Vermont Center on Behavior and Health, University of Vermont, Burlington, VT, USA.
Oxford Neurology Clinic, Oxford, MS, USA.
Headache. 2018 May;58(5):648-660. doi: 10.1111/head.13294. Epub 2018 Mar 9.
Although risk factors for medication-overuse headache have been identified within the general population, most studies have neglected clinical samples. The present study examined the relative and combined associations of these factors with medication-overuse headache in a sample of US adults seeking treatment for primary headache disorders.
Treatment-seeking headache patients provided data on demographics, headache variables, psychiatric variables, use of headache medications, and use of other prescription medications and substances. A classification tree selection strategy was utilized within this cross-sectional study to differentiate between those with and without medication-overuse headache, and a final multivariable model assessed their combined utility.
Forty-three of 164 participants (26.2%) met diagnostic criteria for medication-overuse headache. Relative to non-medication-overuse headache participants, participants with medication-overuse headache reported greater headache-related disability (odds ratio = 1.09, 95% confidence interval = 1.01-1.18), escape and avoidance responses indicative of fear of pain (odds ratio = 1.07, 95% confidence interval = 1.00-1.15), and use of combination medications for headache (odds ratio = 3.10, 95% confidence interval = 1.51-6.36). The final multivariable model differentiated well between the 2 groups (area under the receiver operating characteristic curve = .78; 95% confidence interval = .71-.86).
Items that assess headache-related disability, use of combination medications, and fear of pain help identify patients who are currently overusing acute headache medications and may serve as indicators of treatment progress. Future studies should apply similar analytic approaches longitudinally to identify headache sufferers at risk for medication-overuse headache prior to headache progression.
尽管已经在普通人群中确定了药物过度使用性头痛的危险因素，但大多数研究都忽略了临床样本。本研究在寻求原发性头痛障碍治疗的美国成年人样本中，检查了这些因素与药物过度使用性头痛的相对和综合关联。
治疗性头痛患者提供了人口统计学、头痛变量、精神科变量、头痛药物使用情况以及其他处方药物和物质使用情况的数据。本横断面研究中使用分类树选择策略来区分有和没有药物过度使用性头痛的患者，并使用最终的多变量模型评估它们的综合效用。
在 164 名参与者中，有 43 名（26.2%）符合药物过度使用性头痛的诊断标准。与无药物过度使用性头痛的参与者相比，有药物过度使用性头痛的参与者报告了更大的头痛相关残疾（比值比=1.09，95%置信区间=1.01-1.18）、逃避和避免疼痛的反应（比值比=1.07，95%置信区间=1.00-1.15）以及使用组合药物治疗头痛（比值比=3.10，95%置信区间=1.51-6.36）。最终的多变量模型很好地区分了这两组（受试者工作特征曲线下面积=0.78；95%置信区间=0.71-0.86）。
评估头痛相关残疾、使用组合药物和对疼痛的恐惧的项目有助于识别目前过度使用急性头痛药物的患者，并可能成为治疗进展的指标。未来的研究应该采用类似的分析方法进行纵向研究，以便在头痛进展之前识别出有药物过度使用性头痛风险的头痛患者。
