Migraine is a complex phenomenon that likely has multiple causes. However, current thought on the pathophysiology of migraine is that an initial activation of meningeal nociceptors causes a release of substance P and salcitonin gene related peptide (CRGP) from the nociceptive nerve endings, which in turn causes vasodilation and protein extravasation (Dalkara et al. 2006). This process is known as neurogenic inflammation and/or peripheral sensitisation. In peripheral sensitisation the increased sensitivity of the nociceptors to activation may develop to a degree that, in combination with central sensitisation, they are susceptible to activation by the arterial pulse and head movements (Dalkara et al. 2006). In this way, normal physiological activities may become sources of pain. The causes of the initial activation are still not fully described but several possible mechanisms are discussed below.

Accompanying peripheral sensitisation is the phenomenon of central sensitisation. Activation of trigeminovascular nociceptors leads to an increase in activity in the second-order neurons in the trigeminocervical nucleus (Bolay et al. 2002). Furthermore, activation in the trigeminocervical nucleus occurs with stimulation of both meningeal afferents and those from the greater occipital nerve (a branch of the C2 dorsal root), and that stimulation for just 5 minutes can lead to an increase in response in the nucleus for over an hour to other nociceptive stimuli, including if that stimulation is received from the C2 innervation (Goadsby 2005). This is particularly relevant for manual therapy practitioners with their focus on removal of pain generators from the cervical spine. It is likely that this sensitisation involves the activation of NMDA receptors in the nucleus as in vitro application of sumatriptan, an antimigraine drug, is effective in blocking NMDA receptor activation (Buzzi & Moskowitz 2005).

The cutaneous allodynia is often associated with migraine and is thought to be another symptom of this central sensitisation of the trigeminocervical nucleus. The allodynia usually occurs on the ipsilateral forehead to the headache, but has been documented to occur over the ipsi- and contralateral hands as well (Dalkara, Zervas et al. 2006).

Brainstem dysfunction has also been implicated in migraine, both as a primary migraine generator, and through failure of brainstem pain inhibition mechanisms. Goadsby (2005) reports activation of the locus ceruleus and periaqueductal grey (PAG) in migraine without aura, and that these correspond to areas that have generated migraine-like symptoms in patients when stimulated electrically. Similarly, Goadsby reports excess iron has been found in the PAG of episodic and chronic migraineurs, and lesions in this area have been known to cause migraine. The notion of brainstem dysfunction is further supported by the presence of measured balance deficits, vestibular changes (Akdal et al. 2007; Akdal et al. 2009; Asai et al. 2009), and subclinical cerebellar deficits in migraineurs (Sandor et al. 2001).

The PAG is involved in descending pain control mechanisms, along with the locus ceruleus and raphae nuclei. In 2001, Knight and Goadsby (2001) published research that suggests that it is dysfunction in the PAG that leads to disinhibition of the trigeminocervical nucleus and hence pain.

The combination of brainstem dysfunction along with balance and cerebellar deficits presents many opportunities for intervention by the functional neurologist. This could also explain the anecdotal reports from clinicians who have found that cerebellar and vestibular rehabilitation programmes are useful in the treatment of migraine.

The brain of the migraineur displays different neurophysiological behaviours to that of the normal individual. Evoked potential and transcranial magnetic stimulation (TMS) studies have shown that the cerebral cortex of the migraine sufferer shows an impaired ability to habituate to sensory stimulation across all sensory modalities (Schoenen 2006; Brighina et al. 2009). This is thought to be due to impaired functioning of inhibitory networks in the cortex, and it has been suggested that this dysfunction occurs in a manner similar to the GABA circuit down-regulation that occurs in sensory deafferentation (Brighina et al. 2009). Interestingly, Brighina et al. (2009) report that repeated trains of TMS have been shown to improve the functioning of the inhibitory networks that could have long-term effects. This is consistent with the functional neurology paradigm that it is possible to improve neuronal functioning by repeated stimulation within the tolerance of fatigue, thereby promoting gene expression, protein synthesis, and thus building plasticity.

It is logical to assume that this inability to habituate to sensory stimulation could predispose the cortex to excitotoxic events, and evidence suggests that this is the case (Dalkara et al. 2006). This is further supported by evidence that there is impaired energy metabolism in migraineurs (Lodi et al. 2006), which will predispose to anaerobic metabolism. Anaerobic metabolism is toxic to neurons and the metabolic by-products associated with it such as lactic acid are able to activate nociceptive fibres. Functional neurological interventions to promote aerobic metabolism and to protect from anaerobic metabolism should be considered by the clinician. Interventions that may be considered include:

It is also thought that the migraineurs have immunological changes that predispose them to the activation of the inflammatory cascade, which in turn leads to the migraine attack. These triggers may be infections, foods, or other environmental factors (Longoni & Ferrarese 2006). Any management plan for the migraineur should include the identification and, as far as possible, the removal of the triggers. Headache diaries including diet and environmental exposures can be useful for identifying the triggers.