Is fear right? Applying Gray's two-dimensional neuropsychology of defence to the approach-withdawal model of frontal-brain asymmetry.

Abstract

Two prominent theories have guided decades of research into the biological basis of human fear and anxiety. Davidson's Anterior Asymmetry and Emotion model holds that fear and anxiety derive from a single, withdrawal system, which is in turn aligned with right prefrontal regions (as part of a larger network of ingulate/subcortical structures). In contrast, Gray's Reinforcement Sensitivity Theory (RST), though originally of a similar view, has since been revised to account for an observed functional, behavioural and pharmacological distinction between fear and anxiety which Gray now aligns with two neurobiologically separable systems, the Fight, Flight, Freeze System (FFFS; fear) and the Behavioural Inhibition System (BIS; anxiety). Whilst generating a great deal of empirical interest, Davidson's withdrawal system/right prefrontal hypothesis has to date received very mixed support. It was proposed here that by applying Gray's fear/anxiety distinction to Davidson's model, some of this past inconsistency may be accounted for. Specifically, it was hypothesised that right prefrontal regions should be aligned with fear (FFFS), but not anxiety (BIS). However, although established at an animal level, Gray's BIS/FFFS distinction has received minimal attention within a human experimental setting. Thus testing this right prefrontal/FFFS hypothesis was first contingent on deriving valid methods of distinguishing the FFFS and BIS at a human level. This was initially addressed by exploring the potential relevance of existing self-report trait measures.

Critically, it was observed that Gray's revised RST (which also includes a third system, the Behavioural Activation System, BAS) now closely resembles the three primary dimensions common to most three-factor models of temperament/personality. In particular, extensive conceptual, neurobiological and behavioural overlap was documented between trait models of Gray's three systems (FFFS, BIS and BAS) and Tellegen's Negative Emotionality, Constraint and Positive Emotionality superfactors. Accordingly, Constraint (in particular the Control and Harm Avoidance subscales) and Negative Emotionality (in particular the Stress Reaction subscale) were proposed as potential trait level markers of the revised BIS and FFFS respectively. This was supported in Study 1 (N=190) by observing that Control, Harm Avoidance and Stress Reaction loaded highly and exclusively onto orthogonal BIS (Control and Harm Avoidance) and FFFS (Stress Reaction) factors; two of three factors which emerged when these measures were entered into an exploratory factor analysis with measures of the old (pre-revised RST) BIS construct (i.e., Carver and White BIS scale, CW-BIS; the Trait scale of the Spielberger State Trait Anxiety Inventory, STAIT, Sensitivity to Punishment scale, SP), and existing BAS measures. Somewhat surprisingly, all old BIS measures loaded exclusively on the FFFS factor.