SUMMARY Bipolar disorder is a severe psychiatric disorder that affects about 2.5 million Americans and is the 6th leading cause of disability worldwide. Individuals with bipolar disorder (BD) who were maltreated (MAL) in childhood typically have an earlier onset, more severe course, more comorbidities and poorer response to treatments. What is unknown is whether BD patients with MAL represent an etiologically or pathophysiologically distinct subgroup who may differ in important ways in terms of neurobiology and treatment from non-MAL patients with BD. Thus, it is critically important to attempt to identify the effects of experience and timing of MAL in patients with BD in order to determine if and how MAL experiences lead to a distinct subtype (`ecophenotype') of patients. The ecophenotype hypothesis is most strongly supported by neuroimaging studies that identify brain abnormalities in the MAL ecophenotype that are not apparent in the non-MAL patients. This distinction is critically important as, if true, it implies that efforts to identify the pathophysiological basis and to develop more effective treatments for BD may be different between MAL and non-MAL patients and may markedly be enhanced by focusing on specific subtypes. Thus, understanding how MAL modifies the neurobiology of BD may provide new insights that will benefit MAL individuals with BD who often have a particularly severe form of the disorder. Only a few morphometric studies have addressed this issue in BD and do not lead to firm conclusions. Hence, there is a critical gap in our knowledge and a pressing need to definitively establish whether there are or are not fundamental differences in brain structure between MAL and non-MAL individuals with BD. We have recently reported that MAL is associated with substantial alterations in global network architecture and have developed a network-based model that distinguished between MAL individuals with clinically significant psychiatric symptoms, comparably MAL asymptomatic individuals and unexposed healthy controls with high cross-validated predictive accuracy (N=342). These abnormalities appeared to be specifically associated with MAL and not attributable to major depression, anxiety or attention deficit hyperactivity disorder. What is much less certain is whether MAL can produce a pathophysiologically distinct subtype in far more heritable disorders such as BD. The first aim is to test the hypothesis that there are distinct neurobiological differences between MAL and non-MAL individuals with BD. The second aim is to test the hypothesis that MAL individuals with BD differ from MAL individuals without BD in type and timing of exposure to MAL and to identify sensitive periods when exposure to specific forms of abuse are the most important risk factors for BD. These aims will be accomplished in a cost-effective manner by collecting data on type and timing of exposure to MAL in previously imaged BD subjects (estimated N=250) and comparing t...