Background/Purpose

Outside of a clinical trial setting, higher outcome expectations for exercise, more education, and greater physical activity are inter-related, and outcome expectations may partly explain why education level is associated with physical activity. In clinical trials with exercise interventions, where participants tend to have higher outcome expectations, it would be informative to know if these associations still hold true. Therefore, we assessed if higher education is associated with higher outcome expectations and if outcome expectations mediate the relationship between education and physical activity.

Methods

We conducted a secondary analysis of all baseline data from a randomized trial comparing Tai Chi and Physical Therapy among people with knee osteoarthritis (KOA) as defined by the ACR criteria. Participants completed the Outcome Expectations for Exercise (OES), CHAMPS Activities Questionnaire for Older Adults (CHAMPS), and demographics self-report questionnaires. The OES is a 9-item questionnaire, in which higher mean scores indicate higher expectations for exercise with a range from 1 to 5, CHAMPS is a 40-item questionnaire that measures physical activity, and education was categorized into the following: high school diploma or less, some college/trade school, college degree, graduate school.

We performed an analysis of variance (ANOVA) with OES as the dependent variable to explore the differences among education groups with graduate school as the reference group. We also performed a multivariable analysis to adjust for age and sex.

To examine the role of OES as a mediating factor between education and physical activity, we performed a mediation analysis between education and physical activity with OES as the mediator using a commonly accepted method (Baron & Kenney).

Results

Our analysis included data from 282 participants with an average age of 59.7 years (SD=10.4); 69% female, 51% white, 35% black, 29% working at least part time, 72% completed at least some college, and an average body mass index of 32 kg/m². In the ANOVA, participants with high school or less had lower OES than participants who completed graduate school (Table 1). Higher education attainment was associated with increased OES after adjusting for age and sex (p=0.002). Criteria were not met for the mediation analysis because education did not predict physical activity and OES did not predict physical activity when controlling for education.

Conclusion

Education is associated with OES; however, OES is not a mediator between education and physical activity among individuals with KOA who are enrolled in a clinical trial with an exercise intervention. The lack of mediation may be in part because the participants recruited into this trial had higher OES and most had at least some college education, therefore were not reflective of the general population.