Potential predictors for the efficacy of non-steroidal anti-inflammatory drugs in patients with migraine

Abstract

Objectives: To explore potential predictors of the efficacy of non-steroidal anti-inflammatory drugs (NSAIDs) in patients with migraine. Methods: Consecutive migraine patients were recruited and divided into responders and non-responders to NSAIDs according to follow-up for at least three months. Demographic data, migraine-related disabilities and characteristics, and psychiatric comorbidities were evaluated and used to build multivariable logistic regression models. Subsequently, we generated receiver operating characteristic (ROC) curves to explore the performance of these traits in predicting NSAIDs efficacy. Results: A total of 567 patients with migraine who completed at least three months of follow-up were enrolled. In the multivariate regression analysis, five factors were identified as potential predictors for NSAIDs efficacy in treating migraine. Namely, attack duration (odds ratio (OR) = 0.959; p < 0.001), headache impact (OR = 0.966; p = 0.015), depression (OR = 0.889; p < 0.001), anxiety (OR = 0.748; p < 0.001), and education level (OR = 1.362; p < 0.001) were associated with response to NSAIDs treatment. The area under the curve, sensitivity, and specificity combining these five factors for predicting the efficacy of NSAIDs were 0.834, 0.909 and 0.676. Conclusions: These findings suggest that migraine-related and psychiatric factors are associated with the response to NSAIDs in migraine management. Identifying such key factors may help to optimize individualized migraine management strategy.

Keywords

• Migraine
• Non-steroidal anti-inflammatory drugs
• Efficacy
• Logistic regression
• Receiver operating characteristic curve