All data used for the analyses were simulated based on mean, standard deviation, and percentage contribution values similar to those in #NDNS_yr_7_to_8_statistics.xlsx.
For the food substitution analyses only, real data was used to extract specific means required for the simulation. NDNS Year 7-8 data can be downloaded from the UK Data Service.

All simulations were conducted using R 4.1.1.
The file revised_substitution_simulation_gdt.R contains the data simulation and analyses used in the 'Theory and performance of substitution models for estimating relative causal effects in nutritional epidemiology' manuscript. Results are summarised in results_revised_manuscript_gdt.csv.