Neurologists and patients may be able to predict migraine attacks using a model based on the level of stress from daily hassles, according to research published in the July issue of Headache. The model was well calibrated, but its forecasts were based on participants’ base rates of headache, said the authors. With additional adjustments, the model could enable patients to treat migraine attacks pre-emptively.
Although headache disorders are common, it remains unclear what triggers a migraine attack. Patients have identified many possible triggers, including perceived stress. In people with episodic migraine and chronic migraine, perceived stress is associated with the onset of headache. Researchers previously had not provided evidence that any of the potential triggers could predict a migraine attack, however.
Electronic Diaries Captured Headache Frequency
Timothy T. Houle, PhD, Associate Professor of Anesthesia at Massachusetts General Hospital in Boston, and colleagues conducted the prospective Headache Prediction Study to examine precipitating factors of migraine headache. They recruited participants with episodic migraine who had more than two headache attacks per month and had between four and 14 headache days per month. Secondary headache disorder and change in the nature of headache symptoms in the previous six weeks were among the exclusion criteria.
Timothy T. Houle, PhD
Participants completed morning and evening diary entries daily using electronic systems. In the entries, the participants recorded headaches, headache characteristics, and abortive medications used since the last entry. Participants used the Daily Stress Inventory to assess stress in their evening diary entries. Using these assessments, the investigators examined the frequency of stressors, the sum of the stress impact ratings, and the average stress impact ratings. The primary analysis was the prediction of a future headache attack based on current levels of stress and headache.
Potential for New Treatment Strategies
Dr. Houle and colleagues enrolled 100 participants between September 2009 and May 2014. Five participants dropped out. Approximately 91% of participants were female, and 87% were Caucasian. Mean age was 40. The 95 participants contributed 4,626 days of diary data. In all, 431 diary entries were missing or unavailable for analysis. Participants had a headache attack on approximately 39% of days. Days that preceded a headache were associated with greater stress than days that did not precede a headache.
After estimating a series of models, the researchers found that a generalized linear mixed-effects model using either the frequency of stressful events or the perceived intensity of stressful events fit the data well. The forecasting model had “promising predictive utility” in the training sample and in a validation sample, said the authors. The model had good calibration between forecast probabilities and observed headache frequencies, but had low levels of resolution, meaning that “the forecast probabilities are close to the individual’s long-run average,” said Dr. Houle.
“This appears to be the first evidence that individual headache attacks can be forecast within an individual sufferer, and this finding creates substantial opportunities for additional treatment strategies if the forecasting model can be refined,” said Dr. Houle. “A forecasting model could be used to enhance pharmacologic treatment opportunities, reduce anxiety about the unpredictability of attacks, increase locus-of-control beliefs, and lead to increased self-efficacy assessments about the self-management of migraine attacks.” Neurologists should consider the investigators’ stress model a first step toward headache prediction, and not a final model for widespread clinical use, he added.
Complexities Need Consideration
These data are “fascinating,” but neurologists should consider several complexities as they develop methods for the short-term prevention of predictable migraine, said Richard B. Lipton, MD, Edwin S. Lowe Chair in Neurology at Albert Einstein College of Medicine in New York and Director of the Montefiore Headache Center, and colleagues in an accompanying editorial. First, they must distinguish group-level and within-person analyses of attack predictors. Trigger factors vary from person to person, and within-person analysis may be crucial to prediction and prevention, said Dr. Lipton. Second, in addition to stress, other trigger factors such as premonitory features, self-prediction, and biomarkers also may aid in forecasting attacks. Finally, researchers can measure and model predictors of impending attacks in various ways (eg, lead–lag effects and cumulative effects).
“Houle et al have set the stage for short-term prediction of headaches in persons with migraine as a potential foundation for short-term preventive therapies,” said Dr. Lipton. “To realize the potential of these approaches, we must refine the art of headache forecasting and then test targeted interventions in carefully selected patients.”
—Erik Greb
Suggested Reading
Houle TT, Turner DP, Golding AN, et al. Forecasting individual headache attacks using perceived stress: Development of a multivariable prediction model for persons with episodic migraine. Headache. 2017;57(7):1041-1050.
Lipton RB, Pavlovic JM, Buse DC. Why migraine forecasting matters. Headache. 2017;57(7):1023-1025.