Predictors of increased physical activity in the active for life program

Introduction: Targeting of evidence-based programs can be improved by knowing who benefits least and most. We examined pretest predictors of increased physical activity among participants enrolled in Active for Life. Methods: Participants (N = 1,963) from 9 community-based organizations took part in a 6-month telephone-based or a 20-week group-based behavioral physical activity program and completed a pretest survey; 1,335 participants returned posttest surveys. Interactions tested whether increases in physical activity differed over time, according to baseline characteristics. Results: In the telephone-based program, participants who were younger and less active at pretest and those who had higher pretest social support showed greater intervention effects. In the group-based program, younger participants, those less active at pretest, women, Hispanics/Latinos, heavier participants, and those who reported more health conditions and osteoporosis showed greater intervention effects. Conclusions: Participant response to the 2 programs varied by age, baseline activity level, and other factors. For 6 of the 8 variables associated with differential outcomes, the least active group improved the most, which suggests that the programs worked especially well for participants most in need. Participants who were older than 75 years (both groups) and those who reported lower physical activity social support (in the telephone-based program) on entry did not respond as well and may require alternative or more intensive intervention strategies.