Dr Umar Toseeb Hello, welcome to the Podcast Series for the Association for Child and Adolescent Mental Health, or ACAMH for short. I’m Umar Toseeb, Professor of Psychology. My research focuses on special educational needs and mental health in childhood and adolescence. Today, I’ll be speaking to Dr Bettina Moltrecht and Dr Mauricio Hoffmann, co-Authors of the paper, “Social Connection and its Prospective Association with Adolescent Internalising and Externalising Symptoms an Exploratory Cross-Country Study Using Retrospective Harmonisation,” which has been published in the JCPP. The paper will be the focus of today’s podcast. Bettina, Mauricio, thank you so much for joining me. Dr Bettina Moltrecht Thank you for having us, hello. Dr Mauricio Hoffmann Thank you. Dr Umar Toseeb Shall we start with a round of introduction? Tell us what you do and your research and professional backgrounds? Dr Bettina Moltrecht So, I’m Bettina Moltrecht. I’m a Senior Research Fellow at UCL, so the University College London. I’m based at the Centre for Longitudinal Studies, and there, I’m, sort of, doing population mental health research, and when I’m not there, then I’m based at the Anna Freud Centre, where I’m doing, sort of, more child and adolescent mental health research, around, sort of, digital interventions, sort of, more clinical applied work. Dr Mauricio Hoffmann So, my name is Mauricio Hoffmann. I am Professor of Psychiatry in the Federal University of Santa Maria, South Brazil. I am a Clinical Psychiatrist, as well, active, my research is around child and adolescent mental health with longitudinal studies. I used to work more with educational outcomes, now we are more applied to measurement, and try to test different measures to better detect and predict mental health problems. Dr Umar Toseeb Thank you, both. The paper is about social connection and adolescent mental health symptoms, or mental health difficulties, and one of the things that I liked about it was your conceptualisation of social connection, as in, you didn’t just look at individual factors, you put them under this broad umbrella of “social connection.” So, do you just want to tell us about your definition of “social connection,” what does it mean in the context of this paper? Dr Bettina Moltrecht Yeah, that’s a very good question. We, kind of, used the work by Holt-Lunstad, I hope I pronounced that correctly, because basically there are many discussions around what social connection is. I think, in my head, it’s a fundamental human need. We all need to be socially connected. There’s a big, sort of, challenge in society today, where people are experiencing, sort of, loneliness, they’re disconnected, but we know that it’s, sort of, important for our mental health and, sort of, across the lifespan. And then, while we were trying to, kind of, get into the data that we had available and our research questions, we were trying to understand, what does social connection mean when we have to measure it? And it really, kind of, consists of multiple facts, not – so it’s not just one thing that you can measure, and then we found that paper by Holt-Lunstad, and they have basically, sort of, defined it as – under three different dimensions “the structural, the functional and the quality,” and we thought that fit quite nicely with our research question and the data we have available. At least, sort of, if we talk about the “structural dimension,” you have things like the size of the social network, so how many friends does someone have? Or, in our case, we had something we call “household composition,” so that could be who’s living in your household? Is it siblings, your parents, or maybe even, like, a grandparent? And then the next level is the functional, so that’s just, sort of, do you get social support from your relationships? Maybe if you're a parent and your neighbours support you with childcare, then there’s also a function to that relationship. And then the last dimension is the quality, it’s not just about how many friends we have, but that the quality of our relationships is important. You know, you could have ten different friends, but if they’re – you know, if you don’t have a good relationship with them, or it’s, sort of, with high conflict, then that – you know, it’s a different relationship, or does it – has a different impact on your mental health than two friends that, you know, you know you can rely on and you have a, sort of, a trusting relationship with. So, these are the three, kind of, dimensions there. Dr Umar Toseeb Thank you, and then, in terms of mental health, you talk about “internalising and externalising symptoms,” in the context of this paper, what did those look like? What did you specifically measure in terms of internalising and externalising symptoms? Dr Mauricio Hoffmann So, first, the concept, right, internalising symptoms, there is a broad concept, that put together depressive, anxiety symptoms, usually, right? So, internalising is the process of you don’t expose much in the externalising world, so you feel all in your mind, so that’s the idea, made by Achenbach in 66. And then externalising symptoms will be more the behavioural side of things, like, in this context, we use the SDQ instrument, so if you cheat, you lie, then you do these kind of things that you can observe, right? And internalising symptoms is more, like, fear, crying a lot, these kind of things. So, you use five symptoms each, so you use this instrument to take this aspects of psychopathology, basically. Dr Umar Toseeb Thank you, and I was having a conversation with a colleague recently around “externalising psychopathology,” and I’d written something where we’d talked about “hyperactivity and inattention being forms of externalising psychopathology,” and the colleague said, “Inattention isn't really high – externalising psychopathology.” And I don’t know what to make of that, and I just wondered what your thoughts were? Dr Mauricio Hoffmann You are going in the rabbit hole. So, this – yeah, this is – in terms of – like, there’s a lot of people working in the structural aspects of psychopathology, like, the HiTOP consortium, etc., so people don’t know how to place the inattentive side of things. Like, example, the HiTOP measurement, the HiTOP structure, for example, don’t have a proper place for that. But in terms of factor analysis, you can factor in – like, it tends to converge, the eight symptoms, like, inattentiveness, lack of concentration, and these kind of things, tends to converge to externalising symptoms, but are not really the same thing. So, what we used – like, myself included, but some people, that is actually a separated construct really, and sometimes it’s in the between, like, concentration – like, for example, if you take the symptom of concentration in itself, lack of concentration is a symptom of generalised anxiety disorder, it’s a symptom of depression, so, yeah, and a symptom of ADHD, and it’s not a symptom of conduct disorder or oppositional defiant disorder, these kind of things, so it’s in – more or less in the between. But in terms of factor analysis, you can’t put it to together, we didn’t put that together, we did actually separated the ADHD part of the SDQ. So we only included the ex – proper behavioural – this concept of behavioural symptoms, not the hyperactivity ones, not these kind of things. So, I can find arguments for putting together with externalising or not. In terms of the conceptualisation, I think it’s really don’t belong to the externalising aspects of behaviour. Dr Umar Toseeb Thank you, and you’ve done this study on two different cultural contexts, so you focused on the UK and Brazil. Why these two specifically? And why is it important to look at two very different cultural contexts, with reference to the specific questions that you're interested in? Dr Bettina Montrachet So, I think there were multiple reasons. So, when we started out with the project, we developed this tool called “Harmony.” The idea behind it was that – to really encourage more global mental health research, because especially if we look at young people, we know that 90% of all children and young people live in low and middle income countries, but actually most of our research is conducted in high income countries. And if you, sort of, take that a step further, thinking about interventions, we can’t really expect that, like, taking an intervention that was developed, let’s say, in the UK, that it would work the same way when you place it in Brazil. So, that’s why we, kind of, said, “We need to, kind of, have more of that research that tries to compare different studies, so – or countries, not just studies, but countries, so that we have a better understanding of what works in which countries, where are the differences, where are the similarities?” And why Brazil and UK? Well, one is part of the high income countries and the other one is a low and middle income country, but then there were also the practical, sort of, reasons, because we knew each other and – but I was, sort of, more familiar with the Millennium Cohort Study. I had not heard of the Brazil Cohort Study before, and then Mauricio, well, you had worked with both I think before, and that we thought was a good, sort of, way, bring the two studies together. And I guess another practical reason was that we could align them in some way, so the data – they’re sort of, they’re roughly similar in terms of the age of the participants, when they were born. I mean, there’s slight differences in the study designs, but it seemed possible that we could combine the datasets, as well. Dr Mauricio Hoffmann Yeah, so there is the two types of question, right, Bettina? The more elegant one, which is comparing low and middle income countries with high… Dr Bettina Moltrecht Yeah. Dr Mauricio Hoffmann …countries, but there is a more or less boring one, which is the practical one, it’s, like… Dr Bettina Moltrecht Practical, yeah. Dr Mauricio Hoffmann Yeah, we get together in this grant, and the grant demands to do these kind of things, to have people from different places besides the UK. And Brazil suits the bill quite nicely, especially because it’s the largest country in South America, it is a low and middle income country, and the cohort fits – it’s possible to harmonise with the Millenium Cohort. This is a key thing of harmonisation, right? We had some variable, and especially because data was collected more or less at the same time. So, we don’t have different cohorts really in terms of year of birth. Like, the Brazilian cohort is an accelerated cohort who have different ages, different birth, but the mean is the same as – the mean age, the mean year of being – everybody was born almost matched the same year as the Millennium Cohort. So, this fits the bill quite nicely, and it’s possible then to harmonise and it’s not every chance we have to do these kind of things. Dr Bettina Moltrecht Just to add, we also explored this with – not all of them, but with a Researcher based in South Africa, and there we just had to, kind of, realise that there wasn’t enough overlap in the datasets. So we just couldn’t do it, because they didn’t have enough mental health, sort of, measures in there that we could have done the same thing. So, this was just a very good opportunity for us in terms of the data that was available, to answer, like, this sort of new research question, yeah. Dr Umar Toseeb And you’ve already touched on this already, so let’s expand on this. Data harmonisation, what is that? What does that mean? And what did you do in this study? Dr Bettina Moltrecht I probably should say, we’re talking about “measurement harmonisation” in the first instance, because sometimes – data harmonisation is quite a broad concept, and I – sometimes people confuse it. But we’re basically talking about “retrospective measurement harmonisation,” so that means that, in this case, we’re looking at the two datasets and what measures they have used, and whether they’re even comparable. So, we often, sort of, use this example that if you want to measure depression, there are more than 200 measures that Researchers, Clinicians, could use, to measure depression, so different studies might use a different measurement. So, just because, you know, one study uses measurement A and the other one uses measurement B, that both claim to measure depression, doesn’t mean we can just put them together and compare them. So, that’s why we do the retrospective, sort of, harmonisation. In real life, this, kind of, looks like, sort of, two re – and this is what we did as well, Mauricio and I, we went through the, sort of, metadata, all the, sort of, study item descriptions in both studies, to identify what is common in the first round. We wanted to know what’s actually measured in both, just on a broader concept scale, does Brazil ask about friends? Does the Millennium Cohort Study ask about friends? What does Brazil say about household? What does, sort of, the Millennium Cohort Study say about household members? So, then we went through the PDF files really, and Excel sheets, to figure out what they both have in common. And then you go onto, like, an item level, where you go really, sort of, item-by-item, trying to compare, do they actually say the same thing? So, you know, it could be that one study says, “I have trouble relaxing,” and the other study says, “I’m unable to relax,” or something, so these are similar, but they’re not necessarily the same item. So that’s where the harmonisation comes into play, where we try to, kind of, take an item-by-item approach to say, “How similar are they actually?” And then once you’ve identified the items that you think are semantically similar, then you go to, sort of, the next stage, and that’s where – we’ve described it in the paper as well. So you have, sort of, your – you measure your invariance, are they actually, sort of measuring the same thing in both studies? That’s not the semantic level anymore, so I hope I haven't lost anyone by that description, but… Dr Umar Toseeb And so, just to clarify there, once you’ve got past the semantic bit, where you’ve – on an item level and say, “Are these two items asking about approximately the same thing?” And you’ve gone onto the next level, is that the measurement invariance bit? And is that the same as measurement invariance testing, or is that something else? ‘Cause I think what I was wondering was, when you were going through the review process, I imagine a hardcore quantitative reviewer being, “What is this?” Like, “This is – what – where’s the numbers here? Where the model fit statistics here?” All of those things. So, what’s the process there? So, I think – I understand the qualitative bit of two people or two Researchers agreeing or disagreeing on the semantics of what’s being asked, and then the quantitative bit, what does that involve? Dr Mauricio Hoffmann Just a step back, I told some friends that we did a harmonisation paper, and they said, “What? You are harmonising – facial harmonisation is something that people do in Brazil in the lips and stuff, but you're doing for data, you are making up data?” No, that is not it, right? So, data harmonisation is this process, as Madam Bettina was saying, like, we have different instruments, for example, GAD-7 to measure anxiety and scared. So, Thomas developed this tool in our group, do that automatically, so we put the questionnaires, the metadata there, and then it matches the items that are more similar, based on semantics, using natural language processing, that’s the thing, right? So, this is the process of harmonising instruments at the item level. The other thing is the quantitative part, which is to test if this final measurement are equivalent, so this is the measurement invariance part. For this study, we already have the same instruments in both cohorts, which was the SDQ, so we didn’t need to make harmonisation for SDQ. We just tested if the SDQ is understood the same way in Brazil in the UK, basically. So, differences is in the levels of the SDQ were solely due by the differences in the symptom level, not because people in Brazil were understanding differently the items, or something like that. Then we did our measurement invariance testing. All that are in the supplementary material in the paper. After we did that, we used the gathering variant scores to the analysis, right? So, what is that? That means that the relationship with the item, with the construct, which is internalising symptoms, are the same in both countries, and the levels of difficulty of the item. So, for example, for me to endorse, to answer that item that, “I cry a lot,” that’s totally true, and this is a very difficult item in terms of psychometric speaking. So, what is difficult item? It’s – I need a lot of symptoms to endorse or to respond the item that way, so that’s totally true. This is a threshold parameter which is set to be the same in both countries. So, this is all technical stuff, but to mean that the scores that we have, we ensure that the differences between countries are just because of differences on symptoms, not other – interferences of other things, right? Cultural aspects, blah, blah, blah. That’s the quantitative part and the measurement invariance. I don't know if I… Dr Umar Toseeb Yeah… Dr Mauricio Hoffman …answered… Dr Umar Toseeb …that… Dr Mauricio Hoffmann …the question, but that’s what we did. Dr Bettina Moltrecht And I just want to emphasise, so we’re doing both, so we do the… Dr Mauricio Hoffmann Yeah. Dr Bettina Moltrecht …semantic and the quantitative, you can’t just do one of them, so to say, that would not be the full story. And our colleague, Eoin McElroy, he has published, a few years ago, a very nice paper on how to do it… Dr Umar Toseeb Hmmm. Dr Bettina Moltrecht …step-by-step. Dr Umar Toseeb That’s fine, and you mentioned the “tool” for the first part. Dr Bettina Moltrecht Harmony, yes. Dr Umar Toseeb Is that available for people to use? Dr Bettina Moltrecht Yes, Harmony is for free. It’s an opensource project, so different people can contribute to it, if they’re up for it. We have actually now contributors from across the globe, which is really nice. We have an art package as well, so people can definitely use it, and should. Dr Mauricio Hoffmann These kind of things that we did, like, Bettina and I went through the questionnaires, is, like – have 400 pages of PDF. Dr Bettina Moltrecht Hmmm hmm. Dr Mauricio Hoffmann So, this takes months, right? So, this is – was how it was done in the past, and let’s say Eoin did a lot of that in the past, harmonising a lot of British cohorts’ instruments, so Harmony can do it, and just upload the PDFs or the Excel files or whatever, and do the – do that for you in seconds. That’s the practicality of Harmony. Dr Umar Toseeb Very, very helpful. And then in your paper you talk about “involvement of lived experience experts,” and I’m always fascinated by studies where it’s a secondary analysis of existing data, and then you get people with lived experience involved. Because I think a lot of the time, the lived experience experts are involved in projects that are primary data collection, and then it’s obvious how they can be involved, whereas, in secondary analysis projects, it’s less obvious. So, can you tell us about your experience of what that was like, and what that involvement was about? Dr Bettina Moltrecht It was a bit of a riddle to me why they aren’t involved, because I came from a, sort of, more clinical, applied, sort of, research. I was very used to involving, sort of, lived experience experts in my research, and then, when I moved to the research of using secondary data analysis, I noticed that that wasn’t very commonly done. And I think it’s actually quite straightforward to do that. So, I also wrote a blog post on how – to give people ideas on how they could do that. So, how we did it, in the first instance, so we worked with the National Children’s Bureau in the UK, but then we also conducted workshops in Brazil. And, in the first instance, we just explored with the young people the whole concept of social connection, which sounds very abstract, right? So, we wanted to just, sort of, understand from them what relationships do they have, what relationships are important to them? How do different relationships impact their mental health according to their experience? So, once we, kind of, clarified that, we then showed them the items that we had available in the cohorts, and we let them work with the items to help us, sort of, develop the research question, but, also, you know, they said to us – one item that we hadn’t included at the beginning was moving address. And they, sort of, gave us the idea that we should add that in, because if you as a young person move to another place, then you have to – you're losing friends, you’re – you know, you’re find – you have to find new friends. So – and that’s why we added that one into our analysis. I think it’s very rewarding, and I think they always have very good ideas that we don’t think about. And then we basically did the same workshops in Brazil, and I think, especially with this cross-country, sort of, study, that’s really interesting too, because you have young people speaking from the different contexts, and already there you get a better understanding of what do you do with the data, how do you also interpret the data later on? Dr Mauricio Hoffmann Yeah, yeah, it was a challenge for us, as well, like, at the beginning, we need to include people with lived experience, but we didn’t fit to include them in the tool development, in the Harmony, let’s say. But for the research question of this paper, it fits totally fine, because with them, we derived some of the variables that we look upon in the questionnaires, right? So, for example, the moving part that Bettina said, it was brought up in the UK workshops and the Brazil workshops, people – ‘cause Brazil is the different thing – this is one of the differences that we’re discussing about later on, but, for example, in Brazil, it’s a large country, right? It’s a continental country with very difficult mobility. So, like, in the UK, you can take a train and go to Cornwall or whatever, Leeds, and you can just go to take a train and go, right? So, in Brazil it’s not that way. You have buses, you have planes, which is expensive and not for everyone. So, once you go study in a different city, you’re go – only going to see your parents one or two years later. You stay a long way from home, so – and especially for students that goes to schools. Or when you move, for example, and for military and you’re a son of a military person, you move a lot, right? Or a bank, your father have a job in the bank, you move a lot in Brazil, and you totally lose your connections, and this was brought up here, as well. Another thing was brought up was living with half siblings. It was not an effect that we found in Brazil, but was brought up in the Brazil workshop, because pe – some of the young people said, “Oh, my mother gives more attention to the step-son,” as in, things like that, so some of them felt cast aside. So, this thing was interesting, and we look up, they do have these variables, and these differences was brought up also in the – in this part of the study. Dr Umar Toseeb So, we’ve talked a lot around the context and the rationale and the process of the research, let’s just get into the research questions and the findings. What were you specifically interested in and what did you find? Dr Bettina Moltrecht First of all, finding any social connection factors in the cohorts, in both of them, and how these – the ones that are common across both of them, how these relate to mental health outcomes in adolescents in the two countries. And in the first instance, we looked at it separately in each country, but then we also looked at it combined, so is the diff – is the impact the same in both countries, or are there specific pictures or associations that are very unique to one of the countries, so to say? In the end, with the help of the lived experience experts and the data we had available, we had 12 different social connection factors. Very broadly, in terms of the findings, and to, kind of, say, come back to the, we need to understand what’s similar and different across countries, I would say the one factor that stood out the most is the bullying factor, that that played a big role in both countries. And then there were specific factors which were quite unique to either Brazil, or, sort of, UK adolescents, and then – so, in Brazil, for instance, we had the household size was, sort of, a more unique, sort of, association, which we didn’t find in the UK, so to say. Dr Mauricio Hoffmann Well, yeah, this is the main things that we want to look at, so this step-by-step process in the paper, right? We have to find the variables and then to analyse if the social connected variables relates to this – to our outcomes. And the important thing is that we did that in many ways, and adjusted for important things, like, for sex, for maternal education, we adjusted for prior symptoms, which is quite interesting. So, what we found was these social connectedness variables predict new symptoms that your previous symptoms do not predict, so above and beyond that, and was it – you adjust for these, kind of, things. So, we used also weights to adjust for the representativeness of the samples, and we did that in many angles. Like, as Bettina was saying, we analysed for separated for each country, and we did a meta-analysis as well, and the individual level. This is the more interesting in the main results, that was – Bettina was saying we – is actually what we found in terms of the bullying aspect was – bullying predict more internalising and externalising symptoms in both countries, but also it had a special effect in UK. Dr Bettina Moltrecht Yeah. Dr Mauricio Hoffmann Being bullied or done bullying predicts externalising symptoms in both countries, but more in UK, so we have an interaction effect there. And also we found, for example, in the Brazil, the more people in the house predict more internalising symptoms. This could possibly have socioeconomic interactions there, which we didn’t test, but it’s possible that we have some different hazards that UK, you don’t have. And the bullying thing is very interesting in terms of the discussion, because we found some materials that we mention in the paper that there’s different types of bullying, right? So, in the UK, it’s more frequent the type of bullying that you exclude someone from the social connection, and bullying relates more with excluding that person from the group, right? In South America, Latin America, you have more bullying about physical aspects, but you don’t – you call someone “fat,” but you also call him to play football, right? You don’t exclude them. So, perhaps these differences might explain why bullying in the UK is more problematic than was in Brazil. Dr Umar Toseeb Excellent. Thank you, and for the people who are listening, who are Researchers or Practitioners, what are the implications of your findings? Dr Bettina Moltrecht I think, in the first instance, I would say it supports the idea of – that this type of research, this cross-country comparison, is important, and I think I see this as a first step now. With some of these concepts, and you can probably tell from – you know, when Mauricio shares his understanding of how things work in Brazil, or how, you know, society is set up there, and then if we compare it to, like, UK, even if it’s, sort of, about transport and things like that, it shows you how much difference there is. And this is the first step to identify which broader differences are there, and then we need to, kind of, zoom in further to better understand it. Because then, in the long run, you know, this is what we need to really inform, sort of, interventions and policies and things like that. Dr Mauricio Hoffmann I agree with Bettina. I think there is two main messages from this paper, right? First is the method, right? How to harmonise, how to include people with lived experience to select the variables, how to put data together from different countries, and then do the analysis, in multiple ways, to see – to take an understanding from the data. And I think for – adjusting from prior symptoms is very – really important to understand a bit of causality and trying to exclude reverse causality parts of things. So – and the other thing is the main message. So, for example, we found that move to a different city, which is something that people with lived experience told us that was important, and then we picked up, is related with increased internalising symptoms later on, right? So, child that moves to city – from city, it’s important for both UK and Brazil, and this child should be protected or looked upon. Like, they are at risk for increasing their internalising symptoms, that’s prac – as an example of main message. People that suffer bullying, people that doing bullying, they are at risk for internalising symptoms, and the UK should look that more closely, because it’s strongly linked with the symptoms in UK. I think at least look for some of these factors and try to identify people at risk for higher levels of symptoms, right? And this, as a whole, children in Brazil have more symptoms than the UK. So the latent variables that we looked at, we look at the latent means, Brazil have like a standard dev – a half standard deviation worse internalising symptoms, and a standard deviation higher in externalising symptoms. And children in Brazil are more symptomatic, for many reasons, but, nonetheless, we are predicting people that will increase their symptoms later on, and these social connected variables are important. Like, some, we didn’t found any results, like, parental death, we did not found any results, and widowing, we didn’t found any results, but some we looked and we found, like, half siblings. Brazil have twice half siblings living in the house than in the UK. It’s 25% and the UK. It’s 30 – 13%, so we have these differences, but, nonetheless, these variables are strongly linked with these outcomes anyway. Dr Umar Toseeb In addition, for me, what I liked about this, and what I took away from your paper was the importance of having not only people with lived experiences as part of the co-design process, but people, Researchers, who are embedded in the different contexts, as part of the research team. Because often, I mean, like, I think one of my criticisms with any, sort of, grant that I might review, where there’s some involvement of people with lived experiences is that, well, if you don’t have people as Researchers in the research leadership team who are embedded in the context, you're not asking potentially the right questions. Because if you just have people with lived experiences on your advisory panel or your co-production panel or co-design panel, then you’ve already decided on the questions and you're asking for their opinion. But, actually, if you have people who are embedded in the context at the research team level, then they are involved in designing and asking some questions that you might not think of in the UK context, for example, and I’ve taken that from this, because I think it’s – I just don’t think it can be said enough. It’s about having people who understand the context at all levels of the research, on the research team, in the lived experiences, etc., etc., not just in one aspect. So, that’s what I took from this as well. And what do you have coming up that you’d like to make people aware of? Dr Bettina Moltrecht Yeah, probably quite similar or related to the whole cross-cultural, sort of, question is we’re currently working on developing guidelines on how to guide Researchers to, you know, harmonise and pooled datasets from – studies from different countries, because we found that there are no guidelines yet. So, there are guidelines on, you know, how to generally, sort of, do the harmonisation, but there’s not much guidance yet on how to account for cultural differences, on all levels, right? So, not just in the harmonisation stage, but also when you analyse it and also when you interpret and make sense of the findings, so I – that’s currently what we’re working on. If there are any experts out there that would like to, sort of, share their knowledge or their opinions on how this could be or should be done, then we – yeah, we would be very happy to hear from them. And in terms of Harmony, we are currently exploring it with different countries, so we’re also trying to test Harmony out with a research team in South Africa, because there are different languages, and some languages have the word “depression,” for instance, and some languages don’t, so how do we overcome that? Generally, we’re just trying to expand Harmony as a tool because we think it’s very useful for Researchers, so if anybody’s keen on contributing to that, then we welcome this, as well. Dr Umar Toseeb That will be really interesting to read, ‘cause I think, again, like, the data harmonisation stuff, like you say, there aren’t guidelines out there, so for Researchers it would be very helpful. Can I tell you what I would like to see? Not necessarily… Dr Bettina Moltrecht Yes. Dr Umar Toseeb …from you, but from anyone, is from a place of a person, a Researcher, who uses secondary – or does secondary data analysis, existing datasets, as a day job, like, I use this all the time is, I would like to see a paper that comprehensively talks about the drawbacks and limitations of secondary data, because it’s… Dr Bettina Moltrecht Hmmm hmm. Dr Umar Toseeb …not the answer to everything. Dr Bettina Moltrecht Hmmm hmm. Dr Umar Toseeb And I think that as a person, and maybe you and everybody else as well, as people who are involved in it so much, we might not always see why secondary data isn't the answer to everything. And I sometimes – like, when I’m speaking to students, or potential PhD students, or potential collaborators, I’m, like, “Yeah, secondary data’s great, and I use it all day, every day, but it doesn’t do everything, and there are lots of problems with it.” But, actually, I can never put into, like, a coherent argument, or a coherent structure, what’s wrong with secondary data? So, it would be nice if somebody would do a paper on, “These are the problems and imitations with secondary data,” it does have problems and you can’t do everything with it, but you can do a lot, and there’s lots of positives, as well. And, finally, what’s your take home message for our listeners? Dr Mauricio Hoffmann From this paper, I think the take home messages are it’s not a tutorial paper, it’s not a primary paper nor anything, but we try to give some possibilities of how to answer these type of questions, in terms of methodology, and we tried to pinpoint which people who have some social connectedness factors perhaps are at risk for few years later developing more symptoms, right? So, perhaps looking at – like, policymakers looking at these kind of things, and particularly in some places might be important to prevent higher symptoms, it’s assuming causality, right? But, like, bullying, moving from town-to-town, household compositions, these kind of things, like, housing is an important thing, right? So, peo – to put five people together in a one room place, might be problematic for the children. Dr Umar Toseeb Bettina, Mauricio, thank you so much for taking the time to join us. Dr Bettina Moltrecht Thank you. Dr Mauricio Hoffmann Thank you. Dr Umar Toseeb As always, please visit the ACAMH website, www.acamh.org, and on X, @ACAMH, where you can find out more about Dr Jackson. ACAMH is spelt A-C-A-M-H, and don’t forget to follow us on your preferred streaming platform, let us know if you enjoy the podcast, with a rating or review, and do share with your friends and colleagues.