Present:

Aging Conference: Barbara Resnick is presenting a symposium for the Gerontological Society of America annual conference in Washington, DC in November, 2000. The symposium is entitled
"Measurement of Activity in Older Adults: A Triangulation Approach."

Participants:
B. Resnick,  D.Orwig, J. Magaziner, & S. I. Zimmerman
University of Maryland, Baltimore
Electronic Devices to Measure Activity in Older Adults: Utility of the Step Activity Monitor

C. R. Nigg., C. Padula, P. Burbank, & C. E. Garber*
University of Rhode Island, Kingston, RI; *Brown University, Boston, MA
What Can Measurement of the Stage of Change Add to the Assessment of Physical Activity?

C. M. Castro, A.C. King,  A. L. Stewart*
Stanford University, Palo Alto, CA; *University of California, San Francisco, CA
State of the Art Self-report Measures of Physical Activity in Older Adults

Discussants:
T. Bazzarre, American Heart Association, Dallas, TX
A. Jette, Sargent College of Health & Rehabilitation Services, Boston University, Boston, MA

Abstract:
Concerns about the reliability and validity of methods used to measure physical activity in older adults limits the advancement of knowledge in this area, and the ability to accurately test interventions to increase activity. Although there are many methods available each measures only a part of the entire activity pattern. Validation of these measures is difficult as there is no gold standard with which to compare them. Consequently, a triangulation approach to the measurement of activity is recommended including: an appropriate form of self-reported physical activity, use of an electronic mechanical method of measurement, and consideration of motivational readiness for physical activity.  This symposium will provide a review of  self report mea- sures with a special focus on the CHAMPS and Stanford Physical Activity Recall (PAR), reliability and validity testing of a new Step Activity Monitor (SAM), and appropriate questionnaires to measure motivational readiness in this population. Moreover the triangulation of these methods, and how their combined use best explains activity in older adults will be addressed. The symposium will help identify the most appropriate methods to measure physical activity in older adults.

Book Chapter: Greg Welk is preparing a book on Physical Activity Assessments for Health Related Research and has commissioned Claudio to take the lead on a chapter on "Activity Assessments for Clinical Trials Research". Drafts will be sent to other members of the BCC PA workgroup so the chapter can reflect the collective thoughts of the group.

Web Page Update: The web page includes the mission and goals of the workgroup and a section on the different assessments that are used in the projects.
       

Discussion: While a variety of different instruments will be used across studies it may be possible to have groups try to report common outcome measures. There was some consensus that measures such as minutes of activity (moderate and vigorous) and EE from activity (kcal/kg/day) would be available from most instruments. For sites using logs or other instruments that don’t have these calculations available, the revised version of the Compendium of Physical Activities was recommended as the best source for estimating EE.

 Goal 2: Develop common operational definitions of activity/exercise for cross project use.

Discussion: Common definitions would allow for better cross study comparisons but the type and nature of activity that is being promoted vary across projects. Abby mentioned that the WHO has guidelines for recommended levels of activity across the lifespan and this was viewed as a possible way to establish some definitions.

 Goal 3: Increase our understanding of the measures and specific instruments for measurement of activity/exercise through the use of common measures across studies.

Discussion: Since the last meeting, the workgroup has agreed to recommend the use of a PA staging algorithm for use across all sites. Eleven sites have currently indicated that they would be able to adopt this instrument. Because some of the sites are not as familiar with interpreting this type of data it was agreed that additional information on analyzing and interpreting this type of data would be useful. There was additional discussion about possibly using the staging algorithm to subdivide sample into groups by level of progression or regression. Additional analyses could be used to compare overall activity levels of groups using the primary outcome measures.

There are also five sites currently planning to use the Yale (first two pages) and four sites planning to use the CHAMPS instruments in their projects. It was agreed that efforts should be made to ensure that the protocols used for data collection are similar.

 Goal 4: Increase our understanding of the relationship between demographic variables (age, ethnicity, SES, gender, marital status) and exercise/activity.

Discussion: The BCC has developed common demographic variables that will be collected across all studies. These could be used to look at differences associated with the different target populations. These comparisons could be made with the data from the staging questions or from the use of common instruments and/or common measures in the different projects. Because of potential differences in how activity is perceived in different cultural groups it was recommended that each site include culturally relevant descriptions of activity and/or conduct pilot testing/focus groups for target populations

 Goal 5: Identify instruments for triangulation of measures to increase validity of the findings/measures. (Participants agreed to revise goal to read instruments, not measures.)

Discussion: Projects exchanged practical ideas about how to triangulate or validate measures. Nearly all sites are incorporating multiple measures and/or validation with objective activity data or fitness measures. It was recommended that sites report and analyze outcomes from multiple measures if available. Because measurement error may be less with some objective measures, it may be possible to conduct analyses on subsamples of the population and not sacrifice power.

 Goal 6: Develop a repository of all measures of physical activity used across studies and determine the strengths and weaknesses of each measure.

Discussion: Web site include descriptions of each of the instruments and contact information for further details. Copies of instruments are posted if available. Hard copies of all instruments have been compiled by Terry and are available upon request.

Goal 7: Develop estimates of costs of each intervention for increasing physical activity/fitness levels.

Discussion: The goal reflects a more global effort of the BCC. Detailed data on costs for various treatments would be needed to examine cost effectiveness.