Migraine attacks are not random. Roughly half of people with migraine report that their attacks follow a circadian rhythm, clustering at particular times of day, and many also experience monthly or seasonal patterns that repeat with striking regularity. The term “cyclic migraine” captures this broader reality: migraine is, at its core, a disorder of biological timing. The hypothalamus, which acts as the brain’s master clock, appears to be the primary driver, and its interactions with hormones, sleep, and the autonomic nervous system help explain why attacks arrive in waves rather than at random.
When Attacks Strike During the Day
A large systematic review and meta-analysis pooling data from eight studies found that about half of participants showed a clear circadian pattern to their migraine attacks. The attacks had a broad trough between 11 p.m. and 7 a.m. and tended to peak during daytime hours.1PubMed Central. Circadian Features of Cluster Headache and Migraine: A Systematic Review, Meta-analysis, and Genetic Analysis That said, not every study agrees on when the peak falls. A separate review of clinical studies found that an early morning or late-night peak was the most common finding, but a substantial number of studies showed different timing altogether.2PubMed Central. Circadian Variation of Migraine Attack Onset: A Review of Clinical Studies Some of this disagreement comes from methodological differences between studies, but it also reflects a real fact about migraine: different subtypes may run on different clocks, and individual triggers can shift the window.
What the circadian data tell us, even with the inconsistencies, is that migraine is not evenly distributed across the 24-hour day for most people. If you track your own attacks over months, you will likely see a preferred time window. That window may not match anyone else’s, but its existence points to the same underlying biology: the brain’s timekeeping system is involved in generating attacks.
The Hypothalamus as Migraine’s Pacemaker
The hypothalamus is a small region deep in the brain that coordinates circadian rhythms, hormone release, sleep-wake cycles, and autonomic function. Brain imaging studies have identified it as the structure that lights up before a migraine attack even begins. One longitudinal MRI study tracked a single patient through an entire migraine cycle and found that hypothalamic activation increased in the days leading up to an attack, suggesting a gradual loss of control over pain-processing regions in the brain.3PubMed Central. Migraine attacks as a result of hypothalamic loss of control Another study that scanned a patient daily for 30 days across three spontaneous attacks found the hypothalamus to be the primary motor of migraine, with phase-specific interactions with the brainstem that shifted as an attack approached.4Brain. The migraine generator revisited: continuous scanning of the migraine cycle over 30 days and three spontaneous attacks
This is why migraine premonitory symptoms, the yawning, fatigue, food cravings, and mood changes that arrive hours or even days before pain, make sense biologically. They are all functions the hypothalamus regulates. The attack does not begin with headache. It begins with a subtle shift in how the hypothalamus communicates with the rest of the brain, and the headache is a downstream event.
The Menstrual Cycle and Hormonal Tides
The most familiar cycle in migraine is the menstrual one. Many women notice that their worst attacks cluster around the start of their period, and the reason comes down to a sharp drop in estrogen. Estrogen tends to increase neural excitability, while progesterone and its metabolite allopregnanolone tend to dampen it.5PubMed. Female reproductive steroids and neuronal excitability When estrogen falls rapidly in the late luteal phase, the brain loses a key excitatory input, and the balance between excitation and inhibition shifts in a way that can trigger a migraine in susceptible people.
A cross-sectional study found that women with migraine who still had regular menstrual cycles had higher levels of CGRP (a molecule heavily involved in migraine pain signaling) in their blood and tear fluid during menstruation, compared to women without migraine. Women using hormonal contraceptives or past menopause did not show this difference.6PubMed Central. Sex Hormones and Calcitonin Gene-Related Peptide in Women With Migraine: A Cross-sectional, Matched Cohort Study This suggests that the cyclical hormone withdrawal specifically amplifies the CGRP pathway that modern migraine drugs are designed to block.
There is also evidence that allopregnanolone, a calming neurosteroid derived from progesterone, runs lower in women with menstrual migraine compared to healthy controls. The longer a woman had lived with migraine and the more attack days she experienced, the lower her allopregnanolone levels tended to be.7PubMed Central. Serum levels of allopregnanolone, progesterone and testosterone in menstrually-related and postmenopausal migraine: A cross-sectional study This paints a picture of a brain that is chronically under-protected against excitability, with the monthly hormone withdrawal acting as a reliable trigger that pushes it over the threshold.
Seasonal Patterns and Light Exposure
Migraine also follows longer rhythms. The same large meta-analysis that documented the circadian pattern found a broad seasonal peak in attacks between April and October.8PubMed Central. Circadian Features of Cluster Headache and Migraine: A Systematic Review, Meta-analysis, and Genetic Analysis A study of women living in an Arctic region, where seasonal light variation is extreme, found a clear seasonal fluctuation in migraine with aura, peaking around late May. When the researchers removed attacks that were associated with insomnia from the analysis, the seasonal pattern disappeared, suggesting that the seasonal effect was driven at least partly by light-related sleep disruption rather than light exposure itself.9PubMed. Circannual periodicity of migraine?
Migraine attacks show significant circadian and longer-than-daily rhythms influenced by hormonal levels, light sensitivity, and hypothalamic activation.10PubMed. Time and headache: Insights into timing processes in primary headache disorders for diagnosis, underlying pathophysiology and treatment implications The practical takeaway is that if you notice your migraine burden shifting with the seasons, you are not imagining it, and it is worth discussing preemptive adjustments with your doctor as the higher-risk months approach.
Clock Genes and the Molecular Side
The body’s circadian clock depends on a feedback loop of genes, commonly called clock genes, whose proteins rise and fall in a roughly 24-hour cycle. There is growing evidence that these genes behave differently in people with migraine-related conditions. A study comparing people with vestibular migraine to healthy controls found that the vestibular migraine group had significantly lower expression of several core clock genes, including PER1, PER2, CRY1, BMAL1, CLOCK, and TIM. Among those patients, the ones with more severe symptoms had even lower expression of certain clock genes.11PubMed. Dissecting the Circadian Clock and Toll-like Receptor Gene Alterations in Meniere’s Disease and Vestibular Migraine
This is still early research, but it fits the larger picture: in migraine, the biological clock may not just trigger attacks at certain times. It may itself be running differently, with weaker or less synchronized oscillations that leave the brain more vulnerable to triggers at unpredictable moments. Whether this clock gene disruption is a cause or a consequence of repeated migraine attacks remains an open question, but it helps explain why some people’s attack patterns are more chaotic than others.
What Happens in Sleep Before an Attack
One of the more fascinating findings in cyclic migraine research involves what happens during sleep on the night before an attack. A polysomnography study comparing pre-attack nights to normal nights found several changes: the number of arousals dropped, REM density decreased, and certain brain wave frequencies were suppressed during deep sleep and the first REM period.12PubMed. Polysomnographic findings in nights preceding a migraine attack In plain terms, the brain appeared to be less activatable during pre-attack sleep, as if the cortex was already shifting into a different mode before the pain had even started.
This finding connects back to the hypothalamic model. If the hypothalamus is progressively losing its grip on excitability as an attack approaches, sleep architecture would be one of the first things to reflect that. It also suggests that the “warning signs” of an approaching migraine may extend into sleep itself, which is something you cannot detect by self-report alone but that wearable devices may eventually pick up on.
The Autonomic Nervous System Goes Quiet
Heart rate variability, a measure of how flexibly the heart responds to changing demands, is a common proxy for autonomic nervous system health. During an active migraine attack, heart rate variability drops substantially. One study found that a key measure of heart rate variability was less than half the value seen in healthy controls during an attack, and significantly lower than the same patients’ own readings between attacks. Between attacks, patients’ heart rate variability looked essentially normal.13PubMed Central. Heart Rate Variability Analysis in Episodic Migraine: A Cross-Sectional Study
This tells us that migraine temporarily suppresses the autonomic nervous system’s flexibility, almost like a temporary freeze on the body’s ability to adapt. The fact that this recovers between attacks reinforces the idea that migraine operates in discrete phases: a normal interictal state, a building pre-attack phase, the attack itself, and recovery. Each phase has a different physiological signature.
Timed Prevention for Menstrual Migraine
Because menstrual migraine arrives on a predictable schedule, it is one of the few migraine subtypes where you can time preventive treatment precisely. The strategy, called perimenstrual prophylaxis, involves starting medication a day or two before the expected onset of the vulnerable window and continuing through it. Randomized controlled trials have shown that frovatriptan taken twice daily during the perimenstrual period roughly doubled the chance of staying migraine-free compared to placebo. The incidence of menstrual migraine headache during the treatment period was 67% with placebo, 52% with once-daily frovatriptan, and 41% with twice-daily dosing.14PubMed. A randomized trial of frovatriptan for the intermittent prevention of menstrual migraine Even in women whose menstrual migraines were considered difficult to treat, frovatriptan significantly reduced both the number and severity of attacks.15PubMed. Short-term frovatriptan for the prevention of difficult-to-treat menstrual migraine attacks
Hormonal strategies offer another approach. Supplementing with estrogen during the perimenstrual window can prevent the sharp drop that triggers attacks. Percutaneous estradiol gel, started no more than two days before anticipated menses and used for seven days, has been shown superior to placebo in controlled trials. Transdermal estrogen patches at a minimum dose of 0.1 mg per day are also effective. One important timing detail: starting estrogen supplementation too early, around six days before menses, has been linked to rebound migraine after the supplementation is withdrawn.16PubMed Central. Hormonal Contraceptive Options for Women With Headache: A Review of the Evidence Continuous use of combination contraception, which avoids the hormone-free interval entirely, is another option some guidelines recommend.17PubMed Central. Effect of exogenous estrogens and progestogens on the course of migraine during reproductive age: a consensus statement by the European Headache Federation (EHF) and the European Society of Contraception and Reproductive Health (ESCRH)
CGRP Antibodies and Cyclical Attacks
The newer class of migraine preventives, monoclonal antibodies that block CGRP or its receptor, were not designed specifically for cyclical migraine. But given that menstrual migraine involves a measurable spike in CGRP levels, they have a logical application. A real-world study of 40 women with menstrual migraine found that after six months of CGRP antibody treatment, median menstrual migraine days dropped from five to two per month, pain intensity fell, and attack duration shrank from 24 hours to 8 hours. Nearly 58% of patients achieved at least a 50% reduction in menstrual migraine attacks.18PubMed Central. Effectiveness and Safety of CGRP-mAbs in Menstrual-Related Migraine: A Real-World Experience The antibodies also improved how well patients responded to painkillers when attacks did break through, jumping from about 43% responding at baseline to 95% after treatment. Not every patient benefited: about 13% saw no improvement in either menstrual or non-menstrual attacks.
Migraine Across Life Stages
The cyclical nature of migraine does not stay constant through a woman’s life. Perimenopause, the transitional years before menopause when hormone levels become erratic, is often the worst period. Unstable estrogen and progesterone levels during this time can worsen migraine frequency and make attack timing less predictable.19PubMed. Migraine across the menopausal transition and beyond: A narrative review Menstrual migraine during perimenopause is considered more disabling and harder to treat than non-menstrual attacks.20PubMed Central. Menopause, Perimenopause, and Migraine: Understanding the Intersections and Implications for Treatment
After menopause, the picture splits. Migraine without aura often improves, presumably because the cyclical hormone fluctuations that drove it have ceased. Migraine with aura, however, tends to persist, and it independently raises the risk of stroke and other cardiovascular events.21PubMed. Migraine across the menopausal transition and beyond: A narrative review This divergence is one more piece of evidence that different migraine subtypes run on different clocks, and that the hormonal cycle is the dominant rhythm for some patients but not others.
Melatonin and the Circadian Connection
Given the hypothalamus’s role in both migraine and melatonin secretion, researchers have explored whether melatonin levels differ in people with migraine. Studies have found that nocturnal melatonin levels tend to be lower in adults with migraine compared to healthy controls, though this difference narrows when patients with insomnia and depression are excluded. In a trial, 3 mg of immediate-release melatonin was found to be better than placebo and comparable to a low dose of nightly amitriptyline (a standard preventive) for reducing migraine frequency.22LIDSEN Publishing Inc. (OBM Neurobiology). Effects of Circadian Rhythm on Migraine Therapy This raises interesting questions about whether some of migraine’s cyclical character is driven by weak or poorly timed melatonin signals. Melatonin is cheap, widely available, and well tolerated, which makes it an appealing low-risk option for people exploring the circadian angle of their migraine. The evidence is still modest, but the biological rationale is solid.
Wearable Sensors and Attack Prediction
If migraine follows cycles and produces measurable physiological shifts before pain begins, the logical next step is automated prediction. Several research groups have tested whether wearable sensors worn during sleep can detect an incoming attack. One study using sleep-time data from wearable sensors reported balanced accuracy over 84% for detecting an attack one night in advance, using personal recognition models. However, accuracy varied greatly between participants, and the study included only seven subjects.23PubMed Central. Using Sleep Time Data from Wearable Sensors for Early Detection of Migraine Attacks
A more recent study used a wearable biosensor to monitor electrodermal activity, skin temperature, and movement in ten people. The strongest signals for predicting next-day migraine came from skin conductance and temperature changes during the night. A machine learning model achieved about 81% accuracy in classifying pre-migraine versus normal nights.24PubMed Central. Machine Learning and Wearable Technology: Monitoring Changes in Biomedical Signal Patterns during Pre-Migraine Nights These are early results from tiny sample sizes, but they suggest that the autonomic and thermoregulatory shifts happening before an attack leave detectable fingerprints that a sensor can pick up. The promise of chronotherapy, timing treatment delivery to match the body’s rhythmic vulnerability, depends on getting this kind of prediction right.25PubMed Central. Is it time to think about chronotherapy in migraine If a wearable could reliably warn you 12 to 24 hours before an attack, you could take a preventive medication before the pain cascade fully launches, potentially blunting or aborting the episode entirely.