A Parent General Behavior Inventory Subscale to Measure Sleep Disturbance in Pediatric Bipolar Disorder

Abstract

Sleep disturbance is a reliable marker for differentiating children with bipolar spectrum disorders from those with attention-deficit/hyperactivity disorder. Sleep-related items were extracted from the Parent General Behavior Inventory (P-GBI) to determine whether these items, as a scale unto themselves, demonstrate adequate psychometrics to be useful as a possible endophenotypic marker for bipolar spectrum disorders. From July 2003 to July 2007, 625 youths and their parents completed semistructured Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime version interviews. Parents also completed the 73-item P-GBI. Participants with bipolar spectrum disorders (DSM-IV criteria) scored significantly higher than all other participants on all 7 of the sleep variables (p < .005). On receiver operating characteristic (ROC) analysis, the sleep subscale did substantially as well at discriminating participants with bipolar spectrum disorders as did either of the 2 built-in GBI scales, depression and hypomanic/biphasic (area under the ROC curve = 0.74 vs. 0.75 and 0.77, respectively). The P-GBI sleep subscale, developed in this study, is a reliable measure of a wide range of mood-related sleep problems in youths diagnosed with bipolar spectrum disorders. Sleep disturbance appears to be a promising endophenotype for further clinical investigation, and the P-GBI sleep scale may provide an inexpensive way of quantifying this trait for research. Further research needs to evaluate how parent report compares to objective measures of sleep efficiency, such as actigraphy or polysomnography.