Table 1 Summary of studies included in the review
| Author (year) | Sample | Cognitive measures | Function measures | Results |
|---|---|---|---|---|
| Oldershaw, Lavender & Schmidt [60] |
AN (n = 71) Outpatients. Age over 18 years, M age = 26 | WCST, Brixton, TMT, GEFT | CIA | Cognitive tests did not predict CIA scores. |
| Bentz et al. [9] |
AN (n = 71) 43 first episode M age = 16, 28 recovered M age = 18. HC (n = 41) Age 14–22 years. | D-KEFS: Composite set-shifting measure (shift conditions of Verbal Fluency, Design Fluency and Trail Making), GEFT | ADOS | Composite SS and GEFT did not predict ADOS social function. |
| Hamatani et al. [36] |
AN (n = 21: AN-BP = 13, AN-R = 8). 14 outpatients, 7 inpatients. Adults, M age = 32 years. HC (n = 25). | KWCST, ROCFT | SF-36 |
Greater KWCST DMS associated with lower scores on SF-36 PCS (r = − 0.658, p = < .05). ROCFT CCI 30-min delay recall score associated with SF-36 MCS (r = 0.610, p = < .05). |
| Westwood, Mandy & Tchanturia [94] |
AN (n = 99) Inpatients and outpatients. Age 12–47 years. | WCST, ROCFT, D-Flex | ADOS |
Greater WCST %PE associated with increased ASD symptoms, Welch’s F (2,59.77) = 3.77, p = .029. D-flex increase in cognitive rigidity scores between no vs. high ASD symptoms, p = .018 ROCFT not associated with increased ASD symptoms. |
| Renwick et al. [67] |
AN (n = 100; AN-R = 44, AN-BP = 33, EDNOS-AN = 23) Outpatients. Adults, M age = 24.7 years. | WCST, Brixton, ROCFT |
CIA Reading the Mind in Films Task |
Cluster (17% of participants) show poor performance on WCST, ROCFT CCI and RMIF emotional ToM. No effect of cluster for ROCFT, but overall poor performance. No effect of cluster for CIA. |
| Talbot, Hay & Touyz [79] |
AN (n = 49; 24 acute, 10 weight-recovered, 15 fully recovered) HC (n = 43) Age 19–27 years. | WCST, ROCFT, MFFT | EDQOL, IIP-32 |
WCST PE higher scaled scores (better performance) associated with higher IIP-32 ‘Too dependent’ (r = .305, p = .033), ‘Too aggressive’ (r = .401, p = .004), ‘Hard to be involved’ (r = .313, p = .029) and ‘Hard to be supportive’ (r = .294, p = .040). WCST CC associated with higher IIP-32 ‘Too dependent’ (r = .354, p = .013), ‘Too aggressive’ (r = .334, p = .019) and better financial QOL (r = −.393, p = .007). ROCFT lower CCI and OCI associated with higher IIP-32 ‘Hard to be supportive’ (CI: r = −.336, p = .020; OCI: (r = −.413, p = .004) and lower OCI associated with higher IIP-32 ‘Hard to be involved’ (r = −.343, p = .017). |
| Calderoni et al. [14] |
AN-R (n = 23) HC (n = 46) Age 9–16 years. | NEPSY-II Attention and EF: Auditory response set, Inhibition shifting | NEPSY-II Social Perception: ToM verbal, ToM contextual and Affect recognition |
NEPSY-II Attention and EF no group differences on overall domain but response set shifting test worse in AN vs. HC (t67 = − 2.17, p = .033). No group differences on NEPSY-II inhibition shifting or NEPSY-II social perception domain. |
| Harrison, Tchanturia, Naumann & Treasure [38] |
AN (n = 85; 35 AN-R, 15 AN-BP, 35 recovered) HC (n = 90) Age 18–55 years. |
WCST, Brixton, ROCFT, GEFT, FPT | Reading the Mind in the Eyes test | Higher WCST PE associated with lower Reading the Mind in the Eyes emotion recognition (rs = −.281, p = .05), but not significant after Bonferroni correction. |
| Executive function in daily life | ||||
| Spitoni, Aragonaa, Bevacqua, Cotugno & Antonucci [75] |
AN-R (n = 62) HC (n = 70) Age 16–42 years. | WCST, TMT, ROCFT | BADS | AN WCST PE higher (partial ƞ2 = .03) and TMT-B slower (partial ƞ2 = .10) than HC. AN lower ROCFT CCI than HC at 30s (partial ƞ2 = .13) and 20 min recall (partial ƞ2 = .11). AN BADS accuracy similar, but AN slower than HC across tasks (Rule shift cards time: partial ƞ2 = .14; Key search time: partial ƞ2 = .83, Zoo map planning time: partial ƞ2 = .18; Zoo Map execution time partial ƞ2 = .12) |
| Herbrich, Kappel, van Noort & Winter [39] |
AN (n = 111; AN-R = 90, AN-BP = 21) HC (n = 63) Age 9–17 years. | WCST, TMT, ROCFT, GEFT | BRIEF-SR |
AN-R TMT-4 faster than HC (d = .39) No group differences for WCST, ROCFT or GEFT. BRIEF-SR: ‘Shift’ subscale both subtypes higher than HC (AN-R; d = −.89) and (AN-BP; d = − 1.41) and AN-BP higher than AN-R (d = .57). |
| van Noort, Kraus, Pfeiffer, Lehmkuhl & Kappel [89] |
AN (n = 20) 15 inpatients, 5 outpatients. HC (n = 20) Age 12–18 years. | TMT, ROCFT | BRIEF-SR | No group differences for TMT or ROCFT at baseline. Irrespective of time point, AN less flexible than HC on BRIEF-SR subscales ‘Cognitive shift’ (partial ƞ2 = −.207), ‘Behavioral shift’ (partial ƞ2 = −.230). |
| Stedal & Dahlgren [76] |
AN (n = 20, AN-R = 18, AN-BP = 2). 10 in-, 10 outpatients Age 13–18 years. | D-KEFS: TMT-4, Color-word interference test-4 and verbal fluency-3, Brixton, ROCFT |
BRIEF-SR BRIEF-PF | All neurocognitive test scores, BRIEF-SR and BRIEF-PF were within normal range. |
| McAnarney et al. [54] |
AN-R (n = 24) HC (n = 37) Age 14–20 years. | WCST, CANTAB ID/ED |
BRIEF-SR BRIEF-PF | AN-R no significant differences in WCST PE or CANTAB ID/ED. AN-R greater set-shifting difficulties than HC on BRIEF-PF (M = 54.5 vs. 47, p = <.008) and BRIEF-SR (p = .057). BRIEF-SR ‘Behavioral Shift’ sub-scale scores show larger difference (M = 64.1 vs. 48.5, p = <.008) than ‘Cognitive Shift’ (M = 54.9 vs. 48.7, p = <.008). |