Effects of a computer feedback treatment and behavioral support protocol on drop out from a newly initiated exercise program.

Drop out from newly initiated exercise regimens is problematic. Three treatments intended to support new exercise programs (standard exercise counseling, computer feedback, and behavioral support) were tested to estimate their association with drop out over the initial 3 and 6 mo. Data from a total of 1,336 adults (44% men; M(age) = 41.9 yrs., SD = 9.8) initiating exercise programs at 18 community exercise facilities (six in each of the three treatment groups) were aggregated by facility on a measure of drop out and were then contrasted both within and between groups. After a mixed model repeated-measures analysis of variance was significant overall, planned contrasts were conducted using the Tukey-Kramer test. Significantly less drop out was found at Month 3 for both the computer support (27%) and behavioral support (17%) groups, when contrasted with the standard exercise counseling group (38%). Behavioral support had significantly less drop out than computer feedback at Month 3. The behavioral support group showed significantly less drop out at Month 6 (33%) when contrasted with both the standard exercise counseling (58%) and computer feedback (52%) groups, which did not significantly differ from one another. Limitations and the need to evaluate and extend research on interventions for reducing drop out from exercise programs were discussed.