Data was anonymised through assignment of unique subject IDs and removal of potentially identifying information.