Physical activity and sedentary behaviour in humans and pet dogs

Physical inactivity is a major contributor to non-communicable diseases and many adults and children are insufficiently active to maintain good health. The proportion of children who meet the United Kingdom recommendations for physical activity (at least 60 minutes of moderate-vigorous intensity physical activity each day) has been reported to be as low as 3% for boys and 2% for girls. Systematic reviews on interventions to promote physical activity in childhood have shown that although physical activity is modifiable to some degree most interventions have had only modest and short-term impacts on physical activity. Therefore, novel approaches to physical activity promotion in childhood are required. Dog ownership is a significant societal factor that may be used to encourage and sustain health behaviour change at individual and population levels. A number of observational studies have reported that dog ownership and/or dog walking are associated with increased levels of physical activity. However, evidence is lacking as to whether and how interventions with families and their dogs can be used to promote physical activity. Therefore, the major aim of this thesis was to assess the feasibility, acceptability and potential efficacy of a theory-driven, family-based, dog walking intervention for 9–11 year old children and their families. However, prior to this it was essential to develop ActiGraph cut-points for measuring physical activity intensity in dogs. The ability to measure the intensity of dog physical activity accurately was important as it allows for the effectiveness of dog walking interventions to be tested, therefore another aim of this thesis was to calibrate and cross-validate ActiGraph cut points that can be used to describe physical activity in dogs by intensity. Similarly, no studies have been published previously that assess which factors are related to dog physical activity when measured using ActiGraph accelerometry. It was therefore desirable to explore whether body condition score, breed, age, and neutered status are associated with ActiGraph measured dog physical activity. Furthermore, no published studies have described the spontaneous changes in dog physical activity during substantial weight loss; therefore, another aim of this thesis was to explore changes in physical activity in dogs during a 6 month calorie controlled weight loss programme. Using Receiver Operating Curve analyses Chapter 2 showed that the ActiGraph GT3X can accurately measure the amount of time a dog spends sedentary, in light-moderate intensity physical activity and in vigorous intensity physical activity. The sensitivity and specificity of the cut-points developed when using both the integrated axes and vertical axis accelerometry data were high. Agreement between the accelerometer data and direct observation in the cross-validation subset was also ‘very good’ (as measured by Cohen’s Kappa). This indicates that the ActiGraph GT3X accelerometer is accurate when measuring the intensity of physical activity in dogs, facilitating the use of the ActiGraph GT3X to describe the frequency, intensity and duration of dog physical activity in Chapters 3-6 of this thesis. Chapter 3 shows that, in a sample of dogs of varying breed and body condition scores, obese dogs spend significantly less time in ActiGraph measured vigorous intensity physical activity than ideal weight dogs (6 ± 3minutes/day versus 20 ± 14 minutes/day). Chapter 4 focussed on the factors related with physical activity in the two most commonly registered dog breeds in the United Kingdom, Labrador Retrievers and Cocker Spaniels. Five potential correlates (age, sex, breed, neuter status, body condition score) were tested with associations with ActiGraph measured physical activity. Age and breed were associated with total volume of physical activity, light-moderate intensity physical activity and sedentary behaviour in the final models and age was also associated with vigorous intensity physical activity. Unlike Chapter 3 body condition score was not related with any physical activity variables. Chapter 5 explored the changes in physical activity and sedentary time during weight loss in dogs enrolled in a 6 month calorie controlled weight loss programme. Despite an average weight loss of 15% body weight from baseline there was no marked increase in any ActiGraph measured physical activity variable. Chapter 6 describes the results of the Children Parents and Pets Exercising Together (CPET) Study. CPET was the first exploratory randomised controlled trial to develop and evaluate an intervention aimed at dog-based physical activity promotion in children, their parents and pet dogs. The results show that the CPET intervention was both feasible and acceptable to study participants. Eighty-nine percent of families enrolled in CPET were retained at follow up. Ninety-five percent of intervention sessions were delivered and ActiGraph measured physical activity data were collected for 100% of children, 96% of parents and 96% of dogs at baseline, and 100% of children, 96% of parents and 96% of dogs available at follow up. Despite the apparent feasibility and acceptability of CPET there was no significant change in the primary outcome measure (child physical activity) or the majority of the secondary measures. This thesis shows that the ActiGraph GT3X accelerometer is capable of accurately measuring the intensity of dog physical activity. It also shows that obesity may be related to lower levels of objectively measured vigorous intensity physical activity and the physical activity levels in dogs decline with age and vary by breed. However, it appears that physical activity levels do not increase spontaneously as dogs lose substantial amounts of body weight. Using pet dogs as the agent of lifestyle change in physical activity interventions in children and their parents is both feasible and acceptable; however, the lack of any apparent increase in child physical activity suggests that the intervention may need to be modified in a future, more definitive trial. In summary, the findings of this thesis have important implications for the measurement of physical activity intensity in dogs, the understanding of factors associated with dog physical activity and for the development of dog-walking interventions in children and their parents