Canine atopic dermatitis (cAD) is a chronic allergic skin condition that causes sustained periods of pruritus (itching), inflammation and secondary infections [1
]. Pruritus is an aversive subjective sensation which provokes a response to scratch [2
]. Pruritic conditions in humans, including atopic dermatitis (hAD), have been repeatedly shown to negatively impact upon the quality of life of those affected [2
]; and cAD has been documented to negatively impact the quality of life of both dogs and their owners [5
]. Human patients with pruritic conditions report significantly higher psychological stress compared to non-pruritic controls, which is linearly associated with pruritus severity [7
]. Furthermore, atopic dermatitis in humans is associated with significant psychopathological morbidities [8
]. As such, it is likely that cAD may also have psychological consequences for dog behaviour in terms of stress or anxiety.
There is growing evidence that chronic stress interacts variously with personality, behaviour, health and the immune system in both humans and non-human primates [12
]. Deciding which is the causative agent, and which is the response variable remains debatable, however, and the answer is not likely to be a simple one. Behaviour itself can affect physiological factors, such as immunity, whilst at the same time immunity can affect behaviour [13
]. Lopes [13
] characterises the potential links between immune traits and behaviour as being of three types: (1) Behavioural traits affecting immune traits, for example greater exploratory behaviour leading to a greater exposure to antigens or greater access to food and consequent different gut microbiota; (2) immune traits affecting behavioural traits resulting in behaviour changes indicative of illness; (3) an external or internal mechanism that affects one can simultaneously affect the other, causing changes in both immune response and behaviour. Stress (environmental and psychological) is a factor that disrupts homeostasis, affecting both immune system changes and behavioural changes, with chronic stress being implicated as a risk factor for numerous conditions from cardiovascular disease to cancer [14
]. In humans, chronic stress has also been linked to the development of different psychopathologies, including Alzheimer’s disease, Post-Traumatic Stress Disorder [15
], and depression [16
In addition to impacting upon immune regulation, stress disrupts homeostasis, and whilst acute homeostatic changes as a result of a stressor can be adapted to, chronic disruption leads to distress and disease [17
]. As Hendrix [19
] indicates, since the earliest day of psychoanalysis, skin conditions have been regarded as a sign of psychological ‘distress’. Evidence in support of this supposition now exists, with a growing number of studies showing that the “mind does matter” when it comes to skin conditions [20
], including, but not limited to, atopic dermatitis [21
]. Studies with murine models have shown abnormal skin barrier function can be induced by emotional stress, and can be reversed by administration of anxiolytics [23
]. The link between psychological distress and skin barrier function has also been evidenced in humans, supporting the pathophysiological paradigm of stress as an inducer of impaired epidermal function, which could precipitate or worsen inflammatory dermatoses [20
Potentially complicating matters are individual differences in stress response and ability to cope with stress. Factors such as an individual’s personality type (where personality is defined as consistent inter-individual differences in behaviour) may impact their vulnerability to stress-induced pathologies. Atopic allergies, such as asthma and allergic dermatitis in humans, have been associated with increased scores for emotional sensitivity [25
], lower impulsiveness and greater aggression in relation to irritation [26
]. In dogs specifically, it has been found that dogs that were considered to be “well behaved” by their owners lived longer but no behavioural traits were linked to specific immune conditions [27
]. However, Non-social fear and Separation anxiety as scored by owners on the canine behavior and research questionnaire (C-BARQ) predicted both the incidence and severity of skin problems in adult dogs [27
], although the exact nature of these ‘skin problems’ was not specified.
In contrast to the studies showing causative links between stress and disease status, it has also been suggested that individuals experiencing physical vulnerability, such as injury or disease, may be more likely to pessimistically interpret their environment, which with chronic vulnerability could lead to the development of anxiety disorders [28
]. When considering the link between stress and disease status, it is highly plausible to conclude that the stress experienced from one disease or injury state could lead to secondary comorbidities. Indeed, chronic systemic health conditions, such as chronic pain, cardiovascular disease, and cancer, have been associated with the secondary occurrence of mood disorders, such as depression (for a comprehensive review see
The aim of this study was to ascertain whether, and how, dogs with a diagnosis of cAD (cases) would differ behaviourally from dogs of the same breeds with no skin health problems (controls). Whilst causation cannot be established retrospectively, we tested two competing hypotheses based on what we would expect to see if psychological stress was primarily or secondarily associated with the dogs’ skin allergies: (1) If psychological stress was primarily associated (causative) with skin allergies we hypothesised that dogs with skin allergies would have higher trait-level scores for neurotic/fearful scales, characterised by higher scores on fear-based traits potentially reducing their ability to cope with stressors; and (2) if psychological stress develops secondarily as a result of pruritus, we hypothesised that dogs with more severe pruritus from skin allergies would exhibit more stress-related behavioural problems and lower scores for trainability (a score that requires a dog to attend to its owner and sustain its attention to training tasks) and would not differ on trait-level scores for neurotic/fearful scales.
The results of this study support acceptance of Hypothesis 2, that predicted dogs with skin allergies would exhibit more problem behaviours, lower scores for trainability and no differences for generic fearful or environmentally anxious behaviour. The lack of evidence for Hypothesis 1 supports previous findings from Klink and colleagues [36
] who found no association between pruritus and generally anxious/fearful behaviour or aggression in pruritic dogs. Here, atopic dogs scored higher than controls for a number of everyday problem behaviours (mounting, coprophagia, hyperactive/restlessness, pulling excessively on the lead), comfort-seeking behaviour (attachment/attention-seeking, begging for food), behaviour likely to be directly related to pruritus (self-grooming, allo-grooming, touch sensitivity), and other repetitive behaviour. Atopic dogs also scored lower for trainability, potentially due to reduced focus from pruritus interrupting/distracting them. Rather than generally having anxious/neurotic personality types as per Hypothesis 1, our results could indicate that the dogs diagnosed with cAD may be experiencing low-level chronic stress as a result of pruritus. In support of this supposition, scores on the Edinburgh Pruritus Scale, a measure of itch severity, were associated with the C-BARQ scores in the same manner as case/control status, implying that pruritus could be the underlying cause for the behavioural differences seen between cases and controls. Pruritus severity was associated with lower scores for trainability and was linearly associated with higher scores for chewing unsuitable objects, mounting, coprophagia, hyperactivity/restlessness, pulling excessively on the lead, attachment/attention-seeking, begging for food, self-grooming, allo-grooming, other repetitive behaviour and touch sensitivity.
Self- and environment-directed behaviours can be a consequence of chronic exposure to aversive conditions in animals, and can lead to behaviours, such as excessive grooming (of themselves and others), coprophagia, hyperactivity, overeating and in extreme situations, stereotypic repetitive functionless behaviour [37
]. If the stressor cannot be dealt with, or the motivation cannot be met, animals can express redirected behaviours, such as licking others or inanimate objects or destructive behaviour, where actions are directed towards other stimuli instead [37
]. The results of this study would suggest that excessive licking of people, other animals or inanimate objects as scored on the C-BARQ may be a redirected behaviour, that in this case is directly associated with pruritus (Allo-grooming, Table 4
). Here, pruritus severity was linearly associated with increased scores for multiple redirected, self- and environment- directed displacement behaviours that are likely to be indicative of pruritus-induced psychological stress as has been reported in studies of people with atopic dermatitis [7
]. Some of the starkest differences found in the current study included that cases were 2.73 times more likely to exhibit ‘other repetitive behaviour’ than controls, 2.14 times more likely to display mounting behaviour, and were significantly more likely to excessively self-groom. Similar to these findings, McGreevy and colleagues [33
] found that small dog breeds exhibit an array of stress/anxiety-related behaviour, including higher scores for mounting, touch sensitivity, begging and attachment/attention-seeking, potentially because their short stature exposes them to a greater number of stressors. Importantly, dogs with cAD did not differ from controls for general trait level scores, such as stranger-directed fear, dog-directed fear, separation-related problems or non-social fear, so do not appear to have neurotic/anxious personality types but do exhibit behavioural signs of increased stress.
The results of this study provide evidence in support of the theory that pruritic dermatoses could lead to the development of secondary compulsive disorders (otherwise known as obsessive-compulsive-disorder, OCD) in dogs [39
] (Figure 1
). Compulsive disorder in dogs is characterised by a high incidence of over-grooming/self-mutilation, locomotor behaviours, such as spinning/chasing, hallucinatory problems, such as fly catching and shadow-chasing, pica, and to a lesser extent, coprophagia [40
]. Whilst the dogs in this study cannot be classified as exhibiting OCD at this stage, those suffering from pruritus did exhibit increased redirected and displacement behaviours, reduced trainability (which may be considered a sign of cognitive impairment, e.g., Reference [41
]), and increased comfort-seeking (begging and attachment/attention-seeking). Given these results and what we know about the stress-inducing component of pruritus from human studies [2
], it is likely that psychological stress caused by pruritus could have a pathogenic impact on cAD, however, well-controlled studies would be necessary to confirm a causal relationship.
Due to the fact that chronic stress interacts with health in a bi-directional manner [29
], and the nature of stress as an inducer of impaired epidermal function [20
], it is plausible to conclude that psychological stress-induced by allergic pruritus would further disrupt the skin barrier function, thus, worsening the inflammatory dermatoses. Such a process could mean that the dermatoses becomes self-perpetuating, lengthening the period of a flare even once the initial allergen has been removed. Emotional stress has been implicated in the development and exacerbation of psoriasis in humans, potentially worsening its severity and increasing the time to clearance of a flare [42
]. Evidence from murine studies where dermatitis was induced, demonstrate that psychological stress (in this case social isolation) can prolong pruritus in the form of idiopathic dermatitis from self-scratching, long after the inducement has been ceased, whilst the application of topical corticosteroid to the challenged skin prevented the later onset of idiopathic dermatitis [43
]. Such findings lend support to the therapeutic use of medications that treat the itch and support the skin barrier as a regular practice even between allergen encounters. Additionally, as suggested in Reference [44
], treatment of dermatoses in dogs and other animals should include management of environmental stressors; however, the extent to which this is practiced in veterinary settings is currently unknown. It is also possible that for dogs where environmental stressors and pruritus cannot be controlled, that the use of anxiolytics may result in improved skin barrier function.
Negative affective states are counter-productive for training and learning, with animals that are in negative affective states more likely to judge ambiguity (such as human hand signals or cues) negatively [45
]. Further, Eysenck proposed that individuals experiencing anxiety or stress are in a state of divided-attention, with information irrelevant to the task, competing with task-relevant information for processing space [46
]. It is, therefore, plausible that the persistent sensation of pruritus in dogs affected with cAD divides the dogs cognitive processing abilities, splitting their attention, in addition perhaps to inducing a negative affective state, which could explain the reduced scores on trainability and increased scores for attachment/attention-seeking and food-seeking as forms of comfort-seeking. However, atopic dermatitis is known to disrupt sleep with increasing pruritic severity [10
], meaning that the reduction in trainability seen here could be secondary to sleep disturbances inducing cognitive impairment [48
]. Alternatively, it is possible that some of the associations seen in this study are the result of side-effects from medication prescribed to treat the dermatoses, as dogs with pruritus are much more likely to be treated with antihistamines and glucocorticoids than dogs without pruritus, which both have potential to affect behaviour [36
]. Antihistamines, for example, have a sedative effect and have been shown to induce cognitive impairments in learning and decision-making in humans [49
]. Klinck and colleagues found that dogs treated with glucocorticoids were more likely to be reactive to loud noises and thunderstorms [36
]; however, we found no difference here between cases and controls or pruritus severity for the C-BARQ scale ‘non-social fear’, which includes noise sensitivity. Dogs with epilepsy have also been shown to score lower than healthy controls for trainability and higher for excitability (also measured by the C-BARQ) [41
], as was found here with the dogs with cAD, although no differences in the other behavioural factors found here were reported for epileptic dogs. The current study was limited by a lack of information on any medications the dogs were currently being treated with, as no data was collected for current medication use, only medication efficacy for cases. However, due to the confounding association of use of glucocorticoids and antihistamines with pruritus [36
], only longitudinal studies that evaluate behaviour and pruritus before and after the medication has been used that could potentially impact behaviour, would be able to truly untangle this potential interaction. Given the results presented here, the impact of pruritus and the medications used for treating pruritic conditions on dog behaviour warrants further investigation. Another factor that may have contributed to the reduction in trainability and increase in food-seeking behaviour could be that cases were more likely to be on an ingredient restricted diet. However, we do not know to what extent each dog found their food rewarding, and it is unlikely to explain the full extent of the results here as they were also associated with increasing severity of pruritus, not just case/control status.
In the Labradors, at least part of the behavioural differences seen in cAD cases may be confounded by coat colour. Chocolate Labradors are at greater risk of having cAD [51
], and other pruritic skin problems [52
], and have also been shown to score higher than black or yellow Labradors on traits, such as excitability and lower for trainability [53
]. Any differences in Labradors, due to coat colour should, at least in part, have been controlled for by including the variable ‘breed’ in our models here. This would, however, be an interesting area for future study; at least some of the behavioural differences seen in chocolate Labradors may be associated with their greater incidence of pruritic skin conditions.
Not only are the results of this study important for the welfare and treatment of dogs affected by cAD, but the increased frequency of ‘problem’ or ‘nuisance’ behaviours reported by owners of pruritic dogs could contribute to a break-down in the dog-owner relationship. Problem behaviour and behavioural characteristics of dogs have been shown to be associated with the quality of the dog-owner relationship [54
], and furthermore, behavioural problems are the main reason given for relinquishing animals to shelters [56
]. Combined with the increased financial burden of treating and managing this chronic condition, and reduced quality of life reported by owners of dogs with cAD [6
], the problem behaviours are shown by pruritic dogs in this study could disproportionately increase their risk of being relinquished by their owners.