Background


The central role of places in which physical activity (PA) is done is now widely recognized, so it is important to measure both activity and its location. At present, other than using the very expensive method of doubly-labeled water, if a researcher wants to measure PA in free living humans the most accurate technologies are either combined heart rate and motion (HR+M) sensors or accelerometers. But these devices do not collect data on where the activity occurs. If a researcher wants to know where a person performs physical activity, this information must be collected by means of self report after the PA has occurred. More recently, strategies utilizing ecological momentary assessment (EMA) have been used to sample behavioral experiences, including physical activity, in free living humans while they occur. However, this approach depends upon time- or event-critical sampling of self-reported information and thus continues to depend on self-report from the user to enter the information. Objective measurement of walking and cycling using portable global positioning system (GPS) devices has been successful but GPS data have yet to be combined with highly accurate PA measurement in a way that can be used across settings and populations. Thus, only ad hoc measurement approaches exists that allow researchers to confidently answer questions such as the following:
  1. What percent of a person's total energy expenditure (EE) from PA occurs at home?
  2. What percent of a person's EE from PA occurs in commuting to and from work?
  3. How do individuals vary with respect to their EE from PA given an equal amount of time spent in the same location?
  4. How do individuals with a given medical condition, for example type 2 diabetes or osteoporosis, differ from those without the condition with respect to EE from PA per unit of time spent in a given location?
  5. How do patterns of PA-related energy expenditure (PAEE) differ when individuals move from one location to another?

To answer these questions requires a method of measurement that is capable of simultaneously capturing PAEE data along with continuous monitoring of location. The aim of this research is to develop such a system: a Physical Activity and Location Measurement System (PALMS) comprised of an integrated suite of hardware, software and database solutions that supports real-time capture and subsequent analyses of data on PAEE from a geospatial perspective. If desired by researchers, PALMS will also support ecological momentary assessment (EMA) of psychosocial and contextual factors related to PAEE via cell phone based EMA. PALMS will provide significant advantages over currently-available measurement approaches and contribute to understanding relationships between PAEE, the environment, and health-related factors at the individual and population level.

In sum, the past few years has seen major advances in the ability to measure PAEE in free living humans through the use of wearable technologies and ecological momentary assessment. Concurrently, the field of physical activity research has expanded to include a greater understanding of the importance of the social, cultural and physical context in which PA occurs. Advocates of improved integration of natural and behavioral science research are doing so, because of maturation in sciences in the adoption of more complex study designs, the integration of different types of data (qualitative and quantitative, social and biological, objective and subjective), and developments in methods such as multilevel modeling and spatial data analysis. However, to accomplish this will require a measurement system for PA that builds on the perspectives of all of these sciences -- PALMS.