Psychological Medicine

Original Articles

Neurocognitive impairment in unaffected siblings of youth with bipolar disorder

A. E. Doylea1 c1, J. Wozniaka1, T. E. Wilensa1, A. Henina1, L. J. Seidmana1, C. Pettya1, R. Frieda1, L. M. Grossa1, S. V. Faraonea1 and J. Biedermana1

a1 Pediatric Psychopharmacology Unit, Massachusetts General Hospital and Department of Psychiatry, Harvard Medical School, Boston, MA, USA


Background There is growing evidence for the familiality of pediatric bipolar disorder (BPD) and its association with impairments on measures of processing speed, verbal learning and ‘executive’ functions. The current study investigated whether these neurocognitive impairments index the familial risk underlying the diagnosis.

Method Subjects were 170 youth with BPD (mean age 12.3 years), their 118 non-mood-disordered siblings and 79 non-mood-disordered controls. Groups were compared on a battery of neuropsychological tests from the Wechsler Intelligence Scales, the Stroop Color Word Test, the Wisconsin Card Sorting Test (WCST), the Rey–Osterrieth Complex Figure (ROCF), an auditory working memory Continuous Performance Test (CPT) and the California Verbal Learning Test – Children's Version (CVLT-C). Measures were factor analyzed for data reduction purposes. All analyses controlled for age, sex and attention-deficit/hyperactivity disorder (ADHD).

Results Principal components analyses with a promax rotation yielded three factors reflecting: (1) processing speed/verbal learning, (2) working memory/interference control and (3) abstract problem solving. The CPT working memory measure with interference filtering demands (WM INT) was only administered to subjects aged 12 years and was therefore analyzed separately. BPD youth showed impairments versus controls and unaffected relatives on all three factors and on the WM INT. Unaffected relatives exhibited impairments versus controls on the abstract problem-solving factor and the WM INT. They also showed a statistical trend (p=0.07) towards worse performance on the working memory/interference control factor.

Conclusions Neurocognitive impairments in executive functions may reflect the familial neurobiological risk mechanisms underlying pediatric BPD and may have utility as endophenotypes in molecular genetic studies of the condition.

(Received January 25 2008)

(Revised October 24 2008)

(Accepted October 27 2008)

(Online publication December 11 2008)


c1 Address for correspondence: A. E. Doyle, Ph.D., Massachusetts General Hospital, Child Psychiatry Service, YAW 6-6A, 55 Fruit Street, Boston, MA 02114, USA. (Email:


Aspects of this work were presented as a poster at the 2005 National Institute of Mental Health (NIMH) Pediatric Bipolar Conference in Chicago, IL, USA.