Background: Insomnia is an often seen primary health care problem. Valerian might be an alternative treatment with fewer secondary effects. The aim of this study is to evaluate its effectiveness on insomnia through a meta-analysis of published literature.
Methods: Search for randomized clinical trials (RCTs) of Valerian preparations compared with a placebo on Medline, the Cochrane Library, Embase and Biosis.
Outcomes: sleep-quality improvement (SQ, yes/no), sleep-quality improvement quantified through visual analogical scales (SQS) and the latency time (LT) in minutes until getting to sleep. Three meta-analyses were carried out using inverse-variance weighted random effects models. Heterogeneity was determined with the Q-statistic and was explored through a sub-groups analysis. Publication bias was evaluated using the funnel plot.
Results: Eighteen RCTs were selected; eight had a score of 5 on Jadad's scale. The mean differences in LT between the Valerian and placebo treatment groups was 0.70 min (95% CI, -3.44 to 4.83); the standardized mean differences between the groups measured with SQS was -0.02 (95% CI, -0.35 to 0.31); treatment with Valerian showed a relative risk of SQ of 1.37 (95% CI, 1.05-1.78) compared with the placebo group. There was heterogeneity in the three meta-analyses, but it diminished in the sub groups analysis. No publication bias was detected.
Conclusion: The qualitative dichotomous results suggest that valerian would be effective for a subjective improvement of insomnia, although its effectiveness has not been demonstrated with quantitative or objective measurements. We recommend future investigations oriented toward improving insomnia with other more promising treatments.
Copyright 2010 Elsevier B.V. All rights reserved.