Psychological Medicine

Original Articles

Risk factors prospectively associated with adult obsessive–compulsive symptom dimensions and obsessive–compulsive disorder

J. R. Grishama1 c1, M. A. Fullanaa2a3, D. Mataix-Colsa2, T. E. Moffitta4a5, A. Caspia4a5 and R. Poultona6

a1 School of Psychology, University of New South Wales, Sydney, NSW, Australia

a2 Division of Psychological Medicine, Institute of Psychiatry, King's College London, UK

a3 Anxiety Unit (Hospital del Mar/IAPS) and Department of Psychiatry, Autonomous University of Barcelona, Spain

a4 Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, King's College London, UK

a5 Department of Psychology and Neuroscience, Department of Psychiatry and Behavioral Sciences, Institute for Genome Sciences and Policy, Duke University, Durham, NC, USA

a6 Dunedin Multidisciplinary Health and Development Research Unit, Dunedin School of Medicine, University of Otago, New Zealand


Background Very few longitudinal studies have evaluated prospective neurodevelopmental and psychosocial risk factors for obsessive–compulsive disorder (OCD). Furthermore, despite the heterogeneous nature of OCD, no research has examined risk factors for its primary symptom dimensions, such as contamination/washing.

Method Potential risk factors for symptoms or diagnosis of OCD in adulthood and for specific adult obsessive–compulsive (OC) symptom dimensions were examined in the Dunedin Study birth cohort. The presence of obsessions and compulsions and psychological disorders was assessed using the Diagnostic Interview Schedule (DIS) at ages 26 and 32 years. Individuals with a diagnosis of OCD at either age (n=36) were compared to both a healthy control group (n=613) and an anxious control group (n=310) to determine whether associations between a risk factor and an OCD diagnosis were specific.

Results Childhood neurodevelopmental, behavioral, personality and environmental risk factors were associated with a diagnosis of OCD and with OC symptoms at ages 26 and 32. Social isolation, retrospectively reported physical abuse and negative emotionality were specific predictors of an adult OCD diagnosis. Of note, most risk factors were associated with OC symptoms in adulthood and several risk factors predicted specific OCD dimensions. Perinatal insults were linked to increased risk for symmetry/ordering and shameful thoughts dimensions, whereas poor childhood motor skills predicted the harm/checking dimension. Difficult temperament, internalizing symptoms and conduct problems in childhood also predicted specific symptom dimensions and lower IQ non-specifically predicted increased risk for most dimensions.

Conclusions The current findings underscore the need for a dimensional approach in evaluating childhood risk factors for obsessions and compulsions.

(Received December 10 2010)

(Revised March 05 2011)

(Accepted May 07 2011)

(Online publication June 15 2011)


c1 Address for correspondence: Dr J. R. Grisham, School of Psychology, University of New South Wales, Sydney, NSW 2052, Australia. (Email: