Interventions for fatigue in Parkinson's disease: A systematic review and meta-analysis

The authors sought to review the efficacy of interventions for fatigue in Parkinson's disease. A search was conducted of PubMed, Cinahl, Psychinfo, EMBASE, and Web of Knowledge up to November 2013. Methodological quality was assessed using the PEDro scale. For meta-analyses, studies were weighted on variance. Effect sizes were calculated with 95% confidence interval (CI); overall effect was presented by means of a Z-score; heterogeneity was investigated using the I(2) . Fourteen articles (n = 1,890) investigating drugs and behavioral therapy were eligible. Ten studies demonstrated excellent, three good, and one fair methodological quality. Three articles (investigating amphetamines) were appropriate for meta-analysis, which was performed according to scales used: Multidimensional Fatigue Inventory: mean difference, -6.13 (95%CI: -14.63-2.37, Z = 1.41, P = 0.16; I(2) = 0); Fatigue Severity Scale: mean difference, -4.00 (95%CI: -8.72-0.72, Z = 1.66, P = 0.10; I(2) = 0). Currently insufficient evidence exists to support the treatment of fatigue in PD with any drug or nondrug treatment. Further study is required.