ARL: Current Research

Across areas of investigation, research in my laboratory is guided theoretically by a multilevel, multidimensional model of neural, cognitive, emotional, and behavioral processes. Current basic research and theory on emotion and cognitive processes provide the foundation for these projects and an emphasis is placed on the use of well-validated paradigms and sophisticated psychophysiological indices of key constructs. For example, research on alcohol and fear/anxiety draws on current conceptualizations of emotion (Lang, 1995), and multilevel models of emotion processing incorporating cognitive and neural mechanisms (e.g. Ledoux, 1995). Assessment of emotional response involves psychophysiological measures of emotional valence (fear potentiated startle reflex, facial electromyography) and arousal (skin conductance and heart rate), as well as indices of neuroendocrine response (e.g. salivary cortisol). Moreover, this research employs adaptations of basic animal fear conditioning paradigms that connect my work to animal research on the neural circuitry of fear. Research on intoxicated cognitive deficits utilizes basic cognitive paradigms (e.g., Stroop, flanker, n-back) and cognitive psychophysiological indices (e.g., event related brain potentials such as P3 and Error related negativity) that anchor my research on drug effects to current cognitive neuroscience models on attention networks (Botvinick et al., 1999; Miller & Cohen, 2001, Posner, 1995).

As indicated previously, my current research in my laboratory is organized around 4 research goals. More detailed information about each of these goals is provided below.

1. Alcohol Challenge and the Emotional Response
2. Drug Withdrawal, Affect and Relapse Potential
3. Alcohol Challenge and Cognitive Deficits
4. Mechanisms of genetic risk for alcohol and drug Dependence

[back to top]
Goal #1: Alcohol challenge and emotional response

One primary focus of my research is on the systematic examination of the connection between alcohol intoxication and emotional response, and the delineation of processes that may mediate it. This research incorporates an empirically based conceptualization of emotion as a phenomenon that involves central activation of response tendencies or "action dispositions" that prepare an organism to act. These action dispositions are organized along dimensions of arousal (degree of activation) and valence (pleasant or unpleasant) to represent two primary brain motivation systems: an aversive system governing defensive reactions and an appetitive system governing consummatory and other approach behaviors. It is proposed that these systems are subcortically based, but that through reciprocal connections to higher cortical regions, they can influence and be influenced by more complex cognitive processes such as attention, perception, declarative memory, and imagery. Operating within this framework, my research suggests that alcohol does not directly affect emotional response at the primary subcortical level, but rather that it influences emotion through its impact on cognitive processes.

I am developing a series of parametric experimental studies to investigate the effects of acute alcohol intoxication on fear/stress, and the potentially mediating role of alcohol-induced cognitive processing deficits in these alcohol-stress relations. Applications of state-of-the-art psychophysiological measurement of emotion (e.g., the startle response and other EMG and autonomic indices of emotional valence and arousal) and cognitive processing (event related potential indices such as P3) are a major part of these efforts and have permitted a more precise examination of potentially critical psychological processes than was possible previously.

My initial report (Curtin et al., 1998, Journal of Abnormal Psychology; see included reprint) described a demonstration that the anxiolytic effect of alcohol intoxication on fear-potentiated startle was limited to conditions requiring a division of attentional resources. No reduction in fearfulness was observed among intoxicated participants when threat cues were presented in less cognitively demanding circumstances, or in sober participants regardless of cognitive processing requirements. These results are consistent with the notion that the fear-reducing effects of alcohol were mediated by cognitive deficits caused by the drug. Unfortunately, no measure of the processing of threat or competing cues was available in this experiment. In fact, at that point, no previously published study examining alcohol effects on emotional response had ever directly measured the cognitive processing of emotion-eliciting stimuli associated with affective response.

One way to address this gap in the literature was to incorporate methods for measuring cognitive processes via event related potentials (ERPs) into studies designed to examine alcohol-emotion connections. To this end, I have recently published (Curtin et al., 2001, Psychological Science; see included reprint) a successful demonstration of how alcohol-induced deficits in the processing of threat cues do indeed co-occur with the attenuation of fear reactivity in intoxicated subjects, whereas neither cognitive impairment nor fear attenuation was evident if intoxicated participants were under minimal cognitive load. More specifically, intoxicated individuals exhibited deficits in threat cue processing (as indexed by the P3 event related potential) and subsequent fear potentiated startle to these threat cues only when their alcohol-impaired cognitive capacity was stretched beyond its limits by a task that divided their attention, rendering them less able to process threat cues. In contrast, the responses of sober subjects were relatively unaffected by cognitive demands. Results are consistent with higher cortical mediation of alcohol's effects on fear, and illustrate more broadly how disruption of a cognitive process can lead to alterations in emotional reactivity and adaptive behavior.

Results from these recent studies highlight the merits of a multilevel, multidimensional approach, with its simultaneous psychophysiological indexing of both affective and cognitive processes, to the study of alcohol effects on emotional response. Moreover, they suggest numerous promising directions for further systematic research on the alcohol-emotion nexus. For example, continued examination and refinement of our understanding of the exact nature and parameters of cognitive mediation is clearly needed. Ideally, this should be conducted with reference to both cognitive/psychological constructs and the brain substrates and mechanisms that might underlie them.

I have recently received two years of funding from the National Institute of Alcohol Abuse and Alcoholism to extend basic animal fear conditioning paradigms for use with human participants in an effort to capitalize on recent advances in understanding of the neural pathways involved in fear conditioning within these paradigms. This project proposes to examine potential dissociations in the effect of alcohol on fear conditioning to explicit vs. contextual cues. Research suggests that thalamic-amygdala connections are sufficient to condition fear response to simple, explicit sensory cues (e.g. a colored light stimulus). However, additional processing pathways (most likely involving the hippocampus) are required to condition fear response to a more complex multi-modal stimulus such as the overall experimental context. Dissociation in alcohol's effect on fear response to these two distinct types of eliciting stimuli will implicate the hippocampus in explaining alcohol's effect on fear and anxiety .

In a similar vein, current projects are underway in my laboratory to examine learning phenomena (trace conditioning, latent inhibition, extinction, blocking, overshadowing) within a number of other basic animal fear conditioning paradigms (adapted for use with humans) for which information on the neural substrates of learning are available. For example, trace conditioning involves the delayed presentation of an unconditioned stimulus (UCS; e.g., shock) after termination of the conditioned stimulus (CS). Research suggests that trace conditioning is hippocampally dependent. Latent inhibition involves repeated unpaired presentations of a CS prior to its pairing with the UCS, which results in inhibited/impaired fear conditioning to the CS. Available evidence indicates that latent inhibition is also hippocampally dependent. Interestingly, if alcohol effects on fear are mediated via its detrimental effects on hippocampal function, dissociations in the direction of alcohol's effect on fear response will be observed across these two fear conditioning paradigms. Specifically, alcohol-impaired trace conditioning will lead to reduced fear response, but impaired latent inhibition will lead to facilitation of fear conditioning among intoxicated participants. Such dissociations will help to clarify the heterogeneous effects of alcohol on fear response and begin to address the potential neural mechanisms of these effects.

[back to top]
Goal #2: Drug withdrawal, affect and relapse potential

The negatively reinforcing effects of acute drug use (i.e., drug related reductions in negative affective response to stressors) are an important motive for use and repeated drug use xto cope with stressx may represent one potential path to drug dependence. Interestingly, among drug dependent individuals, drug withdrawal itself may result in tonic (ongoing) elevated negative affect (a negative mood state). In addition to decrements in xbackgroundx mood state, negative affective response to discrete stressors may also be exacerbated during withdrawal. Therefore, attempts to cease drug use may establish an xaffective environmentx in which drug use would be particularly reinforcing due to both of these potentially independent negative affective consequences of withdrawal.

Current research in my laboratory (in collaboration with my graduate student Joanne Hogle) is examining both overall negative mood and discrete negative affective response and subsequent regulation/recovery of that affective response among individuals during withdrawal from nicotine. In an initial study (supported by a sub-award of the Transdisciplinary Tobacco Use Research Center Award from the National Cancer Association), negative affect among xnicotine-withdrawnx and xcontinuing-smokingx dependent smokers (along with non-dependent, occasional smokers and non-smokers required as comparison groups to address specific hypotheses) was examined using measures of self-reported affect, electrophysiological affective response (fear potentiated startle, facial EMG, skin conductance) and HPA axis activation (salivary cortisol). Four specific questions are addressed within this project.

1. Does nicotine withdrawal result in overall, elevated negative mood in the absence of a specific, salient stressor?

2. Is the negative affective response to a specific stressor exacerbated during nicotine withdrawal?

3. Are the temporal features of that affective response (specifically recovery from the negative affect once the stressor is terminated) altered during withdrawal?

4. Are the negatively reinforcing properties of nicotine (i.e., its stress reducing properties) limited to the withdrawal-related negative affect (described in 1-3) or does nicotine reduce stress that is independent of withdrawal related processes?

Preliminary results from this project suggest that nicotine withdrawal impairs the individuals' ability to recover from stress. Specifically, withdrawn and non-withdrawn individuals did not differ in their initial stress response to a cue that predicted potential administration of electric shock. However, after the possibility of shock ended, withdrawn smokers continued to display a robust stress response, in contrast to the quick and adaptive affective recovery observed among non-withdrawn individuals. In addition, a group of occasional smokers ("chippers"), who report relatively frequent nicotine use but display no indications of nicotine dependence, exhibited superior affective recovery (better than all other groups including non-smokers). This intriguing finding suggests that well-developed affect regulation may represent a protective factor from becoming dependent on nicotine.

We are currently conducting follow-up research designed to systematically examine affect regulation among smokers. To the end, I recently sponsored Megan Piper's successful application for a Training Tobacco Scientists Mini-Grant Program from the University of Wisconsin Center for Tobacco Research and Intervention. This award is providing funds for a collaborative project designed to further our understanding about affect regulation during withdrawal.

[back to top]
Goal #3: Alcohol challenge and cognitive deficits

There is a widespread acceptance of the potential for alcohol intoxication to impair cognitive processes (Holloway, 1994). Yet, much remains to be specified and this is a particularly critical time because theorizing about the role of cognitive deficits in understanding intoxicated emotion (e.g., stress response dampening effects) and behavior (e.g., aggressivity, impulsivity, sexual risk-taking) is gaining increasing visibility. Recent research in my laboratory suggests that inebriates have difficulty executing cognitive control required to inhibit dominant responses that are contextually inappropriate.

Current theory defines cognitive control as effortful direction, engagement, and activation of cognitive resources in the selection and processing of task-relevant information for purposes of maximizing performance and minimizing interference on tasks involving a high level of difficulty, complexity, interference, or novelty. Cognitive control biases processing of information toward task-relevant stimuli and responses in order to establish appropriate stimulus-response mappings (MacDonald, Cohen, Stenger, & Carter, 2000). The purpose of cognitive control is to guide, coordinate, and update behavior in a flexible fashion (Braver, Barch, & Cohen, 1999). Cognitive control processes are particularly important to guide behavior under conditions or tasks that involve competition between prepotent and weaker response inclinations. Interestingly, situations characterized by such response conflict are precisely the conditions under which alcohol has been shown to have its most profound behavioral effects. This suggests that alcohol-induced selective impairment of cognitive control processes may provide the mechanism through which alcohol affects behavior. I am currently conducting programmatic research designed to systematically examine alcohol's effect on specific components of cognitive control using electrophysiological indices (e.g., ERPs) within laboratory response conflict paradigms that have advanced our understanding of basic processes in cognitive control.

In an initial experiment, my colleagues and I (Casbon, Curtin, Lang, & Patrick, 2003) examined the impact of alcohol on working memory processes, a component process of cognitive control, using the n-back paradigm. In this paradigm, participants view a series of letters presented one at a time. Participants are required to rapidly press a button to indicate if the current letter matches a preceding letter. Working memory load is manipulated by requiring participants to indicate matches with either the immediately preceding letter (1-back; low working memory load) or the letter two positions back (2-back; high working memory load). In addition, we manipulated response tendency (tendency to respond or withhold response on a given trial) by blocking trials such that some blocks contained 80% matches (requiring frequent response) and other blocks contained only 20% matches (requiring infrequent response). Alcohol and working memory load had interactive effects on task performance such that failures to inhibit the dominant response established by the current block response requirements were most frequently observed among intoxicated individuals during high working memory load. This preservation of the dominant response tendency was observed for both commission errors (responding when no letter match existed) and omission errors (failing to respond letter match did exist). Specifically, commission error rates were elevated among intoxicated individuals in high memory load during blocks that required frequent responding (i.e., dominant response tendency was to respond). Omission error rates were increased during blocks that required infrequent response (i.e., response tendency was to withhold response).

In a second study conducted in the past year (Curtin & Fairchild, 2003), I have employed event related potential (ERP) measures to directly index alcohol effects on various components of cognitive control necessary for adaptive performance in the Stroop paradigm. Alcohol intoxication resulted in impaired performance (increased response time and error rate) only when the task required the execution of the non-dominant response (i.e., color naming) in the context of competing, incompatible word information (i.e., non-matching color word). Examination of ERPs provided information about potential cognitive processes and systems responsible for this failure in behavioral control. Specifically, examination of the P3 component of the ERP suggested that alcohol did not affect selective sensory attention to and identification of color information on incongruent color naming trials. However, alcohol’s effect on subsequent ERP components indexing cognitive control processes (N450 and negative slow wave) indicated that the observed decrements in task performance resulted from a failure of inhibitory mechanisms evoked during response selection and execution among intoxicated individuals.

In a third study, which has just been completed using the Flanker task, manipulations of selective attention and dominant response tendency were employed. Specifically, trials consisted of a string of 5 letters (Hxs and Sxs). Participants made forced choice responses to indicate the center target letter (H or S) while ignoring flanker letters surrounding the target. Flankers were compatible (match target; example: HHHHH) or incompatible (mismatch target; example: SSHSS), with compatible/incompatible trials equi-probable. A dominant response was established by manipulating target letter probability with one response more probable (p=0.80) than the other (p=0.20). Behavioral results indicate that alcohol interacted with the manipulation of dominant response tendency such that the slowest response times and highest error rates were observed when intoxicated participants were required to respond to the infrequent target (and inhibit their prepotent frequent response). In contrast, alcohol did not interact with the selective attention manipulation (i.e., although the expected main effect of flanker compatibility was observed, alcohol did not exacerbate this effect). This effect co-occurred with a significant impairment in Error related negativity among intoxicated individuals. Substantial evidence suggests that the neural generator of ERN is in the anterior cingulate cortex (a structure which has received much attention for its role in the action monitoring component of cognitive control). These results are consistent with my laboratory's hypothesis that alcohol's effects on behavior result for deficits in cognitive control.

[back to top]
Goal#4: Mechanisms of genetic risk for alcohol and drug dependence

Substantial evidence indicates genetic variation in risk for the development of alcohol and drug use disorders. However, the exact mechanisms through which this risk is conferred are unclear. Near-term future work within my lab will identify individuals at risk for the development of alcohol use disorders to determine if differential sensitivity to acute alcohol challenge and withdrawal effects on cognitive processes and emotional response mediate genetic risk. A multidimensional approach to the assessment of risk status incorporating family history of dependence, personality traits (e.g. increased negative emotionality and impulsivity) and potential endophenotypic markers (e.g., decreased P3 of the event related potential waveform) will be employed in this work.

To investigate the role of individual differences in personality in risk for the development of alcohol use disorders, my colleagues and I have recently completed development of a brief form of the Multidimensional Personality Questionnaire (Patrick, Curtin, & Tellegen, 2002; reprint is included). The MPQ (Tellegen, 1982) provides for a comprehensive analysis of personality at both the lower-order trait and broader structural levels. Its higher-order dimensions of Positive Emotionality, Negative Emotionality, and Constraint embody affect and temperament constructs, which have been conceptualized in psychobiological terms. In addition, substantial heritablility has been observed for these components of personality. Whereas the length of the original MPQ was somewhat prohibitive with respect to mass testing, the recently completed brief form can be completed quickly and thus more easily administered to large numbers of research participants. Use of this measure will allow me to assess the degree to which heritable differences in trait emotionality can account for individual differences in negatively reinforcing effects of alcohol, suggesting one avenue for increased risk. Specifically, emotional concomitants of alcohol intoxication among individuals with personality profiles typically observed among individuals with alcohol use disorders (specifically, increased Negative Emotionality and decreased Constraint) will be examined. This work will extend my basic laboratory research on alcohol and emotion to the clinical literature by directly tying it to risk for the development of psychopathology.

Similarly, recent research suggests that reduced a P3 component of the ERP may be an endophenotypic marker of risk for the development of alcohol use disorders. Reduced P3 is observed in select paradigms among alcohol dependent individuals and also among individuals with a family history of alcoholism (ruling out chronic effects of alcohol as an explanation for this deficit). In addition, P3 magnitude appears to be heritable. Given that my research suggests that alcohol effects on emotional response are mediated by acute cognitive deficits, it seems reasonable to suggest that individuals with electrophysiological indication of cognitive dysfunction may exhibit differential emotional response to alcohol in situations where processing deficits result in altered emotional response. Near term future work will begin to address these predictions.

[back to top]