Opiate Addicts Lack Error-Dependent Activation of Rostral Anterior Cingulate Steven D. Forman, George G. Dougherty, B.J. Casey, Greg J. Siegle, Todd S. Braver, Deanna M. Barch, V. Andrew Stenger, Charlene Wick-Hull, Liubomir A. Pisarov, and Emily Lorensen Background: Healthy individuals performing response suppression tasks activate anterior cingulate cortex with occurrence of false alarm error responses to nontargets. Fundamental questions include whether this error-related activation provides a signal contributing to behavioral control and, given generally poorer performance on such tasks by addicts, whether this signal is disrupted in addiction. Methods: We used rapid, event-related functional magnetic resonance imaging to study 13 individuals with opiate dependence and 26 healthy control individuals performing a Go/NoGo task. Results: Compared with controls, opiate addicts exhibited an attenuated anterior cingulate cortex error signal and significantly poorer task performance. In controls, the individual level of event-related anterior cingulate cortex activation accompanying false alarm error positively predicted task performance, particularly sensitivity in discriminating targets from nontargets. Conclusions: The attenuation of this error signal in anterior cingulate cortex may play a role in loss of control in addiction and other forms of impulsive behavior. Key Words: Substance-related disorders, heroin dependence, de- cision making, choice behavior, impulsivity, personality A ddiction is defined by the loss of control over behavioral impulses, specifically, the impulse to use drugs. Current laboratory measures used to study impulsive responding capture either of two basic dimensions derived from animal models of impulsivity. The first dimension, reward-discounting (RD), is considered an inability to delay rewards or the choice of a small immediate reward over a larger delayed reward (Ainslie 1975; Monterosso and Ainslie 1999). Laboratory measures of this dimension of impulsive responding include delay discounting tasks (Bickel et al 1999; Epstein et al 2003; Kirby et al 1999; Madden et al 1997; Mitchell 1999), the Bechara gambling task (BGT) (Bechara et al 1994), and the Balloon Analogue Risk Task (BART; Lejuez et al 2003). The other dimension of impulsive responding has been termed rapid-response (RR) impulsivity (Swann et al 2002). Tasks in this category include the Go/NoGo and the Immediate Memory/Delayed Memory Task. In such tasks, the occurrence of commission (false alarm [FA]) errors is considered to reflect “inability to conform tasks [responses] to an environmental context” (Swann et al 2002). The majority of research relating laboratory measures of impulsivity to addiction has involved reward-discounting tasks; however, two recent reports (Finn et al 2002; Kaufman et al 2003) link RR impulsivity to addiction. Surprisingly, little evidence has emerged directly relating loss of behavioral control to the neurophysiology of addiction. As stated, one prominent cognitive feature characteristic of impul- sive individuals, including addicts, is increased occurrence of false alarm error responses to nontargets on response suppres- sion tasks (Swann et al 2002). Healthy, nonaddicted individuals performing such tasks have repeatedly shown activation of anterior cingulate cortex (ACC) associated with occurrence of FA errors both in imaging (Braver et al 2001; Kiehl et al 2000; Menon et al 2001; Ullsperger and von Cramon 2001) and electrophysi- ological (Falkenstein et al 1990; Gehring et al 1990) studies. A fundamental question is whether this error-related activation provides a signal contributing to behavioral control (Carter et al 1998; Gehring et al 1993) and, given generally poorer perfor- mance on such cognitive tasks by disinhibited individuals (New- man 1987) including addicts (Finn et al 2002), whether this signal is disrupted in addiction. Many theories of cognitive control suggest that the ACC plays a key role in regulating behavior, with dorsal regions primarily responsible for cognitive functions, such as response selection, error detection, and response conflict detection, and rostral regions more involved in affective processing such as mood regulation or pain perception (see Bush et al 2000 for a review). In practice, the separation of “cognition” from “affect” may be impossible, e.g., as Damasio (1994) has suggested that “somatic marking” of events is a necessary component of behavioral control. Hence, the precise contributions of specific ACC regions to behavioral control remain to be determined. The current experiment was designed to test the idea that RR impulsivity in opiate addiction is related to disruptions of ACC functionality. Rapid-response impulsivity can be investigated by applying signal detection theory (SDT) (Green and Swets 1966) on target discrimination tasks. Signal detection theory allows choice behavior to be separated into two component measures. Response bias () reflects the amount of perceptual evidence necessary to decide that a stimulus is a target. Discriminative sensitivity (D') is a measure of the extent to which targets are successfully discriminated from nontargets, which accounts for both the number of targets correctly identified as targets and the number of nontargets correctly rejected as nontargets. Elevated false alarm rates have tended to be equated with RR impulsivity; however, either a “liberal” response bias or low sensitivity could produce such elevated rates of false alarms. Thus, SDT allows one to more clearly differentiate the processes underlying traditional measures of impulsive performance, such as false alarm rates, in From the Department of Veterans Affairs (SDF, GGD, GJS), Department of Psychiatry (SDF, GGD, BJC, GJS, VAS, CW-H, LAP, EL), University of Pitts- burgh, Pittsburgh, Pennsylvania; Sackler Institute for Developmental Psychobiology (BJC), Weill Medical College of Cornell University, New York, New York; and Department of Psychology (TSB, DMB), Washington University, St. Louis, Missouri. Address reprint requests to Steven D. Forman, M.D., Ph.D., Veterans Affairs Pittsburgh HCS, Highland Drive Division (116A5), 7180 Highland Drive, Pittsburgh, PA 15206. Received December 13, 2002; revised September 23, 2003; accepted September 24, 2003. BIOL PSYCHIATRY 2004;55:531–537 0006-3223/04/$30.00 doi:10.1016/j.biopsych.2003.09.011 © 2004 Society of Biological Psychiatry