| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
RESEARCH ARTICLE |
a Department of Occupational Science and Occupational Therapy, University of Southern California, Los Angeles
b Statistical Consultation and Research Center, Department of Preventive Medicine
c Division of Geriatric Medicine, Department of Medicine, University of Southern California Keck School of Medicine, Los Angeles
d Department of Pharmaceutical Economics and Policy, University of Southern California School of Pharmacy, Los Angeles
Stanley P. Azen, Statistical Consultation and Research Center, Department of Preventive Medicine, University of Southern California Keck School of Medicine, 1540 Alcazar CHP 218, Los Angeles, CA 90089-1090 E-mail: sazen{at}usc.edu.
Decision Editor: Toni C. Antonucci, PhD
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
THE number of elderly Americans has increased dramatically in recent years, a trend expected to escalate in the coming decades (Rowe and Kahn 1998
). Because older individuals are at disproportionate risk for chronic disease, functional decline, psychiatric disorder, and other health-related problems (Gatz 1995; Murrell and Himmelfarb 1989), it is critical for society to identify viable interventions that prevent age-related declines in health and functioning. Otherwise, our nation may be faced with an insurmountable health care burden (Gatz 1995).
The Well Elderly Study was a randomized clinical trial conducted from 1994 to 1996 to evaluate the efficacy of preventive occupational therapy (OT) intended to reduce health-related declines among urban, multiethnic, independent-living older adults (Clark et al. 1997; Jackson, Carlson, Zemke, Mandel, and Clark 1998). Significant benefits in health, function, and quality of life resulted from a 9-month OT intervention (Clark et al. 1997). After the conclusion of treatment, participants were followed for 6 months without further intervention and then reevaluated. In this article, we report on this follow-up assessment. We hypothesize that compared with the control groups, the long-term health of the study participants improved with preventive OT.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
]), the short form Geriatric Depression Scale (Sheikh and Yesavage 1986), LaRue Global Health Assessment (LaRue, Bank, Jarvik and Hetland 1979), and Tinetti Balance Examination (Tinetti 1986) was performed for each participant.
A total of 361 participants were recruited from two federally subsidized apartment complexes for older adults, located in or near Los Angeles. To maximize resources at the treatment sites, participants were recruited in two cohorts (143 in Cohort I and 218 in Cohort II) enrolled in the study at different times. The mean (±SD) age of the participants was 74.4 ± 7.4 years, and 65% of the participants were women. Ethnic group representations were Asian (47%), Caucasian (23%), African American (17%), Hispanic (11%), and other (2%). In the Asian group, 66% were tested in Mandarin (Azen et al. 1999). The majority (73%) of participants lived alone and 27% of the participants reported at least one disability. All participants signed an institutionally approved informed consent form prior to randomization.
Participants were randomized into three conditions: an OT treatment group, a generalized social activity control group, and a nontreatment control group. In both the OT treatment and generalized social activity groups, elderly adults engaged in weekly sessions involving 8–10 participants. The OT treatment was administered by registered occupational therapists and focused on helping the older adults to incorporate positive changes within their ongoing lifestyles. Topical foci included health-relevant behaviors, transportation, personal safety, social relationships, cultural awareness, and finances. The overriding therapeutic emphasis centered on achieving a careful understanding of each elder's unique pattern of personal attributes, values, goals, and in-context life circumstances and then working with the elder to design an individually tailored plan for implementing sustainable healthful changes. Methods of program delivery included didactic presentations, peer exchange, and direct experiences and personal exploration (in connection with occasional group outings or supplementary one-on-one therapist–client sessions). On the basis of theory and research in occupational therapy (e.g., Clark et al. 1991; Gauthier, Dalziel, and Gauthier 1987; Kielhofner 1992; Yerxa et al. 1989), the intervention was expected to benefit elderly participants' health and psychological well-being through (a) improving their specific health practices (e.g., exercise, use of joint protection techniques) and (b) increasing their general sense of purpose and meaning via engagement in personally meaningful activity.
In the generalized social activity control condition, older adults participated in craft projects, viewed films, went on community outings, played games, or attended dances. These activity sessions were led by nonprofessionals and were intended to control for activity engagement, social involvement, and general program participation. In the nontreatment control condition, older adults merely received the assessment battery in the absence of any intervention.
Participants were evaluated at baseline and after the treatment period using the RAND 36-item Short Form Health Survey (RAND SF-36; Hays, Sherbourne, and Mazel 1993; Ware and Sherbourne 1992; Ware, Snow, Kosinski, and Gandek 1993), Functional Status Questionnaire (FSQ; Jette and Cleary 1987), Life Satisfaction Index-Z (LSI-Z; Wood, Wylie, and Sheafor 1969), Center for Epidemiologic Studies—Depression Scale (CES-D; Radloff 1977), and the Medical Outcomes Study (MOS) Health Perception Scale (Stewart, Hays, and Ware 1988). Only participants in Cohort II were evaluated by the RAND SF-36.
Following the 9-month treatment phase, an intent-to-treat analysis of the questionnaire outcomes revealed a statistically significant benefit from OT for 10 measures: the FSQ quality of interaction, LSI-Z, MOS Health Perception, and RAND SF-36 bodily pain, physical functioning, role limitations due to health problems, vitality, social functioning, role limitations due to emotional problems, and general mental health scales (Clark et al. 1997). Subsequent to the treatment phase, participants were followed for an additional 6 months (follow-up phase) without further treatment and reevaluated on the outcome questionnaires.
Statistical Analysis
Because there were no statistically significant differences between the two control groups in either post-test (Clark et al. 1997) or follow-up outcomes, the control groups were combined for all analyses. Also, because no cohort main effect was found (Clark et al. 1997), data were analyzed for both cohorts combined. For each demographic and baseline history and physical examination variable, two-tailed tests were conducted for differences between participants with and without follow-up evaluations and between treatment groups (OT vs. combined control) for participants with follow-up evaluations.
For each outcome variable, treatment effects were examined by calculating signed change scores (long-term follow-up minus pre-test score) and then by using analysis of covariance to test for change score differences between the OT treatment group and the combined control group. Covariates included the variables previously found to be related to the change scores at the time of post-testing (Clark et al. 1997). To examine whether the results may have been affected by excluding participants with missing data, additional analyses were conducted in which regression analyses were used to impute values for missing scores (based on participants' post-test outcomes as predictors). Statistical testing was performed at the 0.05 alpha level, using one-tailed assessments to examine whether OT produced more positive mean change outcomes.
To directly determine the extent to which the OT-based benefits at the conclusion of therapy endured over time for the set of 10 measures associated with a significant OT effect at the time of post-testing, Cohen's effect size estimates were calculated (OT treatment group vs. control group, separately for both follow-up and post-test phases) using the adjusted change score means and standard deviations (Cohen 1988). The mean effect size for follow-up assessment was then divided by the post-test effect size mean to derive an overall percentage reflective of the degree to which the post-therapy OT-based gains were retained over the 6-month follow-up interval.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
80 years old (23%); 67% of the participants were women. Ethnic group representations were Asian (50%), Caucasian (20%), African American (17%), Hispanic (11%), and other (2%). The majority (73%) of participants lived alone, 26% of the participants reported at least one disability (the maximum number of reported disabilities was seven), 80% of participants scored good to excellent on the Tinetti Balance Examination, 90% were unimpaired on the M-MMSE, and 75% scored normal on the Geriatric Depression Scale. The average number of medications was three. We contrasted the demographic characteristics and history/physical examination results between the 285 participants with long-term follow-up and the 76 participants who dropped out of the study. Participants who dropped out had lower scores on the Tinetti Balance Examination (p = .04) and LaRue Global Health Assessment (p = .05) and reported taking more medications (p = .03). On average, participants with follow-up evaluations attended more OT and social control group sessions than participants without follow-up evaluations (44–52% vs. 10%, p < .01). We also contrasted demographic characteristics and history/physical examination results between the 96 OT participants and 189 control participants with 6-month follow-up data. No significant between-group differences were present.
Table 1 presents the results of the follow-up intent-to-treat analysis. We report the means of the unadjusted pre-test and 6-month follow-up scores, as well as the covariate-adjusted pre-test to follow-up change scores for each outcome variable. Analyses of covariance of pre-test to follow-up change scores revealed a significant benefit from OT for FSQ quality of interaction (p = .05) and for six of the eight SF-36 scales: physical functioning, role functioning, vitality, social functioning, role emotional, and general mental health (p < .05). We found marginally significant differences for the SF-36 scales bodily pain and general health (p < .10). Analyses based on imputing regression-based scores for missing values revealed a similar outcome, with five of the above seven variables remaining statistically significant beyond the .05 level; two variables were marginally significant at the .10 level.
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Positive follow-up treatment effects were most pronounced in the SF-36 variables, which have high ceilings and are therefore well suited for detecting differences among well elders. A further general tendency was for stronger effects to be present for psychosocial, as opposed to physical, outcome indices. For example, the most significant results were obtained for the SF-36 vitality, social functioning, role functioning, and general mental health scales, whereas only marginally significant or nonsignificant effects were found for SF-36 general health, SF-36 bodily pain, and MOS Health Perception. This basic pattern is consistent with the predominantly psychosocial nature of the intervention, which would be expected to more directly influence psychological health and vitality than physical health. This result is encouraging insofar as an earlier meta-analysis (Okun, Olding, and Cohn 1990) observed that the effects of various interventions on older adults' psychological well-being typically dissipate rapidly over time.
Although not part of the main data analysis, direct comparisons revealed a superior outcome for OT relative to the generalized social activity condition. Across the 10 variables that differentiated OT from the combined control groups at post-test, the mean effect size of direct follow-up comparisons between OT and the generalized social activity condition was 0.33, with six of the SF-36 outcomes significant beyond the .05 level. This result underscores that it is not activity per se that increases health and well-being. Rather, in connection with the character of the OT intervention, activity that is personally meaningful and contextually anchored within elders' everyday lives has the greatest capability to enhance health-related outcomes.
The observation of a durable effect for OT is consistent with the intent of treatment, which was to enable the older adults to permanently embed health-promoting changes into their daily routines on a longstanding basis (Carlson, Clark and Young 1998; Jackson et al. 1998). We speculate that the individualized emphasis of the treatment played an important role in this regard. By considering each elder's personal concerns, values, and environmental resources and limitations, we intended to foster changes that were both intrinsically motivated and contextually feasible within the participant's life, factors jointly conducive to the potential for an enduring therapeutic effect.
Additional research needs to be conducted to evaluate the efficacy of OT with different elderly populations, treatment settings, and logistical approaches to treatment administration. Further, given the intent of the OT program to induce longstanding healthful lifestyle changes, it would be desirable to incorporate a longer follow-up interval in future studies.
|
Acknowledgments |
|---|
Received for publication September 23, 1999. Accepted for publication May 26, 2000.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
