Behavioral and Neural Sustained Attention Deficits in Bipolar Disorder and Familial Risk of Bipolar Disorder

BACKGROUND: Few neuroimaging studies compare individuals affected with bipolar disorder (BP), at high familial risk of BP, and at low risk to identify endophenotypes for BP. None have examined variability in attention, despite promising behavioral work in this area. We used functional magnetic resonance imaging (fMRI) methods uniquely powered to compare the neural correlates of attention variability in these three groups. METHODS: The present study examined 8-to 25-year-old individuals (n = 106) who completed an fMRI attention task: 24 with BP, 29 at risk based on a first-degree relative with BP, and 53 healthy, low-risk individuals. Group differences in intrasubject variability in reaction time were examined, and a sophisticated fMRI analytic approach was used to quantify precisely trialwise associations between reaction time and brain activity. The latter has not been examined previously in BP or risk of BP. RESULTS: Relative to healthy individuals, those with BP or at risk for BP exhibited increased reaction time variability (F-2,F-102 = 4.26, p = .02, eta(2)(p) = .08). Importantly, we identified blunted relationships between trialwise variation in reaction time and brain activity in the inferior and middle frontal gyri, precuneus, cingulate cortex, caudate, and postcentral gyrus (all regions: p < .001, eta(2)(p) > .06) in both at-risk and BP individuals compared with healthy, low-risk individuals. This blunting partially mediated group differences in reaction time variability (beta = .010, 95% confidence interval 0.002 to 0.020, Sobel Z = 2.08, p = .038). CONCLUSIONS: Blunting in key frontal, cingulate, and striatal areas was evident in unaffected, at-risk individuals and in euthymic BP patients. Elucidating such novel neural endophenotypes can facilitate new approaches to BP prediction, diagnosis, and prevention.