Stress and fasting for inducing migraine attacks: A randomized 2 × 2 factorial crossover trial

The objective of this study was to evaluate the independent and joint effects of acute stress and fasting on the likelihood of headache onset in individuals with migraine.

Stress and fasting are two of the most frequently cited triggers of migraine headache, but experimental evidence regarding their individual and combined effects remains limited.

This was a partially randomized, 2 × 2 factorial crossover trial conducted in a controlled laboratory setting in Boston, MA, between March 2017 and August 2020. Participants were 18 to 65 years old with episodic migraine. Enrollment was halted early due to recruitment delays. Participants were exposed to two joint experimental manipulations involving stress and fasting: (1) a randomized 12-h fasting or nonfasting condition, and (2) a fixed sequence administration of the Trier Social Stress Test (stress) or a time-matched control task (control), each administered on separate study visits. The primary outcome was an incident headache attack rated ≥4/10 or treated with abortive or analgesic medication within the 6-h laboratory session.

Among 131 analyzable sessions from N = 74 participants, eight (6%) incident headache attacks were observed. Predicted probabilities of a laboratory headache were 3.8% (95% credible intervals [CrI]: 0.3% to 13.9%) for Control/No-Fast, 5.6% (95% CrI: 0.5% to 21.3%) for Control/Fast, 2.7% (95% CrI: 0.2% to 12.5%) for Stress/No-Fast, and 6.8% (95% CrI: 0.6% to 24.3%) for Stress/Fast. Compared to Control/No-Fast, adjusted posterior odds ratios were 1.54 (95% CrI: 0.39 to 6.5) for Control/Fast and 1.86 (95% CrI:0.48 to 8.1) for Stress/Fast. Secondary outcomes mirrored the primary findings, with the highest cumulative incidence of attacks in the Stress/Fast condition.

Fasting and stress were each associated with a modest alteration in migraine risk, and the combination of both exposures was consistent with a greater likelihood of acute headache onset. Given the wide credible intervals, these findings should be interpreted as preliminary evidence for a possible synergistic interaction between stress and fasting.