Why are some autistic children more likely to experience bullying? A systematic review of risk and promotive factors (Olivia Cheriton, University of Oxford)
Bullying affects the wellbeing of children worldwide, with autistic children at increased risk. Specific factors influencing this are under-researched. The review followed the social ecological framework, focusing on longitudinal studies examining risk/promotive factors for bullying victimisation and perpetration among autistic children. Eight studies met inclusion criteria. Risk factors for being bullied included comorbidities, emotional/mental health and social problems, and school-level factors such as attending mainstream school. Promotive factors, e.g. strong peer relationships and parental support were less frequently identified. A multi-systemic approach to reducing bullying risk factors, whilst enhancing promotive factors, will help parents and teachers support autistic children at greater risk of bullying, and improve their mental health/wellbeing.
Co-developmental trajectories of emotional symptoms and peer problems from childhood to adolescence: Associations with positive and negative mental health outcomes (Aersheng Haidabieke, Fudan University)
Using Millennium Cohort Study dataset, this study applied Parallel-Process Latent Class Growth Models to examine the co-developmental trajectories of emotional symptoms and peer problems from ages 3 to 17 and their associations with positive and negative mental health outcomes at age 17. Five trajectory classes were identified: Primary Peer Problems (Class 1), Co-increasing High Risk (Class 2), Co-decreasing middle risk (Class 3), Primary Emotional Symptoms (Class 4), and Persistent Low Risk (Class 5). Class 5 and Class 3 reported the highest levels of positive mental health outcomes (mental wellbeing, self-esteem, and self-control) and the lowest levels of negative mental health (psychological distress, self-harm, and suicide attempt) at age 17. Class 2 and Class 4 reported the poorest outcomes. Class 1 fell between these classes. Findings highlight the role of individual differences in the co-developmental trajectories and call for tailored interventions to reduce psychological risk and promote well-being.
Physical activity, decision-making and mental health across development: a longitudinal cohort study of sex-specific pathways (Denise Kohlhepp, University of Oxford)
Adolescence represents a critical window of cognitive development and emerging psychopathology with concurrent decrease in physical activity. We aimed to assess multiple potential developmental pathways between physical activity and children and adolescents’ executive function and mental health. This longitudinal study included 14,099 7-to-14-year-old UK Millennium Cohort Study (MCS) participants. We investigated relationships between accelerometer-derived physical activity, decision-making and mental health. For girls, we found that the association between moderate-to-vigorous physical activity at age 7 and emotional symptoms at 14 was mediated via decision-making at age 11. For boys, the association between moderate-to-vigorous physical activity at 7 and decision-making at 14 was mediated via emotional symptoms at 11. No other significant mediation pathways were found.
Trajectories of internalising and externalising difficulties throughout childhood and the exploration of early life predictors and early adult behaviours in the UK Millennium Cohort Study (Jasmine Aherne, University of Manchester)
Objective: To conduct a longitudinal analysis on the UK Millennium Cohort Data to characterise distinct childhood mental health trajectories, identify early life risk factors, and examine behaviours in early adulthood. Method: Childhood mental health status was quantified using the parent reported Strengths and Difficulties Questionnaire (SDQ) from age 3 to 17. The SDQ can be split four types of difficulties: emotional symptoms, peer problems, conduct problems, and hyperactivity. We used a multi-process latent class mixed model to identify joint trajectories of these difficulties on 17,161 children. Early adulthood behaviours were examined from the latest sweeps at ages 17 and 23. (preliminary) Results: Four groups were identified: “late-onset internalising” (18.84%), “stable low” (43.14%), “persistent high” (9.98%), and “decreasing difficulties” (28.03%).
Reducing illiteracy in the UK using risk factor analysis (Katherine Collier, Durham University)
Literacy is a skill that enables access to knowledge and can shape educational and life outcomes. However, the 2023 OECD Survey of Adult Skills found that 18% of adults in England have low literacy levels. In response, this study first reviews how literacy is defined and then uses multiple regression analyses to examine how personal, social and economic characteristics are associated with literacy attainment in UK primary schools, drawing on data from the Millennium Cohort Study. Building on this analysis, literacy risk factors will be modelled to support earlier, targeted interventions. The study will also consider why the identified associations may arise. By exploring the mechanisms linking literacy to personal and socio-economic factors, the study will move beyond simply identifying disparities to understanding how they may be addressed.
Neighbourhood green space and psychological distress among adults in the UK Millennium Cohort Study
(Cedric Galera, Université de Bordeaux)
Background: Green space may benefit mental health, but its effects may differ across socioeconomic groups and environmental contexts. We examined longitudinal associations between neighbourhood green space and psychological distress in adulthood, and whether these associations varied by deprivation.
Methods: We used data from adult participants in the UK Millennium Cohort Study, linked to neighbourhood green space indicators and environmental covariates. Psychological distress was assessed using the Kessler-6 scale. Mixed-effects linear and logistic regression models examined associations with continuous distress scores and moderate-to-severe distress, adjusting for individual, household, neighbourhood, and environmental factors.
Results: Lower exposure to neighbourhood green space was associated with higher psychological distress and increased odds of moderate-to-severe distress. Overall associations for high green space were modest, but protective effects were more apparent among socioeconomically deprived participants, suggesting a potential equigenic effect.
Conclusions: Neighbourhood green space may contribute to reducing adult psychological distress and mitigating socioeconomic inequalities in mental health.
Life Course Socioeconomic Position and health in older adulthood age: A Formal Mediation Analysis in the 1958 British Birth Cohort (Yiling Guo, UCL Centre for Longitudinal Studies)
Childhood socioeconomic position (SEP) is a key determinant of later-life health. To our knowledge, no prospective study has examined whether childhood SEP remains associated with health at the threshold of older age and the extent to which any such association is mediated by adult SEP. Using data from the 1958 British Birth Cohort, we examined associations between childhood social class and a range of health outcomes at age 62. We used interventional causal mediation analysis to assess the extent to which these associations were mediated by adult SEP, including education, occupational class, housing tenure, and income. We found that disadvantaged childhood socioeconomic position remained associated with poorer health at age 62, after accounting for adult socioeconomic position. The degree of mediation varied by outcome and by sex, with adult socioeconomic factors explaining part, but not all, of the association. These findings suggest that population-level improvements in adult socioeconomic position could reduce, but are unlikely to eliminate, later-life health inequalities associated with childhood SEP.
Plasma proteomics reveals early biomarkers of Alzheimer’s disease neuropathology: prospective association analyses in the 1946 British birth cohort (Tahsina Khatun, UCL)
We performed one of the broadest blood plasma proteome-wide association studies of Alzheimer’s disease (AD) neuropathology to date, using plasma p-tau217 as an indicator of AD pathology. Data from the 1946 British birth cohort (N=1,260) was used. Linear regression was conducted to investigate the individual, prospective associations of 10,740 blood-based proteins quantified by the SomaScan 11k assay at ages 60-64 years, with plasma p-tau217 concentration quantified at ages 69-71 years. Predictive modelling was also performed using the full set of proteins.
Association analyses identified two novel proteins that have not been linked to AD before. Prediction analyses showed limited ability of the proteins to predict high plasma p-tau217 ~5-11 years later. Findings provide novel insights into early changes in the blood preceding AD dementia, which warrant further investigations.
Parental involvement in maths at age 7: Evidence from the Millennium Cohort Study (Maria Kireeva, UCL)
Research on parental involvement in maths often generalises participants to the term ‘parent’. I conducted a rapid review of 27 recent studies (2020–2025) which indicated a heavy maternal bias in sampling (80%), frequently overlooking the role of household composition. Situated within post-structural feminist theory, this study uses data from the Millennium Cohort Study to examine how maths help at age 7 varies by child characteristics (sex, ethnicity) and family structure. Specifically, this study uses regression analysis to examine the predictive relationships between household composition, income as well as parental education, and parental help in maths. Comparisons are also made to involvement in reading and writing. By exploring the link between this involvement and children’s maths outcomes, the study aims to provide a more granular understanding of maths help within the home.
Generational Changes in Diet Quality and Mid-Life Blood Pressure: Evidence from the 1946 and 1970 British Birth Cohorts (India Robb, UCL)
Evidence suggests that improvements in cardiovascular health may be stalling in younger generations. High blood pressure (BP) is a major modifiable determinant of cardiovascular disease and a key target for prevention across the life course. This study explores whether generational changes in diet quality (measured using the pyramid-based Mediterranean diet score) are associated with BP in mid-life. Using data from the 1946 and 1970 British birth cohorts, we aim to quantify the relationship between diet quality and BP both within and across generations and assess how socioeconomic inequalities interact with BP. We will investigate whether these inequalities, and the diet–BP relationship itself, differ for men and women. The findings will improve understanding of whether generational shifts in dietary patterns have altered the relationship between diet quality and BP in mid-life.
Changing Faces of Policing: Austerity and the Symbolic Function of the Police in (Re)shaping Social Identity (Tirza Sey, University of Warwick)
This paper examines UK austerity (2010–2019) as a critical juncture shaping policing’s instrumental and symbolic roles under the Cameron and May governments. Drawing on path dependency and life-course theory, it explores how policy-driven socio-economic changes influenced police configurations and their long-term effects on individuals aged 10–23 during this period. Using Millennium Cohort Study (MCS) data, it analyses shifts in attitudes toward the self and the police over time. Logistic regression traces these changes, while multi-group structural equation modelling identifies group-level differences in police affiliation and links to criminal behaviour. Findings underscore how macro-level transformations shape policing, social identity, legitimacy, and pathways into or away from crime.
Accumulated Family Adversities as Moderators of Genetic Associations in Child Development (Kim Stienstra, Leverhulme Centre for Demographic Science)
Gene-environment interaction studies in child development have predominantly focused on socioeconomic status as the key environmental factor. This study broadens the conceptualization of the family environment by examining the cumulative impact of family adversities, including growing up in families with low parental income, job loss, parental separation, and health problems. We test the diathesis-stress model, hypothesizing that family adversity would trigger genetic risks toward developmental problems (lower cognitive skills, more behavioral/psychological difficulties). We leverage the longitudinal design of the Millennium Cohort Study to employ child fixed-effects analyses. This allows us to isolate how changes in the “environmental dose” of adversity moderate the association between genetic factors and (non-)cognitive outcomes.
Mapping Childhood Neurodevelopmental Trait Profiles and Adult Outcomes: Extending ADHD Measurement to Autism and AuDHD in the 1970 British Cohort Study (Yujia Zhang, UCL)
Autism and attention-deficit/hyperactivity disorder (ADHD) frequently co-occur, yet their long-term outcomes remain poorly understood, particularly beyond childhood. Most existing research relies on clinical samples, with limited population-based evidence across the life course. Using the 1970 British Cohort Study (BCS70), my project extends an established approach previously used to retrospectively measure ADHD. Applying data-mining and item response theory (IRT) methods to childhood behavioural items, the project aims to derive dimensional indicators of autistic traits and identify distinct neurodevelopmental profiles, including ADHD-only, autism-only, and combined (AuDHD) presentations. These profiles will then be examined in relation to later-life mental health and cognitive outcomes.
Sex Differences in the Genetics of Depression: Insights from Large-Scale Cohort Studies (David Howard, King’s College London)
Men and women differ in how often depression occurs, how it is experienced, and how people respond to treatment. Understanding the reasons for these differences is important for improving prevention and care. Advances in large scale genetic studies now allow researchers to identify genetic risk factors for depression that may differ between males and females.
This work draws on international collaborations combining data from over 60 cohort studies, including the 1958 National Child Development Study, the 1970 British Cohort Study, and the Millennium Cohort Study. For these studies, additional analysis of the X chromosome was undertaken to maximise use of available data.
This presentation will outline this work and discuss how emerging genetic findings may help inform more personalised approaches to preventing and treating depression.
Is Grey Divorce Leading Blues? Later-Life Divorce Stages and Depression Relation (Özlem Kiliç, Trinity College Dublin)
Grey divorces have become more common across high-income countries, and the UK is one of these countries. However, despite these increases, their consequences on mental health and longitudinal outcomes are poorly studied. In this research, the association between divorce’s stages (-1,0,1) and depression levels will be examined among individuals who experienced grey divorce, in comparison to early divorced individuals, by analyzing the 3 sweeps (Age 42,50, 62) from the National Child Development Study. Building on the panel methods with difference in differences and the hausman test, it has been hypothesized that following the divorce, individuals aged 50+ have an increased depression level compared to early divorced individuals, and this relationship is stronger among women. This research highlights the importance of social and institutional support to later-life divorced individuals.
Gendered disparities in medical career progression: assessing the influence of family responsibilities and societal expectations (Joanne Lawrence, UCL)
Women doctors are outnumbered at senior levels, despite comprising over 50% of UK medical schools’ intake since 1993. If doctors always work full-time there is minimal gendered disparity in career prowess, but part-time doctors are disproportionately women, suggesting some gendered influence. I asked whether part-time work was wholly a free choice.
Using the UK National Child Development Study, I reviewed medical cohort members (CMs) reported hours of work and division of domestic tasks. These showed clear, sex-based differences. Women had less home support; over 80% of men’s partners worked part-time or solely at home (80-90% of women’s partners worked full-time). Women CMs did most domestic tasks, as did male CMs’ partners.
I argue that a woman doctor’s decision to work part-time may be due as much to a need for work-life balance as to any simple wish to prioritise family over career.
Dynamic Family Structure and Offending Trajectories: A gendered analysis using the Millennium Cohort Study (Krystal Li, University of Warwick)
This study examines the relationship between dynamic family structure and later offending trajectories using the MCS data set, with particular attention to gender. Previous research has often treated family structure as a static and binary category, typically comparing children from two-parent and single-parent households. Such approaches risk overlooking the diversity of family arrangements and the dynamic processes through which household composition and economic resources may shape young people’s offending over time.
Drawing on developmental life-course criminology, this project investigates whether different family structures are associated with distinct offending trajectories across adolescence and early adulthood. Rather than treating single-parent families as a homogeneous category, the study distinguishes between two-parent, lone-mother, lone-father and step-parent households, and examines whether these associations vary between boys and girls. The project is also situated within the wider British policy and economic context from the early 2000s onwards, including the New Labour period, the 2008 financial crisis, austerity and the COVID-19 pandemic.
Using longitudinal data, this study aims to identify distinct offending trajectories and explore how early family structure is associated with membership in these trajectories. By integrating family structure, gender and developmental change, the project offers a more nuanced understanding of how family contexts may shape inequalities in offending pathways among young people in contemporary Britain.
Evaluating Links Between Early-Life Social Behaviour and Later-Life Social and Wellbeing Outcomes: A Genetically Informed Multivariate Approach Including Contextual Information (Nicole Ng, Max Planck Institute for Psycholinguistics)
Social interactions shape children’s trajectories across the life course. We investigated how difficulties in developing social abilities (ages 3-17) relate to later-life wellbeing and social health (?17 years) in participants of European ancestry from the Millennium Cohort Study (MCS, N?6541) and 1958 Birth Cohort (N?5349). We modelled exposures and outcomes each as latent variables within a structural model and assessed their relationship using one-sample multivariable Mendelian randomization. Selecting genetic instruments matched for social context (informant, age, and social domain), we found and replicated robust negative exposure-outcome links, spanning up to 30 years. Within the MCS, indirect effects of peer problems acting through low prosocial behaviour were linked to improved wellbeing, consistent with compensatory developmental pathways. Here, we demonstrate that powerful, context-matched genetic instruments enable identification of pathways consistent with causal mechanisms. These findings potentially identify domain-specific and developmental intervention targets that may improve wellbeing and outcomes across the life course.
The impacts of stop and search on young people’s mental health and involvement in crime (Thiago Oliveira, University of Manchester)
Stop and search is widely used by police across England and Wales, yet evidence on its long-term effects on young people remains limited. This project examines whether adolescent exposure to stop and search – both personal experiences and neighbourhood-level policing – affects mental health and later involvement in crime or violence. Using linked data from the MCS, official police records, and neighbourhood datasets, we assess the effects of cumulative exposure to policing on outcomes from late adolescence into early adulthood. Qualitative interviews with 18–24 year olds who experienced stop and search as teenagers provide further insight into lived experiences. Our findings illustrate the long-term implications of stop and search for young people’s development, and provide evidence for practitioners seeking to balance crime reduction with the protection of young people’s wellbeing.
Longitudinal Examination of Perceptions of Social Support by Ethnicity and Sex regarding Wellbeing using Age 23 Sweep of the MCS Data (Abigail Williams-Butler, Rutgers University)
We used Millennium Cohort Study data collected when cohort members were 17 and 23 years old to better understand the relationship between wellbeing and social support across adolescence and young adulthood. In addition, we investigated whether the relationship between social support and wellbeing differed by sex and ethnicity. Preliminary analyses suggest that wellbeing declined between ages 17 and 23, showed a steeper decline for males than females and that wellbeing was significantly related to social support. Although there was some indication that ethnic minorities’ wellbeing declined more than their White peers, there was a lack of evidence for differences in the association between social support and wellbeing by sex or ethnicity. Final analyses will use growth curve modelling to further explore these relationships.
