Current migraine forecasting models represent an important first step in accurately predicting future headache activity, according to a recent investigation. However, to utilize these models in a preemptive treatment paradigm where the risk of headache is treated prior to the actual experience of pain, these models must achieve greater precision with good calibration and generate predictions that are clinically actionable by individuals in their real-time home environments.
A substantial pool of candidate migraine trigger factors could be considered in the creation of forecasting models. However, since mechanistic information about causal factors that precede a migraine attack is not well understood, and such factors are difficult to measure, empirical models that are based on trigger factors that are merely associated with the onset of headache activity are likely to be the focus of forecasting efforts. Of such factors, stress has considerable empirical support and has been used to successfully forecast future headache attacks within individuals over time. At present, however, existing models possess only modest levels of discrimination and lack strong resolution in generated predictions.
Curr Pain Headache Rep . Forecasting migraine attacks and the utility of identifying triggers. 2018;22:62. doi:10.1007/s11916-018-0715-3.