Elsevier

NeuroImage

Volume 19, Issue 4, August 2003, Pages 1439-1448
NeuroImage

Regular article
Increased activation in the right insula during risk-taking decision making is related to harm avoidance and neuroticism

https://doi.org/10.1016/S1053-8119(03)00251-9Get rights and content

Abstract

Decision making and risk taking are interrelated processes that are important for daily functioning. The somatic marker hypothesis has provided a conceptual basis for processes involved in risk-taking decision making and has been used to link discrete neural substrates to risk-related behaviors. This investigation examined the hypothesis that the degree of risk-taking is related to the degree of activation in the insular cortex. Seventeen healthy, right-handed subjects performed a risk-taking decision-making task during functional magnetic resonance imaging (fMRI) using a fast event-related design. This investigation yielded three main findings. First, right insula (BA 13) activation was significantly stronger when subjects selected a “risky” response versus selecting a “safe” response. Second, the degree of insula activation was related to the probability of selecting a “safe” response following a punished response. Third, the degree of insula activation was related to the subjects' degree of harm avoidance and neuroticism as measured by the TCI and NEO personality questionnaires, respectively. These results are consistent with the hypothesis that insula activation serves as a critical neural substrate to instantiate aversive somatic markers that guide risk-taking decision-making behavior.

Introduction

Decision making, i.e., selecting an action from a set of alternatives with an uncertain outcome, consists of several component processes. A particularly important component of decision making is risk taking, which can be defined as the propensity to select an action with the potential for a relatively large beneficial or adverse outcome over an alternative action that results in a relatively small beneficial outcome Slovic, 1987, Mellers, et al., 1997. Risk taking itself, however, can be broken down into several components, including anticipation, reward, and punishment-related processing.

Abnormalities of risk-taking aspects of decision-making behavior have been observed in several psychiatric disorders Mogg et al., 1991, Rahman et al., 2001, American Psychiatric Association, 1994, Rahman et al., 1999, including substance-related syndromes (Rogers et al., 1999a). For example, substance dependent subjects are more likely to select a high gain/high-risk alternative (Bechara, 2001) over a low-gain/low-risk alternative even when the former alternative is associated with a disadvantageous long-term outcome.

Experimentally, risk-taking decision-making behavior appears to be highly sensitive to context. For example, the selection of the risky alternative is dependent on the number of other available outcomes (Weber et al., 1992), on the stimulus context (Mellers and Chang, 1994), and on cultural background of the subject (Hsee and Weber, 1999). One common approach to examine risk-taking behavior is to present subjects with a choice between a sure thing and a gamble (Yates and Stone, 1992). By varying the expected value of each action alternative (i.e., the magnitude of the beneficial effect multiplied by the likelihood of the outcome), one can determine whether subjects are risk seeking (selecting the gamble even when the expected value is lower than the sure thing) or risk averse (selecting the sure thing even when the expected value of the gamble is higher).

Functional neuroimaging studies have shown that risk-taking decision making is critically dependent on the activation of inferior prefrontal cortex Paulus et al., 2001, Ernst et al., 2002, ventromedial and ventrolateral frontal cortex Elliott et al., 1999, Elliott et al., 2000a, Rogers et al., 1999b, anterior cingulate (Elliott et al., 2000a), insula (Critchley et al., 2001), and parietal cortex (Paulus et al., 2001). The anterior cingulate has been implicated in the response selection process when the reward magnitude is altered (Bush et al., 2002), whereas the nucleus accumbens has been shown to activate during anticipation of reward (Knutson et al., 2001). Others have argued that anterior cingulate activation during decision making is related to the degree to which the outcome is uncertain (Elliott and Dolan, 1998), whereas the activation in the nucleus accumbens is due to the calculation of an error signal between an expected and received reward (Pagnoni et al., 2002).

The precise role of the neural substrate underlying risk-taking decision making is not fully understood. The somatic marker hypothesis (Damasio, 1996) has provided a conceptual basis for processes involved in risk-taking decision making and has been used to link discrete neural substrates to risk-related behaviors. This hypothesis poses that external or internal stimuli initiate a state that is associated with pleasurable or aversive somatic markers. These markers function to guide the person's behavior by biasing the selection toward actions that result in an increase in pleasurable somatic markers (while avoiding actions resulting in aversive somatic markers).

The neural systems underlying the somatic marker hypothesis comprise the ventromedial and orbitofrontal cortex, amygdala, insula, and ventral striatum. In particular, the insula acts as a critical interface between affective inputs from limbic structures such as the orbitofrontal cortex, amygdala (McDonald et al., 1999), and anterior cingulate and the attentional prefrontal-parietal network in the processing of somatic states associated with risk-taking decision making (Bechara, 2001). Within the context of the somatic marker hypothesis, the insula has been conceptualized as part of both the “body” and the “as-if” loop system that is critical for the initial representation and the reenactment of somatic markers (Bechara, 2001). In this scenario, increased activation in the insula may signal the strength of the somatic state. If the insula signal is associated with aversive somatic markers, a relatively large activation during a decision-making situation would signal a potentially aversive outcome and may guide the subject to avoid the selection of a risky action alternative.

This investigation examined the hypothesis that the degree of risk-taking is related to the degree of activation in the insula. Specifically, it was hypothesized that a large activation in the insula during a risky response, which would correspond to a potent aversive representation of a somatic state, is associated with a lower propensity to select a risky response. Moreover, if insula activity were related to risk-taking behavior and not to response to punishment, one would expect to observe differential activation during risk-taking trials that were not punished versus those that were punished. Finally, to examine the external validity of this approach, two temperament or personality measures were obtained and correlated with activation in the insula. Cloninger (1987) developed the Temperament and Character Inventory (TCI) to quantify several dimensions of personality. These temperamental dispositions are defined in terms of the basic stimulus–response characteristics and comprise novelty seeking, harm avoidance, and reward dependence. In particular, harm avoidance has been used as a measure of anxiety proneness and reduced risk-taking propensity (Cloninger et al., 1998). In addition, the NEO personality inventory (Costa and McCrae, 1992), a measure of five personality factors, was used to obtain convergent validity that temperamental sensitivity to negative stimuli is closely associated with harm avoidance. In combination, it was hypothesized that a high degree of harm avoidance (i.e., the opposite of risk taking) or neuroticism is associated with a large activation in the insula during a risky response.

Section snippets

Subjects

Seventeen healthy, right-handed subjects (6 females and 11 males) age 38.3 years ± 1.4 (range 27–53) with an average education level of 14.7 ± .5 years (range 11–18) without a life-time history of Axis I DSM-IV disorders based on a structured clinical interview for DSM-IV diagnosis (Spitzer et al., 1992) participated in this study, which was approved by the UCSD Human Research Protection Program. These subjects gave their informed, written consent and performed a Risky-Gains decision-making

Behavioral results

Subjects selected the “safe” 20 response 46% ± 4 of the time, the “risky” 40 response 26% ± 5 of the time, the “risky” 80 response 23% ± 5 of the time [F(2, 32) = 4.75, P < 0.05] and failed to respond 4% of the time. Therefore, subjects selected “safe” and “risky” responses with similar frequencies. As shown in Fig. 2, there was a significant interaction between prior punishment and response type [F(2, 32) = 9.0, P < 0.01] wherein subjects made fewer 40 and 80 (risky) responses after punishment.

Discussion

This investigation yielded three main findings. First, right anterior insula (BA 13) activation was significantly larger when subjects selected a “risky” response versus selecting a “safe” response but also showed significant activation during punishment. Moreover, bilateral insula and left superior parietal lobule activation was larger during nonpunished “risky” responses than during punished trials. Second, in both cases the degree of insula activation was related to the probability of

Acknowledgements

We acknowledge the invaluable help of Larry Frank, Rick Buxton, and Eric Wong in supporting the fMRI experiments. This work was supported by grants from NIMH (R21DA13186, M.P.P.), NARSAD (M.P.P.), and support from the Veterans Administration via Merit Awards (M.P.P. and M.B.S.) and a VISN 22 MIRECC.

References (64)

  • P. Glimcher

    Decisions, decisions, decisionschoosing a biological science of choice

    Neuron

    (2002)
  • M. Liotti et al.

    Differential limbic–cortical correlates of sadness and anxiety in healthy subjectsimplications for affective disorders

    Biol. Psychiat.

    (2000)
  • B.A. Mellers et al.

    Representations of risk judgments

    Org. Behav. Hum. Decision Processes

    (1994)
  • K. Mogg et al.

    Biased cognitive operations in anxietyartefact, processing priorities or attentional search?

    Behav. Res. Ther.

    (1991)
  • F.M. Moresco et al.

    In vivo serotonin 5HT(2A) receptor binding and personality traits in healthy subjectsa position emission tomography study

    NeuroImage

    (2002)
  • M.P. Paulus et al.

    Prefrontal, parietal, and temporal cortex networks underlie decision-making in the presence of uncertainty

    NeuroImage

    (2001)
  • A.R. Peirson et al.

    Relationship between serotonin and the temperament and character inventory

    Psychiat. Res.

    (1999)
  • S. Rahman et al.

    Decision making and neuropsychiatry

    Trends. Cogn. Sci.

    (2001)
  • R.D. Rogers et al.

    Dissociable deficits in the decision-making cognition of chronic amphetamine abusers, opiate abusers, patients with focal damage to prefrontal cortex, and tryptophan-depleted normal volunteersevidence for monoaminergic mechanisms

    Neuropsychopharmacology

    (1999)
  • D.A. Solomon et al.

    Personality traits in subjects with bipolar I disorder in remission

    J. Affect. Disord.

    (1996)
  • V. Starcevic et al.

    Personality dimensions in panic disorder and generalized anxiety disorder

    J. Affect. Disord.

    (1996)
  • I. Tracey et al.

    Noxious hot and cold stimulation produce common patterns of brain activation in humansa functional magnetic resonance imaging study

    Neurosci. Lett.

    (2000)
  • E.U. Weber et al.

    A theory of perceived risk and attractiveness

    Org. Behav. Hum. Decision Processes

    (1992)
  • Diagnostic and Statistical Manual of Mental Disorder (4th Edition): DSM-IV

    (1994)
  • L.R. Becerra et al.

    Human brain activation under controlled thermal stimulation and habituation to noxious heatan fMRI study

    Magn. Reson. Med.

    (1999)
  • A. Bechara

    Neurobiology of decision-makingrisk and reward

    Semin. Clin. Neuropsychiatry

    (2001)
  • S. Blairy et al.

    5-HT2a receptor polymorphism gene in bipolar disorder and harm avoidance personality trait

    Am. J. Med. Genet.

    (2000)
  • G.M. Boynton et al.

    Linear systems analysis of functional magnetic resonance imaging in human V1

    J. Neurosci.

    (1996)
  • C. Buchel et al.

    Amygdala-hippocampal involvement in human aversive trace conditioning revealed through event-related functional magnetic resonance imaging

    J. Neurosci.

    (1999)
  • G. Bush et al.

    Dorsal anterior cingulate cortexa role in reward-based decision making

    Proc. Natl. Acad. Sci. USA

    (2002)
  • T. Canli et al.

    An fMRI study of personality influences on brain reactivity to emotional stimuli

    Behav. Neurosci.

    (2001)
  • C. Cavada et al.

    Posterior parietal cortex in rhesus monkey. I. Parcellation of areas based on distinctive limbic and sensory corticocortical connections

    J. Comp. Neurol.

    (1989)
  • Cited by (614)

    View all citing articles on Scopus
    View full text