline bullying, is associated with greater chances of suicidal ideation and suicide attempt in early adolescents. And this is one of the first studies that really tried to tease apart the digital from the offline world on the same stressor, which is bullying. So, we showed that cyberbullying over and above. So, cyberbullying is bad for mental health, this is something we feel comfortable knowing and saying now. Regarding the actual content and use, the data there, I would say, is still not there. And I know that there's been a lot of debate, and there's been a lot of literature. But to really say that we know what is worst in terms of the digital exposome to the mental health and the physical health of kids, we're still not sure and the data is being collected as we speak.

[Kris Perry]: Your recent study on age of smartphone ownership and correlation to negative physical and mental health outcomes is helping reframe the discussion about how and when to introduce smartphones. Can you tell us about this specific study and its implications?

[Dr. Ran Barzilay]: Sure. So what we did is we leveraged a very unique, precious resource of data from the Adolescent Brain Cognitive Development Study, called ABCD. In a nutshell, this is a study that follows 12,000 American kids from across the United States from age 9 to 10 into late adolescence. The kids are now in the mid to late teens, and the data that's available to crunch now for researchers, such as myself and my team, is up to around age 14. In this paper that we recently published, we looked at the question that the parents were asked, the kids are followed annually. So when the kids were ageing the ABCD study, the caregivers were asked, “Does your kid have their own smartphone?” And if the answer was, “Yes,” the second question was, “How old were they when they got their first smartphone?” So what we did in this research is we took this data and we combined it with data on health outcomes and we tried to answer a very basic questions. When you look at kids at age 12 in ABCD study and you compare health outcomes of those with to those without a smartphone, do you see any difference? Having a nine-year-old, I had my own motivation for this study because my nine-year-old kid – he was eight back then, but we saw that it was coming – wanted their own smartphone because he looks at his older siblings and he knows, because it's a talked about topic in our family, that they got their smartphone when they're around the age of 10. And at the same time, Jonathan Haidt’s book came out, The Anxious Generation, and this whole debate that you cannot ignore with potential concerns about smartphone use and mental health and health of kids. 

So we thought, okay, let's use ABCD study to try to address this question. And what we did in this work is we compared very, I would say generic, simple clinical outcomes to represent both physical and mental health. So we looked at depression, we looked at obesity, and we looked at insufficient sleep. And what we found is that across the board, at age 12, number one, 70% of the kids in ABCD already had their own smartphone, which aligned with other data from other resources in the US. So 70% had a smartphone. And when you compare this 70% to the 30% who were still without a smartphone, the kids with smartphone had consistently greater odds of endorsing depression, of meeting criteria for obesity, and of meeting criteria for insufficient sleep – at that age range, it's defined as sleeping under the nine hours per night. 

So, a major strength of this work is because ABCD study is such a rich resource of information. It includes so much data and so many children that it allowed us statistically to control for gazillion confounders that one can think of – oh, did you find this difference because of this or you found that difference because of that? So just to mention a few, we control for age, for sex, for several socioeconomic variables. We controlled for several parental factors. We controlled for ownership of other devices, like iPads. And we also controlled for pubertal stage because we know that puberty plays a role in all of the outcomes that we looked at and also at the chance that the kid had a smartphone. So this was one finding that I think was very compelling because it was at age 12 – and most kids in America, and I would say globally, have their own smartphone before age 12. So we showed that age 12 is not ideal. 

Now what this paper also allowed us to do is go beyond age 12 and say among the kids that by age 12 have a phone we went to the second question – “what age was your kid when you first gave them their phone?” And we saw that the younger they got the smartphone the more likely they were to have obesity and insufficient sleep at age 12, which may suggest that it's not just by age 12, but it's also how early in development the kids got their own smartphone. And these two pieces of information in that scale were new at the time this was recently published. 

However, the biggest, I would say, strength of this data, of this paper, is that we also had, again – and this is thanks to the amazing ABCD study and the NIH and the researchers who did it and the families who contributed their time to it – it followed kids over time. Now, I'm a very lucky person for many reasons, but one of the reasons is that I have very smart collaborators who are also friends now after a few years working together, and one of them is Professor Samuel Pimentel from UC Berkeley who is a statistician expert in causal inference. What he does is – his expertise is how to use data from observational studies. Not experiments, but just studies that follow people over time, like ABCD, and how to design analysis that improve the level of evidence that we can say this is likely to be causal. And I say it very carefully – likely to be causal – because it's very difficult to say for a hundred percent or with a hundred percent certainty that observational data can give you cause and effect. 

So what he suggested is, “Ran” – when he saw the data he just mentioned at age 12, he said, “That's okay, but what I want you to do now – I want you to take all the kids who by age 12 already have a smartphone in this dataset and just throw them out of the analysis. I want you only to follow the kids without smartphones at age 12, and I want you to show me what happens next year when they are 13 to their health outcomes. Because I'm sure that in this year between age 12 and 13, a big chunk of them will get their smartphones. If you show me that by the end of age 13, even when you control for how they did before they get smartphones, you see an effect for smartphones, then I will start taking this research more seriously.” And I will say that this is a very high bar of evidence. And I was very skeptical that we will see something, but amazingly we did. And we saw that among the kids who did not have smartphones at age 12 – and there were more than 3,000 kids there, it's a large cohort – around half of them received their smartphone by age 13, and the ones that received smartphones had more mental health problem and more sleep problems at age 13. Which, for me, I would say now as a parent, is a sufficient level of evidence to say, “I will not give my nine-year-old a smartphone before they turn 13 just because the data that we were able to generate as a research team suggests that it has bad health outcomes, implications.”

[Kris Perry]: I'm thinking about so many levels of that answer, from how important it is that we support a study like ABCD and thousands of children are tracked over many years and that we're funding NIH and other scientists to collect that data over many years so that over time, new scientists like you come along with new questions. And you're thinking about new impacts on childhood and comparing and contrasting different kids with each other so that we can better understand these sensitive periods, the interaction with digital devices, and how they're doing in real life. So I really appreciated how many variables you brought into that last answer. 

But, more importantly, families have been struggling with this question just like you, who have nine-year-olds, 12-year-olds, 18-year-olds, about when to own a smartphone or what will the pros and cons be of owning a smartphone at certain ages? And we have data that helps us understand that. It's not all anecdotal. This is exactly the information that so many families have been asking for. And you note that parents should pay attention to these findings so that they make the very best decision for their own child. Are you using this in your own parenting with your nine-year-old?

[Dr. Ran Barzilay]: Yes, for him, and he already knows that. People ask me how he takes it, but he's used to it, because he knows the results of this study well before it was published. He will not get his first smartphone before he's 13. And by the way, good for him that it's not me that's making the decision, it's mostly his mother. And we discussed this together and it's just now that we have better data than we had a few years ago when his 14-year-old sibling was nine, we didn't know that. So he got his smartphone around that age, and I tell myself, and I tell other parents who ask me, “Oh, we missed the train – we gave our kid a smartphone.” Well, we did our best and it's okay. It's not detrimental. It's not deterministic. 

So we try, in medicine, to follow the principle of evidence-based medicine. And we can only follow evidence-based medicine guidelines when the evidence is there. And there just has not been strong enough evidence, I feel, to really give a strong recommendation. And again, I'm very careful not to over-claim. We did not do a randomized control, placebo control trial, when we give half of the kids smartphones and half of the kids a placebo phone and compare them. 

But in reality, we know that this is not a study that is really feasible. So at the end of the day, every family, policymaker will need to draw the line and say, “When is the evidence that we have sufficient to do something about it?” I know that for myself as a parent, what I just shared with you is sufficient, but I'm not saying that this is true for everyone. There's definitely different families with different considerations. Difference can come with different levels of maturity. Not every 12-, 13-, 11-year-old is the same. Families can have different considerations. The need for smartphones can be for communication, the need can be for safety reasons. So, each family will need to make their own decision. What we try to do as scientists, as clinician scientists, is generate evidence that can inform families at the individual level and, hopefully, policymakers at the biggest societal level because we have a problem to solve. How do we help kids grow up healthy into adulthood in the digital era? Because we're not going back. I mean, it is the digital era. So we need to help them grow and develop to adulthood with smartphones because they're part of life and technology and be healthy. So this is the science that we try to do.

[Kris Perry]: There's something about 12, 13, 14 that is very significant developmentally. Then products are introduced and widely adopted, and we're basically told by industry that 13 is the ideal time. That's when you should have quite a bit of access to many of the things on the smartphone. In fact, many of those things are social media. What is your research showing about specific uses of social media and their relationship to mental health outcomes?

[Dr. Ran Barzilay]: So I wish you will invite me again, I hope to say next year or in two years, because we're just embarking now on a study that was recently funded exactly to test causal effects of social media use on teen mental health. Currently, the data that we have is weak. And when I say we, I put myself in the bigger group of the science community. It's just not there. And there are many papers that come out with conflicting findings. I would say it's from the data perspective, from the evidence perspective, it's still an open question. I am very, very careful not to overstate what is known and not to understate what is still less known. I don't know, I mean, you made a point about tech companies telling us 13 is the best age. I mean, again, this is not the evidence that we live by, but I cannot make strong claims currently against this suggestion based on what I told you, because, for example, not social media, but just talking on smartphones. And I want to emphasize our paper was on smartphones, not on social media use. We did not have this information to include in our statistical models. Our data shows that until age 12, inclusive of 12, which is under 13, ownership of smartphones seems to have adverse health associations. That's it. And here I pause and I let people who ask me about it digest. And I say, “That's what we know. Now I can tell you some opinions, but I wanna make the distinction when I say, when I mention or describe some opinions or some hypothesis, but I can say that I currently back them up with science.”

I do know that we really need this science. And as I shared, we're very excited that we just got a grant to study that with Professor Pimentel that I mentioned before, using data from smartphones themselves to see how they relate to mental health. But currently, we really don't know. We can guess, but we really don't know. And when we guess, one thing that I want people to remember, we should also guess the potential positive aspects of social media, because it's very, very easy to go to the direction that it will be all just bad. But we should know that there will be some potentially good aspects to it. Connection, interactions, access to knowledge. Now, I know that all of them could potentially be double-edged sword, but oftentimes I feel as if the discussion is really unidirectional and I say until we have the evidence, we should be very, very careful. And we're gonna have some very strong evidence because some countries and some states have really taken some big decisions lately. So I really, really hope that the people and the politicians who have enacted these different policies will put in infrastructure to allow researchers to transparently crunch data and share it with the public. Because maybe they will find that, indeed, under age 16 is very bad to use social media. Maybe, but maybe not, we still don't know.

[Kris Perry]: Well, industry knows. They have the data. We would really all benefit, scientists like you would benefit from having access to that data so that we could make quicker, safer, better decisions for kids. Because we know that they go through these sensitive periods and we want them to have the very best health outcomes possible. So I really appreciate that you're being careful to only state what you know to be true at this point. And we will absolutely call you in a couple of years and have you come back and talk to us about the additional data that you've been able to analyze so that we can better understand these impacts. 

You talked a bit about the challenge of extending limited research findings into clear action steps for decision makers. And I want to dig into that because we care deeply about improving access and policies to protect kids. You've touched on how the data available through the ABCD study and some specific analysis approaches are helping you and your team stretch beyond correlational outcomes. Some people point to the prevalence of correlational research or the conflicting findings that you've mentioned to argue that there isn't enough evidence to determine the relationship between social media and mental health, for example. This stated lack of evidence is then used as a barrier to action or even just clear recommendations like an appropriate first age to get a smartphone, for example. What is your take on that evidence debate in addition to funding additional research? What can we do for kids during this period of lack of knowledge?

[Dr. Ran Barzilay]: I thank you for this question because I think that's the key question at the end of the day. When do we say we have enough evidence with all of its limitations to do something that is more likely to benefit the health of the kids than to harm them? Again, I'm thinking as a physician – first, do no harm. So whatever we want to do, we want to make sure that we increase the chances that it will do good for the kids. Let's take smartphones, for example. It's easier for me to speak about because it's our own individual research that I feel the findings for me as a parent are compelling enough to take an action. We are not going to have a placebo control trial. Giving a kid a smartphone is not a medication. So sometime we need to say – we cut and we take a decision. So I think that, for this specific question about the age of recommendation, this is something that is reasonable to say, that under age 13, the evidence currently suggests that there are adverse health associations. And even though it's not a hundred percent causal evidence, again, for me as a parent, it's good enough. So when people ask me my opinion, I rather tell them what I'm going to do – again, it's not just me, we as a family, me and my wife – what we are going to do with our under 13 kids. So this is one. 

Now, regarding other things or broader context, I would say, society oftentimes looks for a very concrete guideline. And we need to also – to keep some nuance in the messages that we make, that there isn't a magic number. But at the end of the day, it shouldn't be used against taking any action because we can say this is a number that is probably relevant for many people. And maybe for some families, there will be exceptions. There are some concrete things that we can do when we give the kid a smartphone. Oftentimes, I'm being asked by parents, “Oh, we missed the train, we gave our kid a smartphone when he was 10. Is it all lost?”  I say, “No, it's not all lost. You can make certain modifications that we know based on evidence,” but also, I will whisper, “Based on common sense, it's also okay–” I mean, we know that smartphones affect negatively sleep. Keep the child's smartphone  out of the bedroom at night. This is a very concrete, actionable recommendation that I think as a family, if it's decided upon, it's something that I am sure there will be evidence to support it pretty soon. Let's say that. 

Number one – by the way, nothing as good as role modeling for the kids. So I know and I'm not pretending to be perfect at it, but if the kids see that the parents themselves as well put the smartphone aside when they go to sleep at night or when there's family dinner together, that will go a long way with the team. So things like that, I think, can be already translated into action. 

Regarding social media, again, I'm not running away from this even though we don't have the evidence, but it's also important for parents to remember the decision to give their child a smartphone is not synonymous with giving the child unlimited access to what's on the smartphone. So, for example, you can decide that the child will have a smartphone, but you can make sure that they do not have a social media account. Again, I'm not saying we have evidence to support doing that, but I'm saying if you decide as a family, this is important for you, create some set of rules around it. And you can judge based on how your kid manages these rules in other aspects of life to see whether they're ready or not to get a smartphone. For example, do you need each evening to get into a conflict on doing homework, taking a shower, coming to dinner, going to sleep. I mean, all of these things. So if everything is a conflict, it probably implies, I would say, it suggests that introducing the smartphone into the equation will not solve the problem, but rather there will be another conflict. If the child or the teen is in a place where agreements can be made, decisions can be taken together, I am an advocate of also motivational approach with everyone, especially with teens. This is for your health. So my nine year old is still not an adolescent, but getting there, we talked about this study. He was annoyed that he will not get a smartphone for a few years. But you know, of course in our house we talk about this paper and quite a lot and you know, It's for you. I'm not making it up. It's for your health, physical and mental health. Let's see how we can help you grow up in a healthy manner in the digital era. How we can help you develop good tech habits.” It's a process, it's not a one-time thing. What I tell parents, following the smartphone, turning data into action is think of smartphones as a health behavior. How they engage with technology – and I say smartphone because smartphone is a pivot in how we engage with technology. Definitely, when we think of ourselves as adults, everything is on our smartphone. Think of that as something that has direct implications for health. And when you think of it as such, discuss it and communicate it to the child or to the teen. And I think then the conversation is different because it's not, “We are banning you, we're not allowing you.” It's more, “We care for you and we're gonna take some decisions, some of them together to help you develop into a healthy adult in the digital era.”

[Kris Perry]: I cannot believe what a great way you framed that was. It makes it so much easier to understand why you have to be careful when you bring a smartphone into your child's life. It is a health behavior and health is connected to so many other outcomes. 

What do you think should be studied next and most urgently about the digital and non-digital exposome that could help children lead healthier lives in the digital age?

[Dr. Ran Barzilay]: Again, I'll divide it into two. One is what type of data we need, and the other is what we think it is important to study. For the first, we need data from the smartphones themselves. And again, ABCD is amazing, but most of the study population did not provide data from their devices. So it relies on self-report. It can only get you a certain way. I'm not belittling it – I think that our study and other studies, for example, really generated amazing data for society to use, but you can only get so far when you rely on self-report of use. So we need research that uses data that is pulled from smartphone. And you mentioned the tech companies, they will be pivotal in enabling the acquisition of such data. Again, pulling data from smartphone is one thing. It still doesn't solve your – how you measure health in kids. You need to close the loop, but at least the data should be derived from the phone themselves ideally. So that's in terms of what type of data. 

In terms of what in the digital exposome, I'll say the immediate suspect that comes to mind is the use of AI, and specifically, generative AI. And this opened up a huge Pandora's box, we still don't know, but again, I want to highlight it's not necessarily automatically all bad because generative AI has so much potential to transform knowledge, information, also potentially – and hopefully – to mitigate inequalities. It can do all of these things, but it's exactly for this reason that the power it has is immense. And we know that kids interact with it. And we know that kids interact with it in ways that no one imagines they would. They can make it their friend, they confide in it, and it can be very intimate and can be very scary, but it can also be amazing. So this is something that I think we need to get a lot of data on. And we also in my lab are trying to generate this type of data as we speak. 

And again, what in the offline environment amplifies or affects the effect of the digital exposome? So I think this is a great area of exploration, and I'm not familiar of anyone who studies that. So maybe we'll do it as well. I think like in any other research, we  need to understand, we need precision medicine. So we need to understand exactly what is it in one's digital exposome that is more salient for specifically them and not for the neighbor or their friend, but for them? So we need to pass the heterogeneity. Maybe there's some things in the digital exposome that may be good for some kids and bad for other kids. When you think of social media, I love to give this example. When you think of a kid who belongs to a minority or marginalized community and in the offline world, they have no validation. They cannot speak to anyone, but they find it online. That's a lifeline potential. But at the same time, you can think of this person being cyberbullied. So it's very, very important that they identify exactly what are the nuanced exposures in the digital environment that make a difference between individuals. 

So I think that I gave you a topography of where I feel I think the field needs to go. Last thing that I will say is that there's this whole world of using phones as vehicles for interventions and technology as interventions. Again, technology brings concerns, but it also brings tons of opportunities. And we know that we have a mental health crisis in youth, and we know that we don't have enough solutions. And we know that this generation is growing up digitally with a much smaller attention and shorter attention span. So they will most likely benefit more from digital intervention. So we need more, more data on what works, what does not work. 

And last, we have some pretty good data on what works. We need more implementation science data that actually tests how we can implement this good pilot studies that’s shown some proof of concept into scalable, real-world solutions. So we have a lot of work in front of us. I think it's going to be very exciting because it's a very exciting era, of course, given that we'll get some funding to do this research and I think and I hope that policymakers will realize that this is an important area that requires funding for more research.

[Kris Perry ]: Wow. Thank you, Ran, for sharing a powerful framework for understanding how children's health is shaped not by any single factor, but rather by the accumulation and interaction of many factors over time, including those in the digital environment. Dr. Barzilay’s work on the digital exposome helps move the conversation beyond simple screen time debates toward a more nuanced understanding of risk, resilience, and individual differences. 

If you'd like to learn more about the research discussed in this episode or explore evidence-based resources for families navigating digital media, visit childrenandscreens.org. And if you found this episode helpful, please consider subscribing to Screen Deep or sharing it with someone who's thinking carefully about children, technology, and mental health. I'm Kris Perry. Thanks for listening, and we'll see you next time on Screen Deep.