Objective: To provide updated evidence-based recommendations for the preventive treatment of migraine headache. The clinical question addressed was: What pharmacologic therapies are proven effective for migraine prevention? Methods:The authors analyzed published studies from June 1999 to May 2009 using a structured review process to classify the evidence relative to the efficacy of various medications available in the United States for migraine prevention. Results and Recommendations:The author panel reviewed 284 abstracts, which ultimately yielded 29 Class I or Class II articles that are reviewed herein. Divalproex sodium, sodium valproate, topiramate, metoprolol, propranolol, and timolol are effective for migraine prevention and should be offered to patients with migraine to reduce migraine attack frequency and severity (Level A). Frovatriptan is effective for prevention of menstrual migraine (Level A). Lamotrigine is ineffective for migraine prevention (Level A). Neurology Epidemiologic studies suggest approximately 38% of migraineurs need preventive therapy, but only 3%-13% currently use it.1 In 2000, the American Academy of Neurology (AAN) published guidelines for migraine prevention.2,3 Since then, new clinical studies have been published on the efficacy and safety of migraine preventive therapies. This guideline seeks to assess this new evidence to answer the following clinical question: For patients with migraine, which pharmacologic therapies are proven effective for prevention, as measured by reduced migraine attack frequency, reduced number of migraine days, or reduced attack severity? This article addresses the safety and efficacy of pharmacologic therapies for migraine prevention.Separate guidelines are available for botulinum toxin. 4 The 2008 guideline included a Level B recommendation that botulinum toxin was probably ineffective for treatment of episodic migraine. A new guideline is in development. An updated guideline on nonsteroidal anti-inflammatory drugs 5 and complementary alternative treatments has been approved for publication as a companion to this guideline. 5 DESCRIPTION OF THE ANALYTIC PROCESSThe AAN and the American Headache Society participated in the development process. An author panel of headache and methodologic experts was assembled to review the evidence. Computerized searches of the MEDLINE, PsycINFO, and CINAHL databases identified new studies (published in English). The search strategy used the MeSH term "headache" (exploded) and a published search strategy for identifying randomized controlled trials ( Appendices e-1-e-5, reference e1, and tables e-1 and e-2 are available on the Neurology Web site at www.neurology.org.

Among persons with migraine in need of medical care (MIDAS Grade II or greater), only one quarter traversed the 3 steps we proposed to be necessary to achieving minimally appropriate care (consulting, diagnosis, and treatment/medication use). Health insurance status was an important predictor of consulting. Among consulters, women were far more likely to be diagnosed than men, suggesting that gender bias in diagnosis may be an important barrier for men. There were economic barriers related to use of appropriate prescription medications. Public health efforts should focus on improving consultation rates, particularly in the uninsured and diagnostic rates particularly in males with migraine.