Factors related to adherence to an exercise program for healthy adults.

Healthy men (N = 33) and women (N = 73) participated in a 6-month exercise program three mornings per week, and their attendance scores (percent of total classes attended) were related to a variety of physiological, anthropometric, psychological, and demographic variables which were studied. These subjects were also grouped by adherence patterns; 18% attended less than 10% of the classes (early dropouts = EDO), 40% attended between 10 and 50% of the classes (nonadherers = NAd), and 42% attended more than 50% (adherers = Ad). Correlation coefficients between all of the variables and attendance were low. However, certain patterns did emerge. Those who continued the program more than 10% of the sessions tended to be the more physically fit women and less physically fit men. The EDO men and women were more likely to 1) have less stability in the community (less time at present address or occupation), 2) be single, and 3) have no children. Self-motivation scores (SMI) for EDO men were significantly lower, but the correlation between SMI and attendance for all subjects was only r = 0.052. "Blue-collar" men had a greater-than-expected dropout rate; however, educational level did not affect adherence. Health care behavior (including smoking) and previous exercise patterns did not affect attendance. Eleven variables that were related to adherence were selected for further study. The predictive values and sensitivities for these variables ranged from 47-85% and 15-62%, respectively. Using criteria of multiple positive scores did not improve the ability to predict attendance behavior. It was concluded that for healthy volunteers, participant characteristics are not good predictors of compliance to an exercise regimen.