In this evaluation of the timing of improvements in health-related outcomes for participants in the EW program, we observed significant reductions in the percent of participants with each disability risk factor (i.e., depression and physical inactivity) over the first six months of the program. During the second six months, the percent that was physically inactive did not decrease further; however, the percent with depression continued to decrease between six and 12 months. Improvements in risk factor severity were significant for both depression and physical inactivity over the first six months of the program, with additional significant change in the second six months. Secondary outcomes showed significant improvement only during the first six months.
Other year-long health promotion programs have also reported greater improvements from enrollment to six months than from six-month to 12-month follow-up. For example, one study of algorithm-based individual counseling to improve self-management among patients with diabetes demonstrated that almost all of the reduction in hemoglobin A1C (HbA1C), a measure of blood glucose control, occurred during the first six months of the program – i.e., the maximum effect was attained by six months and was sustained through the twelfth month (Aubert et al., 1998
). Similarly, Gill and colleagues found that a six-month PREHAB program reduced disability, although no further gains were observed during months seven to 12 (Gill et al., 2002
; Gill et al., 2004
). A 48-week, cluster-randomized, controlled trial of Tai Chi provides another example of early improvement exceeding later improvement (Sattin, Easley, Wolf, Chen, & Kutner, 2005
). An intense Tai Chi exercise intervention reduced fear of falling among frail older adults with a history of falls (Sattin et al., 2005
) compared to a control group that received wellness education. Fear of falling was similar in intervention and control groups at baseline, and improved after four months in the Tai Chi group, but was not significantly different by study group until subsequent evaluations at months 8 and 12. Notably, in the Tai Chi intervention group, fear of falling improved more from baseline to four months than from four to 8 or 8 to 12 months.
Our finding that more benefits occurred early than late for most participants has important implications from the standpoint of both health promotion programs and the persons served by them. From the program standpoint, more clients can be served by a program of shorter duration, making it more valuable to potential funding partners (e.g., health care systems) in the communities where the program resides. From the participant standpoint, older adults who may be reluctant to join a year-long program may find a six-month program more attractive.
It should be noted that improvements in severity of disability risk factors continued over 12 months of program participation, both for participants who met criteria and for those who did not meet criteria for the disability risk factors we assessed. Improvements from enrollment to twelve months were clinically relevant for participants who met the criterion for physical inactivity at enrollment, and likely clinically relevant for those who met the criterion for depression at enrollment (). For example, going from a mean score of 2.6 to 4.6 on the PACE measure corresponds to a change in the average response from, “I am trying to start to exercise or walk” to an average response of, “I am doing moderate exercise less than 3 times per week.” With regard to depression, a decrease in the mean GDS score from 8.3 to 5.4 indicates less severe depressive symptoms. Thus, although the magnitude of improvements from 6 to 12 months may be smaller, one could argue that clients meeting criteria for physical inactivity or depression upon program entry may benefit from continued participation beyond 6 months. This information has implications for program implementers and for participants themselves: for example, program duration could be tailored to the needs of the individual participant, based on his/her scores on disability risk factors at the time of enrollment.
Our assessment of participants who remained physically inactive at six months suggested that depressive symptomatology may have limited their physical activity. However, about one-third of those participants did improve by 12 months. Although reducing the proportion of participants meeting risk criteria is an important goal, it is also important to acknowledge the value of any improvement that may prevent or slow the rate of physical decline to disability. That is, some individuals who did not improve might have worsened more in the absence of EW participation. Empirical evidence in support of this assertion is available in studies that have involved older adults as control participants: for example, data from the randomized trial of the EW intervention demonstrate that the cumulative incidence of disability in basic activities of daily living (e.g., bathing, dressing, transferring) in the control group was 10.5% at six months and 21.3% at 12 months (Phelan et al., 2004
The EW program is offered to the public at large, and thus participants are self-selected. The authors did not direct recruitment of program participants but simply analyzed the data that were collected from this real-world program (i.e., a program that is not being implemented as part of a research study). This study is limited by reliance on self-report data; however, the outcome measures are valid and reliable, and participants' responses ranged from minimum to maximum on each scale. Additional limitations that bear mention include 1) the lack of complete data for all 355 participants who initially enrolled in the program (295 of 355 or 83% had complete data), and 2) potential selection bias associated with the study of volunteers, in particular the possibility that results may not be generalizable to the overall population of all older adults.
The major strength of this study is that these results should be broadly generalizable to most community-dwelling older adults, because participants enrolled in a community-based health-promotion program and were not recruited for a research study. Another strength worthy of mention is the large number of participants who completed questionnaires at EW sites in several regions (i.e., Northeast, Midwest, West) of the United States.
The EW program may be even more beneficial for enhancing recovery of function among older adults after hospitalization or rehabilitation. Long-term follow-up of participants would be invaluable for determining whether health-related benefits are maintained and what the impact of minimizing disability risk factors may be on health care use and costs over several subsequent years. We conclude that the findings reported herein support a six-month enrollment period in EW for most participants.