In the present review, we aimed to explore and summarize the current literature on neuropsychological difficulties in children and adolescents with migraine, attempting to describe patients' cognitive profile in the interictal phase of the migraine cycle. Moreover, in order to better understand patients' global needs, we investigated the literature data on the association between headache features, duration of disease and cognitive impairment.

Given the results issued from studies in adult patients with migraine, we hypothesized that: (1) neuropsychological difficulties, particularly those involving memory and attention, may be common also in pediatric migraine, and (2) impairment of cognitive abilities may depend on frequency of migraine attacks and duration of disease.

Over the last decades, several authors aimed to study the general intelligence in children with migraine, reporting discrepancies depending on the nature of the tests administered (non-verbal vs. verbal) (Table 1). In 1989, D'Andrea et al. explored the general non-verbal intelligence and reasoning in a small cohort of 20 children with migraine (age range 7–11 years) (37). By using “Raven's Colored Progressive Matrices” (CMP), the authors demonstrated normal cognitive performances and the absence of difference in intelligence levels between patients with migraine and control subjects. Normal general non-verbal intelligence and reasoning were confirmed by Riva et al. (38). In particular, comparing the performance of children suffering from migraine with (n. 17) and without aura (n. 31), the authors did not identify significant differences in the distribution of Progressive Matrices scores between the groups of patients. In a more recent study, Margari et al. evaluated the non-verbal intelligence abilities, including fluid intelligence, attention skills and non-verbal memory, in pediatric patients with primary headache (39). Thirty-five children suffering from migraine (n. 26) or TTH (n. 9) and 23 healthy subjects were evaluated with the cognitive battery “Leiter-3” (57). In line with previous studies, the headache group did not show global cognitive deficits (the IQ mean value in patients was 97.6 ± 11.3). In particular, no statistical significant difference in the non-verbal IQ mean value was found between patients and control group. The study did not find a significant correlation between the attack frequency and duration of headache and Leiter-3 scores.

While results of studies evaluating non-verbal intelligence agree in showing normal values in children and adolescent with migraine, findings obtained from global intelligence tests show a peculiar profile characterized by subclinical difficulties in verbal skills (14, 34, 40, 45). Results, however, are largely heterogeneous (Table 1). Waldie et al. explored the cognitive performances of headache patients from the age of 3–26 years (34). To the best of our knowledge, the authors were the first to evaluate the general intellectual ability of children and adolescents with migraine and TTH, focusing on different and specific cognitive domains (verbal and performance abilities). Using the “Wechsler Intelligence Scale for Children-Revised” (WISC-R) (58) at ages of 7, 9, 11, and 13 years, the authors observed lower verbal skills (in particular verbal comprehension) in patients with migraine compared not only with free control subjects (Verbal-IQ mean score = 100.6 ± 11.4 and 105.7 ± 13.5, respectively), but also with patients with TTH (mean score = 104.4 ± 13.8). Low verbal receptive performances were also found in 3 years old migraine patients tested with “Peabody Picture Vocabulary test” (PPVT) (59), which evidenced a strong correlation with the mean WISC Verbal-IQ (p < 0.001). Data analysis showed that the verbal performance in patients with migraine did not decline with age and remained lower than other groups over the time. On the other hand, verbal expression measured by the “Illinois Test of Psycholinguistic Ability” (ITPA) (60) did not differ between the headache groups. The cognitive abilities were not related to duration of disease, use of medications or duration and severity of migraine attacks. These findings led the authors to suppose that the low verbal abilities evidenced in migraine group could not be imputable to a cause-effects relationship between history of migraine and cognitive performance, but was rather due to a “not-identified” common risk factor presumably occurring in a prenatal developmental phase. As alternative hypothesis, the authors suggested that the lower performances could be due to patients' attention and concentration difficulties (34).

The absence of association between the intelligence profile and migraine features in children and adolescents was also confirmed in a more recent study (45). The authors described the risk of lower verbal performance in young patients with headache diagnosis (20 children with migraine and 10 with TTH, age ranging from 11 to 14 years) in comparison with healthy children. Patients' group showed a normal TIQ (mean = 112.80 ± 13.4), but exhibited a characteristic neuropsychological functioning regarding memory and verbal skills. The authors identified lower scores in the “Working memory” and, though not statistically significant, in “Verbal Comprehension” indexes of WISC-IV test (61). Duration of headache and frequency of the attacks did not influence the results. Moreover, patients with migraine and TTH had a similar cognitive profile.

In pediatric patients with headache, lower resources in verbal skills and in total intelligence were also reported by Parisi et al. (40) who compared the functioning of children diagnosed with migraine without aura (n. 63) and TTH (n.19) with that of a control healthy group (n. 79) by using the WISC-R edition (58). The authors found lower total quotient (TIQ) scores in children with migraine than in control subjects, although no difference between headache groups was found. Moreover, the study confirmed the lower performance in verbal quotient (VIQ), in particular comprehension subscale, in children with migraine and TTH, as compared to control group. On the other hand, no difference between groups was found in the performance scale (PIQ). In disagreement with the previous study by Waldie et al. (34) the authors showed a significant negative correlation between the frequency of attacks and total, verbal and performance intelligence quotients of WISC test and between the age of headache onset and the total intelligence quotient score. It was suggested that cognitive impairment in headache, particularly migraine, may be due to repeated hyperactivation of several cortical or subcortical neuronal networks during the attacks.

A different intelligence profile between migraine and TTH patients was described by Esposito et al. (41). The authors confirmed a normal intelligence development in patients with primary headache, although, in disagreement with previous literature, they showed weak perceptual organization competences in patients with migraine as compared with both TTH patients and healthy controls. Abnormalities in frontal lobe structure and impairment of executive functions have been suggested as causal mechanisms.

Studies focusing on receptive and expressive language in young patients with headache are sparse. To the best of our knowledge, in addition to the above mentioned study by Waldie et al. (34), only two other studies examined specific interictal language abilities in children/adolescents with migraine. In 2016, Costa Silva et al. evaluated semantic and phonological fluency (“Verbal Fluency test”) in migraine patients (n. 28) and healthy children (n. 26) (42). In patients with migraine, there were significant lower scores in semantic verbal fluency (“animals” subtest), but not in phonemic verbal fluency. Petrusic et al. (43) recruited 74 adolescents (age range 13–19 years) who suffered from migraine with aura. Compared with healthy subjects, performances in verbal fluency, but not in language, were lower in patients than in control subjects. The study did not show a significant association between the neuropsychological performance, including language and verbal fluency, and both higher cortical dysfunctions and aura frequency/duration.

Several studies investigated memory abilities in children and adolescents with migraine by using standardized test (Table 1). In an early study, D'Andrea et al. (37) described impaired short and long-term visuo-spatial (“Rey Memory test”) and verbal (“Logical Memory test,” the “t, and 10 word learning test”) memory. Several years later, by using the Kaufman Assessment Battery for Children (K-ABC) (62), Haverkamp et al. (44) found cognitive performance within the normal range in children with migraine (n. 37, mean age = 10 years) and in their siblings (n.17, mean age = 8.81 years). The authors, however, described lower performance in subtests requiring memory competences, as in sequential information processing tasks (SEQ). No neuropsychological vulnerability in patients with aura was revealed. Considering that similar scores were found in both migraine patients and their sibling, the authors hypothesized a sample artifact and concluded that no general risk for defective cognitive development exists in young patients with migraine. Using the structured test “Rey Auditory Verbal Learning test” (RAVLT) (63), deficits in verbal memory and learning were reported (42). Compared with healthy subjects, migraine patient performance was more influenced by distractors. Difficulties in verbal memory and, in particular, registration, consolidation, and recall of verbal stimuli were found. As discussed by the authors, since these functions are associated with the ability to use schemas to seek information and organize thoughts, their impairment may lead to poor school performance. The results of a study by Chiappedi et al. (45) provided confirmatory evidence that adolescents affected by headache (migraine and TTH) may have short-term memory difficulties. Although a specific structured evaluation of memory was not performed, cognitive test WISC-IV showed specific impairment in the “Working Memory” index, especially concerning the subtest “Digit Span,” which provides a measure of the ability to store, manage and manipulate information indispensable for language and learning. The study failed to find statistically significant differences in working memory abilities according to headache diagnosis, frequency of attacks, and duration of headache.

In disagreement with previous studies, Margari et al. (39) reported the absence of significant differences between patients with primary headache and controls in “Forward Memory,” “Reverse Memory” and “Non-verbal Memory” subtests scores, issued from the cognitive non-verbal test “Leiter-3” (57). However, the study described a significant association of frequency of the attacks and headache disability with non-verbal memory abilities, suggesting that higher headache disability may have a significant impact on specific cognitive domains, such as the non-verbal memory.

Focusing on migraine with aura, Petrusic et al. (43) showed significantly lower performance in memory skills in adolescents suffering from migraine with aura compared with healthy subjects. As for language, the authors did not evidence significant association between the presence of higher cortical dysfunctions, frequency and duration of aura and the memory skills (evaluated by the “ACE-R test”) (43, 64).

Data on the role of migraine features on memory abilities in pediatric patients are sparse (Table 1). To the best of our knowledge, in addition to the above mentioned studies (39, 45), only Calandre et al. (32) explored this topic. In a sample of 28 adults and 32 pediatric patients with migraine, the authors described an impairment of short- and long-term visual memory, verbal memory, and working memory, related to frequency of the attacks and headache history. Multiple hypoperfusion areas demonstrated by Single Photon Emission Computed Tomography (SPECT) suggested an involvement of several cerebral areas in both hemispheres.

Research suggests difficulties in attention, processing speed and executive functions during the interictal period of migraine (14) (Table 1). Riva et al. (38) investigated the cognitive performances and behavioral problems in children and adolescents affected by migraine with and without aura (17 and 31 patients, respectively). Normal visual, selective and divided, attention was found in both groups of patients, evaluated with a paper-and-pencil test (“Trail making test,” TMT A and B) (65). However, slower information processing speed, assessed by a computerized simple reaction time to visual stimuli task, was found in migraine patients compared to controls. Frequency of the attacks, but not the duration of the disease, influenced the simple reaction time (38). Confirmatory evidence of altered reaction time in young headache patients was provided by a later study of the same authors (46). The “Conners' Continuous Performance test” (66), evaluating selective and sustained attention, was administered to 62 young patients with primary headache (14 with migraine with aura, 29 with migraine without aura and 19 with TTH) and 52 controls. Faster mean reaction time, associated with highly activated response and errors (more frequent false positive responses), was found in patients compared with controls, suggesting an impulsive response style. The role of emotional factors, such as anxiety, in attentive performance of patients was hypothesized. As in their previous study, no difference was found between the groups of patients, suggesting that migraine and TTH may represent a continuum and share the same pathophysiological mechanisms (22, 24). In a more recent study (42), low performance in attention and executive functions tests was described in children with migraine. In particular, a significant longer execution time (“Trail-Making test”) was found in migraine patients compared with healthy subjects. On the other hand, patients and controls exhibited similar performances in errors. Considering that reaction time explores mainly the speed in detecting and responding to a signal, the authors suggested that its abnormality can be considered as an early and subclinical sign of a neuropsychological impairment, associated with the frequency of headache attacks. In contrast with these findings, another study (44) described similar sequential and simultaneous information processing abilities (evaluated by Kaufman-Assessment Battery for Children) (62) in patients with migraine (with and without aura) and their healthy siblings. These results were replicated by Villa et al. (47), who found no differences between migraine patients (n. 30) and healthy children (n. 30) in a visual reaction time task (“Visual attention test”) (47, 67).

Peculiar attention mechanisms in young patients with migraine were suggested by psychophysiological research (22). Agessi et al. (68) showed impairment of auditory selective attention and physiologic mechanisms of temporal processing. In 2011, we explored both neuropsychological and neurophysiological components of attention in children with migraine (n. 16), TTH (n. 12), and healthy subjects (n. 10) (25). Although a paper and pencil attentive test (“Deux Barrage”) (25) did not evidence any significant impairments and differences in speed and errors between groups, the authors found abnormal neurophysiological mechanisms of spatial attention in patients with headache. Moreover, only in migraine patients, and not in those with TTH, the N140 amplitude increase showed a positive correlation with the neuropsychological performance. We hypothesized that in children with migraine normal attention achievement can be obtained involving a larger amount of “frontal lobe resources,” as compared to healthy children (25).

Though a causal association cannot be maintained, a high prevalence of ADHD symptoms has been described in children and adolescents suffering from migraine (48, 49, 69) (Table 1). Arruda et al. (48) showed a significant higher prevalence of ADHD in children with migraine compared with control subjects. In particular, since this comorbidity mostly involved children with chronic migraine, it was suggested that frequency of migraine attacks increased the risk of ADHD. These results were confirmed by Attygalle et al. (49). The authors described a higher percentage of inattention and impulsive-hyperactive symptoms in patients with migraine than in control subjects. Exploring the opposite association, a high prevalence of migraine (and other primary headache) was described in children with ADHD (70).

We must underline that the association between pediatric headache, in particular migraine, and ADHD is not generally accepted. In a cohort study conducted on 279 high-school students, ADHD prevalence was not different between students who reported headache and those who did not (50). These results were in accordance with a previous study of Pavone et al. (51). In order to explore the prevalence of ADHD and learning disabilities in pediatric patients with headache, a total of 243 medical records of patients (6–18 years) with migraine (n. 107) and TTH (n. 116) were retrospectively reviewed by Genizi et al. (52). Higher rates of ADHD, but also learning disabilities, were found in headache children than in the general population. However, the study did not confirm the relationship between ADHD symptoms and migraine. Supporting a previous study of Mazzone et al. (71), the research revealed a higher prevalence of ADHD symptoms and lower school achievements in children and adolescents with TTH compared to those with migraine.

Learning disabilities are a heterogeneous group of disorders involving difficulties mainly in reading, writing, or mathematical abilities (Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition, DSM-5) (72). These disorders show very often a comorbidity with impairments in cognitive domains, such as memory, attention and processing speed. Few studies investigated the association between migraine and learning disabilities (Table 1). While some of them reported normal abilities in arithmetic and reading in children with migraine (34, 37), other authors described an association between learning disabilities and migraine. Genizi (52) found that about 25% of patients referred to a headache center reported learning disabilities. In particular, compared with children with TTH, those with migraine showed a higher prevalence of learning disabilities (respectively, 26.4 vs. 21.7%). Children with more than 10 attacks/month and long headache duration were more prone to show learning disabilities. Despite these results, children with migraine had good to excellent school achievement. The authors hypothesized that fear of failure and desire for successes, typical of children with headache, may lead to an approach to school characterized by over achievement. Some years later, the same authors (50) showed a higher prevalence of learning disabilities in children with mixed headache (migraine and TTH) compared to those with migraine alone. We have to underline that most studies did not provide a specific neuropsychological evaluation for learning disabilities.

Data on visuo-motor and visuo-spatial abilities of children with migraine during the interictal phase are sparse (Table 1). In a study by Correa Moutran et al., a low perceptual organization (evidenced by WISC-III) was observed (53, 73). On the other hand, no difference in visuospatial skills between migraneurs with aura and controls was found in a later study using “Addenbrooke's Cognitive Examination” test (43). In 2012, Esposito et al. showed a lower performance in motor coordination and visual motor integration evaluation in children with migraine, as compared with controls (54). Using the “Beery Visual-motor integration” test (VMI) (74) and the “Movement Assessment Battery” for Children (M-ABC) (75), the same authors confirmed a low performance in visual motor integration abilities and motor coordination skills in patients suffering from migraine without aura (55). Effectiveness of the Nintendo Wii Fit Plus™ system as a rehabilitative device in these patients was suggested (55). Also, Precenzano et al. showed the efficacy of a specific software training (“Allenare le abilitsuo-spaziali”) for enhancing the visuospatial abilities (56).