Psychological Medicine

Original Article

Profile of neurocognitive impairments associated with female in-patients with anorexia nervosa

L. FOWLER a1, A. BLACKWELL a1, A. JAFFA a3, R. PALMER a4, T. W. ROBBINS a2, B. J. SAHAKIAN a1 and J. H. DOWSON a1c1
a1 Department of Psychiatry, University of Cambridge, Cambridge, UK
a2 Department of Experimental Psychology, University of Cambridge, Cambridge, UK
a3 Ida Darwin Hospital, Cambridge, UK
a4 Leicester General Hospital, Leicester, UK

Article author query
fowler l   [PubMed][Google Scholar] 
blackwell a   [PubMed][Google Scholar] 
jaffa a   [PubMed][Google Scholar] 
palmer r   [PubMed][Google Scholar] 
robbins tw   [PubMed][Google Scholar] 
sahakian bj   [PubMed][Google Scholar] 
dowson jh   [PubMed][Google Scholar] 


Background. Although many studies have reported impairments of neurocognitive performance in patients with anorexia nervosa (AN), these have involved a wide range of assessment methods and some findings are inconsistent.

Method. Twenty-five female in-patients with a DSM-IV diagnosis of AN, identified from three units specializing in the treatment of eating disorders, volunteered for the study. Twenty-five non-clinical control subjects were recruited, matched for age, gender and estimated IQ. Subjects were assessed with a range of computer-administered neurocognitive tasks from the Cambridge Neuropsychological Test Automated Battery (CANTAB), which has been validated in many studies of neuropsychiatric disorders.

Results. The patient group showed significant but moderate impairments (i.e. less than one standard deviation below the mean performance of the control group) on tests of spatial recognition memory, a planning task and rapid visual information processing, while a subgroup of patients (n=14) showed greater degrees of impairments on at least one of these tests. The degrees of impairments did not correlate with body mass index (BMI). No impairments were observed on tests of spatial span, pattern recognition memory, spatial working memory, matching-to-sample, paired associates learning and set-shifting.

Conclusions. The findings, in relation to a mean BMI of 15·3, are compatible with, in general, subtle impairments in neurocognition in AN. However, in those patients with relatively severe degrees of impairments, these may have adverse effects on complex tasks of social and occupational functioning. Further research is needed on the nature of relevant causal mechanisms, including the effects of potentially confounding variables.

(Published Online December 1 2005)

c1 University Department of Psychiatry, Box 189, Addenbrooke's Hospital, Cambridge CB22 2QQ, UK. (Email: