Elsevier

Cortex

Volume 66, May 2015, Pages 81-90
Cortex

Research report
Amygdala signals subjective appetitiveness and aversiveness of mixed gambles

https://doi.org/10.1016/j.cortex.2015.02.016Get rights and content

Abstract

People are more sensitive to losses than to equivalent gains when making financial decisions. We used functional magnetic resonance imaging (fMRI) to illuminate how the amygdala contributes to loss aversion. The blood oxygen level dependent (BOLD) response of the amygdala was mapped while healthy individuals were responding to 50/50 gambles with varying potential gain and loss amounts. Overall, subjects demanded twice as high potential gain as loss to accept a gamble. The individual level of loss aversion was expressed by the decision boundary, i.e., the gain-loss ratio at which subjects accepted and rejected gambles with equal probability. Amygdala activity increased the more the gain-loss ratio deviated from the individual decision boundary showing that the amygdala codes action value. This response pattern was more strongly expressed in loss aversive individuals, linking amygdala activity with individual differences in loss aversion. Together, the results show that the amygdala signals subjective appetitiveness or aversiveness of gain-loss ratios at the time of choice.

Introduction

When making economic decisions, people often deviate from rational behavior (Kahneman & Tversky, 1979). For instance, people tend to overestimate the impact of losing, consequently biasing decisions towards loss aversion: when presented with risky gambles with equal chances of winning and losing, people demand on average twice the amount of potential gains compared to losses in order to accept a gamble (Kahneman and Tversky, 1984, Tom et al., 2007).

Clinical investigations have found diminished loss aversion-bias in amygdala-lesioned patients compared to healthy controls (De Martino, Camerer, & Adolphs, 2010). How the amygdala influence the willingness to accept gambles is not clear. One possibility is that the amygdala are responding to magnitudes of either gains or losses in order to avoid or deal with aversive events (LeDoux, 2000). In line with this view, a recent loss aversion study reported that amygdala activity reflected magnitudes of single losses, but not single gains (Canessa et al., 2013). However, other neuroimaging studies failed to support an involvement of amygdala in the evaluation of single losses or gains (Sokol-Hessner et al., 2013, Tom et al., 2007). These studies rather pointed to other dopaminergic meso-cortico-limbic target areas such as the medial orbitofrontal cortex (mOFC) and ventral striatum which process single gain and loss magnitudes.

The classical view that the amygdala are mainly geared to negative events has been recently challenged by neuroimaging studies showing that the amygdala computes both negative and positive stimulus values during value-based decision-making (Baxter and Murray, 2002, Bermudez et al., 2012, Grabenhorst et al., 2012, Jenison et al., 2011). The “bivalent” coding of value in the amygdala, which is not specific to negativity or positivity of a stimulus, per se, suggests that the amygdala may track other properties of these value stimuli such as task relevance, the impact or consequence of a choice, or the biological salience of a stimulus. Indeed, several lines of work indicate that the involvement of the amygdala in decision-making goes beyond mere value estimation. Functional neuroimaging revealed context dependent activation of the amygdala reflecting whether choices were framed in terms of avoiding losses or seeking gains (De Martino, Kumaran, Seymour, & Dolan, 2006). Further, amygdala activity was linked to choice-related emotions such as ‘relief of a good choice’ or ‘regret of a bad choice’ (Coricelli et al., 2005, Rogan et al., 2005, Sangha et al., 2013, Seymour et al., 2005). Finally, a series of studies suggest that the amygdala play a key role in evaluating task relevance rather than reflecting the absolute magnitude of value (Bzdok et al., 2011, Ousdal et al., 2012, Sander et al., 2003, Wright and Liu, 2006).

Here, we aimed at resolving the role of amygdala in loss aversive decision-making. We used fMRI to map BOLD-responses in the amygdala of healthy subjects, who decided whether to accept or reject “mixed” (gain-loss) gambles. We manipulated only magnitudes of potential gains and losses, while keeping win-lose probabilities equal (i.e., 50%). Participants received no feedback on whether they won or lost, creating a decision-context of maximal uncertainty. Inter-individual differences in the tendency to weigh losses higher than gains were expressed by the individual decision boundary lambda (λ), which represents the gain-loss ratio where subjects on average choose to accept or reject bets with equal probability.

Importantly, in contrast to previous loss aversion studies (Canessa et al., 2013, De Martino et al., 2010, Sokol-Hessner et al., 2013, Tom et al., 2007), each gamble started with a magnitude presentation phase where either the potential loss or gain amount of the gamble was presented alone. This was followed by a decision phase where the full gamble with specified loss and gain amount appeared, and subjects were required to reject or accept the mixed gamble (Fig. 1A). This procedure allowed us to temporally separate neural responses elicited by increasing gain or loss amounts during the first part of the gamble from the assessment of the full mixed gamble defined by the gain-loss ratio during the decision-making phase.

Our study tested two mutually exclusive hypotheses: The amygdala might primarily respond to individual magnitudes of potential monetary losses and/or gains (Belova et al., 2007, Belova et al., 2008, Canessa et al., 2013, Paton et al., 2006, Salzman et al., 2007). Alternatively, the amygdala might integrate both gain and loss-magnitudes into the decision process. The latter hypothesis makes the prediction that the amygdala assesses the value of the full gain-loss ratio, relative to the individual decision boundary λ. If this were the case, amygdala activity might either be tuned to gain-loss ratios far away from the individual decision boundary or preferentially deal with increasingly ambiguous decisions, i.e., gamble ratios that are close to the individual decision boundary.

Section snippets

Subjects

Sixteen male subjects (age range 20–32 years; median age 24 years; 9 right handed) participated in the experiment. Two additional subjects were scanned but not analysed, one due to experimenter error, and one due to missing field measurements.

Subjects were screened for any history of neurological or cardiovascular disorders, contra-indications to MRI-scanning and signed health declarations before study commencement. The study was approved under the ethical protocol KF 01 – 131/03, issued by the

Behavioral results

All subjects weighted potential losses higher than potential gains with a λ-value above 1 (mean λ = 2.16; median λ = 2.08; range of individual λ-values: 1.38–2.92). This is consistent with previous studies, which reported mean λ-values around 2 (Tom et al., 2007, Tversky and Kahneman, 1992). Decision time increased with ambiguity of the bet. The more the gain-loss ratio of a gamble approached the individual decision boundary λ, the more time subjects needed to decide whether or not to accept the

Discussion

The present study sheds new light on the role of the amygdala in loss averse decision-making. We found that the amygdala assessed the gain-loss ratio of a full bet during a gambling task that required subjects to balance the gain and loss of each bet in relation to subject-specific decision boundaries. In contrast, single gain or loss amounts did not modulate amygdala activity. This implies that the amygdala evaluates a mixed gamble as a whole rather than assessing single magnitudes of

Acknowledgments

This work was supported by the The Danish Council for Independent Research in Social Sciences through a grant to Thomas Ramsøy (no. 0601-01361B) and by the Lundbeck Foundation through a Grant of Exellence (no. R59 A5399) to Hartwig Siebner. We thank all the participants for their time. We thank Sid Kouider for helpful comments on the manuscript, Martin Skov for initial discussions, Helle Ruff Laursen for assistance during scanning and two anonymous reviewers for their helpful comments on the

References (39)

  • M.G. Baxter et al.

    The amygdala and reward

    Nature Reviews Neuroscience

    (2002)
  • M.A. Belova et al.

    Moment-to-moment tracking of state value in the amygdala

    Journal of Neuroscience

    (2008)
  • D. Bzdok et al.

    ALE meta-analysis on facial judgments of trustworthiness and attractiveness

    Brain Structure & Function

    (2011)
  • N. Canessa et al.

    The functional and structural neural basis of individual differences in loss aversion

    Journal of Neuroscience

    (2013)
  • M. Carreiras et al.

    An anatomical signature for literacy

    Nature

    (2009)
  • G. Coricelli et al.

    Regret and its avoidance: a neuroimaging study of choice behavior

    Nature Neuroscience

    (2005)
  • M. Davis et al.

    The amygdala: vigilance and emotion

    Molecular Psychiatry

    (2001)
  • B. De Martino et al.

    Amygdala damage eliminates monetary loss aversion

    Proceedings of the National Academy of Sciences of the United States of America

    (2010)
  • B. De Martino et al.

    Frames, biases, and rational decision-making in the human brain

    Science

    (2006)
  • Cited by (15)

    • Aberrant neural signatures of decision-making: Pathological gamblers display cortico-striatal hypersensitivity to extreme gambles

      2016, NeuroImage
      Citation Excerpt :

      Previous studies using mixed gambles with healthy participants found that the separate valuation of gains and losses involve reward-related dopaminergic target regions, specifically the ventral striatum and the vmPFC (Tom et al. 2007). However, when the entire gain/loss gamble is taken into account (i.e., potential gain, potential loss, and the consequences of winning or losing), other studies have found an important role for the amygdala in loss aversion (De Martino et al. 2010; Gelskov et al. 2015). In the present study, we used this task in a population suffering from gambling addiction as a means to gain insight into aberrant value-based decision-making.

    View all citing articles on Scopus
    1

    The senior authors contributed equally to the paper.

    View full text