Abstract
Background: Impulsivity is commonly assessed using the Barratt Impulsiveness Scale (BIS-11). Some studies challenged the reliability of its three dimensional structure and proposed a bi-dimensional structure. Methods: The psychometric reliability of the BIS-11 scale was studied in a sample of 580 euthymic bipolar patients. An alternative structure of the scale was conceived, using confirmatory factorial analysis (CFA) in the first half (N = 290) and cross-validated in the second half of our sample. Associations between the newly defined shortened scale and predefined clinical variables were computed. Results: The original three dimensional structure did not fit in our sample according to statistical criteria in CFA. A 12 items Impulsivity Scale (IS-12) was designed with strong indices of fitting in the first half of our sample and replicated in the second half of our sample. The IS-12 evidences two dimensions: "behavioral impulsivity" and "cognitive impulsivity". Associations between "behavioral impulsivity" and both presence of past suicide attempts and number of suicide attempts were observed. Substance misuse was strongly associated with both subscores of the new scale. Limitations: The rating of the items assessing the two dimensions of the IS-12 is reversed. The population is restricted to euthymic bipolar patients. Conclusions: The Impulsivity Scale assesses two distinct dimensions named behavioral and cognitive impulsivity. It was reliable and valid in our sample and associated with the existence of suicidal behavior and with substance misuse (alcohol and cannabis). Further studies are needed to demonstrate its predictive validity.