BEHAVIORAL SKILLS AND PHYSICAL ACTIVITY OUTCOMES DURING AN 18-MONTH PHONE/MAIL INTERVENTION FOR RECENT COLLEGE GRADUATES

1318 Data from an 18-month phone and mail intervention, Project GRAD (Graduate Ready for Activity Daily), were examined to determine to what extent use of behavior change skills explains change in physical activity outcomes. An innovation of this study was assessing use of skills contemporaneously rather than retrospectively. Recent college graduates (N=170; M age=24.1; 51% female) who had participated in the semester-long GRAD intervention were included in this follow-up study. Behavioral skills included goal setting, decisional balance, self-talk, stress management, time management, enjoyment, social support, activism, and relapse prevention. Use of skills was measured by structured telephone interviews throughout the 18-month intervention. Strength, vigorous, and moderate intensity activities, as well as total physical activity, were assessed using the 7-Day PAR (Blair, 1984). Women reported using a significantly greater number of skills than men (p<.01); specifically, more positive self-talk, decisional balance, enjoyment, social support, and relapse prevention (p<.05). Regressions revealed that women who reported more goal setting and relapse prevention reported more vigorous activity (p<.05). For women, strength, moderate, and total activity were not significantly predicted by the total number of skills used or any specific skill. Men who reported increasing enjoyment reported more moderate intensity activity (p<.05). Surprisingly, men who reported using more relapse prevention reported less vigorous activity (p<.01), and men who reported using more stress management reported less strength activity (p<.05). Results indicate that women may use more behavioral skills than men, and these skills differentially predict physical activity outcomes. This may have important implications for tailoring physical activity interventions for men and women, and may be useful for identifying which components of an intervention may be most related to physical activity outcomes. Supported by NIH #HL49505.