Acquiring a Pet Dog Significantly Reduces Stress of Primary Carers for Children with Autism Spectrum Disorder: A Prospective Case Control Study

Autism spectrum disorder (ASD) is a heterogeneous condition defined by the DSM-5 as a person experiencing persistent difficulties in social interaction in a range of contexts and as showing restricted, repetitive behaviours. These problems must have been evident in early childhood, cause significant impairment in functioning and not be explainable by intellectual disorders or developmental delays (DSM-5, APA 2013). Parenting children with developmental disorders, such as ASD is associated with higher levels of stress, anxiety and negative outcomes (such as depression and social isolation) when compared to parenting typically developing children, or children with other non-developmental disabilities (Dunn et al. 2001; Koegel et al. 1992; Weiss et al. 2013; Wolff et al. 1989). High levels of stress impact not only on the health and wellbeing of the carers themselves, but can also limit the effectiveness of the outcomes of ASD interventions (Robbins et al. 1991; Osborne et al. 2008). As such the assessment of interventions and lifestyle choices that effectively reduce carer stress is a critically important area for research in ASD treatment programmes.

The availability of social support and levels of stress experienced have been associated with successful adaptation of the carer (Koegel et al. 1992; Konstantareas and Homatidis 1989; Weiss et al. 2013). It has been suggested that the type of social support may relate to the effectiveness of stress buffering, with more informal social support (e.g. spouse, family, friends) acting as a more effective stress buffer compared to formal or structured social support (e.g. parenting support groups) (Boyd 2002). Given that informal social support may be an effective remediation tool for reducing stress in carers it is appealing to investigate the potential of companion animals, such as pet dogs to provide informal social support for parents of children with ASD.

It could be proposed that because carers already experience high levels of physical and emotional demands, the demands of acquiring a pet might increase rather than reduce stress. However, there is increasing evidence to support stress reducing and health enhancing benefits of pets on individuals and families (Allen et al. 2001; Friedman and Thomas 1995). Dog ownership has been found to be a positive factor in supporting individuals in difficult times, including children affected by serious illness and death of a parent (Raveis et al. 1993) and in reducing the symptoms of physical and psychological illness in women coping with the loss of a loved one (Barker and Barker 1988). Furthermore, companion animals have been shown to reduce the onset and severity of stress-related conditions (Wilson 1991). Additionally, there is evidence to support the utility of trained assistance dogs as therapy for children with ASD (Solomon 2010; Berry et al. 2013), with benefits such as enhancing family freedom (Burrows et al. 2008), reducing child stress (Viau et al. 2010) and improving the effectiveness of therapy sessions (Silva et al. 2011). These effects are remarkable in themselves for the benefits they provide for the child with ASD, but they may also induce a wider positive impact upon parental stress levels, particularly in circumstances where the dog lives as part of the family as opposed to being a part of structured therapy sessions. Indeed, a study by Burgoyne et al. (2014) report an increase in caregiver competence, although not caregiver strain, in families living with a trained assistance dog.

The limited number of studies which have explored the effects of pet dogs in families with children with ASD, have primarily focused on outcomes related to the child with the ASD diagnosis. For instance, a recent study by Grandgeorge et al. (2012) showed that parents reported an increase in prosocial behaviours in their child with ASD and a reduction in anxiety with the acquisition of a pet (including cat, dog and small furry animal). Such beneficial effects in child behaviours may improve parental stress. Alternatively, because pet dogs interact with the entire family unit it is also possible that the dog may elicit similar stress reducing effects directly in the parent. The only other (known) study looking at the effects of pet (as opposed to service/assistance) dogs reports parental opinions on their perceived benefits and limitations of acquiring a dog in the family (Carlisle 2014). Indeed, it is important that parents acquire a dog after careful consideration of both the potential benefits and negative implications of dog ownership. Due to both the individual nature of ASD and the characteristics of dogs as unique living species it is unlikely that dog ownership will benefit all families in the same way. Therefore, whilst the purpose of this paper is to report data illustrating positive effects of pet dogs for parents of children with ASD we are keen to point out pet dogs as effective ASD therapy is still in its infancy and requires greater scientifically robust evaluations. Our research team has investigated parental expectations of acquiring a pet dog and observed some noticeable disparities between expectations and reality which are important for practitioners and parents to consider when thinking about dog ownership (Wright, Hall, Hames, Hardiman, Burgess, Mills and Mills, under review).

As previously mentioned parental stress levels are thought to be important determiners in the success of ASD therapy programmes (Robbins et al. 1991; Osborne et al. 2008), therefore, one important stage in developing understanding of animal companionship in ASD therapy is to investigate the impact of dog ownership on parent stress. Given the potential combined benefits for the parent, child and whole family, pet dogs may be a widely acceptable and flexible lifestyle change which reduces stress in the carers of children with ASD. Furthermore, there are many potential benefits of dogs as an intervention in this context; they are accessible, socially valid and acceptable in most western cultures, therefore, the aim of this study was to assess the impact of acquiring a family pet dog on the stress of carers of a child with ASD.

The research process was approved by the University of Lincoln’s ethics committee.

We present the results of statistical analysis on parents’ responses to the items on the PSI-SF separately for each sub-scale. Each section starts with the results of the ANCOVA, followed by consideration of whether the two groups had made a clinically significantly change in their responses over the sampling points.

A significant effect of group was seen in three out of four of the measures of stress in the Parenting Stress Index. In the intervention group (those acquiring a pet dog) a reduction of parental stress in the domains of Total Stress (overall stress in the parenting system), Difficult Child (the basic behavioural characteristics of the child) and Parental Distress (the level of stress experienced by the carer in their parenting role), these effects were not observed in the control group. Unsurprisingly given the nature of the intervention, the stress associated with the parent’s perception that his or her child does not meet their expectations did not significantly change in either group (Parent–Child Dysfunctional Interaction subscale).

For all three of the measures in which there were significant reductions in the intervention group, these were evident from the post-intervention period and maintained at the lower level at follow-up, suggesting that dogs may provide a relatively immediate stress buffering effect and that this benefit is relatively enduring. For one measure within the intervention group (Parental Distress) values even fell to within clinical normal levels, but such a dramatic change was not seen in relation to the other measures, which remained well within clinically high ranges. The correlation between the magnitude of the reduction recorded in parental distress scores and baseline parental distress scores is consistent with change within the intervention group being specific and not random, and is supported by no such correlation in the control group. This suggests that pet dogs may be more effective at reducing the stress of carers experiencing higher levels of distress associated with the parenting role. In contrast, the only significant changes apparent within the control group (Difficult Child subscale) appear to have been the result of random change, potentially associated with regression to the mean, following a numerical rise post-intervention.

This study is highly original in at least two ways. Previous studies have examined the effect of interventions on parents (Robbins et al. 1991; Remington et al. 2007; Shields and Simpson 2004; Tonge et al. 2006), but none have focused on the introduction of a family dog on parental functioning. Those studies that have examined the impact of dogs on families such as those used in this study, have focused on the effects on the child (Grandgeorge et al. 2012; Carlisle 2014) rather than their incidental effects of pets on carers of the wider family.

These results suggest that stress in primary carers arising from the behaviour of the child and the carer’s global assessment of their situation but not the severity of their assessment of their child against their expectations, are reduced in the medium term as a result of pet dog acquisition. This impact on carer distress may come about via one or more mechanisms, which may vary between individuals. Firstly, there is evidence to support general but direct physiological stress reducing effects arising from interaction with pets (Odendaal and Meintjes 2003; Nagasawa et al. 2009; Katcher et al. 1983). Dogs may also provide social support, which has been suggested to be particularly effective in reducing stress in carers of children with ASD (Koegel et al. 1992; Konstantareas and Homatidis 1989; Weiss et al. 2013). Therefore, carer stress might also be reduced as a result of the social support provided by pet dogs.

Other possible mechanisms for stress reduction may come about more indirectly for example via changes in physical activity, outdoor access, change in routines or carer time away from child, all of which carers are likely to experience following dog acquisition. Indeed, the acquisition of a pet dog is likely to alter family dynamics such as these, which could reduce parental stress. It could be argued that similar changes in dynamics, and therefore parental stress, could be observed in the addition of another human family member, such as a new baby. However, we do not consider this to be a reliable reflection of the mechanisms underlying the improvements in stress levels. For instance, a study by Allen et al. (1991) showed that stress levels reduced greater in the presence of a dog compared to a friend; indicating the unique processes involved in human-animal interactions as opposed to human interactions. Furthermore, animals are unique in the non-judgemental, unconditional support with which they provide (Kruger et al. 2004). Even less direct mechanisms for stress reduction on the primary carer, include the consequences of the positive effects of the dog on other family members and the child with ASD, which have been reported in other studies (Berry et al. 2013). Accordingly these results pose the intriguing possibility that at least some of the previously reported benefits to the child and family of interventions may be largely subjective arising from reductions in parental stress, rather than real changes in child behaviour.

It is likely that different factors are important in different contexts and while this may pose challenges for assessing or inferring the mechanism behind the changes noted; from a clinical perspective this may indicate that dogs are particularly useful, since their nature means they may provide a personalised and socially valid intervention with the minimum of clinical effort. In this context, voluntary dog acquisition offers a flexible intervention that is economically effective, integrates well into the family, making them potentially very useful for the varied symptoms of ASD, which is highly individual in nature.

One limitation with the study is that it is difficult to make direct comparisons with other studies measuring parental stress levels in ASD programs. Previous studies have reported reductions on PSI outcome measures through implementation of parent-focused early intervention programs (Robbins et al. 1991; Shields and Simpson 2004), direct comparisons with these interventions are difficult since there are a number of demographic and methodological differences. These include child age range, baseline score, version of the PSI used and lack of suitable control groups. An important finding to highlight from this study is that pet dogs appear to have greater stress buffering properties for parents who are experiencing higher levels of parental distress to begin with. This is crucial because it directly targets previously identified issues which show that parents with high stress levels do not respond as well to intervention programs (Robbins et al. 1991; Osborne et al. 2008). In contrast, it appears that parents with high stress levels benefit the most from pet-dog intervention programs. Additionally, the introduction of a dog to the family may not only be beneficial in its own right, but also facilitate increased efficacy of other interventions. These results also imply that pet (and especially dog) acquisition needs to be recorded and its impact either controlled for or considered in other longitudinal studies.

A second limitation of this study concerns the prospective case–control design. Pet dogs are widely considered to be family members by their owners, and so while acquisition of a dog represents a voluntary change, or intervention within the family, the nature of this means a randomised intervention design is neither ethical nor feasible. A prospective case–control design is therefore the most rigorous practical design for assessing the effect of this type of change in lifestyle, with the results strengthened by the use of a previously validated clinical scale (Abidin 1995). In order to gain a wide ranging sample to reflect the large individual differences in the diagnosis of ASD the groups were not controlled for diagnosis or child behaviour. It is noted that a greater proportion of the intervention group were recruited though the PAWS network compared to the control group and although there were more children with a diagnosis of HFA/Asperger syndrome in the intervention group, there is currently no consensus on the relationship between parental stress and severity or presentation of ASD (Pisula 2011), and this did not result in significant differences in other demographics or baseline PSI measures between groups, indicating that the groups were generally well matched. Although a potentially important limitation could relate to the loss of data from carers who relinquished a dog after acquisition given the repeated measures design, this concern can be mitigated by considering that the results relate only to families in which there is a successful and voluntary dog adoption. At 16 % within the first 6–8 months, the rate of relinquishment in this study does not appear to be excessive when compared to typical populations of families acquiring dogs (Mondelli et al. 2004; Patronek et al. 1995; Wells and Hepper 2004).

Other limitations of this study include the measures implemented to confirm child diagnosis and language ability. The diagnosis of ASD is subjective and controversial (e.g. Shattuck and Grosse 2007; Szatmari 2011), however, we chose to take parents’ verbal confirmation that their child had received a clinical diagnosis of ASD, and their subjective rating of their child’s language ability as satisfactory evidence for the purpose of this study. This approach leaves open the possibility of misjudgement, for both under and over representation of their child’s functioning. Future studies could explore the significance of using standardised clinical measures of these factors in relation to the effects of pet dogs in these families to assess whether the strength of the findings is increased or decreased.

While these findings are of enormous potential impact, it has to be recognised that they relate only to the first 6–8 months post dog acquisition and there is a need for longer term follow up. This is a priority, given how promising the results appear, since it is necessary to establish both whether the reduction is maintained and/or continues to help normalise the stress within families in the longer term. It should also be noted that, in this study, the majority of families acquired dogs as young puppies which may also be relevant to the impact seen. However, it would be expected that puppies would place more demands on the carers, in terms of the amount of management and training required, and so while they may be more beneficial in some ways, they may also be more demanding. Despite this, the results of this study indicate that clinicians should support families thinking about getting a dog, so long as realistic expectations are set via programs like PAWS and no specific additional risks are identified.

These results also provide a solid basis to raise other important questions in relation to the potential for dogs to provide stress-ameliorating effects to the wider carer population, for example those caring for sufferers of dementia, or the terminally ill, which is an area of growing economic and political concern.