The investigation of dog walking is an emerging and rapidly growing area of research in the context of an obesity epidemic and the need to find cost-effective strategies for increasing population levels of physical activity. This paper provides evidence of a number of different correlates specific to dog walking behaviour, including demographic factors related to the dog and owner, physical and social environmental factors, and less easily measurable aspects of the dog-human relationship. Overall, the evidence currently suggests that dog walking may be most effectively encouraged through: 1) targeting the dog-owner relationship to increase the sense of obligation to walk the dog as well as the emotional support the dog can provide to the owner; and 2) by the provision of dog-supportive physical environments. These contexts may be the best-buy strategies for future testing of health interventions to increase dog walking amongst dog owners. They may even require implementation together thus acting at multiple levels (both individual and environmental) for the most effective population change . However, the quality of the evidence varies, with a number of different outcomes presented and various methodological approaches that may or may not adjust for the effect of confounding factors. Only one study used a randomised controlled trial design, thus potentially providing a stronger level of evidence of a causal mechanism for targeting the perceived canine need for exercise.
A key dog-related factor to encourage dog walking appears to be the strength of the relationship with the dog, in providing support for enjoyment for walking and a sense of obligation to walk the dog. Although there is evidence for other dog-related factors, studies suggest these disappear after adjusting for the support and motivation provided by the dog for walking [24, 25] or the sense of obligation to walk the dog [26, 42]. However it would be wrong to dismiss dog-related elements as not important; it is likely that dog demographic and behavioural factors contribute to intention to walk and practicalities of walking the dog, that drive the sense of obligation and feelings of support and motivation that arise. For example, these may be stronger for larger dogs than smaller because of differences in perceived or actual exercise needs. More research is required to better understand the correlates of dog walking behaviour, in particular the perception of the amount of exercise that different dogs require. For example, dog owners report that their dog receives ‘adequate’ exercise yet the interpretation of this appears to vary widely as they also report exercising their dogs 0 to >7 times a week . This may provide scope for interventions that change the perception of what a dog needs in terms of exercise and requires further investigation for which in-depth qualitative enquiry may be useful.
The evidence also suggests that interventions that strengthen the relationship between the dog and their owner may be a useful strategy; doing obedience training or simply spending time with a dog can improve the relationship with the dog and the obedience of the dog . However, it is unclear whether walking with a dog leads to a stronger relationship between the owner and their dog or if an existing strong relationship leads to more walking. Likewise, behavioural issues related to the dog (e.g., aggression or fears) may result in less owner physical activity because of decreased motivation to take the dog out in a public place, but may also be caused by it, as inadequate early experiences and socialisation in dogs can lead to the development of behavioural problems .
The design of areas intended for dog walking and how they fulfil dog and owner needs may be an important consideration for future interventions. In order to encourage more dog owners to walk their dogs the recreational areas used for dog walking must be both pleasurable and accessible, as opposed to the common phenomenon of relegating dog access only to the few areas left after other user types have been accommodated . However, it is difficult to tease out cause and effect – do regular dog walkers choose to live closer to high quality parks where they can walk their dog or does living next to a high quality park cause people to walk their dog more? In particular there must be sufficient provision for off-leash walking as this appears to be important to owners; it is perceived that quality of life of the dog would be compromised if dogs could not be walked off-leash in areas . This suggests that an important function of dog walking, particularly off-leash, is enhancing dog quality of life, and thus how a public space fulfils their dogs needs is important to owners.
We also know that walking with a dog is not simply a vehicle for physical activity; it also increases the frequency of interactions with people, especially strangers . However, the effect appears to be dog specific, with breed and age of dog influencing these interactions . In a UK study, 92% of owners noticed seeing the same people and their dogs regularly while walking their dog . In the US, ‘dog parks’ are perceived as providing opportunity to meet neighbours and build community . Pet ownership has also been positively associated with perceptions of neighbourhood friendliness, with pet owners score higher on social capital and civic engagement scales . Dog walking is also a way to spend time with friends; in a UK study 38% of owners reported never walking their dogs with a group of friends and their dogs, but 3% did this every day . If facilitation of social interactions is a potentially strong feature of dog walking practices, there may be a need to acknowledge and encourage this human need when designing interventions and space for dog walking, especially for those owners for which dog walking is a rare opportunity for social interaction (e.g., older adults). It is possible that health promotion activities focussed on increasing owner awareness of the importance to do regular physical activity may sometimes be misplaced; if the physical activity resulting from walking a dog is a secondary outcome to fulfilling other needs of the owner or dog. Owners may be more likely to participate in and sustain physical activity if they actually enjoy it, or feel that it benefits their dog, than if they are simply trying to be healthier and more active for themselves.
This review highlights a number of methodological aspects that have implications for future research and for the interpretation of our findings. The majority of studies to date have been cross-sectional in design which limits the ability to confirm the causal relationships between dog, owner, social and physical environment related factors and dog walking behaviour. Furthermore, the majority of studies did not use objective measures of physical activity. However there is some evidence to suggest that objective measurements of dog walking physical activity (i.e., accelerometers) correlate with self-report measures .
Moreover, many studies did not adjust for the effects of confounding variables. This means that reported findings for some factors may be due to their own correlations with other factors. For example, it is important to control for dog-related demographic variables such as dog size when considering the effects of dog behaviour, and vice versa. Furthermore, the type of dog and way it behaves may also be correlated with the owner-dog relationship. This may explain the sometimes conflicting findings for some of the correlates of dog walking reviewed, for example where sometimes dog behaviour or size appears to have an effect, but not in other studies where an aspect of the relationship or support/motivation provided by the dog is also examined. Furthermore, clear evidence regarding dog-related factors (size, breed, etc.) may also be difficult to gather because they are hard to accurately measure in self-completed surveys and there may not be sufficient power to detect differences, especially in the often small sample sizes used. It is also possible that interactions between dog age and breed as well as a non-linear relationship between dog age and dog walking behaviour may exist and this should be considered in future studies. Few studies in this review sought to adjust for well-known correlates of general physical activity, such as: socio-demographic (gender, age, country of origin, education, occupation, children at home <18 years); perceived physical–environmental; and family social support and intrapersonal factors . Even where multivariable analyses were performed, some adjusted for a wide range of non-significant factors due to their theoretical importance and high face validity (e.g. [24, 25, 58]), whereas others did not clearly report adjustment for other factors (e.g. [23, 42]). Dog walking behaviour outcomes also varied widely between studies with some at the level of the dog and others at the level of the owner (see Table 1).
Finally, future research should include context-specific measures of both the dependent and independent variables . For example, context-specific measures of intrapersonal factors such as ‘intention to walk the dog’ as opposed to measuring ‘intention to walk’ in general (e.g. ) may better capture the factors associated with dog walking behaviour. In addition, it is unclear whether correlates influence dog walking behaviour and/or if they effect intention to walk the dog. As highlighted in our proposed theoretical model (Figure 1) future research should examine the pathways through which the different correlates influence dog walking behaviour. A survey tool has been designed specifically for this type of research (the Dogs And Physical Activity Tool; DAPA Tool) . The DAPA Tool has been shown to be a reliable tool for measuring important attributes and scales relating to dog owners’ physical activity behaviour and the context-specific factors that affect owners’ walking with their dogs. Future studies would be more easily comparable if consistent data collection methods were used, however, the DAPA Tool is likely to require further development for effective use in different contexts and cultural locations than its original design (for example to study children or other cultures than Australia where it was developed).
This review highlights that there has been little explicit research as to what dog walking actually is, to both the owner and the dog; what actually happens on a ‘dog walk’ and what functions it performs. It is recommended that physical activity behaviours are considered separately, in order to study and implement the most effective strategies to fit this particular physical activity context . Previously, dog walking has been considered a leisure-time or recreational physical activity behaviour (e.g., [16, 63, 77–82]); non-exercise related walking ; chores/errands ; and even commuting physical activity . Thus we recommend that in future dog walking is considered in its own right. The dog walking experience also depends on whether the dog is on or off-leash. Most dogs stay fairly close to their owners [35, 54] and a large proportion of the walk is spent sniffing, especially when off leash or if the dog is a ‘gundog’ type . Further research is required to better understand the dog walking experience: its intended purpose (for recreation or transport or just a chore that has to be done); the intensity (light, moderate or vigorous) and pattern (long bouts or numerous short bouts) of the physical activity undertaken with a dog; and how this is affected by factors such as how the dog behaves; which in turn is affected by dog breed, whether the dog is on or off leash as well as the environment the dog is being walked in. Finally, while dog walking has the potential to facilitate increased physical activity in owners, forcing dogs to exercise with us may also raise ethical questions  for example if there is a mismatch between the needs of the dog and the wishes of the owner.
This emerging area of research is also affected by a frequent problem in sociological research; that people, including researchers, already have ‘common sense’ knowledge of what dog walking is to them. However, it cannot be assumed that this is the same across society; there is unlikely to be a single ‘type’ of dog walker and there are likely to be cultural differences even between western countries for example, the use of official ‘dog parks’ (in the USA but not the UK) where owners passively sit or stand is common . Further research in different countries and cultures is required to target the benefits to be gained from dog walking, as multiple strategies are likely to be required; for example, initiation of dog walking, maintenance, or increasing frequency may all benefit from different strategies. Further qualitative and observational research is required to enhance understanding of the phenomenon and its complexity, in particular surrounding the ‘intensity’ of physical activity that occurs during ‘a dog walk’ and the everyday barriers that affect dog owners regardless of their best intentions to walk more. Much of the research to date has focused on why people do walk their dog rather than why they do not, and there has been very little research into how policy affects dog-walking behaviour.
The purpose of this review was to provide information on correlates that may be useful for the design of interventions to encourage dog walking. However, intervention studies need to measure dog walking in the context of overall physical activity, in order to determine whether the intervention increases the physical activity of those who are already physically active, or changes the environment in which physical activity occurs. Although encouraging those owners who already walk their dogs regularly to walk even more is not problematic, the priority for improving population health should be to increase the physical activity of those owners who currently do not undertake physical activity on a regular basis. Intervention studies must also openly report the tool used to deliver the intervention, designed with our model of specific behavioural correlates in mind so that these can be clearly identified and their effects measured. For example it is difficult to evaluate exactly how giving ‘persuasive material about canine health from walking’  is acting on owner’s dog walking behaviour: which particular aspects of canine health is it describing benefits to? (physical? mental?); is it changing perceived exercise requirements?; is it increasing the owner’s overall perception of the valuation of exercise for dog?; is it increasing the sense of dog obligation?; or even using the subjective norm of significant others depending on how the intervention is delivered (e.g.; veterinarians)?
Finally, this review has prioritised the individual health impacts that may be gained by encouraging a person to walk more with their dog. However, the positive impacts of dog owners walking their dogs may also extend to non-owners, through an increased sense of safety in the neighbourhood (see ) as well as sense of community and social capital . It must also be acknowledged that a few studies have reported negative impacts of dog walking or dogs as a barrier to physical activity, for example through concerns about loose/stray dogs and dog waste [88–90] or being an injury hazard through bites or falls [91–94]. There are also concerns over the impact of dog walking on wildlife [53, 95–100]. Any attempts to promote dog walking activity must be done in a manner that is also mindful to the potential negative impacts of dog walking on society; and any attempts to prevent or reduce dog walking and its associated impacts should also be aware of the negative effect this may have on the health of dogs and their owners.