An open-label multicentre trial randomised 279 adults with episodic migraine to twelve weeks of daily self-administered app-based biofeedback or to a waitlist. Mean age was 41 and 92% were women, with about 4.7 monthly migraine days at baseline.
Over weeks 9 to 12, monthly migraine days fell by 0.9 in the intervention group and 0.0 in the waitlist group, a between-group difference of 0.9 days with an interval from 0.3 to 1.5. Migraine-specific quality of life improved by 5.3 points more. Nine device-related adverse events occurred, all mild.
Two things temper this. A waitlist control is the weakest comparator available — it controls for time and regression to the mean but not for attention, expectation or the act of doing something daily, and the control arm's change of exactly 0.0 days is itself a sign that no non-specific effect was captured. And 0.9 days a month is real but small; the conventional threshold for a meaningful preventive response is a 50% reduction, which from a baseline of 4.7 days would be 2.4. This is a reasonable adjunct with essentially no risk, not an alternative to a preventive.
- Waitlist control, so expectation and attention effects are not accounted for
- Open-label with a patient-reported primary outcome — the two combine badly
- 0.9 days a month from a 4.7-day baseline is well short of a 50% response
- Essentially no safety cost, which is the strongest argument for offering it
- 92% women and a mean age of 41 — a narrow slice of the migraine population
The statistics, in plain English
A between-group difference of 0.9 days with an interval from 0.3 to 1.5 excludes no effect, so the finding is real. Statistical significance and clinical significance are separate questions: the smallest difference patients reliably notice in migraine trials is generally taken as around one day a month, and this sits at the boundary. A waitlist comparator also inflates apparent benefit relative to a sham.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for neurology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free