The clinical group were recruited from the South London and Maudsley NHS trust and from recruitment websites, such as the eating disorder charity BEAT website ( The healthy control group was recruited through the University internal recruitment circular email sent to staff and students. The clinical group included individuals aged between 18 and 60 with a formal or self-reported diagnosis of AN at varying stages of treatment and weight restoration. The healthy control (HC) group included individuals with no history of any mental health condition, including eating disorders, and were not taking psychiatric medication at the time of the study (ages between 18 and 61).

A total of 102 participants took part in this study (46 in the clinical group, 56 in the healthy control group). Previous studies using the mnemonic similarity task in novel populations have found clinically relevant effects with group sizes of 20–30 [30, 43, 52]. A compromise power analysis was conducted using G*Power software in order to assess the statistical power of this study, using effect sizes of varying sizes. Based on a small effect size of d = 0.2, a sample of 85 participants (the sample used in the pattern separation main analyses) entered into an ANCOVA model with three covariates yielded an estimated power of 0.58. With a moderate effect size of d = 0.5, the same model yielded an estimated power of 0.83.

Informed consent was obtained from all individuals before study participation and participants were reimbursed with a £5 voucher.