Menu

E = k·S·D·Λ·C

Research · Hypothesis

Cluster Headache, ADHD, and the Dopamine Route

A patient-researcher's hypothesis and the evidence behind it

Author
Daniel Huddleston
Published
Last updated
Version
v1.0

How to read the evidence tags

  • APublished — humanPeer-reviewed human data
  • BPublished — animal, in vitro or abstractPeer-reviewed animal or in vitro data, or a conference abstract
  • CHypothesisProposed mechanism; not yet tested
  • Dn=1The author's own case; one person
  • GAPNot measuredNo study has looked

The question

Cluster headache has been studied against a long list of other conditions. The largest registry study I have found, covering 7,803 people with cluster headache and 387,042 matched controls, measured ten comorbidity categories: hypertension, pulmonary disease, coronary heart disease, migraine, sleep disorders, facial pain, epilepsy, mood disorders, anxiety, and substance abuse.[11] ADHD was not one of them.

I have not found any study that measures ADHD in a cluster headache cohort. That absence matters, because several independent findings in cluster headache point at the same dopamine pathway ADHD involves: raised platelet and plasma dopamine,[1],[4] disturbed tyrosine handling,[3],[4] low platelet monoamine oxidase,[6],[7] and activation of the posterior hypothalamus during attacks,[9] the region that houses the A11 nucleus, the only known source of descending dopamine to the trigeminocervical complex.[8]

The question this page asks is narrow: what happens to the picture if ADHD status is a variable nobody has been controlling for?

What the published evidence shows

Everything below is published work by other groups. Each item carries its evidence tier and a link to the reference list. I have kept the numbers as the authors reported them.

Peripheral dopamine and tyrosine

  • ACluster headache patients had the highest platelet dopamine (0.34 ± 0.36 ng/10⁸ platelets) compared with migraine without aura (0.20), migraine with aura (0.16) and controls (0.10), measured in both remission and active periods. Plasma dopamine was undetectable with that method.[1]
  • APlatelet norepinephrine and epinephrine were lower in all phases; tyrosine increased during attacks.[3]
  • AIn 23 chronic cluster headache patients versus 16 controls, plasma dopamine, noradrenaline and tyramine were several times higher.[4]
  • AReview: dopamine and trace-amine abnormalities across episodic and chronic cluster headache.[5]
  • APlatelet monoamine oxidase activity was reduced in male cluster headache patients in both acute and quiescent phases, later confirmed and shown not to be attributable to smoking.[6],[7]
  • AIn 37 cluster headache patients versus 17 controls, platelet glycine was lower; platelet glutamate and aspartate did not differ.[2]

Hypothalamus and A11

  • AThe posterior hypothalamic gray is activated during cluster headache attacks.[9]
  • AAt 7 tesla, hypothalamic glutamate/tCr was about 17% lower in chronic cluster headache (10 patients versus 11 controls) and remained significant after adjusting for age and sex. The authors favor impaired neuronal energy metabolism over neuron loss. Eight of the ten patients were on verapamil, and the sample is small.[10]
  • BIn the rat, A11 in the posterior hypothalamus is the only known source of descending dopaminergic innervation of the spinal gray matter and modulates trigeminocervical firing. No dopamine receptors were found in A11 itself; D1 and D2 receptors are present in the trigeminocervical complex.[8]
  • BIn vitro and in the rat, cariprazine binds the D3 receptor with roughly ten-fold higher affinity than D2 and acts as a partial agonist. This is pharmacology, not clinical evidence.[16]
  • AA case report: methylphenidate aborted refractory cluster headache attacks within minutes. One published observation pointing the other way along the same dopamine axis.[17]
  • ACase report: a man with chronic cluster headache since age 14, diagnosed with ADHD as an adult, started methylphenidate for ADHD and was headache-free at two weeks; two years later attacks had fallen from up to five a day to about five a week. Vitamin D was started at the same time, and verapamil changes muddied the later course.[18]

Comorbidity and ADHD

  • AVersus matched controls, people with cluster headache had more mood disorders (men 6.7% vs 2.8%; women 9.2% vs 4.9%), anxiety (men 4.1% vs 1.7%; women 7.4% vs 3.3%), substance abuse (men 2.3% vs 0.9%; women 1.3% vs 0.5%), sleep disorders (men 3.3% vs 1.4%; women 3.1% vs 1.3%), and low or medium education (men 77.0% vs 65.1%; women 65.0% vs 55.0%). ADHD was not among the ten categories measured.[11]
  • AIn Norway, age-standardized prevalence rose from 27.0 to 42.5 per 100,000 between 2009 and 2022; the share of women rose from 39.7% to 50.1%; prevalence was higher with lower education.[12]
  • AThe first study to report higher cluster headache prevalence in women than men (ratio 1.2).[13]
  • AIn children, headache was associated with ADHD (OR 2.01, 95% CI 1.63–2.46). The included studies were migraine-dominated and did not analyze cluster headache separately.[14]
  • BA conference abstract screened cluster headache patients for bipolar disorder, the closest psychiatric-trait work in cluster headache I have found. ADHD was not screened.[15]
  • GAPNo study I have found measures ADHD, or stimulant prescriptions as a proxy, in a cluster headache cohort.
Figure 2. Evidence map for the cluster headache, ADHD and dopamine hypothesisCards grouped under three headings. Peripheral dopamine & tyrosine: Platelet dopamine highest in cluster headache (tier A, ref 1); Platelet glycine lower in CH; glutamate, aspartate unchanged (tier A, ref 2); Platelet NE and epinephrine lower; tyrosine up during attacks (tier A, ref 3); Plasma dopamine, noradrenaline, tyramine several-fold higher (chronic CH) (tier A, ref 4); Review: dopamine and trace-amine abnormalities across episodic and chronic CH (tier A, ref 5); Low platelet MAO in male CH, acute and quiescent phases (tier A, ref 6); Low platelet MAO confirmed; not attributable to smoking (tier A, ref 7). Hypothalamus & A11: A11: only known descending dopamine source to the TCC (rat) (tier B, ref 8); Posterior hypothalamic gray active during attacks (tier A, ref 9); Hypothalamic glutamate/tCr ~17% lower in chronic CH at 7T (tier A, ref 10); Cariprazine binds D3 ~10× more tightly than D2; partial agonist (in vitro, rat) (tier B, ref 16); Methylphenidate aborted refractory CH within minutes (case report) (tier A, ref 17); Chronic CH with adult ADHD: headache-free two weeks on methylphenidate (case) (tier A, ref 18); D3 sensing distortion (tier C). Comorbidity & ADHD: Mood, anxiety, substance and sleep disorders higher in CH (registry) (tier A, ref 11); CH prevalence rising; share of women now ~50% (tier A, ref 12); First report of higher CH prevalence in women (ratio 1.2) (tier A, ref 13); Headache associated with ADHD in children (OR 2.01); CH not separated (tier A, ref 14); CH screened for bipolar disorder; ADHD not screened (abstract) (tier B, ref 15); ADHD measured in a CH cohort (tier GAP); ADHD as hidden variable (tier C); Author's case (tier D). Legend: A = Published — human; B = Published — animal, in vitro or abstract; C = Hypothesis; D = n=1; GAP = Not measured.Peripheral dopamine & tyrosineHypothalamus & A11Comorbidity & ADHDPlatelet dopamine highest in clusterheadacheRef 1A · Published — humanPlatelet glycine lower in CH; glutamate,aspartate unchangedRef 2A · Published — humanPlatelet NE and epinephrine lower; tyrosineup during attacksRef 3A · Published — humanPlasma dopamine, noradrenaline, tyramineseveral-fold higher (chronic CH)Ref 4A · Published — humanReview: dopamine and trace-amineabnormalities across episodic and chronic CHRef 5A · Published — humanLow platelet MAO in male CH, acute andquiescent phasesRef 6A · Published — humanLow platelet MAO confirmed; not attributableto smokingRef 7A · Published — humanA11: only known descending dopamine sourceto the TCC (rat)Ref 8B · Animal / in vitro / abstractPosterior hypothalamic gray active duringattacksRef 9A · Published — humanHypothalamic glutamate/tCr ~17% lower inchronic CH at 7TRef 10A · Published — humanCariprazine binds D3 ~10× more tightly thanD2; partial agonist (in vitro, rat)Ref 16B · Animal / in vitro / abstractMethylphenidate aborted refractory CH withinminutes (case report)Ref 17A · Published — humanChronic CH with adult ADHD: headache-freetwo weeks on methylphenidate (case)Ref 18A · Published — humanD3 sensing distortionProposedC · HypothesisMood, anxiety, substance and sleep disordershigher in CH (registry)Ref 11A · Published — humanCH prevalence rising; share of women now~50%Ref 12A · Published — humanFirst report of higher CH prevalence inwomen (ratio 1.2)Ref 13A · Published — humanHeadache associated with ADHD in children(OR 2.01); CH not separatedRef 14A · Published — humanCH screened for bipolar disorder; ADHD notscreened (abstract)Ref 15B · Animal / in vitro / abstractADHD measured in a CH cohortNo study foundGAP · Not measuredADHD as hidden variableProposedC · HypothesisAuthor's caseOne personD · n=1EVIDENCE TIERSAPublished — humanBPublished — animal, in vitro or abstractCHypothesisDn=1GAPNot measured
Figure 2. Evidence map. One card per source in the reference list, plus the hypothesis cards, the author's case, and the gap nobody has measured.

The hypothesis

CHypothesis Everything in this section is proposed, not shown.

I propose that ADHD status may be a hidden variable in cluster headache. Patients who are dopaminergically atypical may respond to dopaminergic drugs differently from neurotypical patients, and the existing studies do not distinguish between them. If that is so, effects that look inconsistent across a mixed cohort could become consistent once the cohort is split by ADHD status.

The claim is conditional on ADHD status. It is not a claim that drugs acting on the dopamine D3 receptor trigger attacks in everyone.

The chain I propose runs like this:

  1. CUntreated ADHD sets an atypical baseline dopamine tone.
  2. BCariprazine, a D3-preferring partial agonist,[16] occupies D3 receptors that act as dopamine-sensing autoreceptors.
  3. CWith those sensors occupied, the dopamine feedback system can no longer read how much dopamine is actually present.
  4. BCThe downstream dopamine signal is distorted, and the A11 nucleus, which supplies the descending dopamine "brake" on trigeminal pain relay,[8] passes that distorted output on.
  5. CTrigeminal pain relay is disinhibited.

A wording point I want to get right: cariprazine does not make dopamine, and nothing from it travels down this pipeline. It sits on the receptor and gives a weaker-than-dopamine signal there. What changes is how much of the body's own dopamine is released, and when.

A second guardrail: the platelet and plasma measurements above come mostly from the gut, kidneys and sympathetic nerves, not from A11. I read them as a parallel signal of disturbed dopamine and tyrosine handling in cluster headache, not as a direct readout of what A11 is doing. The link between the peripheral findings and the central mechanism is itself part of the hypothesis.

Figure 1. The proposed chain from untreated ADHD to a cluster headache attackFive boxes connected by arrows. 1: Untreated ADHD, atypical baseline dopamine tone (hypothesis). 2: Cariprazine binds D3 receptors, which act as dopamine sensors; it binds D3 about ten times more tightly than D2 (published, in vitro and rat; ref 16). 3: Dopamine feedback misreads actual dopamine levels (hypothesis). 4: A11 nucleus, the only known source of descending dopamine to the trigeminocervical complex (published, animal; ref 8); its brake output is distorted (hypothesis). 5: Trigeminocervical complex, pain relay disinhibited, attack (hypothesis). Solid borders are published findings; dashed borders are hypothesis.STEP 1CUntreated ADHDAtypical baseline dopamine tone.STEP 2BCariprazine binds D3 receptorsD3 receptors act as dopamine sensors(autoreceptors). Cariprazine binds D3 roughlyten times more tightly than D2. Ref 16.STEP 3CDopamine feedback misreadsThe feedback system can no longer read howmuch dopamine is actually present.STEP 4BCA11 nucleusOnly known source of descending dopamine tothe trigeminocervical complex (Ref 8). Itsbrake output is distorted (hypothesis).STEP 5CTrigeminocervical complexPain relay disinhibited → attack.
Figure 1. Proposed mechanism. Solid borders are published findings; dashed borders are hypothesis. Badges: A = Published — human; B = Published — animal, in vitro or abstract; C = Hypothesis.

My case (n=1)

Dn=1 One person.

I had never had a cluster headache. In 2025 I started cariprazine (Vraylar) at a starting dose. On the third day the attacks began; I stopped the drug that day and never took another dose.

The bout lasted about five weeks; my last attack was on July 24, 2025. Stopping the drug did not end it: the bout ran its course after a single three-day exposure. Early on, a three-day prednisone course stopped the attacks for a time. A five-day course a few weeks later had no effect. Sumatriptan helped partially, and verapamil was added in the final week.

At the time I didn't know I had ADHD. I was diagnosed after the bout ended and was not on any ADHD medication at any point during it. That is the combination this hypothesis is about: untreated ADHD, a D3-preferring partial agonist, and a first-ever bout within days.

What my case can't show: it is one person, and the link is timing, not proof. I was never re-exposed to the drug. I am a long-time smoker, a known risk factor for cluster headache. About 23 years before onset I had three fractures of the right temporal bone and a ruptured right eardrum. Either could have contributed.

One observation that doesn't fit the usual pattern: alcohol, a common trigger during a bout, did not trigger attacks for me.

Predictions that would test this

CHypothesis Each prediction can fail.

  1. If ADHD is measured in cluster headache cohorts, its prevalence will be higher than in matched controls.
  2. Response to dopaminergic drugs will differ by ADHD status. D3-preferring partial agonists will be more likely to provoke attacks in cluster headache patients who also have ADHD, and stimulants will be more likely to reduce them. Two published case reports point that way: methylphenidate aborting refractory attacks within minutes,[17] and the only published case with both diagnoses, where methylphenidate started for ADHD cut the attacks.[18]
  3. If ADHD-stratified cohorts show no difference in dopaminergic drug response, the hypothesis is weakened. I want to say that plainly: this is the result that would send me back to the drawing board.

The gap

Cluster headache comorbidity studies and whether each measured ADHD
StudyYearCohortComorbidities measuredADHD measured
Møen et al., Cephalalgia[11]20267,803 people with cluster headache vs 387,042 matched controls10 categories: hypertension, pulmonary, coronary heart disease, migraine, sleep, facial pain, epilepsy, mood, anxiety, substance abuseNoGAPNot measured
Riedinger et al., European Psychiatry (Leiden LUCA)[15]2025LUCA cohort screened online; sample size not stated in the abstract1 category: bipolar disorderNoGAPNot measured
Figure 3. The gap. Comorbidity studies in cluster headache and whether ADHD was among the conditions measured.

The table lists cluster headache cohorts I have reviewed that measured psychiatric or medical comorbidity, and whether ADHD was among the conditions measured. So far the answer is no in every row: the largest registry study skipped it, and the Leiden group, which was already looking at psychiatric traits, screened for bipolar disorder but not ADHD. This is the gap I am asking someone with the right data to close.

What I'm asking for

If you hold cluster headache cohort or registry data: please check for ADHD diagnoses, or for stimulant prescriptions as a proxy, and compare against your matched controls. That single query would test the first prediction above. I would be glad to be wrong.

If you live with cluster headache and have been told you have ADHD, or suspect it: I would like to hear from you. Contact: [email protected]

References

Numbered in the order they appear in the evidence map. Each entry links to its DOI.

  1. 1.D'Andrea G, Granella F, Perini F, Farruggio A, Leone M, Bussone G. Platelet levels of dopamine are increased in migraine and cluster headache. Headache. 2006;46(4):585–591. doi:10.1111/j.1526-4610.2006.00407.xA
  2. 2.D'Andrea G, Cananzi AR, Joseph R, et al. Platelet glycine, glutamate and aspartate in primary headache. Cephalalgia. 1991;11:197–200. doi:10.1046/j.1468-2982.1991.1104197.xA
  3. 3.D'Andrea G, Cananzi AR, Morra M, et al. Platelet catecholamines in cluster headache. J Neurol Neurosurg Psychiatry. 1992;55(4):308–309. doi:10.1136/jnnp.55.4.308A
  4. 4.D'Andrea G, Leone M, Bussone G, Di Fiore P, et al. Abnormal tyrosine metabolism in chronic cluster headache. Cephalalgia. 2017;37(2):148–153. doi:10.1177/0333102416640502A
  5. 5.D'Andrea G, Gucciardi A, Perini F, Leon A. The role of neurotransmitters and neuromodulators in the pathogenesis of cluster headache: a review. Neurol Sci. 2019;40(Suppl 1):39–44. doi:10.1007/s10072-019-03768-9A(review)
  6. 6.Glover V, Peatfield R, Zammit-Pace R, et al. Platelet monoamine oxidase activity and headache. J Neurol Neurosurg Psychiatry. 1981;44(9):786–790. doi:10.1136/jnnp.44.9.786A
  7. 7.Littlewood J, Glover V, Sandler M, et al. Low platelet monoamine oxidase activity in headache: no correlation with phenolsulphotransferase, succinate dehydrogenase, platelet preparation method or smoking. J Neurol Neurosurg Psychiatry. 1984;47(4):338–343. doi:10.1136/jnnp.47.4.338A
  8. 8.Charbit AR, Akerman S, Holland PR, Goadsby PJ. Neurons of the dopaminergic/calcitonin gene-related peptide A11 cell group modulate neuronal firing in the trigeminocervical complex. J Neurosci. 2009. doi:10.1523/JNEUROSCI.2887-09.2009B(rat)
  9. 9.May A, Bahra A, Büchel C, Frackowiak RS, Goadsby PJ. Hypothalamic activation in cluster headache attacks. Lancet. 1998;352:275–278. doi:10.1016/S0140-6736(98)02470-2A
  10. 10.Jain A, Rørnes AX, Farestveit A, et al. Altered hypothalamic metabolism in chronic cluster headache patients measured with ¹H-MRS at ultra-high magnetic field. J Headache Pain. 2026;27:172. doi:10.1186/s10194-026-02446-4A
  11. 11.Møen EN, Bjørk MH, Gilhus NE, et al. Cluster headache and comorbidities, income and education: a nationwide registry-based matched cohort study. Cephalalgia. 2026;46(3):1–12. doi:10.1177/03331024261428583A
  12. 12.Møen EN, Bjørk MH, Gilhus NE, et al. The prevalence and incidence of cluster headache: a Norwegian population-based time-trend study. Neurology. 2026;106(11):e214862. doi:10.1212/WNL.0000000000214862A
  13. 13.Hagen K. One-year prevalence of cluster headache, hemicrania continua, paroxysmal hemicrania and SUNCT in Norway. J Headache Pain. 2024;25:30. doi:10.1186/s10194-024-01738-xA
  14. 14.Pan P, Jonsson U, Şahpazoğlu Çakmak SS, Häge A, et al. Headache in ADHD as comorbidity and a side effect of medications: a systematic review and meta-analysis. Psychological Medicine. 2022;52(1):14–25 (published online 2021). doi:10.1017/S0033291721004141A
  15. 15.Riedinger MA, Brandt RB, Knapen SE, et al. Cluster headache and bipolar disorders: phenotypical overlap and comorbidity. European Psychiatry. 2025. doi:10.1192/j.eurpsy.2025.1212B(conference abstract)
  16. 16.Kiss B, Horváth A, Némethy Z, et al. Cariprazine (RGH-188), a dopamine D3 receptor-preferring, D3/D2 dopamine receptor antagonist–partial agonist antipsychotic candidate. J Pharmacol Exp Ther. 2010. doi:10.1124/jpet.109.160432B(in vitro and rat pharmacology, not human data)
  17. 17.Mellick GA, Mellick LB. Cluster headache management with methylphenidate (Ritalin). Headache. 1998;38(9):710–712. doi:10.1046/j.1526-4610.1998.3809710.xA(case report)
  18. 18.McBrien S, Fallon J. Methylphenidate as a potential treatment for cluster headache. Progress in Neurology and Psychiatry. 2022;26(4):16–18. doi:10.1002/pnp.765A(case report)