Bipolar Disorder (BD) is a chronic mood disorder with a prevalence estimated around 1–2%. Bipolar patients may experience social and working residual impairment even during euthymia. Furthermore, specific cognitive deficits, particularly involving working memory (WM), may persist during eythymia as well.

To evaluate the possible presence of cognitive and functional differences between euthymic bipolar subjects vs. healthy controls during euthymia by means of a WM task at fMRI associated with neuropsychological evaluations.

A sample of 30 subjects aged between 20 and 45 years (15 with BD and 15 controls) underwent fMRI examination at 3 Tesla with tasks of working memory (n-back). All participants received a neuropsychological evaluation, inlcuding Stroop Color-Word Interference test, Tower of London, Trail Making Test, Wisconsin Card Sorting Test and Verbal Fluency Test. Comparison tests were performed using statistical software SPSS and SPM5.

The performance of the control group was significantly higher than both at the n-back task and at the neuropsychological tests. The full-factorial analysis of fMRI data showed a hypoactivation in bipolar patients in particular hippocampus and thalamus, associated with increased involvement of areas not involved in the frontal-parietal networks classically associated with WM.

The results seem to confirm the existence of a residual dysfunction during euthymia phase in BD, suggesting two distinct patterns of activation in the two groups studied, both from a neuropsychological point of view and from a neuroimaging perspective.

Bipolar disorder (BD) may be characterized by the presence of psychotic symptoms and comorbid substance abuse. In this context, structural and metabolic dysfunctions have been reported in both BD with psychosis and addiction, separately. In this study, we aimed at identifying neural substrates differentiating psychotic BD, with or without substance abuse, versus substance-induced psychosis (SIP) by coupling, for the first time, magnetic resonance imaging (MRI) and positron emission tomography (PET).

Twenty-seven BD type I psychotic patients with (n = 10) or without (n = 17) substance abuse, 16 SIP patients and 54 healthy controls were enrolled in this study. 3T MRI and 18-FDG-PET scanning were acquired.

Gray matter (GM) volume and cerebral metabolism reductions in temporal cortices were observed in all patients compared to healthy controls. Moreover, a distinct pattern of fronto-limbic alterations were found in patients with substance abuse. Specifically, BD patients with substance abuse showed volume reductions in ventrolateral prefrontal cortex, anterior cingulate, insula and thalamus, whereas SIP patients in dorsolateral prefrontal cortex and posterior cingulate. Common alterations in cerebellum, parahippocampus and posterior cingulate were found in both BD with substance abuse and SIP. Finally, a unique pattern of GM volumes reduction, with concomitant increased of striatal metabolism, were observed in SIP patients.

These findings contribute to shed light on the identification of common and distinct neural markers associated with bipolar psychosis and substance abuse. Future longitudinal studies should explore the effect of single substances of abuse in patients at the first-episode of BD and substance-induced psychosis.