Abstract
One of the major achievements in headache management and research over the past decades is the classification of headache disorders by the International Headache Classification Committee (1,2), allowing standardized assessments of specific headache disorders in various clinical and research settings. Many features and clinical presentation of headaches are included in the criteria and specific algorithms can help clinicians and researches alike to derive a diagnosis. These criteria have also improved the quality of large-scale, population-based, headache studies as the criteria can be implemented in standardized questionnaires (3,4) or telephone interviews (5,6).
While the criteria of main headache forms remain unchanged, specific aspects of the headache classification criteria are revised over time as our understanding of the pathophysiology, clinical presentation, and treatment of headaches subforms advances. The third version of the International Classification of Headache Disorders is expected to appear soon and will provide the newest classification algorithms. For some specific headache forms, however, there is a more profound discussion about the right approach for classification. For example, several definitions for ‘chronic migraine’ exist and new approaches to cover high-frequency headaches have been proposed (7). But, which one should be used? Which one should be implemented when studying the burden of highly frequent headache on the population level?
In this issue of Cephalalgia, Katsarava and colleagues (8) address the important questions: (i) what happens if different classifications of chronic migraine are applied to a population-based setting; (ii) how different are these groups with regard to personal characteristics; and (iii) do these groups differ when compared to patients with episodic migraine of various headache frequencies? They applied a self-administered, standardized questionnaire to a large, population-based sample from Germany covering three chronic migraine definitions: (i) given by the American Headache Society in 2005 (9); (ii) the revised classification given by the International Headache Classification Committee in 2006; (10) and (iii) the so-called Silberstein and Lipton criteria for transformed migraine from 1994 (11). The prevalence of chronic migraine varied substantially from 0.4% (American Headache Society criteria) to 2.0% (Silberstein and Lipton criteria) among the 9350 responders. However, the three groups did not differ according to age, gender, body mass index, alcohol consumption or education. In contrast, many characteristics differed to high- or low-frequency episodic migraine groups.
While some limitations of the study, such as limited power to compare personal characteristics in subgroups, should caution the interpretation, the study provides important information for clinic- and population-based headache research. It is evident, and certainly not surprising, that by applying different criteria to define chronic migraine, somewhat different groups are identified. As the smaller group was a subset of the larger group and the overall personal characteristics did not differ between the three chronic migraine groups, it may be speculated that the various criteria identify subfeatures of chronic migraine but that all patients also had common features. It is further of note that patients with episodic high-frequency migraine had different personal characteristics, suggesting that specific risk factors are determining common chronic migraine features no matter which criteria are applied.
In clinical practice, a few key features will guide a physician to diagnose a migraine or specifics of migraine. As the clinical presentation of migraine is very heterogeneous and changing over time, specific features necessary to fulfil all migraine criteria may be present at one time point but not at another. Introducing the category of probable migraine by the second revision of the International Headache Classification Committee was an improvement for population-based headache research as this category allows some degree of heterogeneity and thus may allow better comparisons between studies.
Applying more strict criteria may help to identify specific subgroups, which are different in some aspects but common features may also exists, as the data by Katsarava and colleagues suggest. While complex diagnostic criteria may have a specific research utility, they may not help to distinguish further risk factor patterns on the population level. Thus, applying too restrictive criteria will make it impossible to study the impact of chronic migraine in the general population and to evaluate the public health consequences. Similarly, too restrictive criteria may hamper practical utilization in general clinical practice, where most patients with headache are seen. It also remains unclear whether complex subcategorization will improve the management of treatment of patients. In general, it is likely that patients with similar broad headache characteristics benefit from similar treatment approaches. Future research will help to unveil which headache features are sufficient to define common risk factor patterns and whether these will translate in modified treatment options.