Clinical manifestations

Presentation and course

Clinical manifestations vary depending on the exposure to or withdrawal of each individual medication or substance. Although many reports of headache induced by medication or substance in the literature do not use ICHD diagnostic criteria, their use is important to prove association.

The clinical manifestations of these headache syndromes can be summarized by considering the diagnostic criteria that describe each syndrome. These criteria are as follows:

The IHS Classification Committee stated that double-blind, placebo-controlled experiments that demonstrate a greater incidence of headache after active drug compared to placebo are the only true basis for attributing headache as a side effect to the medication.

Although many reports of medication-induced or substance-induced headache exist, a consistent clinical profile of these syndromes does not always exist. Although some medications (eg, cyclosporine, dipyridamole, niacin) typically induce headache with migraine phenotype, others (eg, nitroglycerin, calcitonin gene-related peptide, and ethanol) may induce headache with a range of features or tend to precipitate the primary headache type a person typically experiences. Other medications and substances (eg, heroin) may have tension-type headache associated with their use.

Prognosis and complications

Discontinuing the substance exposure or the medication believed to be causing the headache leads to headache termination. Exposure to particular substances, such as cocaine or carbon monoxide, may leave persisting neurologic dysfunction after the acute neurologic syndrome and accompanying headache have resolved.

Headache-related disability in people with medication overuse headache is higher than in people with other headache types. Persons with medication overuse headache are likely to suffer during nonworking days with total or significant disability and to have significantly reduced effectiveness on days spent at work (122).

Clinical vignette

A 39-year-old woman presented with headache occurring at least 6 days per week for the past 4 years. She experienced migraine without aura as a teenager. These attacks resolved by 20 years of age, and she reported only occasional “stress headaches” that were consistent with episodic tension-type headache during her 20s. At 35 years of age, she noted recurrence of her prior migraine attacks and the onset of headache consistent with chronic tension-type headache. No provocation was identified. Initially, over-the-counter medications, including acetaminophen and ibuprofen, controlled the headache. During the next 12 months, the patient’s headache increased in severity, and her intake of nonsteroidal anti-inflammatory drugs (NSAIDs) and acetaminophen increased to as many as eight tablets per day on a daily basis. Her headache severity fluctuated, with headache duration always greater than 8 hours, and averaging 12 hours, each day. The headache was described as holocephalic, with pain described as nonthrobbing and achy in nature. Over the next 3 years, headache severity and frequency increased, and on presentation, the patient was ingesting approximately four different headache medications, including a caffeine-containing compound and a butalbital-containing compound. In the prior year, she used prescription medications twice, for several months each time; these included meperidine and a codeine-containing compound. She even tried marijuana at the insistence of a friend, and initially found relief from her tension-type headache for several hours after smoking it. However, after a few months of marijuana use, she found that her headaches intensified slightly while smoking, and she also reported redness of her eyes, dryness of her mouth, and a warm feeling over her body. She discontinued marijuana use, and the symptoms resolved. The patient’s headaches caused her to miss 5 days of work per month and had caused difficulties with her domestic situation.

Neurologic examination, including funduscopy, was normal. Magnetic resonance imaging of the brain was unremarkable, and routine blood studies were normal.

The patient was hospitalized. The Raskin protocol, using consecutive and recurrent doses of intravenous dihydroergotamine preceded by metoclopramide, was used for 4 days. After several doses without a clinical response, intravenous chlorpromazine was added to the regimen, replacing metoclopramide. After minimal response, intravenous methylprednisolone was also added to the regimen, and the patient became headache-free 6 days after admission. Discharge medications included amitriptyline 10 mg at bedtime to be titrated upwards as tolerated for prevention, naproxen sodium 550 mg up to twice per day as needed for moderate attacks, and rizatriptan 10 mg up to twice per day as needed for severe attacks, with naproxen and rizatriptan usage limited to 10 days per month total. Control was moderately successful over the next 9 months, with headache still occurring 8 to 10 days per month. Her work productivity improved, and she required infrequent over-the-counter analgesics one year after her initial visit, with headache frequency reduced to 5 to 6 days per month.

Biological basis

Etiology and pathogenesis

Alcohol (IHS 8.1.4). Alcohol may provoke headache within 30 to 45 minutes of ingestion. This correlates with peak blood alcohol levels, which suggests that headache is probably not caused by intracranial vasodilation due to this latency period (186). The mechanism by which alcohol causes headache is not fully understood but is felt to be complex and likely involves cortical, subcortical, and brainstem structures (169). The mechanism of “hangover headache” has been debated and is complex. Animal data suggest that the formation of acetate as a result of ethanol metabolism is key (151).

Wine has been noted as a cause of migraine in different parts of the world, with red wine being a more common migraine trigger than white wine in England, and white wine being a more common migraine trigger than red wine in France and Italy (147). White wine with red color added is more likely to induce headaches (147). In a study, persons with migraine who reported red wine to be a trigger were challenged with red wine or vodka; red wine provoked headache in 82% as compared to none who were given vodka; however, neither form of alcohol provoked headache in controls nor in persons with migraine who had not described red wine as a trigger for headache (139). Phenolic flavonoids may be the ingredient in red wine that may trigger a migraine attack (138). Given the inconsistency of alcohol as a trigger (or nontrigger), variability of resultant headaches, cultural factors, and unclear pathogenesis, Dueland concluded in a 2015 review that there appears to be no reason to recommend that patients with headache should avoid alcohol in general; an individualized approach is preferred (56). A web survey of 2197 persons with migraine revealed that 35.6% report alcohol as a trigger, and 25% avoid alcohol altogether due to presumed triggering effects. The most commonly reported trigger was wine, particularly red wine, at 77.8%, but red wine led to acute attacks consistently for only 8.8% of them (167).

Table 1. IHS Diagnostic Criteria for Alcohol and Alcoholic Beverage-Induced Headache: Immediate or Delayed

Antihistamines and histamine (IHS 8.1.7). A number of antihistamines have been associated with headache with migraine features (25). Second-generation antihistamines, such as fexofenadine, (levo)cetirizine, and (des)loratadine are commonly associated with the adverse event of headache (158). Histamine has also been shown to trigger headache with migraine features (13; 25) and cluster features (61). Histamine infusion may cause hypotension followed by a severe, short-lasting headache that is believed to be due to intracranial vasodilation (155).

Histamine inhalation leads to headache that resembles migraine without aura in both persons prone to migraine and controls (127). Histamine may trigger headache via endogenous formation of nitric oxide by vascular endothelial nitric oxide synthase (NOS) (43). It is believed that the H3 receptor plays a major role in histamine and headache (243). IHS criteria for histamine-induced headache are listed in Table 2.

Table 2. IHS Diagnostic Criteria for Histamine-Induced Headache

Caffeine (IHS 8.3.1). Diagnostic criteria for caffeine-withdrawal headache are presented in Table 3.

Table 3. IHS Diagnostic Criteria for Caffeine-Withdrawal Headache

Caffeine is a cerebral vasoconstrictor, and its withdrawal probably leads to vasodilation with vascular congestion (50). In a double-blinded study, patients who normally had moderate caffeine intake were given either placebo or caffeine. Fifty percent of the patients who were given placebo had headaches compared to 6% of the patients who were given caffeine (210). Seventy-five percent of caffeine-dependent teenagers develop caffeine-withdrawal headache (162). Cola intake in children and adolescents has been blamed for daily or near-daily headache. In a study, gradual withdrawal from cola drinks led to complete cessation of all headaches in more than 90% of subjects (98). Next to simple analgesics, caffeine and caffeine-containing tablets are the second most commonly overused drugs in chronic daily headache (119). Caffeine-withdrawal headache may not always be related to high consumption or variation in consumption (211).

Caffeine withdrawal has been blamed for perioperative headache as well. Headache secondary to caffeine withdrawal due to cessation of intake prior to and during the surgical event has been implicated in a review of 287 patients who underwent minor surgical procedures under general anesthesia (83).

Transcranial Doppler measurements of cerebral blood flow velocities demonstrate significant increases in velocity following caffeine withdrawal periods, with subsequent decreases seen within a half hour of caffeine intake. Whether blood flow derangements may play a role in caffeine withdrawal headache or are epiphenomena remains to be elucidated (45).

Caffeine is associated with headache due not only to its withdrawal, but also with regular use, which may be related to its mild analgesic properties, or possibly via mechanisms related to adenosine (206). Caffeine consumption is much higher in patients with chronic daily headache than in population controls prior to the onset of chronic daily headache (199) but not during chronic daily headache. This is particularly true of young women with chronic episodic headaches (199). This may suggest that caffeine intake is a risk factor for development of chronic daily headache or, alternatively, that a drop in caffeine intake after development of chronic daily headache may be associated with a continuing headache syndrome. A population study showed that high dietary caffeine consumption is associated with a modest increase in headache prevalence (87).

Two reviews detail the role caffeine may play in headache, especially migraine (04; 160).

Nitric oxide (NO) donor-induced headache (IHS 8.1.1). Nitroglycerin has been described as a well-known cause of headache, as have amyl nitrate and isosorbide dinitrate (13). The use of endocervical isosorbide dinitrate gel solution for treatment of missed abortion has been associated with development of headache in 60% of patients, compared to 17% of patients receiving misoprostol (12). Exposure to nitroglycerin in the form of dynamite led to headache in munitions workers (202). Fifty percent of patients taking nitroglycerin have noted headache, which may reflect the development of tolerance (46). Nitroglycerin-induced headache has features similar to those of migraine without aura. Nitroglycerin also has induced attacks of cluster headache (61). In persons with cluster headache, cluster attacks may be induced by nitroglycerin during a cluster period (100). Nitric oxide seems to be the key chemical leading to headache, and the reasons for this have been reviewed extensively (166). Functional neuroimaging of migrainous headaches triggered by nitroglycerin revealed that the premonitory phase of the attack was associated with activations in posterior hypothalamus, midbrain tegmental area, periaqueductal grey, dorsal pons, and various cortical areas including occipital, temporal, and prefrontal cortices (143). IHS criteria for nitric oxide donor headache are listed below, with both immediate and delayed forms.

Table 4. IHS Diagnostic Criteria for Nitrate- and Nitrite-Induced Headache

Foods and additives. Many foods and additives have been causally linked to headache, specifically migraine. Many persons prone to headache are able to produce a list of foods that they avoid because they fear getting an attack. However, only a few studies have examined food and food additives, and most of these have been strictly observational. Criteria have been created for headache associated with food components and additives.

Common foods and food additives believed to induce headache have included citrus fruits, chocolate (145), dairy products (174), tyramine (90; 25), sodium chloride (33), sodium nitrate (97), monosodium glutamate (197), aspartame (103), and sucralose (23; 172). This is based mostly on anecdotal case reports, though some diary studies and double-blind, placebo-controlled trials have been conducted. Such studies of chocolate have revealed no statistically significant association with any form of headache (157; 79). The notion that true food allergies may also trigger migraine remains suspected but unproven; however, IgG testing may be of some clinical utility (76).

Monosodium glutamate, commonly found in soy sauces and food additives, has been described as a cause of the so-called “Chinese restaurant syndrome” in as many as 30% of people (190; 13). It has been noted to cause a diffuse, throbbing headache within 30 minutes of ingestion, associated with tightness and burning sensations over the face and chest. A dull, nonthrobbing headache may also occur (190; 246). Similar symptoms have occurred with intravenous injection of monosodium glutamate, and placebo-controlled studies have shown a variety of symptoms in one third of people receiving monosodium glutamate (110; 109). A placebo-controlled, double-blinded study confirmed the phenomenon of the monosodium glutamate symptom complex, a triad consisting of facial pressure, chest pain, and burning in the head and upper trunk (246). A systematic review of available trials concluded that evidence is insufficient to deem a causal relationship given difficulties with blinding and inconsistent results among studies (161). The mechanism of monosodium glutamate-induced symptoms is unknown.

The so-called “hot dog headache” occurs shortly after eating frankfurters or cured meat. Nitrites, added to meat to give a red color, have been implicated as the cause. “Hot dog headaches” are typically bitemporal, nonthrobbing, and moderately severe, often associated with facial flushing (97). IHS criteria place hot dog headache under nitrate- and nitrite-induced headache (IHS 8.1.1).

The synthetic sweetener aspartame, potential etiology of “chewing gum headache” (26), was studied in a double-blind manner and shown not to be associated with headache (200), whereas a similar double-blind study suggested susceptibility to aspartame in patients who self-report it (217). Aspartame has been noted to block the normal response in brain tryptophan, 5-hydroxytryptamine, and 5-hydroxyindoleacetic acid after carbohydrate ingestion, which may explain its ability to induce headache (200; 115), and 8% of migraineurs report aspartame as a migraine trigger (137).

Monoamine oxidase inhibitors have been associated with headache under certain circumstances. When these agents are combined with amphetamines, selective serotonin reuptake inhibitors, ephedrine, other sympathomimetic agents, or foods that have a high tyramine level, they have induced severe headache in addition to hypertension (25).

Hormonal agents. Oral contraceptives and other female hormonal preparations can exacerbate a previously existing headache syndrome (usually migraine) or cause headache (02). Although it has been previously reported that women may develop new-onset migraine, an exacerbation of migraine, or a change in the pattern of migraine during oral contraceptive use and that the attacks frequently remit when the oral contraception is terminated (19), a systematic review was more controversial regarding any association between oral contraceptives and migraine. There was no consistent relationship between oral contraceptive use and headache except in those women with a prior history or family history of migraine. As well, there is no evidence for a dose effect of estrogens on headache. For those women who note headache beginning or worsening in the first month of oral contraceptive use, headache prevalence seems to fall during the second (1 in 3), third (1 in 10), and subsequent months (140).

The pathophysiology of estrogen-induced migraine is unclear, but it may relate to estrogen’s actions on vascular smooth muscle or modulation of vasoactive neuropeptides and neurotransmitters at the intracranial neurovascular junction (19). In addition, disruptions in central prostaglandins, serotonin, and prolactin metabolism may lead to alterations in vasoregulation within the brainstem and changes in cerebrovascular tone, lowering the “threshold” for migraine in prone individuals (201). Data suggest there is interplay of several complex mechanisms modulating nociception (86). Oral contraceptives, estrogen preparations, and progesterone preparations have also been associated with idiopathic intracranial hypertension (80), as has the gonadotrophin analogue leuprorelin acetate (11). The association of oral contraceptives with idiopathic intracranial hypertension remains controversial, but patients reported by Davidson were slim females in contrast to the usual body morphology of obesity in idiopathic intracranial hypertension (51; 22). The mechanism of oral contraceptive-induced idiopathic intracranial hypertension may relate to impaired venous sinus drainage due to occult thrombosis, as cases of venous sinus thrombosis have been reported with oral contraceptive therapy and hormonal replacement therapy (184; 216; 170). Another theory of oral contraceptive-induced idiopathic intracranial hypertension is increased salt and water retention (52).

In addition to headaches induced by its use, estrogen withdrawal following cessation of exogenous estrogens (such as during the pill-free interval of combined oral contraceptives or following a course of replacement or supplementary estrogen) may also induce headache and migraine (19).

Table 5. IHS Diagnostic Criteria for Estrogen Withdrawal Headache

The antiestrogen tamoxifen, used frequently to treat breast cancer, has shown conflicting effects, either alleviating (163) or worsening (148) migraine. It is suspected that the reason for amelioration of migraine in those who benefit from tamoxifen is calcium uptake inhibition or decreased levels of prostaglandin E (163).

Although danazol has been used successfully in the management of catamenial headaches (36; 229), it has also been described as a cause of idiopathic intracranial hypertension (205).

Cocaine (and other illicit drugs) (IHS 8.1.6 and IHS 8.3.2). As many as 60% to 75% of cocaine users report severe headaches related to cocaine use (54). Cocaine has a number of potential side effects, including intracerebral infarction or hemorrhage, tremor, and headache (132; 131; 107; 130; 214). The “crack” form of cocaine has also been associated with headache and with cerebrovascular complications (130). Headache patterns associated with cocaine include a bitemporal, pulsating, short-duration headache that appears soon after cocaine use, a migraine-like headache with binging cocaine use, and chronic daily headache with cocaine withdrawal (137; 54). A hemiplegic migraine syndrome can also occur after cocaine use (137). Severe occipital headache due to subarachnoid hemorrhage secondary to ecstasy use has been reported (15). The basis of headache associated with cocaine and cocaine preparations is probably secondary to sympathomimetic vasoconstrictive effects (54); however, metabolic alterations, including abnormal cerebral glucose metabolism (231) and reduced levels of dopamine D2 receptors (230), have been described. Cocaine withdrawal headache may be due to disrupted serotonin metabolism (54). Rare cases of cocaine-induced vasculitis have been associated with headache (118; 154; 35). Headache in cocaine users has been reviewed (17).

Headache due to illicit drugs may have more specific etiologies. Cocaine- and amphetamine-induced headaches may be related to a rapid cocaine surge, leading to a rapid block of presynaptic norepinephrine reuptake with potent sympathomimetic effects and acute vasoconstriction. Dysfunction of norepinephrine and vasomotor control may also play a role (54). Prolonged cocaine use may lead to secondary presynaptic serotonin depletion, leading to increased severity of headache with cocaine use. Illicit drug-withdrawal headaches may result from a rebound of central functions after CNS depression of various physiologic systems initially modified by the drug (101).

Table 6. IHS Diagnostic Criteria for Cocaine-Induced Headache

Withdrawal from cocaine (54; 231) and amphetamines (54; 25) can lead to migraine-like headache.

Methamphetamine injection can lead to either ischemic or hemorrhagic forms of stroke, which are frequently associated with symptoms of headache. Although uncertain, injection of methamphetamine may produce delayed stroke secondary to vasculitis, which may present with concurrent symptoms of headache (165). Methamphetamine-induced hemorrhagic stroke does not demonstrate any pathological evidence of vasculitis or aneurysm formation (153). Additionally, chemicals and solvents used in the manufacturing of methamphetamine in clandestine drug labs can cause headache in roughly 31% of exposed individuals (116). These chemicals can leach into nearby apartments/structures and remain present in structures even after the manufacturing of the drugs have ceased.

Heroin-induced cerebral arteritis is rare but has been associated with headache (113). A headache that resembles tension-type headache with a diffuse, continuous, and persisting presentation may follow heroin intake.

Opioid withdrawal can be associated with a frontotemporal, pulsating, long-lasting migraine-like headache (53).

Table 7. IHS Diagnostic Criteria for Headache Attributed to Opioid Withdrawal

Sixty-five percent of marijuana-addicted headache sufferers experience pain relief when they smoke marijuana at headache onset (07). Coital headache and stroke related to cannabis use is rare (06). Mild frontal headaches related to marijuana use are reported along with conjunctival irritation, paresthesia, mouth dryness, and sensations of warmth (62).

In a report, psilocybin was reported to cause delayed, transient headaches in nonheadache volunteers in a dose-dependent fashion (105). Psilocybin is a traditional hallucinogen chemically related to LSD (lysergic acid diethylamide), which incidentally was discovered in the process of engineering new migraine medications). Interestingly, it has been reported to be of potential benefit in cluster headache. In this study, not originally intended to examine headache as an adverse event of psilocybin, the researchers noted that delayed headache was frequent (in about 50% to 90% of subjects, dose-dependently), tended to occur around 8 hours after dose administration, and was more intense with higher doses of psilocybin. Very few of the 18 participants had any headache history (one with one to two migraine attacks per month, one with one to five headache attacks per year, two with a few headache attacks associated with perimenopause, and one with a remote history of a 6-month headache). A mechanism for the headache remains speculative, but given the timing may involve NO, histamine, or calcitonin-gene related peptide (CGRP) (known to cause both immediate and delayed headaches), or may be serotonergic in nature.

Toxic gases (eg, carbon monoxide) (IHS 8.1.3). Carbon monoxide poisoning may occur in warehouse workers (63) and divers (41); sources in the home can include the kitchen and the bathroom (57). Both carbon monoxide and carbon dioxide poisoning have been associated with early nonspecific symptoms of headache, irritability, poor concentration, and poor judgment. Chronic carbon monoxide exposure has occurred when faulty heaters or a snow-impeded gas exchange has caused poor ventilation. Headaches associated with carbon monoxide have been described as typically frontal (66%), dull (72%), and constant (74%) and have shown improvement with normobaric oxygen administration (89). One speculated mechanism of headache in this situation is vascular dilation of cerebral venous structures. Head pain correlates with the concentration of carbon monoxide.

Air pollution has also been linked to headaches (159). Although cigarette smoking has been associated with cluster headache, a causal relationship has not been established. However, subsequent data from a survey study revealed that secondhand cigarette smoke exposure during childhood may be linked to the later development of cluster headache (194). Still, the nature of this association has yet to be elucidated and may be due to shared genetic factors. Many with migraine also report sensitivity to tobacco smoke.

Table 8. IHS Diagnostic Criteria for Carbon Monoxide (CO)-Induced Headache

Acetanilid, in excessive amounts, has caused headache due to hypoxemia, as it converts hemoglobin to methemoglobin (25). Nitrous oxide is also a possible cause of idiopathic intracranial hypertension (25).

Acute poisoning with ethyl alcohol, carbon tetrachloride, benzene, arsenic, lead, anticholinesterases, and insecticides has induced headaches of unknown mechanism (207).

Phosphodiesterase inhibitors (8.1.2). Sildenafil, vardenafil, and tadalafil, used to treat erectile dysfunction, act by inhibiting phosphodiesterase 5, an enzyme that degrades cyclic guanosine monophosphate (cGMP). Common adverse effects of sildenafil are headache and flushing, which occur in 16% of men using the sexual enhancement drug (37; 146; 248). Cilostazol, a phosphodiesterase 3 inhibitor, also induces delayed migraine-like headache, purportedly via cAMP release (24; 85).

Table 9. IHS Diagnostic Criteria for Phosphodiesterase (PDE)-Induced Headache

Calcitonin-gene related peptide (CGRP) likely plays a role in the pathogenesis of primary migraine. Infusion of CGRP is noted to trigger immediate or delayed headache in those with migraine, with or without aura (126; 91; 84).

Table 10. IHS Diagnostic Criteria for Calcitonin Gene-Related Peptide (CGRP)-Induced Headache

Another vasoactive peptide, pituitary adenylate cyclase activating peptide-38 (PACAP38), also has been shown to trigger migraine-like attacks in persons with migraine and in healthy controls. The headache tends to be delayed in those with migraine (203). Prostacyclin also has vasoactive properties and is reported to cause both immediate and delayed headache (not necessarily migraine-like) in those with migraine, cluster headache, or no headache history. Activation and sensitization of sensory afferents around extracranial arteries may underlie this phenomenon (173; 237). Another vasoactive prostanoid, prostaglandin E2, also can induce immediate or delayed dull or migraine-like headache in subjects with and without migraine (238; 10), and other prostanoids are reported to have similar, varying effects (239; 240).

Headache due to other substances listed under miscellaneous are also classified as listed in Table 11.

Table 11. IHS Diagnostic Criteria for Headaches Due to Medications Used for Other Indications

Anesthetic agents. Nitrous oxide may precipitate or exacerbate pain in sinus or ear structures during surgical interventions (193). Ketamine and nitrous oxide have been associated with development of idiopathic intracranial hypertension (102). When patients undergoing sinus and ear surgeries are given nitrous oxide, headache may occur. This may be due to increased pressure, as nitrous oxide equilibrates with air in the sinuses and the ear (193). Intrathecal administration of anesthetic agents has led to aseptic meningitis (102).

Anticonvulsants. The anticonvulsant carbamazepine has been noted rarely as a cause of aseptic meningitis (102). In an open-label study of beaded carbamazepine oral therapy in patients with bipolar disorder, headache was the most commonly noted side effect (111). The similar compound oxcarbazepine was reported to cause refractory headache in three patients with epilepsy and migraines who were transitioned off carbamazepine; the headaches abated after discontinuation of oxcarbazepine (178). Valproate withdrawal, but not induction, has been associated with exacerbation of migraine (52).

Gabapentin, an anticonvulsant used in headache prophylaxis, was reported to be associated with headache in 21% of epilepsy patients in an open-label study (20). Lamotrigine, a drug used for bipolar disorder, had headache as its most common adverse effect (31). One report attributed aseptic meningitis with severe headache to anticonvulsant sensitivity syndrome due to lamotrigine (142). The agents lacosamide, rufinamide, and perampanel list headache as a common adverse event (44; 03; 114).

Antibiotics and antimalarials. A number of antibiotics have been implicated in idiopathic intracranial hypertension, including tetracycline (233; 13; 52), trimethoprim sulphamethoxazole, nalidixic acid, ofloxacin, nitrofurantoin (13), minocycline, doxycycline (74), and ampicillin (25).

Drug-induced aseptic meningitis has been associated with cotrimoxazole (117; 81; 25), sulfasalazine, penicillin, amoxicillin, isoniazid, pyrazinamide, ciprofloxacin (81), and cephalosporins (48). The topic has been reviewed in more detail (247).

Linezolid, an antibiotic within the oxazolidinone class of drugs, was associated with headache in a review of trials using the medication (175). Other antibiotics and antimalarials associated with headache include amphotericin, griseofulvin, chloroquine, and ethionamide (207).

Mefloquine syndrome, a posthepatic syndrome due to the antimalarial drug mefloquine, has been associated with a number of nonspecific features that include headache (49).

Anti-parkinsonian agents. Both bromocriptine and dopamine have been associated with migraine-like headaches (52; 25; 69), particularly with higher doses of levodopa. Pramipexole is associated with an increased incidence of headache compared to placebo; this may be due to increased central monoamine sensitivity interfering with antinociceptive processes (179).

Asthmatic agents. Antiasthmatic agents have been implicated as a cause of migraine-like headaches in numerous references. In particular, theophylline, aminophylline, and terbutaline have been noted as a cause of headaches with vascular features (186). However, in a randomized study, the incidence of headache associated with aminophylline use was similar to that of placebo (58). In a randomized controlled trial of beclomethasone and theophylline for mild to moderate chronic asthma, theophylline caused significantly more headaches (188). In a randomized controlled trial, headache was noted to be a common side effect of budesonide, an inhaled corticosteroid (245). One report described three patients who used budesonide for the treatment of Crohn disease and developed idiopathic intracranial hypertension (129). An agent in asthma therapy, zafirlukast, a leukotriene antagonist, was associated with headache in 13% of patients (212). Omalizumab, a humanized monoclonal antibody that prevents binding of IgE to mast cells and other effector cells, was associated with side effects, including ill-described headache, in 15% of patients studied (16).

Cardiovascular agents. Antihypertensive agents associated with headache include atenolol, hydralazine, prazosin (also used in prostatic hypertrophy) (13), and calcium channel blockers, such as nifedipine (25; 244), mibefradil (65), and nitrendipine (77). An open-label study of nifedipine preparations were associated with headache in 20% of patients (244). Nifedipine’s headache-inducing ability is probably due to vasodilation, with potent relaxation of vascular smooth muscle (250). Sudden withdrawal of clonidine, propranolol, or labetalol has precipitated both hypertension and headache (18). Antiplatelet agents and blood flow agents, including dipyridamole and pentoxifylline, have been associated with migraine-like headache and also are suspected to cause headache through a vasodilatory mechanism (25). Amiodarone is a rare cause of idiopathic intracranial hypertension (71). There is a case report of long-duration hypnic headache (a rare disorder affecting older individuals, characterized by regular nighttime attacks of short-lasting, featureless headaches) precipitated by lisinopril withdrawal and reversed after resuming lisinopril (59). Niacin, a lipid-lowering agent, has also caused acute vasodilation and migrainous headache (186). Other lipid-lowering agents from various classes also have headache as a reported adverse event (218). Monoamine oxidase inhibitors have been associated with headache when combined with amphetamines, selective serotonin reuptake inhibitors, ephedrine, other sympathomimetic agents, or foods having high tyramine levels (25).

Gastrointestinal agents. Agents used to treat peptic ulcer disease and gastroesophageal reflux disease, including the histamine receptor (H2 receptor) blockers cimetidine (08) and ranitidine (52), and proton pump inhibitors omeprazole (124) and lansoprazole (52), have been associated with particularly severe, persisting headaches with migraine features. In a double-blinded multicenter trial, cimetidine was associated with headache in 5% of patients, significantly more than placebo (08). Famotidine and ranitidine are rarely reported etiologies of aseptic meningitis (13).

Isosorbide-5-mononitrate, used in therapy of gastroesophageal varices and ulcers in cirrhotic patients, was associated with a significantly increased incidence of headache compared to placebo (75).

Wood creosote, an herbal antidiarrheal and antispasmodic agent, was associated with mild headache in 13% of normal volunteers (121).

Immunomodulatory agents. Vaccination has been reported to be a migraine trigger or an exacerbating factor, particularly in those with existing migraine disease. Vaccinations for polio, measles, mumps, rubella, and hepatitis B have been implicated as etiologies of aseptic meningitis (25). Vaccination against SARS-CoV-2 is associated with a two-fold increased risk of headache within seven days of vaccination (38). This appears to be true across the various vaccine types, suggesting that the etiology of this headache is secondary to a systemic immune response (38).

Intravenous immunoglobulin may trigger migraine, usually in those with migraine disease, but it has more commonly been associated with aseptic meningitis (21).

Cyclosporine as a cause of headache has been described under oncologic agents.

Azathioprine may be a rare cause of aseptic meningitis (13).

Corticosteroids, including withdrawal from corticosteroid therapy, have often been described as causing idiopathic intracranial hypertension (25); they have occasionally been described as causing aseptic meningitis with intravenous methylprednisolone (21). Intranasal corticosteroids may trigger migraine-like headache (180). Withdrawal from corticosteroids can lead to worsening of existing headache syndromes or the development of a new headache syndrome (25).

Headache is a common side effect of interferons used in the treatment of hepatitides and multiple sclerosis; this is often worse in the first month of therapy (189). As many as 40% of patients using pegylated interferon alpha-2b for treatment of hepatitis B complain of headache (227). Ribavirin use has also been found to be associated with headache (164), as has the use of FK-506 and thalidomide (204).

Neuropsychiatric agents. Headache was an uncommon side effect of selective serotonin reuptake inhibitors in a metaanalysis (60). A metabolite of trazodone, M-chlorophenylpiperazine, activates 5-hydroxytryptamine 2B and 2C receptors. Similar activation with selective serotonin reuptake inhibitors may explain headache that occurs with their initiation and withdrawal (IHS 8.4.3) (34). Rapid venlafaxine withdrawal may be associated with severe headache (152).

Benzodiazepines, particularly the shorter-acting ones, can trigger headache (186). In a small, placebo-controlled study, a small number of patients noted headache with preoperative midazolam and lorazepam given nocturnally; however, preoperative anxiety may have played a role in this (150). A possible association between idiopathic intracranial hypertension and lithium has been noted (195). In an open-label study of lithium in children and adolescents with bipolar disorder, 74% of patients noted headache (171).

Oncologic agents. A number of agents used in cancer therapy have been described as a cause of headache. These included retinoic acid, trimethoprim-sulfamethoxazole (13), cimetidine (08), corticosteroids and their withdrawal (25), and tamoxifen.

Cyclosporine, an antineoplastic agent most commonly used for immunomodulation, caused neurologic side effects in 15% to 40% of patients (106). Cyclosporine neurotoxicity, a form fruste of hypertensive encephalopathy, has been correlated with headache, posterior encephalopathy syndrome, visual dysfunction, and changes in level of consciousness (80; 241). The headache induced by cyclosporine often has migrainous features, but a syndrome resembling idiopathic intracranial hypertension has been reported (102). The neurotoxicity associated with cyclosporine is nearly completely reversible with termination of the drug (106).

Tacrolimus, similar to cyclosporine in its immunomodulation, has side effects similar to cyclosporine (106; 155). However, tacrolimus was more commonly associated with the development of headache than cyclosporine in a systematic metaanalysis (234). One third of patients who received parenteral tacrolimus developed headache, whereas oral administration was only rarely associated with headache (73). Tacrolimus-induced encephalopathy may occur in as many as 20% of patients receiving stem cell transplantation, and headache is often a prominent symptom (108).

OKT3 is another immunosuppression agent that has side effects similar to cyclosporine (27). OKT3 has been associated with four different headache syndromes: (1) cytokine release syndrome, a dose-dependent syndrome that occurs within 1 hour of OKT3 administration; it lasts several hours and has clinical features of headache, fever, dyspnea, nausea, vomiting, and chills (182); (2) aseptic meningitis, which occurs in 3% to 5% of patients within 3 days of OKT3 initiation (70); (3) a nonspecific, unclassified headache (27); and (4) the syndrome of headache, neurologic deficit, and cerebrospinal fluid lymphocytosis (HaNDL) (220).

Anagrelide, an agent used to decrease platelet counts and the risk of thrombosis in myeloproliferative disorders, has headache as a side effect; this is hypothesized to be due to vasodilatory effects (176). Capecitabine and 5-fluorouracil can both cause headaches, possibly due to their known propensity to induce vasospasm. The calcium channel blocker diltiazem was reported to abolish capecitabine-induced headache, which may substantiate this theory (185).

The use of 9-cis-retinoic acid, a high-affinity ligand for retinoid X receptors and retinoic acid receptors in the treatment of Kaposi sarcoma and cervical dysplasia, has been associated with intolerable headache at doses greater than 100 mg/m2 per day (01; 05).

Three chemotherapeutic agents (methotrexate, cytarabine, and diaziquone) have been described as a cause of aseptic meningitis (30; 102). Intraventricular gentamicin and antineoplastic infusions have been associated with aseptic meningitis as well.

Intrathecal delivery of liposomal cytarabine in children has led to development of arachnoiditis, characterized by fever, headache, nausea, vomiting, and back pain. Headache is dose limiting in up to 25% of patients, although concurrent dexamethasone may be beneficial in limiting development of arachnoiditis (28).

Radiographic contrast agents. A large number of idiosyncratic reactions have been noted with the use of ionic contrast agents in diagnostic radiographic procedures. One of the more uncommon side effects is headache; more common reactions include anaphylactoid-like reactions and cardiovascular, flushing, and gastrointestinal complications (40).

Intrathecal contrast media has rarely been associated with aseptic meningitis (102).

Oral cholecystographic media, including iopanoic acid and ipodate, may induce gastrointestinal side effects, dysuria, skin rash, and headache (40).

Topical ophthalmic therapy. Phenylephrine, a sympathomimetic agent used to dilate the pupil, has associated side effects of headache, hypertension, tremor, and diaphoresis (128; 82; 235).

A small study of topical prostaglandin F2 alpha for use in glaucoma found that 50% of patients had mild ocular pain or headache (64).

The use of latanoprost 0.005% plus timolol 0.5% in open-angle glaucoma therapy has headache as the most common adverse effect (144).

Vitamins or related agents. Etretinate (29) and retinoids, including isotretinoin, have been associated with idiopathic intracranial hypertension (25), as have both an excess and a deficiency of vitamin A (52). An excess of vitamin A can cause both acute and chronic headache (186). Acute headaches occur 4 to 8 hours after massive ingestion of vitamin A (78). Excessive vitamin A can be caused by ingestion of large amounts of polar bear liver, either as a food fad or as treatment for dermatological conditions. Daily intake in excess of 7.5 mg of retinol (less in children and infants) has led to neurotoxicity. The most common association with headache is idiopathic intracranial hypertension, particularly in infancy (25; 222).

Miscellaneous agents. Perhexiline, benzene hexachloride (215; 228), beta-human chorionic gonadotropin hormone, growth hormone, synthetic luteinizing hormone-releasing hormone, anabolic steroids, and thyroxine (226) have been described as rare causes of idiopathic intracranial hypertension.

Allopurinol, muromonab CD-3, and phenazopyridine may have a role in aseptic meningitis (13).

Methoxamine, a selective alpha-1-adrenoceptor agonist used in urinary stress incontinence, showed a significant incidence of headache in a randomized controlled trial (183). Another urological agent, tolterodine, was associated with headache in a systematic analysis examining randomized trials with the medication (47).

In a small study of primary dysmenorrhea with an open-label, crossover, controlled design, glyceryl trinitrate was associated with a significantly greater incidence of headache than the NSAID diclofenac, with 33% of patients discontinuing therapy with glyceryl trinitrate due to headache (68). The etiology has been related to vasodilation due to the vasorelaxation effects of nitric oxide. Headache associated with glyceryl trinitrate tends to be immediate, along with a delayed migraine attack (225).

Carmustine administration has been associated with headache and flushing, suggesting a vasodilatory effect leading to headache (52).

Intrathecal injection of baclofen is a rare cause of aseptic meningitis (13). Withdrawal of various agents, including prothipendyl, octreotide (249), or benzodiazepines (25), has also precipitated or worsened migraine, with some of these headaches due to medication overuse.

Development of headache following the use of an extended-release dipyridamole and aspirin combination is common (33% to 39%), although these are self-limited, with a demonstrated 69% placebo efficacy at 2 hours (133; 135). Twenty-six percent of patients taking dipyridamole cannot tolerate the therapy, usually due to headache (66). It is possible that headache due to dipyridamole may be dose-related, and smaller doses may produce less common and less severe headache, perhaps increasing compliance (133). In fact, regular dosing of dipyridamole (200 mg orally twice a day) is associated with higher prevalence of headache (39%) when compared to once-daily dosing for up to 14 days followed by twice daily dosing (39).

A study noted that essential oils (camphor, clove, eucalyptus, fennel, sage, tea tree oil, and thujone) in toothpaste caused new onset cluster headache in five individuals; cluster headache resolved after stopping the use of toothpaste with essential oils (149). It is postulated that the mechanism of action for this is stimulation of a receptor, which then results in the release of calcitonin gene-related peptide (CGRP) from the trigeminal nerves (149). This is similar to the mechanism of action suspected after exposure to the Umbellularia californica, also known as the “headache tree,” which induces cluster headache through umbellu-lone contained in its leaves (149).

An Iranian dried flower extract, Echium amoenum (Boraginaceae) has been used in the treatment of anxiety and depression and was associated with development of headache in a small placebo-controlled study (196).

Agents used for obesity management, such as sibutramine (213) and lorcaserin (219), have headache as a commonly reported side effect. Semaglutide, a drug used for both diabetes and weight loss, was shown to cause headache in 15.2% of study participants (242).

Naltrexone, an opioid antagonist, has been associated with headache and somnolence at a dose of 100 mg (104).

Teriparatide is a drug used for osteoporosis that reproduces the 34 N-terminal amino acids of parathormone. It may be associated with headache development in one study (09).

Headache was reported as a delayed (up to 2 days later) adverse reaction in 10 of 50 patients receiving intravenous iron polymaltose (88). In most cases, the headache was at least moderate in intensity with migrainous features, and in some cases, the headache was disabling for several days.

The smoking cessation agent varenicline should be used with caution in persons with migraine, particularly those with underlying mood disturbances because not only can it induce headache, it can have serious neuropsychiatric consequences, including suicidal ideation (92).

Peripheral sensitization with altered neurotransmitter sensitivity may play a role in the mechanisms of headache development due to substance exposure, as may central sensitization to certain substances. Altered pain regulation systems in chronic substance use may play a role in withdrawal headache syndromes. Resetting of the pain control mechanisms may occur with continuous high doses of a substance in headache-prone patients, and subsequent withdrawal may lead to enhanced nociception. Alterations in gene expression, such as with c-fos with intense neuronal stimulation, may lead to enhanced pain perception with substance exposure. Alterations in c-fos, an intermediate gene product, may lead to further changes in other gene products, including neurotransmitters, neuropeptides, and receptors (181).

Nitric oxide-associated headaches may be partly due to the sensitization of perivascular sensory nerve endings via neurogenic inflammation or free radical formation (42). Nitric oxide may also interact with N-methyl-D-aspartate (NMDA) receptors, leading to hyperalgesia (221). Ketamine, an NMDA antagonist, and fentanyl may be beneficial in treating and preventing headaches induced by intravenous nitroglycerin in cardiac patients. Also, a nitric oxide synthase inhibitor was associated with headache relief during spontaneous migraine attacks in a placebo-controlled study (125). Nitroglycerin-induced histamine release may also play a role in migraine development; however, nitroglycerin-induced headache fails to respond to mepyramine, a histamine antagonist (100).

Epidemiology

In general, the epidemiology of medication- or substance-induced headache depends on the substance and the type of headache. Medication as a cause of headache is likely unrecognized and may be the cause of headache in as many as 8% of patients with headache (67; 249). In a German clinical trial center, all adverse events occurring in healthy volunteers over 1 year among all clinical studies were recorded, and headache (2.23%) was the most common adverse event reported (141). In Europe, the incidence of drug-induced headache has been reported as being 5% to 10% of all headaches (156). A random telephone survey conducted in Canada suggested that 1.5% of persons with migraine had medication overuse headache due to analgesics (192). One of the major problems with defining the epidemiology of medication- or substance-induced headache is the inability to prove an association between substance and headache in many cases. Fisher actually refuted the idea of medication-induced headache, but the patients he documented were taking pain-relief medications for arthritis, suggesting that medication-induced headache tends to occur in headache-prone patients (72). The fact that placebo can also induce headache indicates that headache can sometimes occur with the expectation of side effects of a therapy (145). Headache itself is also common, with one study suggesting a one-day point prevalence of 11% of men and 22% of women, so its occurrence with a particular substance may often be coincidental (187). Similarly, primary chronic daily headache occurs in 4% to 5% of the general population, and nearly one third of the population with chronic daily headache had migrainous features suggestive of chronic migraine (198). Despite these problems, some epidemiological data do emerge.

Most headache experts agree that persons with migraine and tension-type headache have a greater risk for drug-induced headache. Patients who use dihydroergotamine daily for arterial hypotension or NSAIDs daily for musculoskeletal difficulties, prophylaxis of stroke, or myocardial infarction have a 2% to 5% risk of developing chronic headache (32; 123). Of patients who present with headache secondary to medication overuse, the initial headache syndrome that led to the medication use is most commonly migraine without aura (71%), followed by migraine without aura coexistent with tension-type headache (11%), migraine with and without aura (10%), and episodic tension-type headache (8%) (191). One predictor of medication overuse headache in patients with chronic migraine is the previous overuse of analgesics, even a decade earlier (251).

Medication overuse headache affects approximately 1.5% of the general population, mostly women. In specialized headache centers in Europe, 4% to 10% of patients have it, and in the United States, it affects 50% to 80% of patients in tertiary centers. Community sampling reveals 18% of patients with chronic tension-type headache and 32% of patients with chronic migraine from acute overuse of medication (55).

Patients with medication overuse headache may have a shorter mean duration of education than patients with uncomplicated migraine, with an inverse association between extent of education and duration of medication overuse headache (14). Low socioeconomic status may also be associated with development of medication overuse headache (14).

In a study, only 8% of children with chronic daily headache could be classified as having probable medication overuse headache, although 16% used analgesics daily and difficulty attaining precise classification may have underestimated analgesic overuse (236).

In a large health maintenance organization, 21% of headache patients managed by primary care physicians were chronic or frequent (more than 14 days per month) users of symptomatic headache medications (232). In this same study, chronic use of over-the-counter medication (15.9%) was twice that of chronic use of prescription medication (7.7%). Chronic medication use was best predicted by the number of headache days at baseline. Using IHS criteria, 90% of persons with migraine reported the use of over-the-counter drugs, and 1.5% of persons with migraine had medication overuse headache resulting from ergotamine tartrate or analgesics (192).

Kulkarni and colleagues also reviewed the topic of medication overuse headache (122).

As many as 60% to 75% of cocaine users report severe headaches related to cocaine use (54).

Carbon monoxide poisoning occurs in warehouse workers (63) and divers (41); sources in the home can include the kitchen and the bathroom (57).

Prevention

In general, headaches due to substance or medication use can be prevented by avoiding exposure or use.

Medication overuse headache can be prevented by patient education and by limiting analgesic use.

Differential diagnosis

The main differential diagnosis is from primary headache syndromes and other causes of secondary headache syndromes. A strong association between medication or substance exposure and headache, along with the absence of headache with avoidance of the substance, gives strength to the diagnosis of medication- or substance-associated headache. If the association between substance and headache is inconsistent, other headache syndromes should be considered and appropriate investigations performed. Acute headache onset in cocaine or parenteral drug abusers may indicate intracranial hemorrhage, stroke, endocarditis, or brain abscess (54).

Diagnostic workup

The most important aspect in the diagnosis of medication- or substance-induced headache is a careful history regarding headache and medication use or substance exposure. An appropriate past history and a normal neurologic examination provide the clinical diagnosis, and no investigations are indicated. If there is no obvious nonorganic cause for headache, clinical reevaluation is necessary. If clinical reevaluation suggests another cause of medical or neurologic illness, appropriate diagnostic testing, including radiographic imaging, is indicated.

Due to the possible risk of intracranial hemorrhage, ischemic stroke, endocarditis, or brain abscess, radiographic imaging of patients who have used illicit drugs may be useful. Urine drug screens help determine or identify illicit drug use.

Patients with the clinical picture of aseptic meningitis require investigation to rule out other causes of meningitis.

Management

Discontinuing exposure to or use of the substance is the best management of medication- or substance-induced headaches. Simple analgesics may be necessary for symptomatic therapy depending on the causative substance or medication. Appropriate medications may be attempted for migraine-like or cluster headache syndromes due to medication or substance. Aseptic meningitis requires appropriate investigations and analgesic medications.

Carbon monoxide poisoning may require treatment with hyperbaric oxygen therapy, which may have some benefit for headache due to carbon monoxide exposure. Sumatriptan was used with good effect for a patient who had carbon monoxide-induced headache (136).

Headaches associated with illicit drug use require discontinuation of the offending illicit drug and may require detoxification in a drug dependence unit.

Headache related to tacrolimus therapy may improve with substitution of sirolimus therapy (224).

Acetaminophen is ineffective in the management of headaches following the use of an extended-release dipyridamole and aspirin combination (135).

Special considerations

Pregnancy

Many of the substances described should be avoided in pregnancy; otherwise, pregnancy does not pose specific challenges in medication- or substance-associated headache in most cases.

Anesthesia

Particular anesthetic agents may be associated with headache as described in the Etiology section. These include nitrous oxide (193), ketamine (102), and intrathecal administration of anesthetic agents (102).