A return to the psychiatric dark ages with a two-system framework for fear

Highlights

• The two-system framework for fear, in which fear is purely subjective, is critiqued.

• The authors propose several negative ramifications of adopting the two-system framework.

• Fear is argued to be an integrated physiological, behavioral and subjective danger response.

Abstract

The past several decades has seen considerable progress in our understanding of the neurobiology of fear and anxiety. These advancements were spurred on by envisioning fear as emerging from the coordinated activation of brain and behavioral systems that evolved for the purpose of defense from environmental dangers. Recently, Joseph LeDoux, a previous proponent of this view, published a series of papers in which he challenges the value of this approach. As an alternative, he and colleagues propose that a ‘two-system’ framework for the study of responses to threat will expedite the advancement of medical treatments for fear disorders. This view suggests one system for autonomic and behavioral responses and a second for the subjective feeling of fear. They argue that these two systems operate orthogonally and thus inferences concerning the emotion of fear cannot be gleaned from physiological and behavioral measures; confounding these systems has impeded the mechanistic understanding and treatment of fear disorders. Counter to the claim that this view will advance scientific progress, it carries the frightening implication that we ought to reduce the study of fear to subjective report. Here, we outline why we believe that fear is best considered an integrated autonomic, behavioral, and cognitive-emotional response to danger emerging from a central fear generator whose evolutionarily conserved function is that of defense. Furthermore, we argue that although components of the fear response can be independently modulated and studied, common upstream brain regions dictate their genesis, and therefore inferences about a central fear state can be garnered from measures of each.

Introduction

Across phylogeny, organisms display characteristic responses to danger, allowing them to avoid predation and other dangers in their environment (Bolles, 1970, Fanselow and Lester, 1988). These responses entail both internal physiological changes including increased heart rate and respiration, and external behaviors such as fight and flight responses (Davis, 1992, Fendt and Fanselow, 1999, Perusini and Fanselow, 2015).

The mental health field has placed great interest in responses to danger (also referred to as defensive behavior) in an effort to understand fear and anxiety disorders, often conceptualized as the body's defensive response exceeding its adaptive function. Owing largely to the relative ease with which behavioral and physiological responses to threat can be evoked in model organisms, as well as the quantitative manner in which they can be measured, we now know a great deal about defensive circuits in the brain (Davis, 1992, Duvarci and Pare, 2014, Fanselow and LeDoux, 1999, Johansen et al., 2011, Paré et al., 2004). This research has already provided us with the ability to predict the efficacy of therapeutic drugs, from benzodiazepines for the reduction of fear and anxiety (Fanselow and Helmstetter, 1988, File and Pellow, 1985, Hart et al., 1998) to d-cycloserine for the augmentation of exposure therapy (Bouton et al., 2008, Bowers and Ressler, 2015, Mataix-Cols et al., 2017, Woods and Bouton, 2006). In addition, studies of the ontology of defensive responses have provided us with information relevant to behavioral therapies; for example, understanding why exposure therapy is liable not to transfer beyond the therapist's office (Bouton, 2002, Bouton, 2004, Bouton et al., 2006).

In several recent and widely publicized papers, LeDoux and colleagues call into question the utility of using autonomic and behavioral responses to danger to make inferences about the associated subjective emotional states of fear and anxiety (LeDoux & Pine, 2016). They argue that autonomic and behavioral responses to threat are orthogonal to the subjective experience of fear (Fig. 1A). Therefore, the terms fear and anxiety should only be used in reference to subjective mental experience, and should be studied accordingly. They propose that the failure to distinguish the systems supporting fear and anxiety from those giving rise to the autonomic and behavioral responses to threat – their ‘two-system framework’ – is one of the reasons that ‘progress has stalled in treatment development for mental disorders’ (LeDoux and Brown, 2017, LeDoux and Pine, 2016, LeDoux, 2017).

Here we contend with this view, and argue that the autonomic, behavioral, and cognitive-emotional responses to danger are best conceptualized as the unitary result of activation of a central fear generator (i.e. one-system).