The perception of odour is certainly an extremely subjective experience, or we would all be wearing the same perfume. Being perhaps the less studied of the senses, the mechanisms behind the way a fragrance is perceived is not yet fully understood. A brief mention here is appropriate for two interesting theories that were proposed in the twentieth century, involving a lock-and-key system and vibrational wavelengths [97], which have not yet been fully developed.

There are several substances whose consumption or inhalation has been popularly related to headaches [98,99,100,101]. Remarkably, Umbellularia californica is a type of tree, commonly known as “the headache tree” [102], which contains umbellulone, a ketone that was reported of being capable of triggering cluster headache-like attacks in a gardener with a history of cluster headaches [103]. It was later discovered that this mechanism was mediated by the activation of the transient receptor potential (TRP) ankyrin 1 (TRPA1) [104,105], followed by the release of calcitonin gene-related peptide (CGRP) [104]. CGRP is also released through the activation of vanilloid receptors, following stimulation with nitric oxide [106] or ethanol [107,108], one of the most relevant cluster headache triggers. TRPA1 has also been involved in the responses to some inhaled chemicals, including the smoke of cigarettes [109], chloride [110,111] hydrogen peroxide-containing substances [111] or formalin, the noxious compound largely used in pain models [112].

It has been reported that up to 70% of migraineurs can develop a headache after the stimulation with some odorants, which happened around 25 minutes following the exposure [113], and there is a case report of migraine improvement following the imposition of mandatory masks in the workplace during the COVID-19 pandemic [114]. Increased sensitivity to smells can be part of the premonitory-like symptoms experienced by migraineurs; therefore, certain smells may be misinterpreted as the trigger for a migraine attack, which might not be a necessary factor for its occurrence [115,116]. As a consequence, the results of studies that assess migraine triggers have debatable interpretations.

Nevertheless, the presence of osmophobia may be related to more florid migraine phenotypes and greater disability, and a scale has been developed recently for the quantification of quality of life related to osmophobia [117]. Migraineurs that present with ictal osmophobia may have more painful headaches [118,119]. Ictal and interictal osmophobia have been associated with a longer history of migraines or high frequency of the attacks, as well as other associated symptoms, such as cranial allodynia [120,121,122], suggesting a central sensitization process [123]. Vomiting can also be more common in the presence of osmophobia [119,121]. Osmophobic migraineurs may also have a higher prevalence of psychiatric comorbidities than those without it [118,124,125,126].

Osmophobia has been proposed as a specific marker, helpful for the diagnosis of migraine [119,124,127,128,129,130,131,132]; however, it is not very sensitive [122]. Around half of the patients with migraines reported an increased sense of smell or reduced tolerability to smells [129,133]. Remarkable examples of patients reporting hyperosmia include the smell of a rose from more than 5 meters of distance, or soap from a different room, and the main scents triggers for osmophobia arose from food, specifically fried food and onions, cigarettes or self-care products, and perfume or paint specifically were reported as triggers [133]. More recently, forty percent of patients with chronic migraine reported osmophobia [134], and a similar number suggested odours or perfumes as potential triggers of a migraine attack [101].

Paradoxically, despite their hypersensitivity to smells, migraineurs have a lower capability for the threshold, identification and discrimination of smells [135,136]. Patients with episodic migraine were found to have a similar olfactory acuity to controls, and furthermore, around one fifth of them developed hyposmia during the attack [137]. Taste abnormalities in migraineurs [133] are a matter of debate [138].

Patients with migraine and osmophobia have neuroanatomical alterations. A significantly reduced volume of the olfactory bulb was observed in 1.5 Tesla MRI, compared to patients with other types of headache [139], and might be more pronounced on the left, in comparison with controls [140]. In migraineurs with reported hypersensitivity to odours, regional blood flow in a study using H₂¹⁵O-positron emission tomography was found to be increased in areas of the left piriform cortex and antero-superior temporal gyrus, as compared to controls, both with and without multiple odour stimuli [141]. During odour stimulation, blood flow was found to be decreased in bilateral fronto-temporo-parietal regions, as well as the posterior cingulate gyrus and right locus coeruleus [141]. Another study using fMRI to compare responses to the smell of roses found higher blood oxygen level-dependent activity in the amygdala and insular cortices of the amygdala and also in the midbrain, particularly the rostral pons. However, the smell of roses did not show significant interictal differences compared to the controls [142]. Activation of the amygdala and orbitofrontal cortex might be related, respectively, with the intensity and valence of the smell emotional experience [143]. The amygdala and cingulate cortex also showed abnormal activation in patients with multiple chemical sensitivity [144,145], which is associated with a high prevalence of headache [146] and was observed in up to 20% of migraineurs [147].

Olfactory hallucinations or phantosmia is a hallmark of temporal lobe epilepsy, and currently a no man’s land when it presents in the form of aura. It is a rare symptom, with a reported prevalence of 0.66% in a headache center [148]. The majority of reported cases had normal electroencephalograms that were, however, taken during the interictal period, and usually respond to antiepileptic drugs.

The reported cases showed that the episodes have an average duration of less than 10 min and the onset occurs prior to the migraine attack [148,149]. Patients with symptoms of phantosmia scanned with FLASH and eco-planar imaging MRI techniques showed increased activation of different brain areas associated with the process of the sense of smell, such as the prefrontal, cingulate, temporal or insular cortex MRI activation was inhibited by typical antipsychotics that perform its activity through a wide range of binding receptors [150]. Peripheral blocking activities can alleviate phantosmia [151].

In 1984, Kayan and Hood described how vestibulocochlear symptoms were frequently reported, in up to 60% of patients with migraine, and these can be important or disabling enough for the patient to be the primary reason for referral to a specialist. The incidence of neuro-otological symptoms for migraineurs seemed homogeneous throughout all ages in males, but had a peculiar distribution in females. For women who reported audiovestibular symptoms only when asked during the study, a positive skew distribution could be observed, with the peak situated in the 3rd decade. However, the female patients whose reason of referral was the presence of disabling audio-vestibular symptoms had a peak in the peri-menopausal 5th and 6th decades. This group with disabling symptoms had a higher incidence in males [28]. They compared 80 patients referred for vestibulocochlear symptoms with 500 patients with multiple sclerosis for benign positional vertigo and Méniere’s [28]. Only migraineurs described cochlear sensations, such as tinnitus, distortion of pitch, or hearing loss [28].

The frequency of migraine in Méniere’s disease is higher than in normal subjects, and phonophobia has a high prevalence in these patients, independently of the presence of migraine headache [152].