@article {Morrise3, author = {Laurel S Morris and Prantik Kundu and Kwangyeol Baek and Michael A Irvine and Daisy J Mechelmans and Jonathan Wood and Neil A Harrison and Trevor W Robbins and Edward T Bullmore and Valerie Voon}, title = {NEURAL CORRELATES OF WAITING IMPULSIVITY: A DIMENSIONAL APPROACH TO ALCOHOL MISUSE}, volume = {86}, number = {9}, pages = {e3--e3}, year = {2015}, doi = {10.1136/jnnp-2015-311750.56}, publisher = {BMJ Publishing Group Ltd}, abstract = {Objective Why do we {\textquotedblleft}jump the gun{\textquotedblright}, speak out of turn, or run a red light? Waiting and stopping are fundamental mechanisms of behavioural control. Here we assess waiting impulsivity, or anticipatory premature responding before target onset, dimensionally across social drinkers, binge drinkers, and alcohol use disorders. We further focus on the characterization of neural correlates underlying waiting impulsivity in healthy volunteers. Special interest falls on the subthalamic nucleus (STN), a major relay structure within the indirect inhibitory pathway of striatal circuitry which also receives hyper-direct projections directly from cortical regions. This rich convergence of cortical inputs together with its principal gating of basal ganglia output implicates the STN as a crucial mediator of more complex control of motor and cognitive function.Method Towards a multi-faceted investigation of waiting impulsivity, we examined its neural correlates using a novel 4-choice serial reaction time task (4-CSRT), tested off-line, and differentiated that network from another underlying well-characterized form of motor impulsivity measured using the stop-signal task. To examine the intrinsic neural correlates of waiting and stopping, we use multi-echo resting state fMRI sequence and independent components analysis (ME-ICA) which has been shown to have up to four-fold enhancements in signal-to-noise ratio relative to conventional fMRI scans. Using machine learning classification, we explored the extent to which STN network connectivity would allow for classification of binge drinkers (BD) and those with alcohol use disorders (AUD) from healthy volunteers. For robust data-driven classifications between groups, we used a support vector machine. Supervised machine learning methods such as support vector machines provide robust characterizations at the individual rather than group level. Its use in classification of neuroimaging data is emerging, with studies demonstrating applications based on structural and functional data, supporting its potential as a diagnostic tool with high translational relevance.Results We show that greater premature responding in humans using the novel 4-CSRT task is associated with decreased intrinsic connectivity in a network involving bilateral STN, bilateral subgenual cingulate and right ventral striatum. These findings are dissociable from motor response inhibition or action cancellation as captured by the SSRT, which was associated with lower connectivity between hyper-direct projections of the right pre-SMA and left STN along with dorsal caudate and STN connectivity. We further show translational potential of these findings in alcohol misuse. In young adult BD subjects at greater risk for subsequent alcohol use problems, there was enhanced premature responding relative to matched healthy volunteers. Furthermore, in healthy social drinkers (healthy volunteers), we show that the degree of alcohol severity correlates negatively with connectivity between the bilateral STN and subgenual cingulate. Using machine learning predictive classification, the degree of STN connectivity predicted the differentiation of BD and AUD from social drinkers (healthy volunteers).Conclusion Our findings implicate dissociable parallel functional systems of the corticostriatal pathway and hyper-direct connections in modulating waiting and stopping. Our findings support a dimensional approach towards alcohol misuse from heavy social drinking to young adult binge drinkers to AUD and suggest the neural correlates of premature responding may be endophenotypic markers of alcohol misuse.}, issn = {0022-3050}, URL = {https://jnnp.bmj.com/content/86/9/e3.52}, eprint = {https://jnnp.bmj.com/content/86/9/e3.52.full.pdf}, journal = {Journal of Neurology, Neurosurgery \& Psychiatry} }