TY - JOUR
T1 - Adults' Physical Activity Patterns Across Life Domains
T2 - Cluster Analysis With Replication
AU - Rovniak, Liza S.
AU - Sallis, James F.
AU - Saelens, Brian E.
AU - Frank, Lawrence D.
AU - Marshall, Simon J.
AU - Norman, Gregory J.
AU - Conway, Terry L.
AU - Cain, Kelli L.
AU - Hovell, Melbourne F.
PY - 2010/9
Y1 - 2010/9
N2 - Objective: Identifying adults' physical activity patterns across multiple life domains could inform the design of interventions and policies. Design: Cluster analysis was conducted with adults in two U.S. regions (Baltimore/Washington, DC, n = 702; Seattle, WA [King County], n = 987) to identify different physical activity patterns based on adults' reported physical activity across four life domains: leisure, occupation, transport, and home. Objectively measured physical activity, and psychosocial and built (physical) environment characteristics of activity patterns were examined. Main Outcome Measures: Accelerometer-measured activity, reported domain-specific activity, psychosocial characteristics, built environment, body mass index. Results: Three clusters replicated (κ =90-93) across both regions: Low Activity, Active Leisure, and Active Job. The Low Activity and Active Leisure adults were demographically similar, but Active Leisure adults had the highest psychosocial and built environment support for activity, highest accelerometer-measured activity, and lowest body mass index. Compared to the other clusters, the Active Job cluster had lower socioeconomic status and intermediate accelerometer-measured activity. Conclusion: Adults can be clustered into groups based on their patterns of accumulating physical activity across life domains. Differences in psychosocial and built environment support between the identified clusters suggest that tailored interventions for different subgroups may be beneficial.
AB - Objective: Identifying adults' physical activity patterns across multiple life domains could inform the design of interventions and policies. Design: Cluster analysis was conducted with adults in two U.S. regions (Baltimore/Washington, DC, n = 702; Seattle, WA [King County], n = 987) to identify different physical activity patterns based on adults' reported physical activity across four life domains: leisure, occupation, transport, and home. Objectively measured physical activity, and psychosocial and built (physical) environment characteristics of activity patterns were examined. Main Outcome Measures: Accelerometer-measured activity, reported domain-specific activity, psychosocial characteristics, built environment, body mass index. Results: Three clusters replicated (κ =90-93) across both regions: Low Activity, Active Leisure, and Active Job. The Low Activity and Active Leisure adults were demographically similar, but Active Leisure adults had the highest psychosocial and built environment support for activity, highest accelerometer-measured activity, and lowest body mass index. Compared to the other clusters, the Active Job cluster had lower socioeconomic status and intermediate accelerometer-measured activity. Conclusion: Adults can be clustered into groups based on their patterns of accumulating physical activity across life domains. Differences in psychosocial and built environment support between the identified clusters suggest that tailored interventions for different subgroups may be beneficial.
UR - http://www.scopus.com/inward/record.url?scp=77956903121&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77956903121&partnerID=8YFLogxK
U2 - 10.1037/a0020428
DO - 10.1037/a0020428
M3 - Article
C2 - 20836604
AN - SCOPUS:77956903121
VL - 29
SP - 496
EP - 505
JO - Health Psychology
JF - Health Psychology
SN - 0278-6133
IS - 5
ER -