Tracing transformation: chronic migraine classification, progression, and epidemiology.

Migraine attacks sometimes increase in frequency over time. Headache experts conceptualize this process with a model that envisions transition into and out of four distinct states: no migraine, low-frequency episodic migraine (<10 headaches per month), high-frequency episodic migraine (10-14 headaches per month), and chronic migraine (CM, >or=15 headaches per month). Transitions may be in the direction of increasing or decreasing headache frequency and are influenced by specific risk factors. Overall, population studies estimate that patients who have low-frequency episodic migraine or high-frequency episodic migraine will transition to CM at the rate of about 2.5% per year. Two longitudinal population studies, the Frequent Headache Epidemiology study and the ongoing American Migraine Prevalence and Prevention (AMPP) study provide longitudinal population data that has defined the rates of and risk factors for transition. Launched in 2004, the AMPP study has followed a sample of >10,000 migraine sufferers annually for 4 years. Cross-sectional data from the Frequent Headache Epidemiology study and the AMPP study show that patients with chronic daily headaches have lower levels of education and household income. In addition, epidemiologic profiles show that CM sufferers tend to be older and have higher body mass indexes. These studies have also assessed a number of potential risk factors associated with the transition to CM. These include baseline high attack frequency, obesity, stressful life events, snoring, and overuse of certain classes of medication. In particular, opiate and barbiturate combination products contribute to migraine progression, and nonsteroidal anti-inflammatory agents are protective in patients with <10 headache days per month. The influence of medication is modified by both headache attack frequency and frequency of medication use. Although depression and anxiety are associated with an increased risk of new-onset CM, the influence of depression is accounted for by migraine disability assessment scale score, whereas the effect of anxiety may be independent of migraine disability assessment scale score. Emerging data on the longitudinal risk of CM suggest that, in a population at risk, CM may be a preventable disorder.