We read with interest the article by Kim et al.1 on a population-based study of 1,431 patients with non-migraine headaches and 507 patients with migraines for the presence of visual aura. Headaches were diagnosed using an internet-based headache diagnosis questionnaire.1 A visual aura was reported in 26% of patients with migraine compared to only 14.5% of non-migraine headaches.1 Among those with non-migraine headaches, those with visual aura reported more headache days and higher headache-related disability than those without aura.1 The study is excellent but has limitations that should be discussed.

The main limitation of the study is that headaches were diagnosed over the internet and not in person.1 Patients with headaches require a detailed individual and family history as well as a thorough clinical neurological examination. In order to distinguish primary from secondary headache types, it is imperative that all patients with headaches undergo cerebral imaging and angiography of cerebral arteries and veins. Because presence or absence of a visual aura and the underlying pathophysiology can be highly dependent on the non-migraine headache subtype, it is important to know how many of the non-migraine headache sufferers have infectious, inflammatory, vascular, traumatic, neoplastic, or structural causes.2

A second limitation of the study is that the current medications that the included patients were regularly taking were not considered to be a causative or contributory factor in the visual aura or headache. There are several illicit drugs that can be complicated by visual hallucinations, such as amphetamines, cocaine, LSD, or ecstasy. There are also numerous drugs that can cause visual phenomena as side effects, such as cephalosporines, steroids, beta-blockers, gabapentin, barbituric acid, ethosuximide, and several others.3

A third limitation of the study is that only the visual aura was examined. Several other types of auras have been reported in migraine and non-migraine headache sufferers. The most common of these are sensory (e.g. tactile disturbances) and aphasic aura. There is also an aura that manifests as muscle weakness, brainstem symptoms (brainstem aura), or retinal manifestations (retinal migraine).4

A fourth limitation of the study is that the sensitivity of the VARS headache score, used to assess presence or absence of visual aura, is only 78.4%.1 Therefore, recognition of the correct diagnosis was limited and 22% of patients diagnosed with visual aura may actually have none.

A fifth limitation is that different types of headaches can occur at the same time. How did the authors rule out that the included patients had different types of headaches, or developed a new type of headache during the observational period?

Although aura is most commonly associated with migraine,1 aura is also a common manifestation of epilepsy, often precedes motor symptoms and is a warning sign for the affected individual. Aura often allows the patient to avoid serious complications (e.g. falls) during the unconscious phase of the ictus. In epilepsy patients, an aura can also occur without consecutive convulsions. Therefore, we should know how many of the included patients suffered from epilepsy in addition to headaches. Knowing the number of epileptic patients is not only important because seizures may be preceded by headaches or, more commonly, postictal headaches may follow seizures, but also because of the possibility that a headache itself represents a seizure.5 Secondary headaches, in particular, can be complicated by seizures and seizures can begin with an aura. It is therefore important that epilepsy was ruled out as the cause of the aura in all included patients. How many of the included patients had epilepsy and how many were regularly taking anti-seizure drugs?

Overall, the interesting study has limitations that put the results and their interpretation into perspective. Addressing these issues would strengthen the conclusions and could improve the status of the study.