• Open Access

Migraine-like Episodic Pain Behavior in a Dog: Can Dogs Suffer from Migraines?


Corresponding author: I.N. Plessas, DVM, MRCVS, Department of Clinical Sciences and Services, Royal Veterinary College, Hatfield, UK; e-mail: iplessas@rvc.ac.uk.


Migraines and other primary headache disorders commonly affect people. There is evidence to suggest that migraines can occur in dogs. In this review, we present a dog with paroxysmal episodes that have a striking resemblance to human migraine, and we give an overview of migraine in people. The current classification, clinical signs, and diagnosis in people are discussed, as well as the anatomy of head pain, pathophysiology, pharmacology, and treatment options.


calcitonin gene-related peptide


cortical spreading depression


familial hemiplegic migraine


functional magnetic resonance imaging


nitrous oxide


trigeminal cervical complex


trigeminal nucleus caudalis

Migraine is a common disabling condition in people and is ranked by World Health Organization as number 19 among all disabling diseases worldwide.[1] Migraine is not a new disease; conditions and clinical signs linked to migraines have been described in Babylonian writings dating back to 3000 bc. For thousands of years, little was known about this condition, which was treated with skepticism and superstition. The word migraine derives from the Greek hemikrania, suggesting that the pain is arising from one side of the skull, but although it is often unilateral, it can be bilateral in 40% of adults and 60% of children.[2] Migraine in people is defined as “recurrent headache disorder manifesting in attacks lasting 4–72 hours.” Typical characteristics of the headache are unilateral location, pulsating quality, moderate or severe intensity, aggravation by routine physical activity, and association with nausea, photophobia, or both and phonophobia.[1] Migraines are common in people with a cumulative lifetime incidence of 43% in women and 18% in men.[3]

The high prevalence of migraines and their significant socioeconomical impact has attracted researchers' attention the last 2 decades. Despite the research that has been done so far, further understanding of this condition is limited by the lack of animal models of migraine. Considering that this condition has never been described in animals as a naturally occurring disease, the creation of an ideal animal model is challenging. Animal models, however, have been developed, and experimentally induced signs suggestive of migraine have been produced in dogs, cats, rats, mice, and other animals.[4-7] None of these model systems in experimental animals, though, can explain all of the features or reproduce all of the clinical signs of migraine in people.[8] Considering the complex and poorly understood pathophysiology of migraine, this might not be surprising.

The lack of specific biologic tests can make diagnosis of migraine in people a challenge. Patients must fulfill specific criteria to be diagnosed with this condition.[1] As animals are nonverbal, such a diagnosis would be even more challenging, but considering the high prevalence of migraine in people, and the fact that some clinical signs suggestive of migraine have been reproduced in experimental animals, it might be possible that animals also have the propensity to suffer from migraines or migraine-like episodes. This review discusses a case report of a dog with unusual clinical signs resembling those of migraine in people. An objective diagnosis of migraine was never made; however, if we could apply the criteria for the diagnosis of migraine in people to this dog, it would have fulfilled them.

Case Example

A 5-year-old female neutered Cocker Spaniel was presented with a history of paroxysmal episodes of vocalization and apparent fear since the dog was 6 months old.

One day to 2 hours before the onset of vocalization, the dog would appear fearful and quiet, and would hide under furniture, avoiding any interaction at home. After this initial period, the dog would start vocalizing, as if she was experiencing pain, and would have a low head carriage. She would remain conscious and responsive, and would refuse to eat or drink. Occasionally, she would show behavior suggestive of nausea (hypersalivation, frequent swallowing, “lip smacking”) or vomiting during the episodes. The duration of the vocalization was initially 2–4 hours progressing over time, so that it could last up to 3 days. The initial frequency was 1–2 episodes per year progressing to 1–2 episodes every month. After the episode of vocalization, she remained quiet for 1–2 days before returning to an apparently completely normal condition. The owner of the dog could not identify any specific triggers of these episodes and they could occur at any time of the day or night. Several veterinarians examined the dog during an episode and could not elicit any focal pain. Treatment with opioids (morphine, methadone), diazepam, acepromazine, and nonsteroidal anti-inflammatory drugs (carprofen, meloxicam) failed to reduce the intensity of the episodes.

On initial presentation to the Royal Veterinary College (RVC) Small Animal Referral Hospital, the dog appeared comfortable with an appropriate level and quality of mentation. Her physical and neurological examinations were unremarkable. Extensive serum biochemical, dynamic bile acid, and hematology profiles did not reveal abnormalities. Assessment by a veterinary behaviorist concluded no evidence of primary behavioral disorder.

Differential diagnoses at this stage included a seizure disorder or a paroxysmal pain syndrome. Magnetic resonance imaging (MRI) of the brain and cerebrospinal fluid analysis were all within normal limits. Thoracic radiographs and abdominal ultrasonography did not reveal abnormalities. A trial treatment with phenobarbital1 (3 mg/kg, PO, q12h) was initiated for 2 months, but there was no response in the frequency or intensity of the episodes. Phenobarbital serum concentrations were within the therapeutic range.

At second presentation to the RVC, the dog had been experiencing an episode of apparent pain for 2 days continuously. She appeared distressed, with constant vocalization, as if in pain, and appeared photophobic and phonophobic. Physical and neurological examinations were unremarkable apart from a low head carriage with no detectable spinal pain. An MRI of the whole spine was performed, which was normal.

With the suspicion of a possible neuropathic pain syndrome or severe headache, paracetamol (acetaminophen)/codeine2 (10/0.2 mg/kg, PO, q12h) and pregabalin3 (3 mg/kg, PO, q8h) were trialed, but also failed to improve her clinical signs after 4 weeks of treatment. A focal seizure disorder refractory to phenobarbitone and pregabalin could not be completely ruled out at that stage, but the clinical presentation resembled more pain arising from the head and cervical area. A migraine-like disorder was suspected and topiramate4 (10 mg/kg, PO, q8h) was trialed, which improved clinical signs markedly. The dog continued to experience these episodes, but the duration would last only 1–3 hours and the intensity was dramatically reduced to the point that the dog would only seem quiet, no longer vocalize, and be keen to go for walks, eat, and drink as normal. The dog no longer appeared to be photophobic and phonophobic. Interestingly, the owners reported that if they failed to give topiramate quickly enough and they administered it when she had already started vocalizing, it took longer to recover (6–7 hours); however, the intensity improved within 30 minutes of administration, she would appear more quiet, stop vocalizing and hyperventilating, and would eventually go to sleep. It was decided to give topiramate only when the dog was experiencing the episodes. Eighteen months after institution of this treatment, she continued to respond well. The frequency of the episodes has reduced from 2 episodes/month before the initiation of topiramate, to 1 episode every 2–3 months. The owner perceives that the dog now has a good quality of life, whereas before treatment with topiramate quality of life was considered poor enough that the owners were considering euthanasia.


The unusual presentation of the aforementioned dog has striking resemblance to migraine attacks in people. To the authors' knowledge, a condition with similar clinical signs has not been described in veterinary literature, so we summarize the current literature from human and experimental medicine on migraine.

Classification and Clinical Signs of Migraines in People

The International Classification of Headache Disorders (revision 2)[1] defines 4 primary and 10 secondary groups of headache. Primary headaches include migraine, tension-type headache, trigeminal autonomic cephalalgias, and “other primary headache” (a heterogeneous grouping of primary headaches). Migraine is further divided into 6 major subtypes[1]:

  • Migraine without aura
  • Migraine with aura
  • Childhood periodic disorder
  • Retinal migraine
  • Complications of migraine
  • Probable migraine

Migraines are typically recurrent severe headaches associated with autonomic clinical signs.[9] They can be episodic (less than 15 days per month) or chronic (occurring 15 days or more per month).[2] They can manifest with aura (20%) or without.[2] Aura is a transient phenomenon, which appears usually gradually over 5–20 minutes and lasts less than 60 minutes. It is followed by a headache attack and is characterized by visual (most common), sensory, motor, or both effects. When the aura is not followed by a headache attack, it is called an acephalgic migraine.[10] In some people, a premonitory (or prodrome) and headache resolution (or postdrome) phase can proceed and follow, respectively, a migraine attack with more vague clinical signs of hyperactivity, hypoactivity, depression, craving for specific foods, repetitive yawning, and other less specific clinical signs.[1] The actual headache attack can last 4–72 hours, have unilateral location, pulsating quality, moderate or severe intensity, and they can be associated with autonomic signs (nausea), photophobia, or phonophobia.[1] The aforementioned case was experiencing clinical signs that could reflect all 4 phases, although distinguishing a premonitory phase from aura might not be possible in animals.

Diagnosis in People

Diagnosis of migraine can be challenging. The Headache Classification Committee of the International Headache Society has created the following diagnostic criteria[1] to help the clinician to distinguish migraine from other types of headaches:

  1. At least 5 attacks fulfilling criteria B through D. For migraine with aura, 2 attacks are sufficient for diagnosis
  2. Headache attacks lasting 4–72 hours
  3. Headache has at least 2 of the following characteristics:
    • 1.Unilateral location
    • 2.Pulsating quality
    • 3.Moderate or severe pain intensity
    • 4.Aggravation by or causing avoidance of routine physical activity (eg, walking or climbing stairs)
  4. DDuring headache, at least one of the following:
    • 1.Nausea, vomiting, or both
    • 2.Photophobia and phonophobia
  5. ENot attributed to another disorder

Although these criteria are intended for people and were created after decades of studying and treating humans with this condition, it is interesting to note that if we could apply them to the aforementioned dog, she would fulfill all of them. The dog had experienced numerous attacks (more than 5), the episodes of vocalization could last from 4 hours up to 3 days, the episodes were severe enough to prevent physical activity, as she was refusing to go for walks. She appeared to be photophobic and phonophobic, and had clinical signs suggestive of nausea, occasionally vomiting. Extensive diagnostic investigation could not identify any known cause that could explain her clinical signs.

Diagnostic imaging cannot currently be used for diagnosing migraines. However, functional MRI (fMRI) has shown activation (hyperoxia and blood volume increase) of the red nucleus and substantia nigra during a migraine attack[11] and increased photoreactivity of the occipital cortex during the interictal period in migraineurs.[12] Diffusion-weighted MRI has shown microstructural white matter alterations in migraine patients[13] and Gentile et al[14] have reported MRI findings of cerebral vasogenic edema in a patient in status migrainosus (migraine headache lasting more than 72 hours). Although diagnostic imaging cannot diagnose migraine, it is important to rule out structural brain lesions, and fMRI studies can provide valuable information to improve our understanding of migraine pathophysiology.

Anatomy of Head Pain

The brain parenchyma is insensate. However, the meninges, meningeal blood vessels, cerebral arteries, venous sinuses, cranial nerves, cervical roots, and other head structures (ie, eyes, ears) are innervated and can cause pain.[2] Intracranial structures rostral to the tentorium cerebelli are innervated by the trigeminal nerve, and those caudal by the cervical nerves C2 and C3, facial (CN VII), glossopharyngeal (CN IX), and vagus (X) nerves.[2] The dura mater and meningeal blood vessels have both sensory and autonomic innervations. Unmyelinated afferent C fibers innervating peripheral structures pass through the trigeminal ganglion, enter the pons and synapse on second order neurons at the trigeminal nucleus caudalis (TNC). The TNC extends from the medulla oblongata into the cervical dorsal horns in a functional continuum that includes a cervical extension. Together the neurons from the TNC and its cervical extension form the trigeminal cervical complex (TCC), which is implicated in the largest proportion of primary headaches.[2, 15] Fibers, from the TCC in return, project through the quintothalamic tract and after decussating in the brainstem, form synapses with the thalamus and collaterals to the autonomic nuclei in the brainstem and the hypothalamus. Thalamic neurons project to the somatosensory cortex, but also to areas of the limbic system. In addition, the TNC has polysynaptic connections to the parasympathetic superior salivatory nucleus in the pons. Efferent parasympathetic axons project from this nucleus through the pterygopalatine ganglion and innervate meningeal vessels and contents of the nasal sinuses and eyes.[16] Brain imaging studies suggest that important regulation of the trigeminovascular nociceptive input comes from the dorsal raphe nucleus, locus ceruleus, and nucleus raphe magnus.[15]

Migraine Pathophysiology

Over the past few decades, migraine pathophysiology has attracted interest and has sparked a debate on whether it is a vascular disorder or a primary neuronal disorder. It is now clearer that the vascular changes are considered secondary to a primary brain excitability and a sensory dysmodulation disorder.[2] There are certain triggers that have been associated with a migraine attack in people. These include bright lights, loud noise, certain odors, certain foods (aged cheeses and other) or delaying a meal, hormonal changes, head trauma, and nitrous oxide (NO).[2]

A phenomenon named cortical spreading depression (CSD), which proceeds the headache attack, seems to play an important role in migraine pathophysiology, and there is strong association between this phenomenon and the visual aura that 20% of patients experience before a migraine attack.[17] It is believed that CSD is taking place in the cerebral cortex, cerebellum, and hippocampus.[2] During this phenomenon, the intracellular calcium concentration rises and calcium waves are generated, throughout glia, changing vascular activity. A wave of depolarization moves across the cerebral cortex and NO, arachidonic acid, protons, and potassium are released extracellularly. As a consequence, trigeminal neurons of the trigeminovascular pathway release calcitonin gene-related peptide (CGRP), substance P, and neurokinin A, which causes meningeal vasodilation (after an initial phase of vasoconstriction) and inflammation (sterile neurogenic inflammation). This inflammation activates the first-order trigeminal neurons (peripheral sensitization), which is thought to carry pain signals centrally.[18] After activation of the TNC, the superior salivatory nucleus can be activated too through its polysynaptic connections and subsequently parasympathetic fibers innervating the dural vessels release acetylcholine, NO, and vasoactive intestinal polypeptide. Clinically, the patient experiences pounding pain at this stage, which can be associated with miosis, ptosis, red eye, lacrimation, nasal congestion, and rhinorrhea.[2] If the attack is not treated at an early stage, then it can progress and second- (trigeminothalamic) and third-order (thalamocortical) neurons can be activated (central sensitization or wind-up phenomenon) and clinically can be manifested as cutaneous allodynia (facial/neck pain that occurs spontaneously or in response to nonpainful stimuli).[2]

Migraine is typically hereditary and is usually polygenic. Studies on twins indicate a 34–54% genetic influence on the development of migraines.[19] Despite this influence, no single genes predisposing to common forms of migraine have been identified yet. The only single gene-related headache disorder is a form of migraine, named familial hemiplegic migraine (FHM). FHM is a rare subtype of migraine with aura with an autosomal dominant inheritance pattern and is characterized by transient hemiparesis or hemiplegia during the aura phase of a migraine attack. The remainder of the clinical signs of this condition is similar to the other forms of migraine.[2, 19] Mutations in 3 different genes (FHM1, FHM2, and FHM3) have been identified in families with FHM and have been associated with brain hyperexcitability and CSD, as a result of increased glutamate action.[2]

Migraine and epilepsy share many clinical features and underlying pathophysiologic mechanisms. The initial event preceding CSD is increased cellular excitability associated with localized epileptiform discharges.[20, 21] Glutamate seems to be a critical mediator of hyperexcitability in both focal seizures and migraine. Also, epidemiological studies demonstrate that epilepsy and migraine are comorbid conditions that might have a shared genetic basis.[21]

Pharmacology and Treatment

In 1938, Graham and Wolff demonstrated that intravenous administration of ergotamine during migraine attack decreases the amplitude of arterial pulsations, which happened to coincide with the reduction in headache pain.[2] This inspired the vascular theory that an initial cerebral vasoconstriction followed by an extracranial vasodilation leads to a migraine attack.[22] This theory was the basis for research for many decades until recently. Over the years, the migraine research focused on the development of other vasoconstrictive drugs such as the triptans (eg, sumatriptan), but at the same time, it became apparent that other drugs without any vasoconstrictive properties, such as ibuprofen and CGRP antagonists, were equally effective.[2]

It is now believed that ergot drugs, as well as triptans, bind to many different sites, which can affect neurotransmitter release. These drugs primarily bind to serotonin receptors 5-hydroxytryptamine 1D and 1B (5-HT1D and 5-HT1B).[2] The 5-HT1D receptors are presynaptic and they can be located on trigeminal neurons and the TNC. Activation of these receptors prevents the release of CGRP from the trigeminal neurons, whereas activation of 5-HT1B receptors of blood vessels causes vasoconstriction. The latter receptors in the TNC also decrease central sensitization.[2]

Serotonin plays an important role in the pathogenesis of migraine. Blood serotonin levels decrease at the beginning of a migraine attack, while its metabolite 5-hydroxyindoleacetic acid (5-HIAA) levels rise.[23] Administration of serotonin depleters during a migraine attack causes aggravation of the clinical signs, whereas administration of serotonin on the other hand can abort the attack, but it can cause severe adverse effects, including those resembling carcinoid syndrome (flushing, diarrhea, heart failure, bronchoconstriction).[2] Triptan drugs are basically modified serotonin molecules that cause fewer adverse effects.[2]

Other drugs that are used mainly for chronic migraine can increase the CSD threshold and subsequently the number of CSD events that trigger migraine attacks. Drugs that have this mode of action include topiramate, valproate, amitriptyline, methysergide, and DL-propranolol. These drugs can also be used as prophylactics.[2]

Topiramate has been an effective drug for treatment of chronic migraine[24-26] and for prevention of episodic migraine in people.[27-32] There is also evidence that topiramate inhibits the trigeminovascular pathway in cats.[5] Topiramate has multiple mechanisms of action.[33] It can inhibit voltage-gated sodium channels, high voltage-activated calcium channels, and glutamate-mediated neurotransmission at a-amino-3-hydroxy-5-methylisoxazole-4-propionic acid and kainate receptor subtypes. Topiramate can also enhance GABAA-receptor–mediated chloride flux and modulate trigeminovascular signaling.[4] It remains unclear which of these mechanisms of action contributes the most to the treatment and prevention of migraine, but research on pathophysiology and genetics of migraine suggests dysfunction of multiple systems, so topiramate may contribute in multiple ways.[27]

In Search of an Ideal Animal Model

Migraine has a vast welfare and economical impact on people and significant research has been done to investigate the complex pathophysiology and discover effective drug treatments. A great limitation in all this research is the lack of an animal model that can replicate all the features of migraines.[8] Understanding of migraine pathophysiology has improved markedly over time, and can now be divided into many components, and then recreated within animal models.[8]

There are 3 groups of animal models for migraines: vascular, neurovascular, and molecular/genetic animal models.[8, 34]

The vascular models use vasoconstriction or vasodilation of relevant vessels (carotid arterial bed, arteriovenous anastomoses, and other meningeal arteries) to study vascular changes during a migraine attack and the effect of drugs on them. Pig and dog models have been used in the past, by constricting the external carotid artery.[35] The major limitation of these animal models is that it does not reflect the complexity of a migraine and focuses only on the vascular changes, which, as aforementioned, are an epiphenomenon and not the main cause.[8, 36]

The neurovascular models focus on many different aspects of migraine pathophysiology. One of them targets the expression of a specific protein called Fos (from Finkel, osteogenic, sarcoma)[8] within the TNC as a marker of trigeminal nociceptive stimulation. Although Fos protein can be expressed at low levels in many normal cell types, including neurons and glial cells, it is thought that continuous or deregulated expression might result in cellular transformation.[37] Fos protein expression within the TNC is induced by applying mechanical, electrical, or chemical stimuli in tissues innervated by the trigeminal nerves. This has provided us with valuable maps of functional migraine pathways.[8] A major limitation of these Fos gene expression models is that they can only be as good as the stimulus that drives their expression. Four different substance P-neurokinin-1 receptor antagonists failed to improve or prevent migraine attacks, despite them blocking Fos expression in the TCC.[38-40] Other models in this group target the CSD as a model for migraine aura. Finally, intravital microscopy (studying the dural vasculature in vivo through a small cranial window with the help of microscopic video) has also been used with or without laser Doppler flowmetry, which studies meningeal blood flow.[41-43]

The last group of animal models (molecular/genetic models) include mice with the FHM gene (FHM1) mutation that have been developed either naturally or transgenically, and they express a variety of clinical signs, including ataxia and seizures.[8]

In summary, if animals suffer naturally occurring migraines, and experience most, if not all the features of migraine in people, they would be ideal models for study. If dogs do suffer from migraines or other primary headache disorders, they could provide invaluable information about the pathophysiology of this complex condition.

The current review of the literature does not confirm the existence of migraines in dogs. The dog presented here has shown a combination of unusual clinical signs, which could be suggestive of a migraine or another severe headache disorder. Other differentials for this dog include focal seizure disorder refractory to phenobarbitone and pregabalin, or an unusual somatic pain syndrome refractory to opioids and nonsteroidal anti-inflammatory drugs. Electroencephalography was unfortunately not able to be performed on this dog. The dog had apparently normal mentation during the episodes, had no focal or generalized involuntary movements, and failed to respond to trialed antiepileptic drugs (phenobarbitone and pregabalin). Although this does not rule out a seizure disorder, we thought that a migraine-like disorder could possibly be a better explanation of this dog's clinical signs. The apparently marked response to topiramate, which has been used in humans for management of chronic migraine or prevention, may also support our suspicion of a migraine-like disorder. Many drugs have been used for management of acute migraine in people, including triptans, which are thought to be more effective.[10] In our case, we avoided using triptans because of the lack of safety, pharmacokinetics, and efficacy data in dogs.

The diagnostic approach of similar cases should include taking a detailed history (including a diary of events and video footage), and thorough physical, ophthalmic, and neurological examination. Any foci of apparent pain should be investigated. A behavioral disorder should be ruled out by consulting a clinical behaviorist. An extensive extracranial (blood work, abdominal, and thoracic imaging), intracranial (MRI, cerebrospinal fluid analysis) investigation, and electroencephalography, where applicable, should be performed. If there is suspicion of a seizure disorder, an antiepileptic drug trial could be performed, bearing in mind that there are refractory seizure disorders. If the aforementioned investigation is unremarkable and antiepileptic/analgesic medication fails to improve the clinical signs, then a primary headache disorder should be considered. In such cases, the diagnostic criteria for primary headache disorders in people could be used as a guide. Further understanding and characterization of such conditions and clinical treatment trials will be necessary to validate this approach.

The aim of this review was to summarize the current literature and to create awareness that migraine-like disorders may exist in dogs, and that there are medications available that may be used for the management of similarly frustrating presentations. For centuries, migraines were treated with substantial skepticism and superstition, and it is only during the last few decades that it has been recognized as a severe disabling disease. Further case reports, studies, and an “open mind” will be required to determine whether these conditions are likely to occur in our patients. Even today, migraine and other headache disorders can be difficult to diagnose in people, and will likely be more problematic in animals despite a thorough and accurate history taking and extensive physical and diagnostic examinations. A better understanding of these potential conditions may help solve many disorders characterized by perplexing “episodes” and perhaps provide us with an ideal animal model for primary headaches in the future.


The authors thank the referring veterinarian and the owner of the dog, as well as all the staff of the Royal Veterinary College Small Animal Referral Hospital who contributed to the dog's procedures and care. The authors gratefully acknowledge the Research Office of the Royal Veterinary College for assessing the manuscript according to the Royal Veterinary College's code of good research practice (authorization number: VCS_00519).

Conflict of Interest Declaration: Authors disclose no conflict of interest.


  1. 1

    Epiphen; Vetoquinol UK Limited, Buckingham, UK

  2. 2

    Pardale-V; Dechra Limited, Shropshire, UK

  3. 3

    Lyrica; Pfizer Limited, Kent, UK

  4. 4

    Topamax; Janssen-Cilag Ltd, Bucks, UK