Capturing the Full Spectrum of Migraine Symptoms in Children and Adolescents

Two recent studies using an approach called Cluster Analysis of Migraine-associated Symptoms (CAMS) suggest improvements to current diagnostic criteria for migraine.

To diagnose migraine, and to differentiate it from the multitude of other headache conditions, doctors rely on a classification system called the International Classification of Headache Disorders (ICHD). Now in its third edition, the ICHD-3 criteria have been developed over decades to provide simple and clear diagnostic guidance for physicians.

However, while the ICHD is largely effective at capturing most cases of migraine, new research suggests that future iterations of this system would benefit from expanding the scope of symptoms considered for a diagnosis. That would potentially capture more cases and also provide a truer reflection of the burden of migraine, including in younger individuals.

Across two related studies, researchers led by Carlyn Patterson Gentile, Children’s Hospital of Philadelphia, US, used a new approach, called Cluster Analysis of Migraine-associated Symptoms (CAMS), to better understand the spectrum of migraine symptoms in children and adolescents. In the first study, a retrospective cross-sectional investigation that relied on two headache registries from pediatric tertiary care centers, the group considered a number of migraine symptoms not currently part of ICHD-3, including global neurological symptoms like difficulty thinking and vestibular symptoms like ear ringing, among others. This approach identified symptom clusters that allowed for a better understanding of the burden of migraine, and how symptoms varied according to age and sex.

Using the same participants from the first study, the second study used CAMS to critically examine current ICHD-3 migraine criteria. While CAMS confirmed that most ICHD-3 criteria are well founded, it also captured more individuals with migraine than the current criteria would, showing that many of those with probable migraine would now receive a definitive migraine diagnosis.

“Overall, these were very well-done studies. It’s rare to see a sample of youth with migraine this large, that is this well characterized,” according to Serena Orr, University of Calgary, Canada, who studies migraine in children but was not involved with the new studies. “This was a quite rigorous, well-done, unique data set with a lot of strengths, including the fact that it was multi-centered,” she added.

Both studies appeared in the November/December 2024 issue of Headache.

Digging into the databases
Currently, the ICHD-3 diagnosis of migraine requires that patients experience sensitivity to both light (photophobia) and sound (phonophobia), or have nausea and/or vomiting. However, there are other neurological symptoms of migraine that are often highly burdensome to patients but are not well represented in the ICHD.

“I am always struck clinically by the diversity of neurological symptoms that are reported by children and adolescents with migraine. There appears to be these clusters of symptoms that tend to occur together that are not really reflected in the ICHD. These associated neurological symptoms can be as disabling, or more disabling, than the pain itself,” explained Patterson Gentile.

To investigate this further, in the first study, she and her colleagues took advantage of having access to two large pediatric headache patient databases from the Children’s Hospital of Philadelphia (CHOP) and the Cincinnati Children’s Hospital Medical Center (CCHMC). Using CAMS, which the researchers had developed in previous work, they were able to examine the presence or absence of a range of symptoms that occur with migraine, and ask if one symptom is present, how likely it is for another to be present.

“We found with CAMS that if photophobia is reported, it’s highly likely that phonophobia will also be reported. But we also found that as migraine burden gets worse the diversity of symptoms increases, including the appearance of cognitive and vestibular symptoms. In a way, it’s another measure of burden beyond the frequency and severity of headaches,” according to Patterson Gentile.

In the first study, the authors validated CAMS on the CHOP and CCHMC databases. Ninety-six percent of 10,721 patients (median age of 13 years; 67% female) exhibited a migraine phenotype, and were retrospectively examined for the presence or absence of 11 migraine-associated symptoms.

Photophobia (80%), phonophobia (75%), and nausea (65%) were the most frequently reported symptoms, all of which are included in the ICHD-3 migraine criteria. But following these was an array of non-ICHD-3 symptoms including vision changes, dizziness, lightheadedness, difficulty thinking, neck pain, and spinning sensations, which were reported by approximately 25% to 50% of participants.

Mapping out symptoms
Using the CAMS approach, the researchers then mapped out how closely individual symptoms were related to one another, focusing on three dimensions that can capture clusters of symptoms. The first dimension indicated a large clustering of symptoms, showing that participants who reported one symptom were more likely to report multiple symptoms.

The second dimension revealed a clustering of ICHD-3 migraine symptoms, as well as a mostly separate clustering of non-ICHD-3 migraine symptoms. Osmophobia (sensitivity to smell) was the only non-ICHD-3 migraine-associated symptom appearing with ICHD-3 symptoms.

The third dimension separated symptoms of sensory hypersensitivity from vestibular symptoms and vomiting. Here, for example, symptoms of sensory hypersensitivity like photophobia and phonophobia were very likely to co-occur and were separate from vestibular sensitivities like spinning and vision changes, which were also likely to co-occur.   

Next, the authors sought to examine differences in CAMS dimensions across demographics and particular headache characteristics. Interestingly, the number of migraine-associated symptoms increased in adolescence. In particular, before age 12, males and females had the same number of symptoms. But after age 12, the number of non-ICHD-3 symptoms increased in females compared to males.

Orr believes such findings seem right when viewed in the context of brain development.

“As the brain changes and grows, it makes sense that you would see changes from pediatric to adolescent populations, and prior research has shown that migraine symptoms change as youth transition from childhood to adolescence,” according to Orr.

However, she also posited that the differences between younger and older kids may be limited by the language ability of younger children to express what they are feeling, as well as by their higher rates of nausea and vomiting, which have been shown in prior studies. “Is part of the reason for fewer reported associated symptoms in young children due to the fact that they are experiencing more nausea and so don’t notice whether smells bother them or that they are dizzy?” asked Orr.

In terms of migraine burden, overall, having more migraine-associated symptoms was associated with greater headache severity and headache-related disability. Notably, a larger number of non-ICHD-3 symptoms was associated with greater headache frequency, but lower headache severity, compared to participants with ICHD-3 symptoms.

The Cluster Analysis of Migraine-associated Symptoms (CAMS) analysis for combined sites. The presence (bubbles) or absence (black dots) of 11 associated symptoms plotted in CAMS symptom number (x-axis), CAMS migraine International Classification of Headache Disorders (ICHD) symptoms (y-axis), and CAMS sensory hypersensitivity versus vestibular and vomiting (color axis) space combined across sites. The size of the bubble represents the proportion of participants who reported that symptom. Dotted lines have been added to visually represent data groupings (indicates no symptoms on x-axis; and the split between ICHD and non-ICHD symptoms). Credit: Image and caption from Gentile et al. Cluster Analysis of Migraine-associated Symptoms (CAMS) in youth: A retrospective cross-sectional multicenter study. Headache. 2024 Nov-Dec;64(10):1230-43. CC BY-NC 4.0

Rethinking current criteria
Armed with a broader perspective on migraine symptoms and the burden of headache in children and adolescents, the researchers then critically evaluated the current ICHD-3 criteria for migraine in these populations, in their second study.

The ICHD-3 for migraine consists of criteria A-E, which describe the various symptoms and characteristics of a migraine. For a diagnosis, criterion A states that a patient must have five attacks fulfilling criteria B-D. Criterion D states that a patient must, at the least, have photophobia and phonophobia, or nausea and/or vomiting, during a headache.

Of the 10,721 participants who were included in the CAMS validation study, 9,017 were included in the ICHD-3 evaluation study. Ninety-nine percent of these participants received a diagnosis of migraine or probable migraine. A probable migraine diagnosis (903 participants in total) was when participants failed to meet criterion D. Notably, 51% of these probable migraine patients reported two or more associated symptoms previously captured by the expanded CAMS list.

“From shared clinical experience, it’s known that if an individual receives a probable migraine diagnosis, they will respond to migraine treatments, although treatment availability for these patients is much more limited, and probable migraine can be just as disabling as migraine,” explained Patterson Gentile.

The researchers next investigated how altering migraine criterion D would impact migraine diagnosis. First, they found that allowing for photophobia or phonophobia as qualifying symptoms, rather than photophobia and phonophobia (as is required in ICHD-3), recategorized 56% of probable migraine cases as having migraine. Photophobia and phonophobia regularly clustered together in CAMS, but it was common for a patient to report the more prominent symptom in patient questionnaires, even if in fact both symptoms were present upon further investigation by a provider; simplifying the criteria would aid in diagnosis.

Second, allowing for inclusion of lightheadedness, the most common non-ICHD-3 symptom from CAMS, and requiring two of three associated symptoms to fulfill criterion D, 20% of probable migraine cases met diagnostic criteria for a migraine diagnosis. Lastly, including both lightheadedness and spinning sensation, and requiring two of four symptoms for criterion D, recategorized 26% of probable migraine cases into migraine.

Based on these findings, Patterson Gentile and colleagues propose changing criterion D in future ICHD-3 iterations in children and adolescents, to allow for photophobia or phonophobia to meet the associated symptom requirements for migraine; this casts a wider net for a full migraine diagnosis. Interestingly, they stopped short of proposing inclusion of any of the additional associated symptoms found in CAMS.

Explaining this recommendation, Patterson Gentile said that “ultimately, from a diagnostic standpoint, we want to identify people who will benefit from migraine-specific therapies. Of course, these criteria need to be easy to use, so including a list of extra questions to account for these associated symptoms doesn’t really do that. We had strong evidence for switching the and to an or, but don’t currently have it for the other symptoms. Perhaps we will with further research.”

Orr tends to agree with this proposal, particularly for pediatric populations.

“In pediatrics, because of the difficulty in eliciting the symptoms, particularly in younger kids, where the verbal aspect is harder, being more permissive is beneficial. Additionally, their data show that the majority of patients had both symptoms [photophobia and phonophobia], yet there was a proportion that only had one, but it was still otherwise clustered with associated symptoms, [so it] would make sense to classify [these patients] as [having] migraine,” she said.

Image credit: 123RF Stock Photo/ninalisitsyna

References
A critical appraisal of the International Classification of Headache Disorders migraine diagnostic criteria based on a retrospective multicenter cross-sectional headache registry study in youth.
Gentile et al.
2024 Nov-Dec;64(10):1217-29.

Cluster Analysis of Migraine-associated Symptoms (CAMS) in youth: A retrospective cross-sectional multicenter study.
Gentile et al.
Headache. 2024 Nov-Dec;64(10):1230-43.