Migraine Progression: A Systematic Review

Background Migraine is a common and often debilitating neurological disease. It can be divided into episodic and chronic subforms based on the number of monthly headache days. Because only a subset of individuals with episodic migraine (EM) progress to chronic migraine (CM) over any given time period, understanding the factors that predict the new onset of CM or “migraine progression” may provide insights into the mechanisms, pathophysiology, prevention, and treatment of CM. In this review, we identify and summarize studies that report risk factors associated with the new onset of CM or related chronic headache diagnoses, group these risk factors and report the strength of evidence for the identified risk factors. Objective To conduct a systematic review of studies that identify risk factors for the new onset of CM or related chronic headache diagnoses such as transformed migraine (TM) and chronic daily headache (CDH). Methods Herein we summarize the findings of studies of risk factors associated with the new onset of CM/TM, CDH, or related diagnoses from the English language literature published before March 2018. The PubMed database was searched for relevant studies. Longitudinal studies with follow‐up data and case‐control studies were included in this qualitative synthesis. We report methodology, analytic criteria, and results for each manuscript and for the parent study. Next, we review the strength of evidence for each of the identified risk factors using a modified version of AB Hill’s criteria for causation and rank evidence as fair, moderate, or strong. We categorized risk factors as nonmodifiable, modifiable and based on putative mechanisms. We further categorized risk factors into sociodemographics, lifestyle factors and habits, headache features, comorbid and concomitant diseases and conditions and pharmacologic treatment‐related. Finally, we review theories of the pathophysiology underlying the development of new onset chronic migraine or increasing attack frequency. Results The PubMed search yielded 1870 records after duplicates were removed. Nine additional records were identified through expert consultation and other methods (eg, citations found as references in manuscripts identified in the literature review and through communication with the authors of manuscripts included in the review). The 1879 manuscripts were screened against the inclusion and exclusion criteria and 109 were found to be potentially eligible. Of 109 full‐text articles