Also available as PDF (figures and tables appear in the PDF).
Introduction: The behavioral disorder, attention deficit hyperactivity disorder (ADHD) is typically associated with human children, but its clinical manifestation in dogs has recently been investigated indicating that certain gene variants (e.g., the DRD4 gene) contribute to its expression. Objectives: In this study, the aim is to strengthen the understanding of the causes of ADHDlike behavior in dogs by investigating the potential non-genetic correlates of the condition. Methods: In particular, the presence of ADHD-like behavioral symptoms across different dog breeds and groups were examined with respect to various social and physical factors using a diverse set of questionnaires (ADHD, anamnesis, and personality and interview) given to the dog owners of two dog schools and one dog club (n=61) as well as an existing, unpublished database that was established previously (n=60). Results: Our results showed that various social factors (e.g., degree of affectionate behavior, number of social contacts, and the duration of period of separation) and physical parameters (amount of play time as puppies as well as frequency and duration of walking, possibly castration) were associated with ADHD-like behavior across all the dog breeds, with the Akita breed being noticeably calmer than the other dog breeds. Conclusion: Although, ADHD-like behavior does not appear to be a heritable trait in an evolutionary scale, the results reported in this study combined with existing genetic information about this trait indicated that the expression of ADHD-like behavior in dogs would appear to depend on a classical gene-environment interaction as is the case with many neurological disorders in humans. The conclusion is directed towards making recommendations to improve the social and physical situation that might minimize the potential for the development of ADHDlike symptoms in animals that are genetically predisposed to this condition.
Keywords: ADHD; Dogs; Dog breeds; Environment; Phylogeny; Social bonds; ADHD; Dogs; Dog breeds; Environment; Phylogeny; Social bonds
Attention deficit hyperactivity disorder (ADHD) is a behavioral disease that is most commonly associated with and investigated in humans and in children in particular.1,2 It is characterized by a suite of symptoms that include inattention, increased motor activity, and a tendency to show Page 96 an aggressive behavior or impulsivity 3-5 and affected individu als often show poorly adjusted social behavior. 6 The diagnostic and statistical manual of mental disorders (DSM-5) distinguishes three presentations of ADHD depending on the predominant symptom pattern for the past six months 7: combined presenta tion (presence of the core features hyperactivity/impulsivity and inattention; presence of ≥6 symptoms/feature), predominantly inattentive presentation (≥6 symptoms of inattention, ≤6 symptoms of hyperactivity/impulsivity) and predominantly hyper active/impulsive presentation (≥6 symptoms of hyperactivity/ impulsivity, ≤6 symptoms of inattention). Existing research on ADHD has concentrated largely on its causes as well as on the prevention and treatment of its clinical expression.5,8,9,10 As is the case for most other human psychological diseases, ADHD has a genetic component such that specific genes increase the risks of developing the disease; however, it is also assumed that gene-environment interactions play an important role in this context.11 The recent extrapolation of ADHD research to the domestic dog, which, like humans, is a social mammal on account of which some authors have argued that this species has the potential to display aspects of human social behaviour. 12 Humans and dogs also share prosocial qualities as well as maintain social bonds.13,14 Additionally, dogs and children also show some degree of similarity with regards to their association with the social environment and the behavioral responses related to it13,15,, which is of clinical relevance in the context of ADHD. Indeed, in contrast to rodent models and related laboratory experiments, dog models consider social relationship as a factor, one that was highlighted by Vas et al16 as an important factor in shaping human social relationships or in understanding the hyperactivity of children. Thus, altogether, dogs appear to provide an adequate model for (some) psychological diseases in humans, especially because the latter can affect different dog breeds.17 With regards to research on ADHD-like behavior in dogs, several studies have posited a genetic underpinning for this disease. For instance, increased activity/impulsivity values in German Shepherds are associated with a specific allele of the tyrosine hydroxylase gene TH.18 In addition, as in humans, polymorphism of the dopamine receptor gene DRD4 has been associated with aggressive or inattentive behavior in dogs as well as general nervousness.15,19,20,21 Inattentive behavior has also been associated in Belgian Shepherds (Malinois) with a polymor phism of the dopamine beta hydroxylase gene DPH or the dopamine transporter gene DAT22 and might more generally also be influenced by variants of the oxytocin receptor gene or genes of the brain opioid system in addition to epigenetic effects, which appear to be associated with behavioral patterns that are directed towards humans.23-25 However, in addition to genetic factors, social and physical factors also appear to have an important influence on the origin of ADHD-like behavior in dogs. In a review article based on the findings from human and laboratory animal studies, Szyf et al 26 emphasized on the importance of diverse environmental factors for the brain and for triggering long-lasting behavioral changes. Social play, for example, helps juvenile mammals improve their mental and behavioral functions9 and, in their absence, a facilitation of ADHD must be assumed. 27 Even the simple act of dogs playing with other dogs can help reduce separation-related behaviour.28 Moreover, early mental and physical activity helps provide shelter against neurological diseases. 29 Associated with this behavioral parameter, castration with its resulting decrease in testosterone values, can evoke ADHD-like symptoms in dogs30,31 such that the age at which castration is performed plays a significant role in determining the clinical manifestation of ADHD.32 Finally, Hense33 describes the other factors closely related to hyperactivity in dogs, in particular the rearing conditions, treatment received from the owner, and the existence of stressful conditions.
In this study, a comprehensive examination of potential non-genetic factors that might influence ADHD-like behavior in dogs was undertaken. A combination of questionnaires was used which directly assessed ADHD-like behavior in dogs as well as the basic personality traits of animals (see Jones and Gosling 34; Turcsán et al35) and other information from its current and past living conditions that might be relevant to the development of ADHD. Although, this study focussed on non-genetic factors, it is relevant to ask whether ADHD-like behavior was heritable on a longer, evolutionary timescale by examining the differences in its prevalence among breeds or groups of breeds (based on kennel club or genetic classifications). With a focus in particular on the breed Akita Inu, which, despite being characterized as calm,36 was described as increasingly nervous by some of the owners. Specifically, the following hypotheses were tested: 1. Social and physical factors in the environment of individual dogs influenced the degree of ADHD-like behavior exhibited by the dog.
2. The factor of castration influenced the degree of ADHD-like behavior exhibited by the dog.
3. Scores of categories of behavior in the ADHD and Personality questionnaires (motor activity or inattention; different dimensions of personality) were correlated.
4. Groups of related dog breeds differed with respect to the degree of exhibited ADHD-like behavior.
5. ADHD-like behavior in dogs is a trait that is heritable over evolutionary timescales among different dog breeds.
The degree of ADHD-like behavior in different dogs (Tables 1-3) was assessed using one or more different questionnaires. The ADHD questionnaire of Vas et al 16 required the dog owners to score the everyday behavior of their dogs, allowing for the collection of information about the degree of motor activi - Page 97
Table 1. Database Collection–main Sample Information.
Individual Breed/mongrel Age in years Sex Castration 1 Kromfohrländer 3 m n 2 Mongrel 2 m y 3 Nova Scotia Duck Tolling Retriever 2 m n 4 Labrador Retriever 3.5 m n 5 Pointer 2.5 m y 6 American Hairless Terrier 1 m n 7 Mongrel 4 m y 8 American Pitbull Terrier 2.5 m y 9 Entlebucher Mountain Dog 2.5 m n 10 Boxer 3 m - 11 Mongrel 2.75 m y 12 Labrador Retriever 2 m n 13 Leonberger 2 m y 14 Mongrel 0.5 m n 15 Great Swiss Mountain Dog 3.5 m y 16 Labrador Retriever 1.5 m n 17 Border Collie 1 m n 18 Mongrel 3.5 m y 19 Mongrel 9 m y 20 Parson Russell Terrier 3.5 m n 21 Mongrel 1 m n 22 Vizsla 6 m y 23 Mongrel 6.5 m y 24 Mongrel 4 m y 25 Mongrel 0.75 m n 26 Australian Sheperd (miniature) 3.5 m y 27 Labrador Retriever 1 m n 28 Hovawart 1 m n 29 Catalan Sheepdog 3 m n 30 Mongrel 12.5 m y 31 Mongrel 5 m n 32 Mongrel 0.75 m n 33 Doberman 2 m y 34 Bull Terrier (miniature) 2 m n 35 Poodle 3.5 m y 36 Mongrel 1 m n 37 Berger Blanc Suisse 2 m n 38 Mongrel 2.5 m y 39 Hunting Terrier 1.75 m y 40 English setter 2 m n 41 Yakutian Laika 2 m n 42 German shepherd 4.5 f n 43 Mongrel 1 m n 44 Austrian Pinscher 3.5 m y 45 Mongrel 3 m y 46 Boxer 2 f n 47 Parson Russel Terrier 3 m n 48 Border Collie 1.5 m n 49 Black Russian Terrier 1.75 m y 50 Portuguese Podengo 4 f y 51 Rhodesian Ridgeback 1.75 m n 52 Australian Shepherd 4 m n 53 Mongrel 1 f n 54 Catalan Sheepdog 3 m n 55 Catalan Sheepdog 5.5 m y 56 Chihuahua 4 f n 57 Mongrel 2 m n 58 Border Collie 5.5 m n 59 Mongrel 7 m n 60 Mongrel 0.5 f n ty and inattentiveness of the animal in different situations. The Personality questionnaire of Turcsán et al 35 measures the level of response with respect to the four distinct dimensions of the personality of the dog, namely calmness, trainability, dog sociability, and boldness, with the latter correlating to the factor extraversion in the Five Factor Model. 35,37 As with the ADHD questionnaire, the Personality questionnaire represented the behavior of the dog in the terms of a score (0=behavior was unincisive, 1=behavior was distinct at a maximum).
Table 2. Dog-School Subjects – Main Sample Information.
Individual Breed/mongrel Age in years Sex Castration 1 Australian Shepherd 4 m n 2 Cão da Serra de Aires 2 m n 3 Mongrel 4.5 f n 4 Irish Wolfhound 2.75 f n 5 Mongrel 1.75 f n 6 Mongrel 1 m n 7 Mongrel 8 f y 8 Mongrel 3 m n 9 Mongrel 8.5 m y 10 Mongrel 7 f n 11 American Bulldog 6.5 f y 12 Mongrel 9 m y 13 Berger Blanc Suisse 4 m y 14 Schnauzer 1.5 m y 15 Mongrel 8 f n 16 Bernese Mountain Dog 1.5 m y 17 - 6 m y 18 - 10 w n 19 Jack Russell Terrier 3.5 f y 20 Australien Shepherd 1.5 f n
Table 3. Akita Dogs - Main Sample Information.
Individual Age in years Sex Castration 1 0.5 m n 2 3 f n 3 2 f n 4 2.5 m n 5 4 f n 6 9 f y 7 10.5 m n 8 8 m n 9 9 f n 10 1.5 f n 11 1.5 m n 12 2.5 m n 13 2.5 m n 14 2 f y 15 2 m n 16 13 f y 17 7 m n 18 5 f n 19 2 m n 20 4 f n 21 0.5 f n 22 6.5 m n 23 3 f n 24 5 f n 25 2 f n 26 5 f n 27 3 m n 28 1.5 m n 29 2 f n 30 7 m n 31 6 f y 32 10.5 f y 33 7 f y 34 12 f y 35 0.5 f n 36 7 f n In addition, we also applied results from two questionnaires where, unlike the previous two, their scientific validi ty has not yet been established. The Anamnesis questionnaire (Gansloßer and Strodtbeck, unpublished records) mainly generated information about the earlier and current life situations of dogs. Finally, an Interview questionnaire was developed with the help of dog trainers as an extension of the Anamnesis questionnaire to try and detect ADHD -related factors about which hardly or no information could be gathered. The questionnaire was divided into an introduction and three further sections with questions about the early environment and the mother-child-re lationship of the animal, the current environment of the animal, and the level of physical exercise (e.g., training, play or the participation in dog schools and/or puppy hours). All the questionnaires have been included as the supplementary material (in below).
Figure 1. Comparison of Motor Activity Values for Four FCI Dog Groups.
Data Collected from Akita Dogs.
Kruskal-Wallis ANOVA: c2=26.643, n=68, p<0.001. Kruskal-Wallis ANOVA:c2=23.846 n=64, p<0.001.
Figure 2. Comparison of Motor Activity Values for Four Molecular Genet- ic Dog Groups (after Parker et al 39). The Ancient-Asian group has been Represented by the Collection of Akita dogs.
Figure 3. Neighbour-joining tree after vonHoldt et al.40 Allele-sharing Phylogram of Individual SNPs for Breeds and Wolf Populations.
Data were then collected from different sources using one or more of the questionnaires. First, the “collective data base” (Gansloßer and Strodtbeck, unpublished.; www.einzelfelle.de) comprised a data set with information from different dog breeds and mongrels (n=60) that was collated using the ADHD, Anamnesis and Personality questionnaires. The dogs being studied were heterogeneous with respect to their age, gender and breed. Secondly, a database of “dog school subjects” was compiled from the records of animals of different breeds and mongrels (n=20) that belonged to collectives from two dog schools in the northern part of the German province of Lower Saxony. In these dog schools, the dog owners placed their pets into one of the two different personality categories: active and nervous versus calm and/or relaxed. Data were collected using the ADHD and Interview questionnaires. Again, the investigated animals presented with a heterogeneous character with respect to their age, gender and breed. Finally, an Akita-specific database (n=36) was compiled with the help of the Japan Akita e.V . dog club in Germany and scores for motor activity and inattention were obtained using the ADHD questionnaire. For some analyses, not all questionnaires for a data set could be used because of missing answers and the sample sizes were adjusted accordingly.
Statistical analyses were performed using the IBM SPSS Statistics v.24. All the tests used a nominal alpha value of 0.05 corrected for multiple comparisons using a Bonferroni correction. This procedure was employed on a per hypothesis basis for each factor that was involved in two or more correlations and in between-group comparisons (see below).
Motor activity and inattention scores as well as scores for the questions 1 to 13 from the ADHD questionnaire) were correlat ed with factors from case-study questionnaires (i.e., Anamnesis and Interview questionnaires) to identify the social and physical correlates of ADHD-like behavior in dogs. Important social factors determined were the duration of petting bouts, the duration of periods of separation and the number of social contacts established. The latter was defined as a non-isolated dog-dog or dog-human contacts of at least a week-long duration. Physical factors included the duration and frequency of walks as well as the frequency of play time. In addition, nominal factors like castration (hypothesis 2) were also determined.
Both the ADHD and Personality questionnaires attempted to assess the degree of ADHD-like behavior in dogs, albeit using different criteria and perspectives, with the ADHD questionnaire representing a more direct assessment. Using Spearman rank correlation, it was tested how strongly each of the four personality dimensions in the Personality questionnaire correlated with the scores for motor activity and inattention scores in the ADHD questionnaire.
Two analyses were performed to determine ADHD-like behavior in dogs which had a heritable component over the evolutionary timescales, potentially resulting from its potential genetic underpinnings. First, the thorough bred dog subjects from the collective and Akita databases were pooled into groups following the grouping systems of the Fédération Cynologique Internationale, Thuin, Belgium38 and that of Parker et al39 which was based on a cluster analysis of molecular genetic information. For both the systems, it was assumed that the groups roughly reflected the degree of relatedness between the different breeds. Breeds that were not listed in either system were excluded from the analyses, with the exception of Akitas, which were included among the Spitz and primitive types (FCI) or Ancient-Asian (molecular genetic) groups. FCI groups or molecular genetic dog groups were compared among themselves for motor activity and inattention scores from the ADHD questionnaire using a Kruskal-Wallis ANOV A test. Only groups with n≥6 were used in the analyses: FCI groups – Sheepdogs and Cattledogs (n=16), Pinscher, Schnauzer and associated dogs (n=10), Terriers (n=7), Spitz and primitive dogs (n=35); and molecular genetic groups – Ancient-Asian Group (n=35), Herding-Sighthound Group (n=9), Mastiff Terrier Group (n=8), Hunting Group (n= 12). Next, it was determined if ADHD-like behavior showed phylogenetic signals in so far as its presence was significantly clustered in the molecular evolutionary tree of the different dog breeds in vonHoldt et al. 40 For these analyses, the presence of ADHD-like behavior was coded as a present/absent categorical variable based on the motor activity and inattention scores from the ADHD questionnaire responses for the collective database. Breeds for which any ADHD-like behavior was indicated were scored as (1=present); all other breeds, including those for which no information was available, were scored as 0. For the latter, it was assumed that ADHD-like behavior represented such an obvious disorder that the absence of any information in this regard for the breed meant that no symptoms for it were obviously present. Significant clustering of ADHD-behavior among the dog breeds was determined using the D measure of Fritz and Purvis41 as was implemented using the caper package v0.5.2 in R.42 Briefly, if D was not significantly different from 0, the trait was not randomly distributed within a tree and was assumed to have a phylogenetic implication, whereas there was no phylogenetic signal if D was not significantly different from 1.
The results of the present study indicated that specific ADHDlike behavior in dogs indeed correlate with numerous social and physical factors (Table 4). In particular, social factors including the duration of affectionate behavior following adoption, the number of social contacts, the duration of period of separation, and dogs that slept alone at night, all correlated significantly with increased ADHD-associated symptoms (Table 4). A negative correlation was also observed for the physical factor of the frequency of play time, whereas increase in either the frequency or duration of walks correlated significantly with increased inattention (as was measured by ADHD question 11). Castration also had an effect so far as castrated dogs displayed a significantly higher motor activity value relative to the uncastrated ones (Table 4).
Although, existing literary records of this condition on dogs is often lacking in this regard, an analysis of the human and mammalian literature supports the findings of many results, even if only at a more general scale. For instance, in their re- Page 100
Table 4. Summary of the Statistical Analyses Yielding Significant Correlations for Hypotheses 1 to 3 from the Introduction.
Hypothesis Factor - Factor / Group comparison Database Spearman’s rank correlation coefficient (rs) / Test statistics p-value (a after Bonferroni correction) 1 – Social factors Duration of affectional behavior (after adoption) – inattention Dog school subjects -0.56 (n=19) 0.013 (0.025) 1 – Social factors Number of social contacts – values for ADHD question 2 (Velocity for gaining or losing interest) Collective database -0.525 (n=20) 0.017 (0.025) 1 – Social factors Number of social contacts – values for ADHD question 3 (Concentration problems in tasks or games) Collective database -0.522 (n=20) 0.018 (0.05) 1 – Social factors Duration of separation period – values for ADHD question 5 (Problems to keep calm) Collective database -0.328 (n=49) 0.023 (0.025) 1 – Social factors Duration of separation period – motor activity Collective database -0.29 (n=49) 0.044 (0.05) 1 – Social factors Sleeping situation (alone/not alone) –inattention Collective database Mann-Whitney U=71,5 (n=37 and n=8) 0.023 (0.025) 1 – Physical factors Frequency of play time (until 6 months) – motor activity Dog school subjects -0.561 (n=17) 0.019 (0.025) 1 – Physical factors Frequency of walks – values for ADHD question 11 (Hectic or Problems in solving tasks) Collective database 0.297 (n=52) 0.034 (0.05) 1 – Physical factors Duration of walks – values for ADHD question 11 (Hectic or Problems in solving tasks) Collective database 0.29 (n=54) 0.035 (0.05) 2 – Castration Castration (yes/no) –motor activity Collective database Mann-Whitney U=594,0 (n=59) 0.005 (0.025) 3 – Personality Calmness –motor activity Collective database -0.52 (n=60) 0.00002 (0.0125) 3 – Personality Calmness –inattention Collective database -0.499 (n=60) 0.00005 (0.0125) 3 – Personality Sociability – inattention Collective database -0.326 (n=59) 0.012 (0.025) 3 – Personality Trainability – inattention Collective database -0.334 (n=59) 0.01 (0.017) Page 101 view article on human and laboratory studies, Szyf et al26 clearly indicated that factors existing in the social environment can indeed evoke long-lasting brain and behavioral changes. It is also known that mammals have a so-called panic-grief system 43 that promote the development of social bonds, 9 and also make them vulnerable to depression via social rejection or loss. 9,43 These symptoms are closely related to the synthesis of receptors involved in the development of the emotional system of the dog before the onset of sexual maturity.33 Dogs possessing a well-developed receptor system tend to achieve a feeling of social security more quickly than others. 33 Additionally, existing literature emphasizes on the possible stress effects on dogs resulting from restrictions in their social (or spatial) settings.44 In this regard, the negative correlation between many of the social factors and inattention values as outlined above becomes understandable. For instance, longer phases of affection ate behavior towards a (young) dog can release high amounts of oxytocin in it, 45 with oxytocin being an important regulator of social dynamics.46-48 Not only does oxytocin influence the developing brain of humans it also plays a role in establishing the mother-child relationship,49 towards supporting the regulation of negative emotional states as well as to enhance the motivation for building and facilitating new social interactions.50 However, oxytocin is only one such compound; it should be noted that many other chemical substances in addition to it (e.g., opioids, prolactin, corticoliberin or glutamate) have affect the origin of social relationships.51 The implied social situation of the dog can also influence its behavior as was shown by both the sleeping situation and the duration of the period over which the dogs were left unattended. Dogs that slept alone or were left unattended for longer periods were considered to be undergoing different forms of social isolation, something that was shown to increase stress levels in humans, 52 mammals in general 53 and guinea pigs and chickens54 and could contribute towards the higher inattention values in the former group and decreased motor activity scores in the latter (Table 4). Both the variables indirectly impacted the number of social contacts. 55-57 Thus, dogs might have preferred sleeping near the owners or other pets, just as Rütten and Fleißner58 observed that pack mates in African wild dogs were close to one another when resting or sleeping. Furthermore, although dogs generally remain calm during the period of separation from their owners (regardless of the duration of this period),59 those separated over a longer time show greater physical activity when their owners returned, 60 potentially reflecting on their increased perceived social isolation during the prolonged separation period. Thus, although short periods of separation did not seem as detrimental–and might have positive benefits if the results of Raineki et al61 showed that short periods of separation in rats from the mother could decrease anxiety or stress-related behavior in young rats (so-called neonatal handling) which could also be explained with respect to dogs–long-lasting, repeated periods of isolation impaired the welfare of the dog.60 Perhaps less obvious is that the physical environment of the dog also contributes to its psychological well-being. Although many studies have documented the general positive effects of physical activity (e.g., Mattson62), it remains undisputed that recovery phases will also be needed. Thus, the positive correlations that were observed between the walking activity and ADHD-like symptoms might have been derived following an overstimulation of the dogs compared to their natural behavior, with many studies indicating that dogs were largely inactive over their daily schedules. For instance, feral dogs invest only 12% of the day in travel activity.63 Further details were provided by Bloch64, who stated that dogs spend 71.4% of the day in inactivity, with the active hours (3.3 to 4.7 hours; 1.6 hours more in juveniles) being dominated by territorial behavior, feeding, or vigilance. Analogous results have been observed for guard dogs, which were found to be inactive for 70±23% of an eight-hour period65 and free-ranging Indian dogs, that spent only 15.7% of the day walking and more than half being inactive.66 In the light of these results, different walking opportunities and the pleasant anticipation associated with them could result in a permanently elevated level of arousal or unrest in some dogs and could be problematic if the dogs were already overtly active.33 In contrast, more playing opportunities for young puppies (<six months of age) appear to have a positive effect in that such individuals show less motor activity and so are less hyperactive. It is well accepted that juvenile play, which is widespread among mam mals, has a positive influence on the development of the juvenile mammalian brain and the developmental process in general.9,67,68 Furthermore, it is assumed that the inadequate play opportunities in childhood can contribute to the development of ADHD and depression in humans.9 Finally, castration in dogs correlated with increased motor activity values, as was explained by the findings of Salmeri et al 69. In association with this observation, Farhoody and Zink70, Lisberg and Snowdon71 and Zink et al72 established a link between castration and an increase in the risk for the development of fear, anxiety, uncertainty and hyperactivity, possibly because stress and fear symptoms were more difficult to regulate because of the decrease in testosterone levels. 30 However, other studies have found castrated dogs to be less hyperactive than uncastrated ones,31,73 therefore, it is possible that other factors might be involved here as well.
Existing literature based on human findings, where Parker et al74 and Nigg et al75 reported a certain degree of association between some basic personality dimensions and ADHD, the results of the present study showed several personality traits in dogs that could be correlated with ADHD-like behavior. In particular, the ADHD-like trait of inattention correlated negatively with each of the personality traits of calmness, sociability, and trainability (Table 4), such that motor activity and calmness also showed a negative correlation to each another (Table 4).
Indeed, many of the correlations that were found were hinted at on studies conducted on human subjects. For instance, Nigg et al75 established a correlation between increased aggression (which was linked to lower agreeableness or, in the current context, sociability) and heightened hyperactivity-impulsivity (here, potentially inattention). Moreover, calmer individuals showed fewer symptoms of negative emotional states. 76 The latter included anxiety and depression, which was linked to increased inattention75 and, in the context of this study, perhaps decreased trainability in dogs because their ability to concentrate on specific tasks was considerably affected. Naturally, one cannot conclude too much from these results because it is yet to be understood how many of these traits would be applicable for dogs given that it remains unknown whether similar personality traits in humans and dogs underlie the same psychological (e.g., emotional) mechanisms. In addition, dogs themselves are heterogeneous with respect to behavioral parameters which often differ between the conventional (e.g., FCI) or genetic dog groups (e.g., Ancient-Asian group).35 Heritability of ADHD-Like Behavior Although, the present study highlights on the possible non-genetic determinants of ADHD-like behavior in dogs, numerous studies also point to a genetic predisposition for this condition (e.g. Faraone et al 77; Burt 78). Being aware that many psychological diseases find their origin in gene-environment interactions (as in humans11), it was of great interest to observe if certain breeds or groups of closely related breeds were more disposed to develop ADHD-like behavioral traits in comparison to the others, reflecting on the heritability of this trait on an evolutionary timescale. In this context, differences between the breeds concerning the occurrence of certain alleles linked to ADHD-like behavior have been documented,20 which have also been assumed to lead to differences on the behavioral level.20,79 It is also known that certain breeds are more vulnerable towards developing anxiety relative to the others 80 which could lead to ADHD-like behavior (e.g., inattentive behavior caused by increased anxiety).
Following the comparison of groups of different dog breeds to one another, the results indicate a significant differ ence among groups, with significantly lower motor activi ty values for Akitas compared to all other groups, whether with the FCI Spitz and primitive group (Figure 1 and Table 5; Kruskal-Wallis ANOV A c2=26.643, n=68, p<0.001) or with the molecular genetic Ancient-Asian group (Figure 2 and Table 6; Kruskal-Wallis ANOV A: c2=23.846, n=64, p<0.001). Similarly, significant differences between the groups with respect to their Page 102 inattention scores were found for both the FCI (Kruskal-Wallis ANOV A: c2=9.239, n=68, p<0.05) and molecular genetic groups (Kruskal-Wallis ANOV A: c2=8.637, n=64, p<0.05), although pairwise comparisons only revealed one difference and that between the Ancient-Asian and Hunting groups for the latter (p=0.043). Together, these results support the idea that motor activity and inattention values are positively correlated for some dog breeds (groups), which, in turn, depend on personality traits as discussed above.
In addition, these results also fit well with the general description of Akitas as “calm”, 36 but do not match the state ments of some Akita owners who have evaluated their dogs as becoming increasingly nervous. However, it must be remem bered that the statements of the Akita owners about their dogs’ behaviour must be viewed somewhat critically because of their potential subjectivity. An owner normally has a special relationship to his/her dog and therefore could overestimate the severity of any apparent behavioural changes. It is also conceivable that the impression of some owners has more to do with the fact that Akitas might be acting more “nervously” at club meetings, where the animals are often kept closely to each other. Nevertheless, the opposite perspective is also worth considering in that other groups do have significantly higher motor activity values than do Akitas. Bradshaw et al81 indicated that herding dogs (e.g., the FCI Sheep and Cattle Dog group) tend to show an increasingly aggressive behavior, which could result in higher motor activity values; the same reason might apply for the molecular genetic Hunting group. Similarly, the Terrier group is characterized by traits such as high energy level, excitement, reactivity, but also boldness,35,82-84 which could also translate into higher motor activity.
However, these differences do not translate more generally across dogs with respect to the D-value of Fritz and Purvis41 for the presence of ADHD-like behavior based on the phylogeny as was observed in Figure 3 was 0.94, which was significantly different from 0 (p=0.01), but not from 1 (p=0.36). Thus, this trait was distributed randomly on the phylogenetic tree indicat ing a lack of any phylogenetic signal/heritability at this scale. The slight difference in these two sets of results might be derived from Akitas being noticeably calmer than other groups of dog breeds, but also because the phylogenetic analyses were more fine-grained so far as the different dog breeds were not lumped together in a limited number of categories. Moreover, although the breed groups created by kennel clubs (e.g., American Kennel Club, NY , USA or Fédération Cynologique Internationale, Thuin, Belgium) showed some behavioral differences among themselves, these were established on the basis of external similarities or undefined scientific criteria 85 rather than evolution ary relatedness. In addition, because our questionnaire sampling is “opportunistic” (i.e., data are only generated for those dogs suspected of having ADHD-like behavior), it does not represent a random sample of information from all dog breeds. Thus, ADHD-like behavior might be underrepresented in both hyper active or rare breeds because it is either held as being “normal” for the breed in the former case or there are simply not enough animals to obtain a positive result in the latter case.
In the present study, numerous non-genetic (social and physical) factors that might correlate with ADHD-like behavior in dogs have been indicated. Our initial findings naturally are subject to the limitations associated with any questionnaire study (e.g., representativeness of the sample, potential perceptional bias on the part of the owners, and, for two of the questionnaires in this case, scientifically validity the questionnaire itself). In addition, as with any correlation study, the possibility of an unknown, third causal factor (e.g., Baron and Kenny86) cannot be excluded; however, it seemed clear that ADHD-like behavior in dogs, as with many neurological diseases in humans 11, could be derived from a gene-environment interaction. As such, this study indicates the possible actions that could be implemented for those animals or breeds that possess a genetic predisposition to ADHDlike behavior to minimize the likelihood of its manifestation in these animals. Generally, these actions revolve around providing a stable, affectionate, and enriching social environment for the animal together with adequate (but not excessive) exercise and play time, particularly in the case of younger animals. Although many breeds have been bred to possess traits that might underlie or facilitate ADHD-like behavior (e.g., terriers with their high energy levels, excitement, reactivity, and also boldness), it is not necessary that the individual animals actually display symptoms Page 103
Table 5. Comparison of Motor Activity Values for the FCI Groups – Summary of the Significant Results.
Group – Group comparison Corrected p-value Sheegdogs & Cattle dogs – Spitz & primitive types 0.001 Pinscher, Schnauzer & others – Spitz and primitive types 0.001 Terriers – Spitz & primitive types 0.010 Table 6: Comparison of Motor Activity Values for the Molecular Genetic Dog Groups – Summary of the Significant Results.
Group – Group comparison Corrected p-value Herding-Sighthound group – Ancient-Asian group 0.009 Mastiff Terrier group – Ancient-Asian group 0.022 Hunting group – Ancient-Asian group 0.001 of this disease given the right social and physical environment.
We thank the dog schools Friesland and Schoone for coordinating meetings of owners and dogs. We also would like to thank all the voluntary dog owners. Another thank you has to go to the members of the club Japan Akita e.V ., with whose help we could gain Akita dogs for that project. In that regard, we are grateful for Mrs. Bettina Schmidt-Pinnekamp, who was always available as the main contact person. We thank Prof. Dr. Franz Bairlein (University of Oldenburg) for his support and useful discussions and for acting as the co-examiner for Nikolai Hoppe’s Master thesis upon which this publication is based.
The authors declare that they have no conflicts of interest.