Primary headache disorder is characterized by recurrent headaches which lack underlying causative pathology or trauma. Primary headache disorder is common and encompasses several subtypes including migraine. Vestibular migraine (VM) is a subtype of migraine that causes vestibular symptoms such as vertigo, difficulties with balance, nausea, and vomiting. Literature indicates subjective and performance-based cognitive problems (executive dysfunction) among migraineurs. This study compared the magnitude of the total effect size across neuropsychological domains to determine if there is a reliable difference in effect sizes between individuals with VM and healthy controls (HC). An additional aim was to meta-analyze neuropsychological outcomes in migraine subtypes (other than VM) in reference to healthy controls.

This study was a part of a larger study examining neuropsychological functioning and impairment in individuals with primary headache disorder and HCs. Standardized search terms were applied in OneSearch and PubMed. The search interval covered articles published from 1986 to May 2021. Analyses were random-effects models. Hedge’s g was used as a bias-corrected estimate of effect size. Between-study heterogeneity was assessed using Cochran’s Q and I2. Publication bias was assessed with Duval and Tweedie’s Trim-and-Fill method to identify evidence of missing studies.

The initial omnibus literature search yielded 6692 studies. Three studies (n=151 VM and 150 HC) met our inclusion criteria of having a VM group and reported neuropsychological performance. VM demonstrated significantly worse performance overall when compared to HCs (k=3, g=-0.99, p<0.001; Q=4.41, I2=54.66) with a large effect size. Within-domain effects of VM were: Executive Functioning=-0.99 (Q=0.62, I2=0), Screener=-1.15 (Q=3.29, I2=69.59), and Visuospatial/Construction=-1.47 (Q=0.001, I2=0.00). Compared to chronic migraine (k=3, g=-0.59, p<0.001; Q=0.68, I2=0.00) and migraine without aura (k=23, g=-0.39, p<0.001; Q=109.70, I2=79.95), VM was the only migraine subgroup to display a large effect size. Trim-and-fill procedure estimated zero VM studies to be missing due to publication bias (adjusted g=-0.99, Q=4.41).

This initial attempt at a meta-analysis of cognitive deficits in VM was hampered by a lack of studies in this area. Based on our initial findings, individuals with VM demonstrated overall worse performances on neuropsychological tests compared to HCs with the greatest level of impairment seen in visuospatial/construction. Additionally, VM resulted in a large effect size while other migraine subtypes yielded small to moderate effect sizes. Despite the small sample of studies, the overall effect across neuropsychological performance was generally stable (i.e., low between-study heterogeneity). Given than VM accounts for 7% of patients seen in vertigo clinics and 9% of all migraine patients, our results suggest that neuropsychological impairment in VM deserves significantly more study.