Preliminary Evidence That Melatonin Is Not a Biomarker in Children and Adolescents With Episodic Migraine

Anthony Berger, PhD; Jessica Litwin, MD; Isabel Elaine Allen, PhD; William Qubty, MD; Samantha Irwin, MB, MS, FRCPC; Maggie Waung, MD, PhD; Amy A. Gelfand, MD


Headache. 2019;59(7):1014-1023. 

In This Article

Abstract and Introduction


Background: To date, there have not been reliable biomarkers to identify impending migraine episodes. A prior study in adults with migraine demonstrated a reduction in the urinary metabolic substrate of melatonin (urinary 6-sulfatoxymelatonin; aMT6s) during a migraine. The aim of this study was to examine whether evening urinary melatonin metabolite levels could predict migraine the next day in children and adolescents with migraine.

Methods: Twenty-one children and adolescents with migraine (aged 5-17 years) were recruited to this observational study conducted at UC San Francisco to provide urine samples for 10 days and maintain a prospective headache diary during the same period. Nightly melatonin metabolite 6-sulfatoxymelatonin in urine was assayed and results from nights preceding migraine were compared to nights preceding a non-headache day.

Results: Mean (±SD) aMT6s levels the night prior to a migraine attack were 56.2 ± 39.0 vs 55.4 ± 46.6 ng/mL (P = .915), and mean melatonin metabolite levels the night following migraine were 55.5 ± 46.9 vs 57.0 ± 37.7 ng/mL (P = .841). However, in post hoc exploratory analyses, aMT6s levels were lower the night before a migraine in those who experienced aura or premonitory symptoms.

Conclusion: While urinary melatonin metabolites do not predict migraine attacks in children and adolescents overall, they may be predictive in those who experience premonitory phase symptoms as part of their migraine attacks.


Migraine is a common neurologic disease among children and adolescents. By age 10, the prevalence of migraine in the U.S. is approximately 5% and it increases throughout adolescence.[1] Migraine is associated with missed attendance and poor performance in school.[2] According to the Global Burden of Disease study, migraine is the number one cause of years lived with disability.[3]

The ability to predict approaching migraine attacks in children and adolescents would potentially allow for treatment interventions that could prevent attacks entirely and avoid disability.

The hypothalamus is a critical brain region implicated in migraine pathophysiology.[4] It exhibits heightened activity during the premonitory phase of a migraine attack.[5] The premonitory phase precedes the onset of headache by up to several hours to a full day[6] and occurs in approximately two-thirds of children and adolescents with migraine.[7] Many premonitory phase symptoms overlap with known functions of the hypothalamus, including changes in mood, appetite, food cravings, level of wakefulness, etc. The hypothalamus also influences melatonin synthesis and release by a feedback loop involving the suprachiasmatic nucleus (SCN), which can be entrained by light input.[8]

Melatonin is a hormone produced by the pineal gland where it is released into the bloodstream, allowing it to interact with multiple organ systems.[9] Melatonin is critical for the functioning of the human circadian rhythm and sleep. Melatonin secretion is suppressed by environmental light, and is increased in the dark.[9] Melatonin is already used to treat conditions such as epilepsy,[10] jet lag,[11] and sleep disorders.[12] It has been used to treat cluster headache, a primary headache disorder with a typical circadian pattern, since the 1970s[13] and more recently has been used to treat primary stabbing headache.[14] Melatonin's role in the pathophysiology of migraine may involve actions on melatonin receptors in the hypothalamus.[15,16] anti-inflammatory effects,[17,18] or through its effect on improving sleep.[19,20] Notably, melatonin has been shown in a recent double-blind placebo-controlled trial to be safe and effective for migraine prevention in adults.[21] As both premonitory symptoms and melatonin production are influenced by the hypothalamus, it is possible that melatonin levels in children and adolescents who experience premonitory symptoms may differ from those who do not.

The identification of biomarkers during the premonitory phase that can predict the onset or severity of a migraine would enable clinicians and patients to potentially prevent or ameliorate an attack and would also guide the use of early abortive therapy through improved understanding of migraine pathogenesis. Previous work in adults has demonstrated that melatonin levels are decreased on a migraine day compared to a non-headache day.[22] Therefore, the present study evaluated the hypothesis that decreased overnight melatonin levels would predict migraine incidence the following day in children and adolescents with migraine.