Publications

Neiders, I., Mežinska, S. & van Haren, N.E.M

Mental health prediction tools are becoming more advanced, aiming to identify individuals who may be at risk of developing serious mental illnesses such as schizophrenia, bipolar disorder, or major depression – potentially even before any symptoms appear. While these tools offer the promise of earlier intervention, they also raise important ethical and social concerns. To better understand these issues, Neiders and colleagues conducted a comprehensive review of existing literature. They reviewed and analysed 129 academic articles from various fields to identify common themes related to ethical and social concerns. The review found that, while predictive tools for severe mental illness offer potential benefits, such as early detection and the possibility of preventive care, they also raise significant ethical and social challenges. These include the risk of misinterpreting results, which could lead to unnecessary anxiety, stigma, or even discrimination. The use of artificial intelligence (AI) in these tools adds complexity, especially when algorithms lack transparency or are trained on biased data, potentially reinforcing existing inequalities. People’s reactions to learning they are at risk can vary widely, sometimes negatively affecting their mental well-being, identity, or life choices. Additionally, there is a clear need for better education, communication, and professional guidelines to ensure these tools are used responsibly and ethically. While prediction tools could help prevent or delay serious mental illness, several ethical questions must be addressed before they are widely implemented. These include ensuring fairness, protecting privacy, avoiding harm, supporting informed decision-making, and considering differences in mental health care systems and cultural contexts. Future research should focus on gathering real-world perspectives from people at risk, their families, and healthcare professionals to better guide the ethical use of these tools. Currently, a new publication is in preparation, detailing the findings of a survey conducted among child and adolescent psychiatrists, focusing on ethical concerns related to the use of AI-based prediction models in clinical practice.

Allegrini AG, Hannigan LJ, Frach L, Barkhuizen W, Baldwin JR, Andreassen OA, Bragantini D, Hegemann L, Havdahl A, Pingault JB

This study examined how genetic risk for neuropsychiatric conditions can contribute to emotional and behavioural difficulties in children, such as anxiety or restlessness, across generations. The focus was on understanding both direct genetic inheritance and indirect influences that come from the family environment shaped by parents’ own traits. Data from nearly 15,000 genotyped families in the Norwegian Mother, Father and Child Cohort Study (MoBa) were analysed by testing polygenic scores. A polygenic score is an index of a person’s genetic predisposition to a specific trait or condition. It is calculated by adding together the small effects of many genetic variants across the genome, thereby providing an overall measure of genetic risk, rather than focusing on a single gene. By testing associations in these polygenic scores in parents and children, and relating these to maternal reports of children’s difficulties, the study aimed to distinguish inherited genetic effects from environmental influences linked to parental genetic risk. The findings showed that children’s emotional and behavioural challenges are influenced not only by the genes they inherit, but also by aspects of the environment associated with their parents’ genetic profiles. This means that even if certain risk genes are not passed on, the family environment influenced by those genes can still affect a child’s development. Some of these effects were broad, influencing a range of difficulties, while others were more specific, e.g. linked only to symptoms of hyperactivity. These new findings are consistent with the presence of risk pathways that can be targeted early in development to disrupt the transmission of psychiatric risk. Future research will need to confirm these findings using different study designs and independent samples. The aim is to better understand whether the observed indirect genetic effects are truly due to so-called genetic nurture – where parents’ genetic traits shape the environment they provide to their children – and, if so, to identify which aspects of that environment have the greatest impact.

Poortman SR, Jamarík J, Ten Harmsen van der Beek L, Setiaman N, Hillegers MHJ, Barendse MEA, van Haren NEM

Dahl Askelund A, Hegemann L, Allegrini AG, Corfield EC, Ask H, Davies NM, Andreassen OA, Havdahl A, Hannigan LJ

Budin-Ljøsne I, Fredheim NAG, Jevne CA, Kleven BM, Charles MA, Felix JF, Flaig R, García MP, Havdahl A, Islam S, Kerr SM, Meder IK, Molloy L, Morton SMB, Pizzi C, Rahman A, Willemsen G, Wood D, Harris JR

Bakken NR, Parker N, Hannigan LJ, Hagen E, Parekh P, Shadrin A, Jaholkowski P, Frei E, Birkenæs V, Hindley G, Hegemann L, Corfield EC, Tesli M, Havdahl A, Andreassen OA

Poortman SR, Setiaman N, Barendse MEA, Schnack HG, Hillegers MHJ, van Haren NEM

Children of parents with severe mental disorders, such as bipolar disorder or schizophrenia, are at increased risk of developing a mental disorder themselves. These children show structural brain alterations, but it is still unclear how these differences in brain anatomy develop. To study and compare this non-linear development over time, Poortman and colleagues used brain scans (magnetic resonance imaging, MRI) on children and adolescents with at least one parent diagnosed with bipolar disorder or schizophrenia, as well as on those with parents without these disorders (control group).             
In the brain, the outer layer, known as the cortex, primarily consist of grey matter, while white matter is located closer to the centre and connects between grey matter regions. Children of parents with schizophrenia showed different development over time of the thickness of the cortex and volume of white matter compared with children in the control group. Specifically, cortical thickness of the middle temporal gyrus, a brain structure important for face recognition and reading, developed differently in children of parents with schizophrenia compared to the control group. However, this difference was no longer observed after correcting for the overall thickness of the brain. Moreover, regardless of their age, children of parents with schizophrenia showed smaller total brain volume compared to children of parents with bipolar disorder and those in the control group, and a thinner cortex overall compared to children in the control group. These findings suggest that familial high risk of schizophrenia is related to smaller brain structures and altered patterns of brain development. It is important to follow up on children of parents with schizophrenia as they grow up, beyond the typical age of illness onset for mood and psychosis-spectrum disorders. This will help us better understand two key areas. First, how brain development is linked not only to high familial risk of mental disorders, but also to how severe the symptoms might be, and whether these patterns can help predict who might develop mental disorder later in life. Second, it could help identify resilience factors that protect against mental disorder, which could lead to more personalized interventions.

Raballo A, Poletti M, Preti A

Wootton RE, Dack K, Jones HJ, Riglin L, Madley-Dowd P, Borges C, Pagoni P, Roth C, Brantsæter AL, Corfield EC, Stoltenberg C, Øyen AS, Davey Smith G, Ask H, Thapar A, Stergiakouli E, Havdahl A

Simon R. Poortman, Marjolein E.A. Barendse, Nikita Setiaman, Martijn P. van den Heuvel, Siemon C. de Lange, Manon H.J. Hillegers, Neeltje E.M. van Haren

Offspring of parents with severe mental illness, such as bipolar disorder or schizophrenia, are at higher risk of developing psychiatric illnesses themselves. Reasons for this could be genetic predisposition and the increased burden of environmental stress. Recent research has found that young children of parents with severe mental illness have disrupted brain connections, but it is still not understood how these brain networks develop in this high-risk group. To study and compare this development over time, Poortman and colleagues used brain scans (magnetic resonance imaging, MRI) on children and adolescents with at least one parent diagnosed with bipolar disorder or schizophrenia, as well as on those with parents without mental illness. They created detailed maps of brain connections and analysed these maps to understand how the brain’s structure and connectivity differ between these groups. The study revealed that children of parents with schizophrenia showed a subtly different development of brain connectivity, especially in terms of strength and clustering, compared to children of parents with bipolar disorder and those without mental illness. In particular, the overall efficiency of their brain connections developed differently compared to children of parents without mental illness, and their local brain connections developed differently from those of children with parents who have bipolar disorder. These findings suggest that children with a familial high risk of schizophrenia show unusual patterns in the way their brain networks develop over time. It is important to follow children from these offspring at high risk for mental illness studies beyond the usual age when mental illnesses start. This could help to understand how their brains develop in ways that either increase their risk or help protect them from mental disorders. This new knowledge could lead to new strategies for preventing mood and psychotic disorders, reducing risk, and strengthen resilience.

Hughes AM, Torvik FA, van Bergen E, Hannigan LJ, Corfield EC, Andreassen OA, Ystrom E, Ask H, Smith GD, Davies NM, Havdahl A. 

Rodriguez-Ayllon M, Neumann A, Hofman A, Vernooij MW, Neitzel J. 

Defina S, Woofenden T, Baltramonaityte V, Tiemeier H, Fairchild G, Felix JF, Cecil CAM, Walton E. 

Mulder RH, Kraaij R, Schuurmans IK, Frances-Cuesta C, Sanz Y, Medina-Gomez C, Duijts L, Rivadeneira F, Tiemeier H, Jaddoe VWV, Felix JF, Cecil CAM

van Houtum LAEM, Baaré WFC, Beckmann CF, Castro-Fornieles J, Cecil CAM, Dittrich J, Ebdrup BH, Fegert JM, Havdahl A, Hillegers MHJ, Kalisch R, Kushner SA, Mansuy IM, Mežinska S, Moreno C, Muetzel RL, Neumann A, Nordentoft M, Pingault JB, Preisig M, Raballo A, Saunders J, Sprooten E, Sugranyes G, Tiemeier H, van Woerden GM, Vandeleur CL, van Haren NEM

Children of parents with mental illness are more likely to develop a mental illness themselves. This so-called intergenerational transmission of mental illness is not given adequate attention in clinical settings, diagnostics, or childcare. This results in delays in identifying mental health issues in young children, missing opportunities for prevention through protective measures and resilience building. This is where the FAMILY project steps in. The EU-funded FAMILY project is a collaboration between researchers from Europe and the US with the goal of understanding why, how, and when mental illnesses are passed from parents to children. The project focuses on changes in the brain, the epigenome, and genetic and environmental risks, comparing children of parents with and without mental illness, and using relevant animal models for research. The project also uses modern technologies like artificial intelligence and machine learning to build a prediction model to help understanding risk for and resilience against mental illness. This model is supposed to estimate the likelihood of a child developing a mental illness if their parents are affected. In addition, the FAMILY project looks at the social and ethical issues related to predicting risks. Overall, this work aims to prepare clinics and hospitals for the potential future use of predictive tools.

This consortium paper summarises the FAMILY project aims to achieve three main objectives:

• advance our understanding of why, how, and when severe mental illnesses are passed down in families and identify the best timing for preventive and intervention measures,
• create statistical models that help predict which children are more likely to develop mental illnesses at specific times given certain risk and resilience factors, and
• provide insights into the social and ethical implications of predicting mental health risks.

Kalisch R, Russo SJ, Müller MB.

Click here to watch the 15-min video summary.

Oblong LM, Soheili-Nezhad S, Trevisan N, Shi Y, Beckmann CF, Sprooten E

In recent years, researchers have widely used Genome-wide association studies (GWAS) to explore how millions of small genetic variations across our DNA influence the individual structure and function of the brain. However, these genome-wide associations are challenging to interpret because they result from combinations of many complex biological processes, influenced by both genetic variants and the environment. To address this, our study introduces a novel method called “genomic independent component analysis” (genomic ICA). This approach transforms genome-wide brain associations into simpler, more reproducible structures termed “genomic components”. First, we optimized the genomic ICA algorithm to ensure the quality of these components. Next, we assessed whether these components could be reproduced in independent samples and compared their performance with traditional GWAS outputs. Finally, each component was analysed, to check whether there are links with certain aspects of structure and function of the brain. Our analysis revealed improved reproducibility of genomic components compared to traditional GWAS results. Moreover, we identified specific combinations of genetic variants that collectively influence distinct aspects of brain structure and function, such as cortical thickness and white matter structure. Overall, our study shows that we have successfully developed a new, data-driven method that can transform large gene-brain association data into simpler structures, which reflect the joint influence of genetic variants on distinct brain features.In future analyses, this can help to better understand how genetic variants interact with each other and with the environment, and to gain insights into the biological processes underlying brain related conditions. Next, we plan to use a method called polygenic scoring, to learn more about individual differences in the genetic influences on the brain, and how this relates to mental health. This will help us understand how risk and resilience for mental health problems are passed down in families.

Luo M, Walton E, Neumann A, Thio CHL, Felix JF, van IJzendoorn MH, Pappa I, Cecil CAM.

Arzate-Mejia RG, Carullo N and Mansuy IM

The term epigenome originates from the Greek term “epi”, meaning “above” the genome. It consists of specific chemical compounds that modify or mark the genome, dictating its function, locations and timing. These marks, known as methylation, are separate from the DNA itself and can be passed on from one cell to another during cell division, as well as from one generation to the next. This paper aimed to review and summarize current experimental evidence in mice, regarding the effects of chronic stress on the epigenome and the expression of epigenetic modifiers in brain cells. To accomplish this, the authors conducted a systematic literature review. Arzate-Mejia et al. found that exposure to stress, particularly when chronic, can induce changes in the epigenome of brain cells. These changes are cell-type specific, affect different brain regions, and are influenced by genetic background, sex, and developmental time of exposure. At the functional level, changes to the brain epigenome have been correlated with changes in basal gene expression. Some cases propose that they can impact stimulus-dependant transcriptional responses in the brain and prime genome activity. Consequently, they serve as a potential source of molecular susceptibility of future regulatory responses.
In summary, chronic stress in rodents, caused by physical or social challenges, can alter the epigenome and how genes are controlled in brain cells. Further research is needed to track how the brain’s epigenome changes over time after experiencing stress. This specifically should entail performing time course analyses to evaluate changes to the brain epigenome at different time points after stress exposure. Furthermore, modelling epigenetic modifications in specific regulatory elements by using CRISPR tools is essential to demonstrate if specific changes cause certain effects. Finally, using multimodal experimental approaches to study epigenetic changes would greatly benefit the understanding of how stress affects the epigenome.

Cecil CAM, Schuurmans IK.

Suleri A, Rommel AS, Neumann A, Luo M, Hillegers M, de Witte L, Bergink V, Cecil CAM

Giacomo FD, Strippoli MF, Castelao E, Amoussou JR, Gholam M, Ranjbar S, Glaus J, Marquet P, Preisig M, Plessen KJ, Vandeleur CL

The aim of the study conducted by Di Giacomo et al. was to investigate the factors that might contribute to an increased risk of developing mood disorders in offspring of parents diagnosed with bipolar disorder. To identify risk factors, a discordant-sibling design was used – emphasis was placed on comparing differences in early mental disorders, temperament, personality traits, and coping mechanisms between siblings within the same family. Offspring who developed bipolar disorder or major depressive disorder were compared with at least one brother or sister, who did not manifest either condition. This so called “sib-pair approach” is a methodology that takes variations among siblings and similar factors such as shared genetics and environmental influences into account. Importantly, the information collected originated directly from the offspring, rather than relying solely on data provided by the parent with bipolar disorder. The statistical models applied revealed differences in three dimensions of the Dimension of Temperament Survey-Revised (DOTS-R) version: those who later experienced mood disorders scored higher in “Rhythmicity for daily habits” (displaying more regularity in daily routines), “Task orientation” (demonstrating persistence and reduced distractibility in tasks), and “Approach to novelty” (exhibiting a greater inclination to explore new things) compared to their siblings without mood disorders. Surprisingly, the observed scores were higher, contrary to the expected lower scores. These higher scores could indicate increased vulnerability to mood disorders, but they may also be associated with mood swings before the actual disorder starts or strategies used to cope with them. In the future, data from similar studies need to be combined worldwide, given that sibling-pair studies generally suffer from low sample size.

Lazar-Contes I, Tanwar KT, Arzate-Mejia RG, Steg LC, Ulrich Feudjio O, Crespo M, Germain P-L and Mansuy IM

This study aimed to investigate the gene activity patterns (transcriptomics) and chromatin accessibility in early postnatal and adult spermatogonial cells (SCs), and to understand the molecular mechanisms driving their maturation during sperm cell development. Transcriptomics and chromatin accessibility in SCs were studied in mice using RNA-sequencing and ATAC-sequencing. RNA-sequencing provides information about the transcriptome, specifically what genes are active, while ATAC-sequencing gives insights into the DNA accessibility. To further understand the regulatory mechanisms of SCs, transcription and chromatin accessibility at postanal and adult stages was compared.          
In early postnatal life, SCs exhibit specific gene activity patterns (transcriptional signatures), with high expression of genes linked to cell cycle regulation, stem cell proliferation, transcription, and RNA processing. In contrast, adult SCs prioritize pathways related to localized cell-to-cell signalling (paracrine signalling), mitochondrial function, niche communication and oxidative phosphorylation for energy creation. Genes that show different levels of activity between different stages of SC development are very specific to each stage and show stage-specific changes in transcription. Examining chromatin accessibility further revealed that many regions with different accessibility between stages have characteristics of enhancer elements, which help regulate gene activity. Interestingly, only a small portion of these differentially accessible regions overlap with genes that show different activity levels, suggesting that mechanisms other than chromatin accessibility may also play a role in controlling gene activity in SCs.
Summarizing, these new findings indicate that during early postnatal development, SCs show unique gene expression patterns related to cell cycle regulation and stem cell proliferation, while in adulthood, they prioritize pathways involved in paracrine signalling, niche communication, and mitochondrial functions. Furthermore, this study suggests that gene expression in SCs is not only regulated by changes in chromatin accessibility, but other mechanisms, such as epigenetic modifications and post-transcriptional regulation, may also be involved. Further studies could focus on identifying the main factors controlling gene expression and regulation in SCs at different stages. This includes understanding the signalling pathways involved and the impact changes might have on sperm production. Additionally, it would be important to investigate whether the regions showing chromatin accessibility in early postnatal stages can integrate environmental information that persist into adulthood, thus serving as a form of cellular memory.

Kaupper CS, Blaauwendraad SM, Cecil CAM, Mulder RH, Gaillard R, Goncalves R, Borggraefe I, Koletzko B, Jaddoe VWV. 

Neumann A, Ohlei O, Küçükali F, Bos IJ, Timsina J, Vos S, Prokopenko D, Tijms BM, Andreasson U, Blennow K, Vandenberghe R, Scheltens P, Teunissen CE, Engelborghs S, Frisoni GB, Blin O, Richardson JC, Bordet R, Lleó A, Alcolea D, Popp J, Marsh TW, Gorijala P, Clark C, Peyratout G, Martinez-Lage P, Tainta M, Dobson RJB, Legido-Quigley C, Van Broeckhoven C, Tanzi RE, Ten Kate M, Lill CM, Barkhof F, Cruchaga C, Lovestone S, Streffer J, Zetterberg H, Visser PJ, Sleegers K, Bertram L; EMIF-AD & ADNI study group. 

Hughes DE, Kunitoki K, Elyounssi S, Luo M, Bazer OM, Hopkinson CE, Dowling KF, Doyle AE, Dunn EC, Eryilmaz H, Gilman JM, Holt DJ, Valera EM, Smoller JW, Cecil CAM, Tiemeier H, Lee PH, Roffman JL.

Walton E, Baltramonaityte V, Calhoun V, Heijmans BT, Thompson PM, Cecil CAM

Arzate-Mejia RG, Mansuy IM

Chromatin serves as the packaging material for genetic information inside the nucleus of cells. It consists of a complex made up of DNA and proteins. The DNA tightly wraps around the proteins, condensing the long DNA strands to fit into the cell’s tiny nucleus. While a lot is known about the dynamics of chromatin during programmed cellular processes such as development, the role of chromatin in experience-dependent functions is not completely clear yet.
This paper aimed to review current evidence supporting how chromatin is involved in the maintenance of traces of prior brain cell activity and to describe potential mechanisms of establishment and functional implications in health and disease. To do this, the authors conducted a systematic literature review. The key findings indicate that environmental stimuli can induce lasting changes in the chromatin of different cell-types, possibly affecting gene activity. Although evidence specific to brain cells is still limited, the current data suggests that chromatin may act as a form of cellular memory for brain cells, storing information of past activities in the brain. 
This means that the events we experience in life can actually alter how genes are controlled in the brain. These changes might influence the brain’s capacity to store information from past events. Going forward, Mansuy’s team plans to dive deeper into understanding how early-life trauma affects the molecular makeup of chromatin in brain cells using a mouse model of maternal separation.

Charlotte A. M. Cecil, Alexander Neumann, Esther Walton

This review discusses the application of epigenetics, particularly DNA methylation, in child and adolescent mental health. It emphasizes the progress, challenges, and the importance of more research in this area that takes the different stages of development into account. Cecil and colleagues reviewed various existing studies in the field of psychiatric epigenetics, and shared their thoughts on the findings by providing this narrative review and commentary. They discovered that even though the field of psychiatric epigenetics has grown a lot, research on mental health in children and adolescents is still not as advanced. This is surprising, given that most psychiatric disorders have their origins in early development, making the time period between pregnancy and young adulthood extremely important to study. This review emphasizes that by bringing “timing” into the focus of epigenetic research, we could enhance the ability to predict risks early on and better understand the underlying mechanisms of child and adolescent psychopathology. In the future, the field should move towards larger-scale, collaborative and developmentally-oriented international initiatives to better understand the complex relationship between DNA methylation and mental health over time.