The neural correlates of subjective pleasantness

Abstract

Processing of subjective pleasantness is essential in daily life decision making, particularly in the context of cognitive and environmental factors. Pleasure is mediated by a neural network and this network has been suggested to be the biological basis of pleasure including a whole range of different modalities and domains of pleasantness. This quantitative meta-analysis of brain imaging data focuses on studies 1) based on correlations between self-reported judgements of pleasantness and brain regions and investigates whether 2) immediate (during scanning) versus subsequent judgements (after scanning) differ in brain activity.

We investigated concurrence across 40 studies reporting brain regions correlated with self-reported judgements of subjective pleasantness (attractiveness, liking or beauty) by means of activation likelihood estimation (ALE).

Positive correlates of subjective pleasantness were found in mOFC, ventromedial prefrontal cortex, left ventral striatum, pregenual cortex, right cerebellum, left thalamus and the mid cingulate cortex. Negative correlates were found in left precentral gyrus, right cerebellum and right inferior frontal gyrus. A comparison of studies with subjective pleasantness judgement during or after scanning revealed no significant differences in brain activation.

We conclude that subjective pleasantness judgements are directly related to brain regions that have been described as part of the reward circuitry (mOFC, ventral striatum). The results suggest that the evaluation of likability or pleasure is an automatic process and that it is neither elicited nor enhanced by instructions to report the outcome of these judgements.

Introduction

In daily life we frequently encounter situations in which we have to make choices based on our subjective pleasantness processing, e.g. when buying fruits in a grocery store. In order to decide between different options the brain computes the values associated with stimuli and integrates it with motivational, cognitive and contextual needs (Kable and Glimcher, 2009). We might therefore choose oranges instead of cherries because we strive for something sweet but are thirsty at the same time. Neuroimaging research has made advances in understanding how the brain implements positive value processing and represents subjective pleasantness. In particular the medial prefrontal cortex (orbitofrontal (mOFC) and ventromedial prefrontal cortex), the ventral striatum and the amygdala have been implicated in reward processing (Burgdorf and Panksepp, 2006, Grabenhorst and Rolls, 2011, Liu et al., 2011, Peters and Büchel, 2010). In order to systematically explore brain regions involved in subjective pleasure processing we conducted a meta-analysis using the activation likelihood estimation (ALE) approach (Eickhoff et al., 2009, Laird et al., 2005, Turkeltaub et al., 2002). ALE allows disclosure of statistically significant concordance of activated voxels across numerous studies while controlling for chance clustering. By seeking concordance at the voxel level, ALE tests for statistically reliable clustering of activations in standardised locations, which avoids spatial distinction errors and incongruence of labelling across studies that can befall narrative-based reviews.

A recent ALE meta-analysis reported common brain activation of positive-valence neuroaesthetic processing across different sensory modalities in the right anterior insula (Brown et al., 2011). This meta-analysis included multiple studies in which positive valence was only implicitly assumed and not substantiated by participants' subjective reports. The analysis comprised investigations in which pictures of participants' offspring were shown and compared with pictures of other peoples' offspring. Alternatively, pictures of naked individuals compared with dressed individuals were presented on the assumption that naked bodies do elicit positively valenced emotions (e.g. Bocher et al., 2001, Ferretti et al., 2005, Leibenluft et al., 2004). Since many of the studies lack the association with participants' subjective pleasantness it is not clear what participants actually experienced during scanning. In order to address this problem we present a quantitative meta-analysis on coordinates of neuroimaging studies in which participants indicate how pleasant or attractive they regard certain experiences such as odour, taste, music or visual stimuli. We were particularly interested in the neural correlates of subjective pleasantness across various domains; we therefore selected studies in which 1) participants themselves judged pleasantness and 2) brain signal was reported as correlations or contrasts based on these individual judgements.

Recently research has been undertaken to explore whether choices or obtainable rewards need to be real in order to activate brain regions such as mOFC and ventral striatum. Bray et al. (2010) demonstrated that the experience of real as well as imagined rewards engages the mOFC. Furthermore, Kang et al. (2011) have shown mOFC and ventral striatal activation during real as well as hypothetical choices. From these studies one may conclude that activation within pleasure related brain regions does not depend on the reality of the outcome. Within the present meta-analysis we want to take the exploration of boundary conditions of hedonic brain activation one step further by asking the question whether the instruction to think about pleasantness or likability during scanning is required to elicit mOFC activation. Since it has been proposed that the mOFC represents the affective or so called goal value of stimuli (Grabenhorst and Rolls, 2011, Kringelbach, 2005) we set out to investigate whether mOFC is only engaged when an explicit judgement is required. In order to explore to what extent these areas are dependent on explicit instructions and to rate subjective pleasantness, we subdivided studies into those that afforded judgements during scanning itself and those in which the judgement was assessed before or after scanning.