Philosophical aspects of neurobiological research on anxiety and anxiety disorders
Published in: Philosophical aspects of neurobiological research on anxiety and anxiety disorders. Current Opinion in Psychiatry, 17 (2004), 457-464.
Purpose of review:
Neurobiological research literature on anxiety and anxiety disorders of the previous year (April 2003 – June 2004) is investigated from the perspective of four conceptual questions: (1) is pathological anxiety an animal, a human or a layered phenomenon with aspects of both? (2) is anxiety a unitary or a non-unitary phenomenon? (3) what is the relationship between different aspects, or ‘components’, of anxiety? (4) what is the relationship between scientific and clinical concepts of anxiety?
Recent findings: These questions are investigated in research papers in the areas of fear conditioning, delineation of neural pathways underlying aspects of anxiety (and anxiety disorder), and developmental plasticity. With respect to question (1) most authors are inclined to a view in which both animal and human aspects are recognized and combined in some way. Methodological restrictions limit the claims of researchers. Underneath the surface, however, reductionistic tendencies may exist. With respect to question (2) authors are predominantly inclined to a non-unitary view, against the background of model of multiple overlapping brain circuits. With respect to issue (3) the language of causal interaction between reified subsystems is avoided; leaving however other issues undiscussed, such as the issue of the unity of brain functioning and the relationship of neural function and subjective experience. The dominant framework seems to be non-eliminative physicalist. With regard to issue (4), clinical versus scientific understanding, there is a widely acknowledged conceptual gap between these two; however, newer methodologies enable a broader understanding, by taking into account that environmental factors play an important role in brain development and that study of regional circuitry is needed above study of neural loci.
Summary: There is a paucity of papers on purely conceptual issues. The rich phenomenology of anxiety should serve as an incentive to deeper understanding. Conceptual rigor and refinement may have an important impact upon fundamental empirical research.
Professor Gerrit Glas MD PhD, Zwolse Poort, P.O.Box 110, 8000 AC ZWOLLE, The Netherlands.
Tel: ++31 38 4565 847 or ++31 6 1091 4513.
The study of anxiety, more than of any emotion, has been a touchstone for sound thinking and conceptual clarity. It is evident why this is the case: the phenomenology of anxiety is elusive; its manifestations display an enormous variety; linguistic meanings are not sharply delineated and show considerable overlap; and there is a wide variety of disciplines, preclinical and clinical, in which the study of anxiety is undertaken – a variety which asks for creativity and interdisciplinary research.
In this article I will first give a short introduction into some of the ‘large’ questions surrounding anxiety research. Then I will review those facets of recent neurobiological research on anxiety and anxiety disorders which in some way or other come close to a discussion of these questions. In the conclusion I will evaluate these developments by relating them to the ‘larger’ questions previously mentioned.
Classical questions in anxiety research
Any theory of anxiety has to face the following conceptual issues and/or questions:
1. Should anxiety be considered as an animal reaction or as a typically human emotional response or as a layered phenomenon with both animal and specific human elements?
If anxiety is basically seen as an animal response, how are the typical human aspects related to this animal ‘core’? This question becomes even more difficult when applied to anxiety disorder. Do pathological forms of anxiety represent remnants of some ‘old’, animal way of brain functioning which has escaped from cortical control? And if so, does this mean that we should deny any psychological or existential relevance to the fears of anxiety disorder patients?
This set of questions has been addressed by scientists who base their views on Paul MacLean’s work on the so-called triune brain (for a reappraisal see Lambert [1*] and Ploog [2*]. A clear example of this approach can be found in an early paper by Jack Gorman et al. , who hypothesized that the symptoms of panic disorder (with agoraphobia) should be subdivided into three subsets representing dysfunctions of each of the three parts of the triune brain. The panic attack itself, especially its physiological and motor manifestations, were to be seen as the expression of abnormal brainstem functioning, anticipatory anxiety as a reflection of limbic system overactivity, and avoidance behavior as the result of activation of the frontal cortex. In this conceptualization a hierarchical view of brain functioning seems to emerge – one in which brainstem functioning is kept under control by the limbic system and the limbic system by the prefrontal cortex. Recent theorizing – see next section – casts doubt on this more MacLean type of conceptualization of brain functioning.
2. Is anxiety a unitary phenomenon or the result of the functioning of different, overlapping brain circuits?
In older emotion theories, such as circumplex, early developmental and differential models, the unitary view prevails. At present, the non-unitary view seems by far dominant. Peter Lang, psychophysiologist and one of the first advocates for a non-unitary view, argues that fear should not be seen as “some hard phenomenal lump”, but as a conglomerate of “three loosely coupled response systems”, i.e., physiological activation, fear related behavioral acts, and verbal reports of distress [4, 5]. Psychopathological conditions like agoraphobia are characterized by desynchrony between these systems.
LeDoux discerns two fear-mediating neuronal systems, the amygdala system and the hippocampal system . The amygdala system represents the quick and dirty part of the alarm system, which means that it reacts to a wide variety of fear-related stimuli. The hippocampal system is narrowly tuned to more specific fear-provoking cues and may give rise to explicit memories. Because both systems function in parallel, fear-eliciting situations may occasion at the same time both uniform physiological and motor reactions and specific behaviors related to more explicit memories and evaluations.
3. How should the relationship between the different aspects or ‘components’ of anxiety be conceptualized – as a causal relationship or as a hierarchical or parallel relationship regulated by some underlying pattern or structure?
This question inevitably leads into the philosophical conundrum of the mind-body relationship: how are motor and physiological responses related to subjective experience? Does evaluation of the situation lead to emotional responses (as cognitive theorists suggest) or is it the other way around and are feelings nothing but the ‘felt perception’ of bodily changes (like William James and Antonio Damasio affirm)? And, more basically, are there causal relationships at all between these ‘components’ or features of anxiety, or should the construction of such causal relationships be considered as being based on a gross conceptual misunderstanding, i.e., the unjustified reification of (groups of) functions?
Bennett & Hacker [**7] defend the latter view. In their important and powerful book on the conceptual underpinnings of neuroscientific research they argue against Antonio Damasio (and others), who in their view seem to accept causal relationships between mental images and thoughts on the one hand and emotional reactions on the other hand.
Although their presentation of Damasio’s views is a little one-sided, it should be admitted that Damasio construes a kind of temporal-causal series in which mental images give rise to bodily reactions (autonomic, motor, endocrine), which by peripheral feedback contribute to the generation of feelings. Feelings, according to Damasio, consist of a fusion between the original mental images, generating the bodily reaction, and the ‘on-line’ registration of these bodily changes [8, 9].
Bennett & Hacker build forth on a rich tradition of Anglo-Saxon philosophy of mind, in particular the work of Anthony Kenny  . Their attack on Damasio is conceptual. Kenny and others maintained that emotions are to be distinguished on the basis of the object of emotion and the circumstances in which they occur, and not on the basis of bodily responses or the causes of these responses. What makes the blushes blushes of shame instead of blushes of love are not the blushes themselves, nor the images or thoughts causing them, but the circumstances and the object of the emotion. The object of emotion should be distinguished from its cause. Mary’s blushes may be caused by an innocent remark, the object of her emotion may be something different, for instance her love for Pete which she wishes to hide and might become obvious by the remark. Objects of emotion refer to situations that are intrinsically (internally; non-contingently; intentionally) tied to the emotional state; causes are externally (contingently) related to the emotions they occasion. Causes and objects may overlap, however.
In order to understand the meaning of a particular emotional state one should first gain a clear picture of the object of emotion. Causes provide the additional information, for instance about the threshold for the emotion to arise at all. Objects of emotions are formed against the background of more global notions of the self. This implies that by understanding the objects of emotion we are not merely understanding the emotion, but first of all the person having that emotion.
What is the relevance of the distinction between object and cause for the discussion about the relationship between ‘components’ of emotion? The most important point to make here is that emotions are primarily qualified by their object and not by the nature of one of their components or the relationship between these components. The primacy of the object in the definition and the understanding of emotion highlights that – from an epistemological perspective – the distinction between components of emotion comes afterwards and can only be made on the assumption of a preexistent understanding of what it is to have that particular emotion. So, the primacy of the object in the definition of emotion highlights the abstract nature of many theoretical accounts of emotion.
A theorist may have good reasons for isolating (abstracting) a particular aspect or component of emotion. Scientific understanding can not do without such abstractions. However these abstractions and the constructs that are based upon them should not be identified with the emotion itself. Construing causal relationships between components of emotion easily leads to substantialization (reification) of these components. Instead of being used as conceptual devices which are meant to bring order into an overwhelmingly complex range of phenomena, these constructs are held for quasi-entities having an existence in themselves. This, subsequently, leads to all kinds of pseudo-problems, for instance the question how the reified components are causally related.
4. How are clinical concepts of anxiety related to scientific concepts of anxiety?
Given the important role of abstraction in the construction of scientific hypotheses, one may wonder how contemporary anxiety research relates to the clinical practice of treating patients with anxiety disorder. Not only do theories abstract from the biographical and social embedding of anxiety disorder; by addressing only one or two aspects of the phenomenon of anxiety, they rupture the coherence with other aspects. At the level of scientific theory formation this does not need to be a problem, however, it may become a problem when scientific knowledge is translated back to the level of clinical practice. How does panic relate to firing of neurons in the brain stem; and what has checking behavior to do with dysfunction of the cortico-striato-basal ganglia system? Contemporary neurobiological research focuses, not only on entire circuits in the brain, but also on molecular and genetic processes mediating mood and behavior. This tendency to study processes at a microscopic level increases the distance between fundamental research and clinical practice. The DSM, which was meant to serve as a clinical diagnostic and research tool, was an attempt to fill this gap. The capacity of the DSM to serve both ends has proven to be limited. Endophenotypes, experimental paradigms, risk factors and psychobiological dimensions (genetic, temperamental, statistical) are the new concepts found out to bridge the distance between basic research and clinical reality. How successful are these concepts? Are there new developments to be welcomed?
Conceptual aspects in recent neurobiological literature
Let us take a closer look at the neurobiological literature in the period covered here (April 2003 –June 2004). This review covers only a selection of the enormous literature and is limited to studies which in some way address human anxiety and touch the conceptual questions raised above. The emphasis will, therefore, be on reviews and on articles with a more extended theoretical discussion.
Fear conditioning in animals and humans
An excellent example of the attempt to bridge the gap between animal and human studies is offered by Grillon & Baas’s review of the modulation of the startle reflex by affective states [7**]. The startle reflex is an automatic, cross-species response to sudden, intense stimulation with the likely purpose to facilitate flight and/or to protect the body. The importance of this reflex for the experimental study of behavior lies in its sensitivity to habituation, sensitization, sensorimotor gating and affective modulation. The startle response can be studied as a baseline variable reflecting constitutional (genetic) and acquired (e.g. sensitization) factors. Startle responses can also be used as an objective measure for the valence of an organism’s affective state. This is so, because the amplitude of the startle response varies with the affective state of the organism, with characteristically high amplitudes in aversive states like fear. This sensitivity for affective modulation is important because psychophysiological research has been impeded by the lack of independent measures for the organism’s emotional state. Classical measures such as heart rate and/or skin conductance are non-specific and merely reflect a state of physiological arousal. Variations in the magnitude of the startle response can be used to infer the organism’s emotional state. This is one of the reasons why the fear-potentiated startle reflex has become one of the most often used experimental methods to study conditioning processes in humans and animals. A third area of investigation concerns the use of the startle reflex as an index of sensorimotor gating. It has been noted that the startle response can be inhibited by stimuli presented shortly before the startle-eliciting stimulus. This so-called prepulse inhibition is thought to reflect a kind of protection of preattentive processing, i.e. a functional low-level gating of information processing in the presence of stimuli deserving the organism’s attention. Prepulse inhibition is diminished in conditions like schizophrenia and obsessive-compulsive disorder.
After a discussion of methodology and of variations in experimental design Gillon & Baas point at the importance of context conditioning, which is probably mediated by the bed nucleus of the stria terminalis (BNST). Aversive conditioning may involve not only conditioning to aversive stimuli, but also to the context in which these stimuli occur. Context conditioning is reflected by an increase in baseline startle when subjects return to an experimental situation where they were confronted with unpleasant stimuli. Some aspects of anxiety (e.g. general apprehension, worry) may better be modeled by context conditioning, whereas other aspects of anxiety (e.g. hypersensitivity to specific cues) are better understood by cued fear conditioning. Context conditioning might be (one of) the mechanism(s) mediating generalized anxiety disorder.
Posttraumatic stress disorder is a more difficult case, because data on the exaggerated startle reflex are mixed and seem to depend on the type of trauma, age of traumatization and length of the illness. Gillon & Baas suggest that the initial hypersensitivity to fear-related cues subsides in the course of years and may be seen as the result of sensitization or an inherited vulnerability. There are, at present, no prospective studies with regard to the second possibility. With respect to sensitization, however, there are a number of investigations supporting the idea that the exaggerated startle response in (relatively) early PTSD is a result of progressive sensitization caused by prolonged trauma-induced stress. This ‘tonic’ reaction is, then, distinguished from the ‘phasic’ heightened startle response in prolonged PTSD, which is caused by trauma-related stimuli or stressful environments implying a form of context-conditioning.
In other anxiety disorders, there does not seem to exist a global hyperresponsivity to non-specific stressors. Generally speaking, in anxiety disorder potentiation of the startle reflex is limited to cues that are specific. Because most anxiety disorders lack this specificity, they also lack potentiation of the startle reflex. The most notable exceptions to this are the small animal and blood/injury phobias. Anxious apprehension is a feature that cuts across the anxiety disorders and can be seen as a risk factor for the development of anxiety disorder. Girls (and adults) at risk do not show a fear potentiated startle (for specific cues), but their baseline startle appears to be elevated. These children have a tendency to react fearfully to mild situational stressors, suggesting that their fears are contextually conditioned. Data from groups of high-risk boys show the inverse pattern: normal contextual fear and increased fear potentiation to threat.
The authors conclude that startle has great potential as an investigative tool, but has limited validity as a diagnostic tool. This seems an acknowledgement of the gap between experimental design and clinical reality. As an investigative tool the startle reflex is of considerable interest for the study of mechanisms that are similar between different types of psychopathology, such as anxiety disorder, mood disorder, addiction and personality disorder. Refinement of the experimental paradigm has made it possible to distinguish between cued fear and context conditioning. The authors make suggestions with respect to the relevance of this distinction for the development and manifestation of PTSS and other anxiety disorders.
A point of discussion might be that the authors do not account for the difference between apprehensions directed at imaginary and real situations; and that they do not give an account of how ‘mental’ worries could induce the same changes in brain circuitry as repeated ‘physical’ stress.
The startle reflex is only one of a number of experimental designs to study mechanisms underlying animal and human anxiety. An extension of this approach to social behavior can be found in a study by Haller and coworkers who investigated the social behavior of both individually and group-housed rats after exposure to either electrical shock or social defeat [12*] Effects to electric shocks were not influenced by housing conditions. Group-housing did and individual housing did not protect the rats from the effects of social defeat (repeatedly being attacked by a larger rat). This article is illustrative of a tendency in animal research to include environmental factors as variable in experimental designs and to concentrate on ongoing behavior instead of on acute reactions. Another variations in the startle paradigm can be found in a study by de Jongh et al. [13*], who investigated the differential effects of light-enhanced and fear potentiated startle in mice. The effects in the light enhanced condition could be diminished by administration of CRH.
In studies in humans the fear conditioning paradigm has been investigated with respect to its genetic basis. Hettema et al. studied fear conditioning in monozygotic and dizygotic twins [14*]. Moderate heritability was demonstrated for habituation and acquisition and extinction of conditioned responses to evolutionary fear-relevant stimuli. Statistical analysis suggested that two sets of genes underlie the trait of fear conditioning: one that most strongly affects nonassociative processes such as habituation (but also extending to acquisition and extinction) and another related to associative conditioning. Like the startle reflex and (other) animal models twin studies of fear conditioning can best be seen as attempts to define endophenotypes which reflect processes more proximal to the mechanisms that are considered to be responsible for clinical conditions.
The statistical analysis of variance between mono- and dizygotics can be supplemented by molecular biological insights, for instance about the role of the serotonin transporter gene promotor. The study of Melke et al. [15*] on serotonin receptor 3A (HTR3A) polymorphism offers one of the many examples of this approach, with as main finding an association between HTR3A polymorphism and the temperamental dimension of harm avoidance. Earlier studies revealed no association with polymorphism for genes coding for 5-HT2A and 5-HT2C.
Dissection of neural pathways in humans
Contemporary anxiety research is strongly inclined to a non-unitary view of anxiety. Anxiety disorders are seen as representing the activity of a spectrum of functionally overlapping brain circuits.
The experimental induction of panic is one of the oldest approaches to carve out pathophysiological mechanisms underlying anxiety disorder. Substances like m-chorophenylpiperazine (mCPP), sodium lactate, yohimbine and cholecystokinine have been administered both in animals and humans to induce anxiety. By blocking the effects of these compounds with selective agents investigators have tried to unravel the neural pathways to anxiety and receptor subtypes involved in these pathways. A review of studies on experimentally induced panic is given by Ströhle ; cf. also Maron et al. . The review of the mCPP-literature by Gatch [18*] shows that behavioral data at present point to 5-Ht1B and 5-HT2C receptors as the main receptors which mediate the behavioral effects of mCPPP. A major limitation to the use of mCPP as a model of anxiety is the broad range of 5-HT receptor subtypes at which this compound is active. Because mCPP activity at one receptor subtype is sufficient to generate the full anxiety response, selective antagonists are not effective at blocking the effects of mCPP. This means that the classical approach (selective blocking of receptor subtypes after challenging with an anxiogenic agent) does not contribute to further understanding of the different pathways to anxiety. Substitution of mCPP by selective agonist is described as an alternative to identify receptor subtypes contributing to discriminative stimulus effects.
A very promising picture emerges from an important neuroimaging study by Mataix-Cols et al. in which symptoms of three dimensions of OCD (washing, checking, and hoarding) were provoked with stimuli (pictures) in both patients and controls [19**]. In the washing condition patients showed greater activations in the bilateral anterior cingulate and orbitofrontal regions. The whole picture of activation suggests that this type of behavior is mediated by regions that are involved in the processing of emotions, specifically disgust. Checking behavior showed heightened activity in bilateral subthalamic regions, including some brain stem nuclei, right putamen/globus pallidus, right thalamus and various dorsolateral cortical regions. Central to checking is a dysfunction in the system for attentional and motor behavior as well as the inhibition of unwanted impulses. Hoarding experiments reveal increased activation of the left precentral/superior frontal, fusiform and right orbitofrontal gyri. Activity in the right orbitofrontal region is negatively correlated with response to pharmacotherapy and positively with response to cognitive behavioral therapy. Increased activation of visual areas in controls compared to washers and checkers suggests that controls direct their attentional resources more to visual detail than to the emotional salience of the pictures that were presented. Controls also showed increased activation of the left inferior prefrontal regions in washing and checking, indicating that they are more successful in the suppression of negative emotions such as anxiety.
Rauch, Shin and Wright [20**] give an extensive overview of neuroimaging studies of amygdala function in anxiety disorder. This approach favors recognition of similarity and differences between neural pathways involved in the genesis of different anxiety disorders. The authors highlight the different cognitive activation probes involving paradigms of three types: face stimuli (masked and non-masked), emotionally valenced stimuli and classical fear conditioning. The masked face paradigm provides a potential tool for exploring amygdala functioning relatively dissociated from the top-down influences of the (pre)frontal cortex. It also enables the study of habituation profiles. These studies, together with the extant literature on neuroimaging have provided increasing evidence for differential responses within right and left amygdala. Right amygdala functioning appears to show a more temporally dynamic pattern of response (probably associated with some of the anxiety disorders) whereas the left amygdala shows a more stable response based on emotional valence of stimuli (with a possible association with depression). Very generally, data thus far provide considerable evidence for the amygdala to play a dominant role in posttraumatic stress disorder and social anxiety disorder. PTSD, the authors conclude, can be conceptualized as a fear-conditioning process superimposed over some diathesis such as amygdala hyper-responsivity, anterior cingulate deficiency, hippocampal deficiency, or exaggerated susceptibility to stress. In social anxiety disorder neuroimaging data show hyperresponsivity of medial temporal lobe structures during symptom provocation and in response to human face stimuli, suggesting a hypersensitivity for the assessment of threat to human faces. Data in panic disorder point to activation of insular and motor striatal regions during symptomatic states and reduced activity in widespread cortical territories. This pattern of activation may reflect a fundamental abnormality in monoaminergic neurotransmitter system functioning at the level of the brainstem, giving rise to abnormalities of metabolism, hemodynamics, and chemistry in widespread cortical areas. There is some support for involvement of hippocampal and amygdala dysfunction. Neuroimaging studies in OCD point at a specific role for the orbitofrontal-caudate circuit (intrusive thoughts and ensuing compulsive behaviors) and a non-specific role for the anterior cingulate cortex. The authors accord to the prevailing view which hypothesizes that pathology at the level of the striatrum (specifically, the caudate nucleus) leads to insufficient thalamic gating, resulting in orbitofrontal and anterior cingulate hyperactivity (compulsive behavior and anxiety, respectively). The authors conclude that neurobiological research should evolve, and in fact is evolving, beyond a focus on individual loci in the brain.
Research in children and in the developing animal brain suggests that environmental factors may exert a crucial influence on the development of neural circuitry underlying anxiety and other stress responses [21*] . From a conceptual point of view the maturing brain is a topic of utmost interest because it allows the longitudinal study of brain-environment interactions in both humans and animals. In his introduction to this area of research in a special issue of the Annals of the New York Academy of Sciences Lederhendler [22*] mentions a number of areas of topical interest: attachment and the effects of separation; risk assessment and sense of safety; activation, inhibition and emotion regulation. The author argues for dimensional above categorical distinctions and for contextually sensitive and interdisciplinary research. Pine [23**] reviews the interface between basic neuroscience and childhood pathology in the area of trauma and trauma-related disturbances. Manipulations in rat pups during specific developmental periods may produce long-term changes in the stress response system, including the HPA axis. These changes, interestingly, mimic alterations after direct administration of corticotrophin releasing factor (CRF). Amygdala lesions have opposite effects depending on the developmental state of the organism. Refinement in pathways connecting the amygdala with different parts of the temporal association cortex may offer a partial explanation for the reversal of the lesion effects. The inferior temporal cortex seems implied in the categorization of stimuli and is fundamental for the acquisition of language. Children suffering from abuse and/or neglect not only show emotional abnormalities but also underdeveloped language skills. The foundation for this underdevelopment may have been laid very early in life.
Developmental plasticity is an immensely rich and fruitful area of research. Clinical research on the effect of attachment patterns on the development of anxiety disorders is just beginning to emerge [24*].
discussion and conclusion
The neuroscientific study of anxiety has become an immense domain of interdisciplinary research. In this review we could only give a fragmentary glimpse of what is going on in the field. In spite of the existence of many excellent review articles, papers exclusively devoted to conceptual issues are rare, although the importance of these issues is often acknowledged. With respect to the conceptual questions which were raised earlier we can formulate the following conclusions.
Ad 1 (is pathological anxiety an animal, a human or a layered phenomenon with aspects of both?)
Curiously enough, this question is seldom raised explicitly; for exceptions see Nelson [25*] and Glas [26*]. Most authors seem inclined to the view in which both animal and human aspects are recognized and combined in some way. The overall impression is that research has simply not progressed far enough to draw conclusions. Neuroscience is just beginning to unravel some of the pathways underlying some aspects of the full phenomenon of human anxiety. It is for methodological reasons that most research is on relatively simple mechanisms that are presumed to be shared by animals and humans.
Underneath this learned ignorance, however, one sometimes senses a more reductionistic atmosphere. To mention only one example, what to think of this passage: “The brain states and bodily responses are the fundamental facts of an emotion, and the conscious feelings are the frills that have added icing to the emotional cake”? This is a quote from Joseph LeDoux’s admirable The emotional brain (p. 302) . I have chosen it because it is criticized by Bennett & Hacker [7**] for its identification of emotions with brain states (and bodily responses) (Bennett and Hacker 2003, 208-209). Emotions do not evolve as brain states, Bennett & Hacker say, rightly I think. It is the other way around: brains evolve in such a way as to make it possible for animals (or humans) to respond affectively to their environments. Emotion words apply to creatures not to brain and/or bodily states. Notify that this implies that animals can have emotions with an (intentional) object. However, humans also have emotions animals can not have (fear of bankruptcy, for instance).
Ad 2 (anxiety unitary or non-unitary?)
All authors are strongly inclined to a non-unitary view of anxiety and not only for methodological reasons. This was most clearly illustrated in the OCD-neuroimaging study by Mataix-Cols et al. [19**], in which the overlap between the different presumed brain circuits not only was reflected in behavior and attitudes, but also in the experiential realm, i.e. in the overlap between anxiety and disgust. What is it, given the enormous variation at both the phenomenal and the neural level that connects all these phenomena in such a way that it is still meaningful to use the word anxiety? Common to all anxiety is the reference of the term to danger, imagined or real, and even in the absence of awareness of danger. Language users like those general terms. Stress is another example (interestingly enough a metaphor derived from physics). Once the limitations of these general terms become apparent and more precise labels are available, it can be predicted that these more precise terms will be used when needed.
The boundaries between scientific, clinical and common language are continuously shifting. So, it seems wise to keep both options open: to build up a conceptual apparatus which is more precise and admits overlap between subtle differences of meanings, while at the same time allowing general concepts to persist.
Ad 3 (relationship between different ‘components’ of anxiety)
It is remarkable to notice that most researchers refrain from construing causal connections between reified components of anxiety. They seem inclined to a view which favors a conception in terms of overlapping circuits or networks. A conceptually different approach is exemplified by the phenomenon of context conditioning. Here the circuits do not overlap. There is, instead, one circuit with same mechanisms and, yet, different behavioral phenomenologies, depending on the context the organism inhabits.
Two large questions remain, however.
One question regards the issue of unity. What gives coherence and, in the end, unity to all these circuits? I can only mention this question and I do it because it is often evaded (not, however, by Damasio and Hurley) [see 8, 9; and 27 respectively].
Another important issue concerns the relationship between neural functioning and subjective experience. In the neurobiological literature accounts of this relationship are typically philosophically naïve. In so far as the question is addressed at all, researchers seem inclined to a form of non-eliminative physicalism. Since this position seems philosophically unstable , much more work has to be done. Most importantly the question has to be addressed how ‘higher’ order functions (conscious thoughts, volitions, imagination) exert their influence at ‘lower’ functional levels (tissues, cells, molecules). This age-old question can not be solved by neuroscience, however, neuroscience can add new and more precise input to the discussion and, by doing so, offer an indispensable contribution.
Ad 4 (scientific and clinical concepts)
Most researchers appeared to be aware of the large distance between animal data and pathological anxiety in humans. Suggestions to diminish this distance were similar to what can be heard elsewhere in the literature: more research on endophenotypes and on dimensional constructs, avoidance of grandiose extrapolations, and reminders of the complexity of the subject under study. Two important methodological recommendations emerged in our discussion: to take environmental factors more systematically into account in animal research; and to focus not on individual foci but on regional circuitry and its embeddedness in brain functioning as such.
To conclude, there is a richness in the phenomenology of anxiety in both humans and animals which goes beyond all theoretical reconstruction. This richness will always challenge us to look for new explanations and to not prematurely identify our concepts with reality itself. We have to listen to our patients and be sensitive to the depth of what they go through .
*1 Lambert KG. The life and career of Paul MacLean: A journey toward neurobiological and social harmony. Physiology and Behaviour 2003; 79:343-349.
Very readable overview of life and work of this eminent physician and researcher. Highlights MacLean’s emphasis on the social importance of the brain.
*2 Ploog DW. The place of the Triune Brain in psychiatry. Physiology and Behaviour 2003; 79:487-493.
Valuable discussion of the concept of the triune brain as early conceptualization of a form of evolutionary psychology and/or psychiatry.
3 Gorman J, Liebowitz MR, Fyer AJ, Stein J. A neuroanatomical hypothesis for panic disorder. American Journal of Psychiatry 1989; 146: 148-161.
4 Lang PJ. A bio-informational theory of emotional imagery. Psychophysiology 1979; 16:148-173.
5 Lang PJ. The cognitive psychophysiology of emotion: fear and anxiety. In: Tuma AH, Maser J, editors. Anxiety and the anxiety disorders. Hillsdale/London: Lawrence Erlbaum Ass, 1985, 131-170.
6 LeDoux J. The emotional brain. The mysterious underpinnings of emotional life. New York: Touchstone, Simon & Schuster, 1996.
**7 Bennett MR, Hacker PMS. Philosophical foundations of neuroscience. Malden, Oxford, Melbourne, Berlin: Blackwell, 2003.
Fascinating landmark study on the conceptual basis of the neurosciences. Four parts: historical and conceptual roots of philosophical problems in neuroscience; analysis of human faculties (perception, cognition, will, belief, emotion et cetera); consciousness; and methods.
8 Damasio AR. Descartes’ Error. Emotion, Reason, and the Human Brain. New York: AVON Books, 1994.
9 Damasio AR. The feeling of what happens. Body, emotion and the making of consciousness. London: Vintage, Random House, 1999.
10 Kenny A. Action, emotion, and will. London: Routledge, Kegan and Paul, 1963.
**11 Grillon C, Baas J. A review of the modulation of the startle reflex by affective states and its application in psychiatry. Clinical Neurophysiology 2003; 114:1557-1579.
Excellent, well written and thorough review article, investigating the possible meaning of the startle reflex methodology for almost all psychiatric disorders.
*12 Haller J, Leveleki Cs, Baranyi J, Mikics É, Bakos N. Stress, social avoidance and anxiolytics: a potential model of stress-induced anxiety. Behavioural Pharmacology 2003; 14: 439-446.
Gives an idea of possible variations and extensions of startle reflex methodology.
*13 De Jongh R, Groenink L, van der Gugten J, Olivier B. Light-enhanced and fear-potentiated startle: temporal characteristics and effects of α-helical corticotropin-releasing hormone.
Biological Psychiatry 2003; 54:1041-1048.
Attempt to differentiate fear and anxiety by investigation of the time-scale and amenability for CRH influences.
*14 Hettema JM, Annas P, Neale MC, Kendler KS, Fredrikson M. A twin study of the genetics of fear conditioning. Archives of General Psychiatry 2003; 60:702-708.
Important study with refined statistical methodology addressing an issue that until now almost exclusively has been investigated in animals.
*15 Melke J, Wesberg L, Nilsson S, ea. A polymorphism in the serotonin receptor 3A (5HT3A) gene and its association with harm avoidance in women. Archives of General Psychiatry 2003; 60:1017-1023.
Classical example of a study of a possible connection between gene polymorphism and a character trait. One of the first studies implying a possible role for the 5-HT3 receptor.
16 Ströhle A. Die experimentelle Provokation von Panikattacken als humanexperimentelles Angstmodell [The experimental induction of panic as a human experimental anxiety model]. Nervenartzt 2003; 74:733-739.
Incomplete overview of studies in the experimental induction of panic. Argues for a role for atrial natriuretic peptide as a possible inhibitor of CRH.
17 Maron E, Töru I, Vasar V, Shilk. The Effect of 5-Hydroxytryptophan on cholecystokinin-4-Induced panic attacks in healthy volunteers. Journal of Psychopharmacology 2004; 18: 194-199.
Administration of the 5-HT precursor 5-hydroxytryptophan protects women more than men against the panicogenic effects of CCK-4.
*18 Gatch M. Discriminative stimulus effects of m-chlorophenylpiperazine as a model of the role of serotonin receptors in anxiety. Life Sciences 2003; 73:1347-1367.
Valuable overview of virtually all literature on behavioral effects of mCPP. Repeats itself now and then.
**19 Mataix-Cols D, Wooderson S, Lawrence N, Brammer MJ, Speckens A, Phillips ML. Distinct neural correlates of washing, checking, and hoarding symspom dimensions in obsessive-compulsive disorder. Archives of General Psychiatry 2004; 61:564-576.
Intruiging study in which distinct neural correlates of washing, checking, and hoarding are found in a neuroimaging design using cognitive probes.
**20 Rauch, SL Shin LM, Wright CI. Neuroimaging studies of amygdala function in anxiety disorders. Annals of the New York Academy of Sciences 2003; 985: 389-410.
Thorough overview, which does not limit itself to amygdala function.
*21 Gross C, Hen R. The developmental origins of anxiety. Nature Reviews Neuroscience 2004; 5:545-552.
Reviews the evidence for a genetic basis for anxiety disorder (serotonin transporter polymorphisms, MAO-A gene abnormalities, twin studies) and emphasizes the combined influence of genetic and environmental factors.
*22 Lederhendler II. Behavioral neuroscience and childhood mental illness. Annals of the New York Academy of Sciences 2003; 1008:1-10.
Describes global developments and gives suggestions with respect to possible future directions of research in infant psychiatry.
**23 Pine D. Developmental psychobiology and response to threats: relevance to trauma in children and adolescents. Biological Psychiatry 2003; 53:796-808.
Excellent review of the neurobiology underlying the processing of threat and the role of amygdala, the medial temporal lobe, and the prefrontal cortex
*24 Myhr G, Sookman D, Pinard G. Attachment security and parental bonding in adults with obsessive-compulsive disorder: a conparison with depressed out-patients and healthy controls. Acta Psychiatrica Scandinavica 2004; 447-456..
*25Nelson, CA. Can we develop a neurobiological model of human social-emotional development. Integrative thoughts on the effects of separation on parent-child interactions. Annals of the New York Academy of Sciences 2003; 1008:48-54.
One of the very few articles which explicitly addresses the issue of the usefulness of animal data for clinical practice. Gives a model of overlapping methodologies in different species (mice, apes) and humans.
*26 Glas, G. Anxiety – animal reactions and the embodiment of meaning. In: Fulford KWM, Morris K, Sadler J, Stanghellini G, editors. Nature and Narrative. An Introduction to the New Philosophy of Psychiatry (International Perspectives in Philosophy and Psychiatry) Oxford/New York: Oxford University Press, 2003, 231-249.
I apologize for using my role as a reviewer for bringing my own work to the attention of the reader. The reason for doing so is that this is one of the very texts which try to develop a philosophically informed framework enabling the connection and interaction between animal and human research.
27 Hurley S. Consciousness in action. Cambridge (Mass): Harvard University Press, 1998.
28 Kim, J. Mind in a physical world. An essay on the mind body relationship and mental causation. Cambridge (Mass): MIT Press, 1998.
*29 Harris JC. Anxiety (Angst). Archives of General Psychiatry 2004; 61:15-16.
Describes the psychological and existential aspects of anxiety in the life and some of the works of Norwegian painter Edvard Munch.