On this page you will find the script corresponding to the paper “Longitudinal associations between adolescent adversity, brain development and behavioural and emotional problems” by Ayla Pollmann (1), Dr. Divyangana Rakesh (2) and Dr. Delia Fuhrmann (1).
Affiliation:
Institute of Psychiatry, Psychology & Neuroscience, King’s College London, London SE5 8AF, United Kingdom
Neuroimaging Department, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, London, UK
Please cite the paper if you are re-using some of the code provided here.
Abstract:
Adolescent adversity is thought to have lifelong implications for mental health. One potential pathway through which this may occur is neurodevelopmental changes. This study investigates the interplay between adverse adolescent experiences, brain development, and behavioural and emotional problems.
We employed a random intercept cross-lagged panel model to investigate longitudinal relations within the ABCD study (N ≈ 12.000, 9-15 years, United States). The model included structural white matter connectivity (measured by fractional anisotropy), adversity (family conflict and low neighbourhood safety) and behavioural and emotional problems.
We found a positive association between family conflict and behavioural and emotional problems: family conflict was related to increased behavioural and emotional problems at 10 – 12 years (β = 0.06, p = 0.002), and vice versa (β = 0.18, p < 0.001). At 12 – 14 years, behavioural and emotional problems were also related to increased family conflict (β = 0.20, p < 0.001). The neighbourhood environment was related to behavioural and emotional problems and white matter connectivity. At 10 – 12 years, low neighbourhood safety was related to lower levels of white matter connectivity (β = -0.04, p = 0.041) and vice versa (β = -0.17, p = 0.024). It was also associated with more behavioural and emotional problems (β = 0.05, p = 0.015) and vice versa (β = 0.12, p < 0.001). Behavioural and emotional problems were positively associated with neighbourhood perception for adolescents with more friends (χ²(1) = 9.82 p = 0.02).
Overall, our results highlight the importance of considering adversities in their socio-environmental contexts for understanding adolescent brain development. They also highlight that adversity in adolescence may shape later emotional and behavioural problems.
Analysis Overview
We will examine neurodevelopmental trajectories in adolescents exposed to adversity and investigate the influence of resilience factors on emotional and behavioural problems in adolescence. This page includes the following analyses.
Random intercept cross-lagged panel models (RI-CLPM):
Family conflict
Perceived neighbourhood safety (with parental perception of neighbourhood safety as a control)
The analysis is based on the code by Mulder and Hamaker, which can be found here.
Split-Sample approach
We first developed the models using a smaller part of the data (20% of the entire sample), then cross-validating the remaining, larger subset (80%). An independent researcher chose a random subset of the 20% exploratory sample. The analytic scripts were developed, and the model convergence was tested using this subset. The remaining 80%, also known as the confirmatory sample, were used to conduct the primary analysis.
Ayla Pollmann - 2024