اضطراب خصلتی، ترس صفت و عاطفی: چشم انداز مطالعات غیر انسانی

Current clinical classifications, and many psychologists, do not distinguish categorically between anxiety and fear. Likewise, the normal language use of these and related terms suggests considerable overlap. However, ethology, behavioural pharmacology and neural analysis all suggest that anxiety and fear are fundamentally distinct at the functional and neural levels – while co-occurring and interacting with each other in many situations. As states with distinct and coherent neural control, anxiety and fear should give rise to distinct personality factors of fear-proneness and anxiety-proneness. However, additional factors are suggested by neural, epidemiological, pharmacological and genetic data, which suggest that a higher order factor of emotionality (potentially equivalent to neuroticism) will contribute to trait variance in both fear and anxiety, and a lower order factor of panic-proneness will contribute to fear, in addition to factors specific to fear or anxiety, per se.

The editors of this special issue on anxiety asked the authors to answer, each from their own perspective, a series of questions. I was asked to apply the perspective of animal studies. These questions included the following, on which I will focus most below: 1. What are the major differences between trait and state anxiety? 2. To what extent does anxiety (and its disorders; e.g., GAD) differ from fear (and its disorders; e.g., phobia): are fear and anxiety qualitatively or quantitatively different? 3. What are the functional similarities and differences between anxiety and neuroticism? It turns out that the non-human literature can provide fairly clear answers to these questions, but only on the condition that the questions are first rendered answerable by experimentation. The first task of this paper, therefore, must be to make clear the confusion of meanings in the normal uses of the words “fear” and “anxiety” and also to point out a lack of clarity in this area as to the role of “traits”. These issues are brought into sharp focus by the third question, above, where one might wonder whether it is about state versus trait anxiety, or about anxiety versus fear, or about trait anxiety versus some trait that is a developmental risk factor for trait anxiety. The remainder of this paper will then look at what can be gleaned from the rat literature. This provides us with quite restricted meanings of “fear” and “anxiety”. It also points us to biological factors that could underlie “traits” and so provides insight into the types of entity that might need to be characterised in the human literature. The overall approach is conceptual and so points the way to animal literatures of potential interest to the human personality researcher rather than attempting proper reviews of those literatures. It should be noted that the questions asked do not directly link to any particular current theory of personality. This paper should be read in the same way. It deals with elements that could form part of, or identify wholly with, presumed factors in a variety of theories that all link to defensiveness of one form or another. It could be argued that if all such theories ensured a tight link to the biological variables discussed here they would necessarily become integrated into a single theory.

Perhaps the most important message from the animal literature is that fear and anxiety are fundamentally distinct and can be defined functionally. That is, each is a cluster of reactions that has evolved piecemeal with the retrospective “function”, strictly teleonomy (McNaughton, 1989 and Pittendrigh, 1958), of either escape from (fear) or approach to (anxiety) danger. Provided the words “fear” and “anxiety” are used in conformity with this definition, our original questions can be answered in a straightforward manner. To what extent does anxiety (and its disorders; e.g., GAD) differ from fear (and its disorders; e.g., phobia): are fear and anxiety qualitatively or quantitatively different? We have just answered this straightforwardly for state anxiety and state fear. They are qualitatively different and, in a sense, functional opposites. That said, they share a number of autonomic reactions – but quite a few of these are also shared to some extent with many other emotions. The distinction between fear and anxiety also extends to the clinic – where the original distinction between panicolytic and anxiolytic drugs was derived. What are the major differences between trait and state anxiety? Here we must first be clear as to what is meant by trait anxiety, and this issue is addressed by the next question. But it is clear that the GABA-A receptor mediates state changes that, at the least, impact on the level of anxiety experienced. Taking what is controlled here as potentially a subtype of anxiety process, if not anxiety itself, the distinction between state and trait is clear. The state of the chloride channel (the extent to which it is open) depends primarily on the presynaptic release of GABA ( Fig. 3C, D). If there is no release there will be no response ( Fig. 3A and B). The trait aspect, that is the sensitivity of the GABA receptor, depends on the levels of endogenous benzodiazepines and so the amount to which their receptor is bound. By itself this produces no response ( Fig. 3B), but in the presence of a fixed quantity of GABA it alters the size of the response (compare Fig. 3C with D). Here we have a clear distinction between trait and state aspects of what have to be the same factor. Whether the label “anxiety” should be applied to these state/trait entities is a matter of both future research and linguistic preference. What are the functional similarities and differences between anxiety and neuroticism? This question, I believe, must be taken as contrasting trait anxiety with neuroticism since (see previous paragraph) it does not really make sense to compare a trait variable with a state variable. Both trait anxiety and neuroticism are potentialities that under appropriate circumstances are reflected in particular levels of state anxiety and, in the case of neuroticism, fear. High trait anxiety, functionally, will lead to high levels of passive avoidance and related anxiety-specific output. It involves an increase in gain, by definition. Neuroticism is less clearly defined. However, if we see it as linked to emotionality, it would appear to involve an increase in the dynamic range of responses. So, our tentative answer, using definitions derived from the rat work discussed in this paper, is that, at times, high trait anxiety and high neuroticism will produce similar behaviour (say in the upper portion of a Yerkes–Dodson curve). However, in the same task with other parameter values, they are likely to produce opposite effects; and the range of tasks (anxiety) in which trait anxiety will have effects will be narrower than that (anxiety + fear) for neuroticism. What is the most appropriate level of explanation to understand human anxiety? Is it possible, or indeed desirable, to attempt to understand anxiety at all levels of explanation (e.g., evolution, DNA, brain, endophenotype, and behaviour)? In this paper I have used arguments at many levels of explanation. These are most trustworthy when several levels concur. The key point is that analysis of, for example, the pharmacology of rats does give us testable predictions (such as the prediction, now confirmed3 in humans, that clinical phobias are insensitive to anxiolytic drugs); and often, provides us with functional perspectives (for example, classifying fear as the set of adaptive reactions that promotes avoidance of a danger; and anxiety as the set of adaptive reactions that promotes approach to danger) that can be directly transferred – as with the ethological analysis (also now confirmed in humans). The benzodiazepine receptor (which is linked, inter alia, to anxiolytic action) is present, with similar functionality, in goldfish. I would argue that it is not only possible, and not only desirable, but also essential to understand anxiety at all the applicable levels of explanation (at the moment we can safely exclude quantum mechanics as being at too low a level to be useful). Moreover, it is essential to try to ensure that the explanations at each level are consistent with those at the other levels. A failure of such consistency indicates a clear flaw in the analysis at one or both levels involved.