Sleep and migraine

What people report

Among 1,207 clinic patients, 49.8% named sleep disturbance and 32.0% named sleeping late as a trigger (Kelman 2007). The Migraine Trust lists too little sleep, too much sleep and a change in sleep pattern. All of this rests mainly on recall.

Too little sleep

The clearest measured data come from one study of 98 adults with episodic migraine. They wore a wrist actigraph and kept twice-daily diaries for 6 weeks, giving 4,406 days and 870 headaches, with adjustment for caffeine, alcohol, activity and stress (Bertisch 2020).

Sleep measure	Day after the night	Second day after the night
Short sleep, 6.5 hours or less	Not linked	Not linked
Poor self-rated sleep quality	Not linked	Not linked
Low sleep efficiency (fragmented sleep)	Mixed, see below	Higher odds, diary measure OR 1.39 (95% CI 1.06 to 1.81)

Reports of the study describe the higher odds on the second day for both diary and sensor measures. On the first day, sensor-measured fragmentation went the other way (OR 0.64 to 0.74). The results do not all point in one direction, and six weeks of data do not rule out a link in individuals. An earlier prospective study reported more headache after two consecutive nights of 4 hours or less, by self-report (reported in Neurology Reviews 2018). A review lists sleep disturbance among the most common reported triggers (Marmura 2018).

Too much or irregular sleep

The evidence here is weaker. A 2013 letter states that longer sleep raises attack risk about as much as sleep deprivation, and suggests that a change in usual habits may be the main factor, but it gives no data of its own (Vollono 2013). Retrospective studies report that nearly half of patients name too little or too much sleep (Neurology Reviews 2018). The sources for this page contain no prospective study that measured long sleep directly.

Which comes first

Sleep and migraine affect each other, so a link in your records has several possible explanations.

• Sleep as a marker. A meta-analysis of 32 studies found worse self-rated sleep in adults with migraine than in controls, with a larger effect in chronic migraine. Comparisons of this kind cannot show direction, and the authors call the relationship poorly understood (Stanyer 2021).
• Tiredness as an early symptom. In a 3-month electronic diary study of 97 people recruited because they reported warning symptoms, feeling tired and weary was the most common premonitory symptom, in 72% of attacks with warning features (Giffin 2003). A poor night before an attack may reflect an attack already under way.
• Attacks that cut sleep short. The American Migraine Foundation says nearly half of attacks occur between 4:00 and 9:00 am.

How to check it in your diary

1. Log bedtime, wake time and night waking, not only hours.
2. Flag nights that differ from your usual, in either direction, so that long nights are visible too.
3. Note whether an attack woke you. That night is then an outcome, not an exposure.
4. Separate a late morning from what it delays. Sleeping in can also mean a late coffee and a late breakfast.

For the general method, see how to test a suspected trigger.

When to see a doctor

If you often wake with headache, mention it to your doctor. ICHD-3 lists sleep apnoea headache as a separate diagnosis.