An evolutionary approach to the analysis, assessment and treatment of behaviour problems in companion animals

*Dr. Daniel Mills dmills@lincoln.ac.uk, Gill Sheppard

* Author for all correspondence

The nature of behaviour and behaviour problems

Understanding the nature of behaviour problems is essential to developing a rational basis for their treatment. Behaviour problems arise as a result of an interaction between factors relating to the current environment and developmental factors within a patient of a given state. Not all behaviour problems represent dysfunctional, abnormal or maladaptive behaviour since "the problem is the not the animal’s behaviour per se but rather the problem that this behaviour poses for its owner" (Askew 1996). Broadly speaking behaviour problems may be divided into behaviours which are adaptive but inconvenient for the owner, those which are derived from attempts to behave in an adaptive way in a suboptimal environment and those which are truly maladaptive e.g. seizure activity.

The causes of behaviour may be investigated at a proximate or ultimate level (Mayr 1961). Proximate explanations relate to the environmental stimuli and mechanisms within the individual which bring about the physical expression of the behaviour; they describe the pathological and aetiological processes with which veterinary surgeons are familiar in normal clinical practice. The ultimate explanations describe why such proximate processes should come about; they explain the function or adaptiveness of the processes. Tinbergen (1963) suggested that a full explanation of behaviour was to be found at four distinguishable but complementary levels: phylogeny (evolutionary history), ontogeny (development within an individual), mechanism and adaptive value. This paradigm has now become a central tenet to ethology. Unfortunately, within the field of animal behaviour therapy, there has tended at times to be a fragmentation of this unifying model with emphasis being given to particular levels according to the philosophical standpoint of the therapist, which may be medical, psychological or more purely ethological. Clinical ethologists tend not to adhere rigidly to a single model as it is generally accepted that internal and external causes interact in the development of a disorder. However, one type of cause does tend to be emphasised at the expense of other possible explanations. (Sheppard & Mills 1998).

Ethological explanations which focus on the ultimate factors governing behaviour are widely used for the category of problems which consist of adaptive, species typical behaviours which are inconvenient for the owner, like urine spraying in cats and hierarchical aggression in dogs (Borchelt and Voith 1982). However the potential for an ultimate analysis of behaviour problems which have traditionally been thought of as psychopathologies, is a much more recent phenomenon. These problems tend to be analysed from a behavioural or medical perspective which concentrates on proximate explanations of behaviour (Mills 1997).

Criticisms of existing models

Many of the difficulties that are faced in investigations of psychological problems in animals have been apparent for much longer in the field of human psychiatry.

Kraepelin proposed the first comprehensive system for the classification of mental disorders in 1896 and his view still holds enormous influence in the field. His beliefs are the core of the biomedical model and its associated problems. He suggested that psychological disturbances represented a pathology in much the same way as physical problems, and that diagnosis should be based on the description of symptoms. This view rapidly gained popularity as his work was soon supported by the experiments of Krafft-Ebing (1897) who demonstrated that mental illness could have an infectious cause, in the case of syphilis (Rosenhan and Seligman 1995).

Emphasis on the description of symptoms is the main feature of the Diagnostic and Statistical Manual of Mental Disorders (4^th edition). This is atheoretical with regard to the aetiology and development of disorders and was developed to facilitate clinical practice and communication, with clinical utility its highest priority (American Psychiatric Association, 1994). Individuals are categorised within the same diagnostic group on the basis of similarity in signs and symptoms. Descriptive criteria determine diagnoses, resulting in a classification system that is mostly based on superficial similarities in signs and symptoms. The system may provide precise criteria for diagnoses but not precise boundaries. Diagnoses can be based on meeting a certain number (but not necessarily all) of a list of criteria. For example, dysthymic disorder requires the presence of two symptoms from a list of six. The diagnostic categories may thus ignore and conceal important individual biological differences between patients with the same diagnosis (Dubovsky and Butler, 1995).

DSM-IV is based on the classical nosological approach and leads to explanations of disorders being sought in terms of:

• Clusters of co-occurring signs and symptoms
• The definition of disorder
• Biological markers and causes

Some authors (e.g Maas and Katz, 1992) have suggested that the classical nosological approach has been a hindrance to investigations to identify specific causes of psychiatric disorders; a field which has produced a very low return for the amount of research invested into it. The problems that have emerged may be due to an incompatibility between the nature of psychiatric disorders and the form of nosological categories (Clark et al, 1995).

The nosological approach aims to develop a series of categorical diagnoses that are mutually exclusive and jointly exhaustive (Goldberg and Huxley, 1992). Since disorders are viewed as discrete entities with discrete bases that are usually assumed to be biological (Kendell, 1982), most research aims to identify specific brain dysfunctions as causal factors. However, psychiatric disorders are not discrete in nature; symptoms are not all-or-nothing phenomena. They occur with varying degrees of severity and appear to exist on a continuum, blending with normal behaviour. Categories impose arbitrary and imprecise boundaries on such a continuum and may overlap or leave gaps that are not covered by the diagnoses. This creates difficulties in fitting clinical presentations to diagnostic criteria. It may also result in diagnostic categories that reflect quantitative differences rather than qualitative ones. Additional categories may increase the coverage of the continuum and decrease the likelihood of gaps alternatively they may be used to create smaller categories and more specific diagnostic criteria. However, additional categories increase the likelihood that diagnoses overlap, resulting in comorbidity when only one disorder is present (Frances et al, 1991). They also increase the difficulty in categorising borderline cases that do not fit easily into a diagnostic group. Despite many years of refinement, DSM-IV still requires "dustbin categories" to ensure that all clinical presentations can be accommodated within the classification system. For example, "depressive disorder not otherwise specified" is a category that can include all of the possible clinical presentations of depression that do not fit elsewhere. Thus the categories of DSM-IV do not reflect discrete biological processes.

Investigations of the aetiology of psychiatric disorders tend to assume that diagnostic categories are valid. It is more likely that such categories have only face-validity and investigations of these diagnostic groups are unlikely to identify causes of disorders.

It is likely that disorders exist as dimensions that reflect the interaction of multiple factors rather than the action of a single discrete cause. The growing acceptance that disorders occur on dimensions is indicated by the increased use of terms such as "spectrum" and "continuum" in psychiatric literature.

The rise of behavioural psychopharmacology has offered the possibility of using pharmacological probes to help elucidate the biological bases of various disorders. The inferences drawn from such experiments must be made with caution and cannot easily be used to form a nosological framework. Response to treatment does not necessarily correlate with a change in the causal mechanism of the problem (Lader 1991). Treatment may be only symptomatic and thus indirectly supportive of the pathological organ. Lader (1991) illustrates this point well by reference to the use of diuretics to treat congestive heart failure. Understanding the action of diuretics on the kidney, does not help us to understand the processes behind the failing heart. Similarly, heavy sedation may prevent a problem response from occurring without addressing the mechanism behind the behaviour (Cooper and Mills 1997). By way of a more subtle example of this logical fallacy, benzodiazepines were the standard anxiolytic against which newer preparations are assessed, but there is no evidence that there is a specific benzodiazepine receptor with a specifically anxiolytic function (Petersen et al 1986), rather it has a more general effect on the chloride ionophores of inhibitory GABA receptors. Many other compounds, like tricyclic antidepressants and monoamine oxidase inhibitors (Shader and Greenblatt 1983), which do not act through benzodiazepine receptors also have anxiolytic properties. Psychological phenomena, including disturbances, tend to be the product of a dynamic interaction between several central systems which result in a co-ordinated response. Until the neurophysiological mechanisms behind a condition are fully elucidated we must consider the possibility that response to a particular type of antidepressant, antipsychotic or anxiolytic may be operating at only a phenomenological and not necessarily a specific mechanistic level. The doses of drugs prescribed and the systemic levels which result, are also many times in excess of any natural biological level. It is then to be expected that their effects will extend beyond the specificity of any known drug-receptor interaction. In this context, the action of drugs may be far from correcting any supposed biochemical imbalance underlying the problem, but rather chemically flooding the system to oppose a behavioural tendency. In which case drug therapy is again not treating the cause but rather stimulating a system which will mask the symptoms. It is not surprising therefore that treatments are not specific to one disorder but are partially effective across a range of disorders. None are consistently effective in the treatment of a diagnostic group, no matter how narrowly it is defined (Kendell, 1989). Other concerns relating to the use of the pharmacological approach to understand the biology of psychological problems relate to the lack of a predictable relationship between disorder and treatment response and prognosis. Thus the diagnosis is of limited value to the choice of treatment and predicted prognosis (Clark et al, 1995).

Since psychological disorders in animals appear to share a similar inherent structure and mechanism to human psychiatric disorders, it is likely that investigations of the former will be equally unsuccessful unless a more appropriate approach is rationalised.

A proximate factor is not necessarily a primary cause. A neurophysiological change that accompanies a disorder may be a response to another neurochemical change or a response to an environmental event. For example, Raleigh, McGuire, Brammer and Yuwiler (1984) investigated the association between social status and serotonin levels in vervet monkeys. High-ranking male vervet monkeys have serotonin levels that are, on average, almost twice as high as those of low-ranking males. A high-ranking male that is displaced displays behaviours that appear similar to those associated with depression in humans, and serotonin levels decline to that of a low-ranking male. In this example the neurophysiological change follows the environmental event.

Behaviour problems consist of clusters of behaviours and emotional states, most of which are not specific to one disorder. If proximate causes are identified they could be associated with one or more signs and may also be associated with other disorders. The neurological processes underlying such general states as depression, anxiety or phobic responses could be normal adaptive mechanisms rather than a pathological process in the normal sense of the word. For example, fearful behaviour in a dog is regarded as a disorder if it is too prolonged, too frequent or occurs in the absence of an appropriate trigger. The difference between "normal" and "abnormal" is subjective with such an approach and unsatisfactory as a foundation for a clinical science (Mills 1997).

Mechanistic investigation is in danger of identifying the normal mechanism that instigates a fear response as opposed to the primary cause of what makes it inappropriate, which may relate to endogenous or exogenous individual factors.

The evolutionary approach

Biomedical and pharmacological models of psychological problems emphasise the internal causes of disorders; by contrast, the behavioural and sociocultural models emphasise external causes. In any case, each model implicates a specific type of proximate cause in the development of behaviour problems and provides explanations on that basis alone. Proximate explanations that give equal consideration to the influences of internal and external factors have greater value but cannot provide a full understanding of the causes of a disorder, as they do not evaluate the potential role of ultimate factors involved. The importance of ultimate factors such as function in the classification and treatment of ethological behaviour problems is well recognised, and the evolutionary approach to psychiatry extends this to psychological phenomena. The different levels of explanation proposed by Tinbergen (1963) are interdependent in the construction of a paradigm for assessing behaviour problems and so should have equal consideration. An evolutionary framework enables us to integrate these different factors in a coherent manner and on a sound basis, since the theory of evolution is the central tenet of biology.

The evolutionary approach to psychological change focuses attention on the interactions of an animal with its environment and investigates their potential functional value as well as the nature, source and degree of any perceived suboptimality. Two stages of psychological evaluation can be recognised:

1. evaluation of the functional value at an evolutionary level of the psychological process being evoked
2. evaluation of the functional capacity in situ of the process involved

The adaptive value of the processes of fear and anxiety in helping an animal avoid or prepare for a noxious event are obvious, but other psychological processes which may also feature as "psychological problems" may also have adaptive value. Price et al (1997) hypothesise that depressive states are adaptive mechanisms that enable individuals to cope with defeat in social competition and to adjust to a low social rank. In these circumstances, a depressive response assists an individual in deferring attacks from higher ranking individuals and helps in the recruitment of social support from other members of the group, thus minimising the impact of defeat and maximising the coping potential of the individual. However, depression when alone might suggest a different function, such as withdrawal from uncontrollably oppressive features of the environment (McGuire and Troisi 1998).

Behavioural and emotional responses are often considered to be disorders when responses are too intense, too prolonged, too frequent or when they appear to occur in the absence of an appropriate triggering stimulus. These responses could however be adaptive in other contexts such as alternative genetic combinations, different stages of the animal’s development, the opposite sex, or alternative environments (Nesse and Williams, 1997). In which case the behaviour and psychological processes behind it cannot be considered pathological. Normal population variation means that individuals in a given population differ in the degree of baseline optimality of specific traits. This is the raw material for natural selection and evolution proceeds as a result. The sensitivity of different systems may need to vary for optimal adaptation in different environments; consider the cosseted pet with its feral neighbour for an extreme contrast. If suboptimality is due to a mismatch between the animal’s adaptive range and the environment in which it is placed the prognosis for psychological recovery is better than when there is a genuine dysfunction of the trait. In the latter case, there is a real neurological disturbance and the prognosis is considered much poorer (McGuire and Troisi 1998)

When a large proportion of functional capacities are highly flexible an individual can adjust to and live successfully in a wider range of environments than when functional capacities are more limited. However, the latter does not necessarily exhibit any behavioural disturbance. For example, a dog with limited functional capacities may be able to live in a quiet, rural environment without displaying any signs of a behaviour problem, but in a noisy inner city area it may be unable to adjust so well. This does not inevitably lead to any form of pathology but it may lead to behaviour problems associated with a specific psychological state. We can identify the two categories of response described above. Firstly the dog may use unacceptable or concerning strategies to control its environment and help it to cope. This might include increased aggression to repel strangers or a depressive withdrawal. If this is identified, functional treatment should not seek to control the behaviour per se but rather address the problem of compromised adaptability which has led to it. Alternatively the mismatch between the animal and its environment may be addressed in order to help the dog to cope in a more acceptable way. Even if the pathways involved in the control of the behaviour become sensitised and the response generalised to a wider range of stimuli, it still maintains a functional form and so is not considered pathological.

If the situation is so prolonged or intense that it exhausts or defeats the coping mechanisms, then we may start to see a truly dysfunctional behaviour. In this case we have a genuine psychopathology, with the behaviour no longer structured in a functional way. Prognosis in these cases is poor.

Thus disorders may represent attempts to behave adaptively in the face of limitations and in other contexts be signs of an overtaxing of these and a true dysfunction.

The evolutionary approach is not a radical departure from that used currently by other clinicians but provides further information for the management of cases where psychological factors are significant. It is to be hoped that with this approach a functional ethogram can be constructed against which an objective assessment of behavioural pathology can be made. The approach has also helped in owner counselling and a more rational application of psychopharmacolgy. For example, Mrs W. contacted the Animal Behaviour Clinic for advice concerning her two neutered male Border Terriers who had started fighting each other since they were nine months old, some six months previous. The smaller dog (R) apparently initiated the attacks despite being consistently defeated by its larger sibling (H). These would tend to occur only when the dogs were on the lead. After a fight R would withdraw from H, but stay within sight of the owner and appear "depressed". The owner admitted to feeling sorry for R and was wondering whether or not he should be rehomed, as she felt he no longer enjoyed life. Traditional treatment strategies emphasise the need to reinforce the dominant and subordinate dogs’ positions but this is difficult for owners to implement when they already feel guilty about the subordinate’s quality of life. Alternatively, antidepressants may be considered to improve the well-being of the subordinate dog. However, in this case, when it was explained that the depressed behaviour was most probably functioning as a care-soliciting behaviour designed to recruit owner support which maintained instability in the unit, which was leading to these fights, the owner quickly complied with the treatment strategy proposed. R was ignored whenever he showed the depressed response and the owner made sure that H was never compromised in the presence of the other dog, for example by being on the lead when R was free. The owner was encouraged to allow the dogs to play competitive games like tug’o’war, since H reliably won these by virtue of his size. After two weeks, there had been only one fight, when H was startled whilst on his lead. The owner responded as directed after the dogs had been separated and 8 months later there have been no further fights. The owner also noted a marked improvement in the mood of the R within these first weeks. In this case an evolutionary approach also predicts that pharmacological intervention would be contraindicated, as serotonin re-uptake inhibitors suppress submissive behaviour (McGuire and Troisi 1998). Thus treatment of the depression would probably extend the range of situations when fights were instigated. Whilst the evolutionary approach may not alter the treatment offered in this particular type of case it provides a much more satisfactory explanation of the situation than any behavioural or medical model. It helped improve owner compliance and avoided the potential misapplication of pharmacotherapy. The pattern of response seen in R, has since been noted in several other cases of sibling competition seen by the author, and these cases have been managed equally effectively.

It is accepted that further research is necessary in order to investigate the hypotheses that such an approach generates about the nature of the psychological state of patients with supposed psychological disturbance. However, it is suggested that this approach has the potential to explain the success and limitations of the systems adopted and proposed by different experts within the field.

References

American Psychiatric Association (1994) The Diagnostic and Statistical Manual of Mental Disorders IV American Psychiatric Association. Washington, DC:

Borchelt P.L., Voith V.L., (1982) Classification of Animal Behavior Problems. In: Voith V.L., Borchelt P.L., (eds) Veterinary Clinics of North America ( Small Animal Practice) 12: 571-586

Clark, L.A., Watson, D. and Reynolds, S. (1995) Diagnosis and classification of psychopathology: challenges to the current system and future directions. Annual Review of Psychology 46: 121-153.

Cooper J.J., Mills D.S. Welfare considerations relevant to behaviour modifications in domestic animals In: Mills D.S., Heath S.E. Harrington L.J. (eds) Proceedings of the First International Meeting on Veterinary behavioural Medicine, UFAW, Potters Bar 164-173

Dubovsky, S.L., Butler, L.D. (1995) Casebook and Study Guide: Abnormal Psychology. W.W. Norton and Company, London.

Frances, A.J., First, M.B., Widiger, T.A., Miele, G.M., Tilly, S.M., Davis, W.W. and Pincus, H.A. (1991) An a-z guide to DSM-IV conundrums. Journal of Abnormal Psychology 100: 407-412.

Goldberg, D. and Huxley, P. (1992) Common Mental Disorders: A Bio-Social Model. Routledge, London.

Kendell, R.E. (1982) The choice of diagnostic criteria for biological research. Archives of General Psychiatry 39: 1334-1339.

Kendell, R.E. (1989) Clinical validity. Psychological Medicine 19: 45-55.

Lader M. (1991) Animal Models of anxiety: a clinical perspective In: Wilner P. (ed) Behavioural Models in Psychopharmacology: Theoretical, industrial and clinical perspectives. Cambridge University Press, Cambridge.

Maas, J.W. and Katz, M.M. (1992) Neurobiology and psychopathologic states: are we looking in the right place. Biological Psychiatry 31: 757-758.

Mayr E. (1961) Cause and effect in biology. Science 134: 1501-1506

McGuire M., Troisi A. (1998) Darwinian Psychiatry. Oxford University Press, Oxford

Mills D.S. (1997) Separating a dog’s bite from its owner’s problem: - conceptualising behaviour problems. In: Mills D.S., Heath S.E. Harrington L.J. (eds) Proceedings of the First International Meeting on Veterinary behavioural Medicine, UFAW, Potters Bar 7-9

Nesse, R.M. and Williams, G.C. (1997) Evolutionary biology in the medical curriculum - what every physician should know. Bioscience 47: 664-666.

Peteren E.N., Jensen L.H., Drejer J., Honore T., (1986) New perspectives in benzodiazepine receptor pharmacology. Pharmacopsychiatry 19: 4-6

Price, J., Sloman, L., Gardner, R., Gilbert, P., Rohde, P. (1997) The social competition hypothesis of depression. In: S. Baron-Cohen, (Ed.) The Maladapted Mind: Classic Readings in Evolutionary Psychopathology. Psychology Press, Hove, 241-253.

Raleigh, M.J., McGuire, M.T., Brammer, G.L. and Yuwiler, A. (1984) Social and environmental influences on blood serotonin concentrations in monkeys. Archives of General Psychiatry 41: 405 – 410.

Rosenhan D.L. Seligman M.E.P., (1995) Abnormal Psychology (3^rd edn) W.W. Norton and Company, New York.

Shader R.I., Greenblatt D.J., (1983) Some current treatment options for symptoms of anxiety. Journal of Clinical Psychiatry 44: 21-29

Sheppard G., Mills D.S. (1998) Veterinary clinical ethology: concepts, history, terminology and future. In: Ibanez, M., Dominguez C. (eds) Etologia Clinica Veterinaria. Proceces Print, Madrid

Tinbergen N. (1963) On aims and methods of ethology. Zeitschrift fur tierpsychologie 20: 410-433