A Possible Explanation of Cluster Headaches: A Disruption of the Dehydroepiandrosterone to Testosterone Ratio in Vulnerable Individuals (This is an explanation of the mechanism. Low DHEA and low testosterone in cluster headaches have been noticed before this treatise.)

Copyright 2013 James Michael Howard, Fayetteville, Arkansas, U.S.A.

(New added support: bottom of page)

Cluster headaches (Chs) are relatively rare. I think a possible explanation of their cause and possible treatment, not cure, may be found in my usual explanation of pathology, that is, the “secular trend.” I think cluster headaches are a disorder exposed by the reductions in dehydroepiandrosterone (DHEA) and testosterone which naturally occur beginning around age twenty-five. A disruption of the DHEA to testosterone ratio may expose malfunctioning genes or congenital effects. Behaviors, which increase the natural reduction of these androgens, will participate in triggering cluster headaches when these androgens begin to decline in vulnerable individuals.

(My detailed explanation of the relationship of DHEA and testosterone is available: “DHEA, Estradiol, Testosterone, and the Relevance of Their Ratio …The Androgen Receptor …and the Secular Trend,” at: http://anthropogeny.com/Androgen%20Receptor%20and%20Secular%20Trend.htm . I recommend reading this.)

I think a combination of melatonin, DHEA in different amounts during the day and night along with discontinuation of smoking and alcohol consumption may provide some relief for these individuals.

As you read further, keeps the following in mind: Treatment for cluster headaches often includes sumatriptan (Imitrex) which has been found to decrease prolactin (J Psychopharmacol. 2013 Jan 16. [Epub ahead of print]: “The 5-HT1D/1B receptor agonist sumatriptan enhances fear of simulated speaking and reduces plasma levels of prolactin,” de Resende, at al.). “Prolactin and ACTH responses to the drug [sumatriptan] were significantly (P < 0.05) reduced in patients with cluster headache, both in the active and in the remission period.” (Cephalalgia. 2003 Jun;23(5):354-60) (I explain the importance of prolactin in cluster headaches below; please read further.)


A connection of testosterone levels with cluster headaches is not new with me. I think it is the combination of reduced DHEA and testosterone levels that are involved in exposing a vulnerability to cluster headaches that may be more affected by loss of DHEA.

The decline of testosterone may be the key factor, though, loss of DHEA may exacerbate this. Chs occur much more often in men than women, about twice as often. Some suggest that this ratio of Chs in men to women may be falling. Since my work suggests that testosterone is increasing in women, it fits that testosterone may be involved. (This sounds contradictory at this time but I will explain the consequences of higher testosterone and its decline.) I think the consequences of the secular trend are heavily related to an early onset of testosterone within individuals which results in an early decline of testosterone. (The peak of testosterone production occurs earlier in individuals because of the secular trend.) The usual onset of Chs follows the decline of DHEA and testosterone. I think a significant amount of human pathology occurs because of the loss of gene activity caused by the loss of the effects of DHEA and testosterone.

A pattern may be identifiable in phenomena often associated with cluster headaches. (The following is derived from a list of phenomena that “may trigger cluster attacks.” Alcohol and smoking reduce testosterone in men. It is my hypothesis that alcohol and smoking increase DHEA levels during prolonged use. However, this long-term increase in DHEA increases the onset of the natural decline of DHEA during the life span. It is this loss of DHEA that participates in exposure of pathological conditions with aging. The earlier onset of this decline of DHEA caused by alcohol and smoking is the main causal mechanism of their use. High altitude reduces testosterone in men. Bright light may decrease melatonin production which may reduce DHEA levels (to be explained below). Exertion in men who are experiencing the reduction of testosterone of the secular trend may exhibit characteristics of a “hypogondal” man. Exercise actually reduces testosterone in the “exercise-hypogandal male condition” (J Endocrinol Invest. 2008 Oct;31(10):932-8). Heat reduces testosterone. Nitrites reduce testosterone. Cocaine use eventually reduces testosterone levels.

The beneficial effects of oxygen are simply the beneficial effects of oxygen. In my preparation for this paper, I found that blood vessel dilation is involved in cluster headaches.  I assumed this could result in ischemia of the vessels involved.  Therefore, less oxygen is available to the tissues involved.  Subsequently, I discovered that oxygen deprivation causes pain and one source said that cluster headaches, among some other causes mentioned, is, indeed, an "oxygen deprivation headache."  So, the oxygen treatment fights the pain of oxygen deprivation.  Since DHEA might stimulate constriction, therefore, opening due to the mechanical effect of stimulating the walls of the vessels involved, DHEA should help with this too.

This is a good synopsis of cluster headaches and sleep, specifically REM sleep by James Weintraub, Michigan Head Pain & Neurological Institute: “There has been an association of these attacks [cluster headaches] to REM sleep and research has noted an increase of sleep apnea in patients who experience cluster headaches. In fact, in one report, patients who have cluster headaches and sleep apnea were also noted to have experienced bed-wetting and night terrors during childhood. In patients with cluster headaches who have sleep apnea, it is suggested that they may be experiencing an episodic decrease in oxygen during the night that can precipitate the headache attacks. In patients who have chronic cluster headaches, it is noted that two-thirds of the attacks usually occur during the last two hours of sleep.” (DHEA, testosterone, prolactin, and melatonin are not mentioned in the full article from which this is taken.)

It is part of explanation of sleep (link below) that sleep begins with very low levels of DHEA, which allow sleep to occur. At this time melatonin is high because the low DHEA allows the largest release of the night from the pineal gland. Melatonin triggers a response release of prolactin which stimulates DHEA to maintain brainstem function. As sleep continues, this cycle of melatonin – prolactin – DHEA gradually increases DHEA production which reaches its maximum which then results in consciousness. I think these periodic increases in DHEA activate the brain without producing consciousness but do cause the slight brain activity which is called “REM sleep.” Because of this mechanism, I have deduced that any approach to insufficient DHEA production to maintain the brainstem results in an adrenergic reaction which stimulates DHEA which then re-institutes brainstem activity. (This may be why some people cannot maintain sleep.) My point is that this response occurs when DHEA is very low which I have stated may be the cause of cluster headaches. I suggest this is why REM sleep is connected with cluster headaches. This is a time of such low DHEA that negative consequences of this low DHEA occur, besides cluster headaches. Sleep apnea, bed-wetting, night terrors, I suggest, are also caused by dangerously low levels of DHEA during sleep which sometimes stimulate a significant adrenergic reaction to avoid brainstem shutdown and death.

Cluster headaches present on one side of the head. A substantial amount of research has demonstrated / supported evidence of laterality of the brain related to testosterone levels.

Plasma levels of testosterone have been reported in cluster headaches (Cephalalgia March 1983 vol. 3 no. 1 41-44). Some men experienced relief from cluster headaches with testosterone treatment (Headache. 2006 Jun;46(6):925-33). Other studies failed to find a positive treatment effect of testosterone in chronic sufferers (Cephalalgia. 1993 Aug;13(4):258-60). Antiandrogenic therapy has been beneficial (Int J Clin Pharmacol Res. 1988;8(1):21-4).


As far as I can determine, using “Pubmed,” DHEA has not been measured in cluster headaches. Since prolactin is a direct and specific stimulator of DHEA production, I suggest a strong case may be made for a cyclical feedback regulation of prolactin and DHEA. Prolactin (PRL) levels have been measured in Chs.

“In males the nocturnal PRL peak was blunted during remissions as compared with that in cluster periods and that in control individuals. The 24-h mean PRL levels were lower during remission and cluster periods than in the controls.” (Cephalalgia. 1987 Mar;7(1):43-54) This may indicate that DHEA has increased which reduced PRL levels and corrected the DHEA to testosterone ratio. Note: From above, sumatriptan (Imitrex) reduces prolactin and this may mean that the beneficial effects of sumatriptan are caused by increases in DHEA.

In 1985, I produced an explanation of sleep involving melatonin, prolactin, and DHEA ("Sleep, Melatonin, DHEA, AIDS and Sudden Infant Death Syndrome," at: http://anthropogeny.com/Sleep%20and%20SIDS.htm ).  This explains that the function of sleep is to stimulate DHEA upon awakening.  When DHEA reaches a certain level, it stimulates awakening. The function of sleep, involving melatonin and prolactin, is the production of DHEA. Melatonin and prolactin levels change during sleep to assure maintenance of a proper levels of DHEA during sleep to maintain the brainstem to maintain breathing and the cardiovascular system. Since DHEA is low during sleep, cluster headaches often occur at night. This would be a time of vulnerability to excessively low levels of DHEA and may explain why cluster headaches are rare; if one's DHEA declines too low, I suggest one may die.

Melatonin is directly involved in production of DHEA. Melatonin has been found to be reduced during cluster headaches.

Furthermore, it is also my hypothesis that cortisol (stress) evolved to counteract the positive effects of DHEA on neuronal function.  That is, cortisol reduces initial motivation produced by DHEA to continue during "fight or flight" circumstances.  An increased cortisol to DHEA ratio also reduces the effects of DHEA in tissues other than the brain. Cortisol has been found to be connected with cluster headaches. Hight cortisol adversely affects the function of DHEA.

“The cyclic nature of cluster headache warranted a study of the 24-hour rhythms of serum cortisol and melatonin. They were both altered during cluster periods as compared with periods of remission and healthy controls. The 24-hour mean and maximal cortisol levels were higher and the timing of the cortisol minimum was delayed as compared to the same patients in remission. Although there was no relation between the cortisol and melatonin levels and headaches, the rise of cortisol following many attacks might in part represent an adaptive response to pain. The nocturnal melatonin maximum was lower during cluster periods than in remission. This finding, and the dysautonomic signs during attacks, may reflect a change of the vegetative tone in a hyposympathetic direction.” (J Neurol Neurosurg Psychiatry 1987; 50: 207-213)

The cyclic recurrence of cluster periods and the regular timing of headache occurrence in cluster headache (CH) induced us to study the circadian secretion of melatonin and cortisol in 12 patients with episodic CH, during a cluster period, and compare them with 7 age- and sex-matched healthy controls. Blood was sampled every 2, h for 24 h. All subjects were confined to a dark room from 22.00 to 08.00. Plasma melatonin levels were significantly reduced in CH patients (repeated measures ANOVA p < 0.03; mesor p < 0.02), and the cortisol mesor was significantly increased (p < 0.03). Amplitudes and acrophases did not differ between the groups. Individual cosinor analysis showed that 4/12 (33.3%) CH patients had no significant melatonin rhythm, and that 5/11 (45.5%) had no cortisol rhythm. Group analysis of cosinor revealed significant rhythmicity of melatonin and cortisol secretion in both groups. In controls, the timing of melatonin and cortisol acrophase significantly correlated with each other, indicating that the biorhythm controllers for the secretion of these hormones were synchronized. Such correlation was not found in the CH patients; mesor, amplitude and acrophase of melatonin and cortisol did not correlate with duration of illness, duration of headache in course, or time since last headache attack.” (Cephalalgia June 1995 vol. 15 no. 3 224-229 )

“A fall in nocturnal plasma melatonin occurs in patients with cluster headache, suggesting that melatonin may play a role in the promotion of attacks. During a cluster period, we administered melatonin to 20 cluster headache patients (2 primary chronic, 18 episodic) in a double-blind placebo-controlled study of oral melatonin 10 mg (n = 10) or placebo (n = 10) for 14 days taken in a single evening dose. Headache frequency was significantly reduced (ANOVA, p<0.03) and there were strong trends towards reduced analgesic consumption (ANOVA, p<0.06) in the treatment group. Five of the 10 treated patients were responders whose attack frequency declined 3–5 days after treatment, and they experienced no further attacks until melatonin was discontinued. The chronic cluster patients did not respond. No patient in the placebo group responded. There were no side effects in either group. Although the response rate is low, melatonin may be suitable for cluster headache prophylaxis in some patients, particularly those who cannot tolerate other drugs.” (Cephalalgia November 1996 vol. 16 no. 7 494-496 )

“The circadian changes in testosterone (T) and Cortisol secretion and morning luteinizing hormone (LH) levels were evaluated in nine episodic cluster headache (CH) patients in active phase and in seven healthy volunteers, with collection of blood samples every 2 h for 24 h. CH showed a significant reduction of the 24-h integrated mean T value (mesor) (4.4 + 1.1 ng/ml; × ± SD) in comparison with controls (6.6 ± 0.8 ng/ml) (P < 0.01). Both groups had plasma T circadian rhythm with peak values in early morning, but in CH single cosinor analysis showed its absence in three out of nine CH patients. The rhythm showed an acrophase delay of 101 min in CH. Both patients and controls had a significant circadian rhythm of plasma Cortisol concentration. CH patients, however, showed an acrophase delay of 106 min and significantly increased concentrations from 1200 h to 2000 h. Morning LH values were similar in the two groups. The reduced secretion of plasma T in CH patients in the active phase coexisted with an acrophase delay of its circadian rhythm. A similar delay was found in 24-h plasma Cortisol levels. We suggest that stress accompanying attack expectancy in the active phase is the mechanism behind the elevated plasma Cortisol levels. This in turn could reduce T concentrations, acting at the testicular level. These disturbances in internal chrono-organization support the hypothesis that cluster headache is basically a dyschronic disorder.” (Cephalalgia March 1986 vol. 6 no. 1 29-34 )

“Cluster headache is a stereotypic headache disorder marked by short-lasting bouts of severe unilateral head pain and associated autonomic symptoms. Almost pathognomonic of this condition are nocturnal attacks that usually occur during the first random eye movement sleep phase of the evening. Melatonin levels have been found to be decreased in cluster headache patients. A lack of melatonin secretion may predispose the cluster sufferer to nocturnal and, possibly, daytime attacks. Leone et al. demonstrated that melatonin could rapidly alleviate cluster attacks, but only in episodic cluster patients. We report two chronic cluster headache patients who had both daytime and nocturnal attacks that were alleviated with melatonin.” (Cephalalgia December 2001 vol. 21 no. 10 993-995 )

New Support:

“Conclusions The majority of patients with migraine or cluster headache do not have sexual activity during headache attacks. Our data suggest, however, that sexual activity can lead to partial or complete relief of headache in some migraine and a few cluster headache patients.” (Cephalalgia April 2013 vol. 33 no. 6 384-389 ).

Prolactin increases following sexual activity and is 400% higher following intercouse than following masturbation. (Biol Psychol. 2006 Mar;71(3):312-5)

I suggest the foregoing supports my contention that prolactin and, therefore, DHEA decreases cluster headaches.