Type of Document	Dissertation
Author	Cox, Janet Gray
URN	etd-10292007-093659
Title	Is Exercise an Evidence-Based Intervention for Clinical Depression in Older Adults: A Meta-Analysis of Randomized Studies 2000-2006
Degree	Doctor of Philosophy
Department	Social Work, College of
Advisory Committee	Advisor Name Title Bruce A. Thyer Committee Chair Charles C. Ouimet Committee Member Charles R. Figley Committee Member
Keywords	Older Adults Elderly Physical Activity Depression Exercise Randomized Evidence-Based
Date of Defense	2007-09-19
Availability	unrestricted
Abstract ABSTRACT Promoting physical activity is consistent with many models and perspectives on social work, such as the empowerment perspective, self-efficacy theory, resilience, the strengths perspective, good health, hardiness, and self-determination, yet it has received little attention in social work literature as an effective intervention. Among older adults, exercise has been shown to reduce costs for health care, limit injury, decrease heart disease and obesity, improve diabetes, and result in a more active life style. Moreover it has been shown to elevate mood. Recent neuroimaging studies indicate that exercise expresses neural functions similar to anti-depressant medication without the side effects. Findings of this meta-analysis found a significant overall pretest-posttest mean change effect size of -0.46 (p<.01), adding to a preponderance of evidence that exercise can prevent or remit symptoms of major depression—a prevalent global disorder, a major contributor to a massive burden of disease (CDC, 2004), and a disorder seen by almost every venue of social work. This meta-analysis included randomized experimental studies of exercise and follow-up with clinically depressed samples of older adults conducted 2000-2006. Clinical depression was defined according to DSM-IV-TR and ICD-10 criteria. Twenty-six independent samples describing nine studies met criteria. The null hypothesis proposed exercise would have no effect upon major depression, as defined by statistically significant differences and meaningful effect sizes. A comprehensive literature search was conducted and features of each study were coded independently by two coders, with an inter-rater reliability of 0.97. Analyses looked at overall effect, group of assignment, duration, treatment studies and follow-up, outcome measures, type of exercise, age: young-old versus old-old, and standards and protocols used to determine clinical effectiveness. All effect sizes were computed as standardized pretest-posttest mean-change measures. Weighted analyses adjusted for variable sample size. In the fixed effects model analysis, each effect size was weighted by its inverse variance; the random effects model employed a method-of-moments estimator. The significant overall pretest-posttest mean change effect size was a moderate -0.46 (p<0.0001) for treatment versus control and comparison group effects. This means that on average older adults with clinical depression who exercised had -0.46 of a standard deviation reduction in clinical depression from baseline compared to individuals who did not exercise. The size of the effect is consistent with findings of related meta-analyses. Treatment versus control (-0.46, p<.01) showed a larger effect size than a non-significant treatment versus comparison group (-.18) effect. Exercise was as effective as other standard treatment with reduction in levels of depression in all treatment groups and little differences among groups. Effect sizes were larger (-0.52, p<.01) during the treatment phase then follow-up (-0.30, p<.01); follow-up showed an effect from continued exercise up to two years. The clinician-administered Hamilton Rating Scale for Depression (-0.73, p<.01) and the Geriatric Depression Scale (-0.46, p<.01) showed a larger effect than a non-significant finding from the Beck Depression Inventory (-0.19). The smaller effect was thought to reflect a confound from calculation of control and comparison samples as an amalgam. No group differences were found for type of exercise: aerobic (-0.35); resistance (-0.45); or combined (-0.69); all treatment mean change effects were significant at <0.01 alpha level. Old-old adults (-0.49, p<.01) were as likely to respond to exercise as young-old adults (-0.40, p<.01). The high standard of design requirements may have elicited slightly more conservative effect sizes; randomization and diagnostic constituency were held constant. All effect sizes in design investigation were significant; no differences were found between those studies with and without the intent to treat or blinding features. The role of design features in meta-analysis needs further investigation. Both specific and overall findings among the studies of the analyses suggest that like cardio protection, a lifestyle change of ongoing regular moderate exercise can effect remission from clinical depression in older adult clients and protect from relapse. Findings indicate that efforts by social work researchers and practitioners are worthwhile at the micro, mezzo, and macro levels to provide professional pre-education, develop exercise interventions within practice, educate clients and others about the antidepressant nature of exercise, promote accessibility, and advocate community and regulatory improvement. Other studies suggest a neural link between exercise and remission of depression, and further intervention research should include interdisciplinary teams drawing from venues of social work, neuroscience, medicine, and exercise to explain the causal mechanism, identify the populations at risk, and investigate the prophylactic effect.
Files	Filename       Size       Approximate Download Time (Hours:Minutes:Seconds)     28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access    CoxJFall2007.pdf 1.67 Mb 00:07:44 00:03:59 00:03:29 00:01:44 00:00:08