Background
Evidence suggests that parents play an important role in shaping children’s physical activity levels through behavioral modeling and support. Physically active parents are more likely to have physically active children. However, research indicates that on average, parents and children spend less than 5 minutes per day engaging in physical activity together. Increasing joint parent-child physical activity could have health benefits for both children and parents alike. Girls and older children spend more time performing physical activity with a parent than boys and younger children. Yet, time spent in joint parent-child physical activity does not differ by the weight status of the child. Understanding where and in what types of situations that joint parent-child physical activity occurs for these subgroups could inform interventions in this area.

Objectives
Research examined the locations of joint physical activity in parent-child pairs who both wore an accelerometer and Global Positioning Systems (GPS) device over the same 7-day period. The objectives were to (1) classify the locations of joint parent-child physical activity according to primary land use type (e.g., residential, commercial, open space), and (2) determine whether the locations of joint parent-child physical activity vary by age, gender, ethnicity, annual household income, and weight status of the child and parent.

Methods
The sample consisted of parent-child pairs living in Southern California. Children ranged in age from 8-14 years. An Actigraph accelerometer and GlobalSat BT-335 GPS device collected activity and geographic positioning data, respectively, using 30-sec. epochs across 7 days. Overnight hours (11pm-5am), school time (8am-3pm on weekdays), time spent in motorized transit (> 32 kph), and periods of accelerometer non-wear (> 60 min. of consecutive 0 activity counts) were removed. Whether behavior could be classified as moderate-to-vigorous physical activity (MVPA) was defined using age-specific thresholds for the accelerometer. Linear distance between the parent and child for each 30-sec. epoch was calculated using geographic coordinates from the GPS. The analytic dataset consisted of 30-sec. epochs exhibiting “joint” parent-child MVPA, defined as MVPA performed by parents and children at the same time and in the same location (< 50m. apart). In Geographic Information Systems (GIS), land use data provided by the Southern California Association of Governments was used to assign land use classification to GPS data points of joint parent-child MVPA epochs. Multilevel multinomial logistic regression analyses tested whether age, gender, ethnicity, annual household income, and weight status of the child and parent predicted the likelihood of joint parent-child MVPA bouts occurring in each land use type as compared with residential locations (reference group).

Results
Of the 291 parent-child pairs with sufficient accelerometer and GPS data, 260 pairs had at least one epoch of joint MVPA (6,597 joint MVPA epochs across the entire analytic sample). Children were 53.1% female, 41.5% Hispanic, and 35.5% overweight or obese with a mean age of 11.1 years (SD = 1.47). Parents were 88.1% female and 70% overweight or obese. The largest proportion of joint parent-child MVPA occurred in residential locations (35%), followed by commercial/services/office /industrial (e.g., retail stores, restaurants, personal services, private health club/gym, motels) (24%), open space/recreational/agricultural/ vacant/construction (e.g., parks, gardens, wildlife preserves) (20%), educational institutions (e.g., schools) (14%), public facilities (e.g., government, health care, religious, libraries, community centers) (7%), and mixed/other land uses (e.g., roads, water) (1%). The locations of joint parent-child MVPA differed by child weight status (Adj. Wald F = 3.83, df = 10, p< .001). As compared with children who were obese, normal weight and underweight children were more likely to perform joint physical activity with their parent in residential locations (39% versus 25%) and on land classified as open space/recreational/agricultural/ vacant/construction (25% versus 5%). In contrast, children who were obese were more likely to engage in joint physical activity with their parent in commercial/services/office /industrial locations (29% versus 21%), at educational institutions (30% versus 9%), and public facilities (11% versus 5%) than children who were normal weight or underweight. The locations of joint parent-child MVPA were unrelated age, gender, ethnicity, and annual household income of the child and parent.

Conclusions
The study represents one of the first attempts to use objective measures (GPS and accelerometer) to understand where children and their parents engage in physical activity together. Results suggested that residential locations were common venues for joint parent-child activity. The locations of joint parent-child physical activity taking place away from home, however, varied by the child’s weight status. Normal and underweight children were more likely to engage in physical activity with a parent at parks or other types of open space; whereas obese children were more likely to perform joint activity with a parent on school grounds, public facilities (including community centers), and commercial locations (including private health clubs/gyms). Future research is needed to understand whether these patterns represent different preferences for types of activities or amenities.

Support/Funding
National Cancer Institute #R01-CA-123243 (Pentz, PI).