Longitudinal Structural MRI Analysis: Statistical Scripts

Overview

This repository contains the R script used for the statistical analysis in the study titled "The Impact of Bullying on Brain Development Across Adolescence and Early Adulthood: A Longitudinal Structural MRI Analysis." The study investigates the effects of bullying on brain development during critical periods of growth using longitudinal structural MRI data.

Contents

• LME_Model_Metrics.R: This R script performs various statistical analyses, including data cleaning, modeling, and visualization, to assess the impact of bullying on brain development. It includes:
  • Loading necessary libraries.
  • Reading and preprocessing the dataset.
  • Running Linear Mixed-Effects (LME) models.
  • Analyzing model outputs.
  • Generating summary tables and visualizations.

Prerequisites

This script relies on several R packages, which need to be installed before running the script. You can install them using the following command:

install.packages(c("gamm4", "mediation", "tidyverse", "lmerTest", "knitr", "lavaan", "psych", "MBESS", "greybox", "lme4", "haven", "dplyr", "ggplot2", "ggseg", "effects", "rstatix", "broom", "ggpubr", "effectsize", "ggeffects", "plyr", "methods", "stargazer", "sjPlot", "nlme", "data.table", "MuMIn", "lmtest", "officer", "readxl", "RColorBrewer", "dvmisc", "qwraps2", "AICcmodavg", "simr", "emmeans", "stats", "mutoss", "powerlmm"))

Data

Raw Data Available:

https://github.com/imagen2/imagen_mri

The dataset used in this study is expected to be loaded from an external file (e.g., IMAGEN_MASTER.xlsx). Ensure that the dataset is correctly formatted and located in the appropriate directory as specified in the script.

Usage

1. Prepare your environment: Make sure all required libraries are installed.
2. Load the data: Adjust the file paths in the script to point to your local dataset.
3. Run the script: Execute the script in R or RStudio. The script will:
   • Load and preprocess the data.
   • Fit the Linear Mixed-Effects models.
   • Generate and save the statistical outputs and visualizations.

Results

The script generates various outputs, including:

• Model summaries: Detailed statistics for the fitted models.
• Plots: Visual representations of the results, such as effect sizes and predicted values.

Contact

For any questions or issues, please contact Michael Connaughton at michaelconnaughton@rcsi.com.