- Email: [email protected]
Pediatric migraine equivalents
Original Articles
Pediatric Migraine Equivalents: Occurrence and Clinical Features in Practice Waleed A. Al-Twaijri, MD* and Michael I. Shevell, MD, CM*† Migraine equivalents of infancy, childhood, and adolescence are recognized periodic, paroxysmal syndromes without associated headache that are thought to be migrainous in etiology. Five such equivalents are presently recognized. Their clinical features and relative frequency in ambulatory pediatric neurology practice have not been well documented. Utilizing a comprehensive, standardized computer database, the occurrence of these migraine equivalents in a single pediatric neurology practice together with their observed clinical features were documented over an 8-year period. Of a total of 5,848 patients in the database, of whom 1,106 were migraineurs, 108 patients (1.8% of total, 9.8% of migraineurs) were identified to have migraine equivalents. The following distribution among migraine equivalents was observed: benign paroxysmal torticollis 11 (10.2% of patients with migraine equivalents), benign paroxysmal vertigo 41 (38%), abdominal migraine/cyclical vomiting 20 (18.5%), acephalgic migraine 31 (28.7%), and acute confusional migraine 5 (4.6%). In each type, with the exception of benign paroxysmal torticollis and acute confusional migraine, females clearly predominated, and in all types a strong positive family history of migraine was elicited (68%–100%). There was variation in the age of onset of a particular equivalent with considerable overlap observed. Coexisting more typical migraines were observed in from 10% (benign paroxysmal torticollis) to 70% (abdominal migraines/cyclical vomiting) of the cases. In conclusion, pediatric migraine equivalents occur with relative frequency in ambulatory practice, possessing discrete clinical features that have a clear relationship to more typical migrainous phenomena. © 2002 by Elsevier Science Inc. All rights reserved.
Al-Twaijri WA, Shevell MI. Pediatric migraine equivalents: occurrence and clinical features in practice. Pediatr Neurol 2002;26:365-368.
From the Departments of *Neurology/Neurosurgery and †Pediatrics; McGill University; Division of Pediatric Neurology; Montreal Children’s Hospital; Montreal, Quebec, Canada.
Communications should be addressed to: Dr. Shevell; Room A-514; Montreal Children’s Hospital; 2300 Tupper Street; Montreal, Quebec, H3H 1P3 Canada. Received September 27, 2001; accepted December 5, 2001.
© 2002 by Elsevier Science Inc. All rights reserved. PII S0887-8994(01)00416-7 ● 0887-8994/02/$—see front matter
Introduction Migraine and its variants are frequent causes for referral for neurologic assessment in pediatric patients [1]. Over the latter half of the twentieth century, clinicians have recognized discrete migraine syndromes occurring in infancy, childhood, and adolescence that lack a prominent headache component [2,3]. These specific syndrome complexes have been labeled migraine equivalents, migraine precursors, or periodic syndromes of childhood [3]. They are thought to bear a pathogenetic and causal relationship to more typical childhood migraines by their paroxysmal or periodic nature, similarity to more well-established adult migraine syndromes, backdrop of a strong family history of migraines in the affected child, and the later evolution in the individual child to more typical migraines [3]. At present, five such migraine equivalent syndromes have been generally recognized in pediatric patients: benign paroxysmal torticollis [4], benign paroxysmal vertigo [5], abdominal migraine/cyclical vomiting [6], acephalgic migraine [7], and acute confusional migraine [8]. The literature on these migraine equivalent syndromes largely consists of case reports and series limited to a description of each of the recognized subtypes on its own. There are no reports yet on their relative frequency of occurrence within the context of ambulatory pediatric neurology practice and their frequency within the context of migraines encountered as a whole. In addition, summarizing, comparing, and contrasting the clinical features observed for each subtype in a single series will both
Al-Twaijri and Shevell: Pediatric Migraine Equivalents 365
Table 1. Frequency distribution of migraine equivalent subtypes identified (N ⴝ 108) Number of Patients Identified (% of Total Number of Migraine Equivalents)
Migraine Equivalent Subtype Benign paroxysmal vertigo Acephalgic migraines Abdominal migraines/cyclical vomiting Benign paroxysmal torticollis Acute confusional migraines
41 (38%) 31 (28.7%) 20 (18.5%) 11 (10.2%) 5 (4.6%)
provide refinement of their individual clinical recognition and illustrate their individual distinctiveness along the migraine spectrum. Materials and Methods Patients were identified through systematic review of a comprehensive standardized computer database of the ambulatory general pediatric neurology practice of a single pediatric neurologist (M.S.). The database includes demographic, diagnostic (neurologic and nonneurologic), and treatment information on all patients observed in the context of this practice. Information is entered into the database by the neurologist at the time of initial assessment and modified, as necessary, depending on clinical evolution and the results of laboratory testing at each subsequent follow-up visit. As part of this general ambulatory practice, children were observed in three separate locales: (1) hospital-based private office practice, (2) hospital-based neurology outpatient clinic in conjunction with rotating house staff (neurology and pediatric), and (3) a suburban private office practice situated in a large pediatric group clinic. This practice profile does not include a specialty “headache” clinic or solicit in particular referrals of patients with migraine. The above database in its entirety was searched for all records containing the diagnosis of one of the five possible migraine equivalents (benign paroxysmal torticollis, benign paroxysmal vertigo, abdominal migraine/cyclical vomiting, acephalgic migraine, or acute confusional migraine). The search was limited to those children observed for their initial neurologic assessment between July 1991 and June 1999 (8 years inclusive). All such records identified in this comprehensive search had their case files retrieved and reviewed systematically to yield the results presented below. The diagnosis of a migraine equivalent and its specific type was made by the initial assessing neurologist. In the absence of a biomarker for any of these migraine equivalents, diagnosis was based on the presence of the typical clinical features specific to a migraine equivalent subtype thus far noted and the absence of any positive laboratory testing that may suggest another possible etiology for the child’s symptomatology [3].
Results A total of 5,848 patients were entered into the database for the 8-year inclusive period that was the time frame of Table 2.
the data search. A total of 1,106 migraineurs were identified, representing 18.9% of the patient population seen in the ambulatory pediatric neurology practice. A total of 108 children with a migraine equivalent were identified, representing 1.8% of all patients entered into the database and 9.8% of those identified as migraineurs. Thus approximately one in 10 of all migraineurs were suffering from a recognized migraine equivalent subtype. The frequency distribution of migraine equivalent subtypes identified is presented in Table 1. Forty-one children had benign paroxysmal vertigo, 31 had acephalgic migraines, 20 had abdominal migraines/cyclical vomiting, 11 had benign paroxysmal torticollis, and only five had an acute confusional migraine. From a frequency perspective, benign paroxysmal vertigo represented 38% of all migraine equivalents; acephalgic migraine, 28.7%; abdominal migraine/cyclical vomiting, 18.5%; benign paroxysmal torticollis, 10.2%; and acute confusional migraine, 4.6%. Thus two thirds of all migraine equivalents diagnosed were accounted for by the subtypes of benign paroxysmal vertigo and acephalgic migraines taken together. For the 20 patients with abdominal migraine/cyclical vomiting, six had presented initially with recurrent vomiting, whereas 14 had presented with episodes of abdominal pain. Table 2 contains the distribution among migraine equivalent subtypes of sex, age of onset (range and mean), frequency of family history of migraines, and recognition of coexisting more typical migraine syndromes (i.e., migraines without aura [common migraines] and migraines with aura [classical migraines]). Among migraine equivalents as a whole, there was a preponderance of females (63 females [58.3%] and 45 males [41.7%)]. In all subtypes, females exceeded males with the exception of a relatively even distribution among the benign paroxysmal torticollis and acute confusional migraine subtypes. There existed a strong family history of migraine in all subtypes ranging from a low of 65% (abdominal migraine/cyclical vomiting) to a high of 100% (acute confusional migraine). With the exception of benign paroxysmal torticollis, there was a wide range in the age of onset. Three subtypes demonstrated the potential for onset in infancy (benign paroxysmal torticollis, benign paroxysmal vertigo, and abdominal migraine/cyclical vomiting). The other two subtypes (acephalgic migraine/acute confusional migraine) began in either mid-childhood or adolescence. The average age of onset for each of these subtypes in ascending order was as follows: benign paroxysmal torti-
Frequency distribution of pertinent clinical features among the various migraine subtypes
Migraine Equivalent Subtype
No. of Patients
Gender (M/F)
Mean Age of Onset (range)
Family History of Migraine (%)
Coexisting Typical Migraine Syndrome (%)
Benign paroxysmal vertigo Acephalgic migraines Abdominal migraines/cyclical vomiting Benign paroxysmal torticollis Acute confusional migraines
41 31 20 11 5
17/24 12/19 8/12 6/5 2/3
5 yr (1-13 yr) 6.8 yr (5-14 yr) 6.5 yr (9 mo-13 yr) 1.17 yr (2 mo-3 yr) 10.9 yr (6.5-15 yr)
28 (68%) 22 (70%) 13 (65%) 8 (73%) 5 (100%)
4 (10%) 15 (48%) 14 (70%) 2 (18%) 1 (20%)
366
PEDIATRIC NEUROLOGY
Vol. 26 No. 5
collis, benign paroxysmal vertigo, abdominal migraine/ cyclical vomiting, acephalgic migraine, and acute confusional migraine. There was a wide range of overlap in terms of the age of onset between the different migraine equivalent subtypes. For children with abdominal migraine/cyclical vomiting, those who presented initially with cyclical vomiting had an average age of onset of symptoms of 3.6 years (range ⫽ 1-7 years), whereas those whose initial symptom consisted of abdominal pain had an average age of onset of 7.6 years (range ⫽ 9 months to 13 years of age). Coexisting more typical migraine syndromes (migraines with or without aura) were recognized relatively rarely (i.e., less than one fourth of cases) in three of the subtypes: benign paroxysmal vertigo, benign paroxysmal torticollis, and acute confusional migraine. In the other two subtypes (abdominal migraine/cyclical vomiting and acephalgic migraine), coexisting typical migraine syndromes were observed relatively commonly, that is in more than one half of the cases of these subtypes noted in this series. Discussion Five discrete migraine equivalent syndromes have thus far been recognized to occur within the pediatric patients. Benign paroxysmal torticollis [4,9,10,11] typically is thought to begin during infancy and presents with torticollis, with or without associated pallor, vomiting, or behavioral changes. The torticollis is episodic, may occur to either side, and usually lasts between 4 hours and up to 4 days in duration. Typically the frequency and duration of these episodes decline as the child gets older, and by early to mid-childhood the episodes have resolved in their entirety. Benign paroxysmal vertigo [5,12-15] typically has a later onset of between 1 and 3 years of age. The episodes are brief, usually lasting between 1 and 5 minutes, and are characterized by the appearance of unsteadiness, pallor, and sometimes evident fear. The verbal child is able to describe a sensation of spinning or “vertigo.” Typically these episodes resolve within 2 years after their onset. Abdominal migraine and cyclical vomiting [6,16-18] typically begin in mid-childhood between 4 and 10 years of age. Often these episodes have recognized specific precipitants and are characterized by crampy, periumbilical abdominal pain or pronounced nausea and vomiting with a high frequency of emeses of small volume. Frequently there is associated pallor, and the episodes typically are longer than 1 hour. There is usual resolution of this migraine equivalent within 2 years after onset. Acephalgic migraines [7,19] also begin typically in mid- to late childhood between 5 and 12 years of age. This migraine equivalent is characterized by the occurrence of typical migrainous auras, predominantly visual in type, featuring sometimes micropsia or metamorphopsia without associated headache. These auras typically last less than 10 minutes. The final recognized migraine equivalent subtype is that of acute confusional migraine [8,20,21].
This condition typically begins in late childhood to adolescence. It may be the initial presentation of a migrainous predisposition and is often preceded by minor head trauma. The youngster appears confused and agitated and may have pronounced memory disturbances. Headaches are only rarely recognized within the context of the acute symptomatology that can last up to approximately 8 hours in duration. Our series highlights the relative frequency of migraine equivalents. These discrete syndrome complexes taken together as a whole represented close to 10% of all migraineurs identified in an ambulatory general pediatric neurology practice. Given the relative paucity of literature reports on these individual syndromes, this is larger than expected. Furthermore, these entities are likely underdiagnosed because the symptoms associated with them acutely may be underrecognized by primary practice providers as being migrainous in origin [3]. The frequency of these equivalents observed in our series has implications with respect to the training and education of both future and current practice providers. This occurrence is essential for the accurate recognition of these discrete syndromes because their proper diagnosis is both reassuring given their typical benign evolution and also effectively limits unnecessary testing when classic features are present [3]. It is interesting to note the relative frequency within the practice profile of each of these migraine equivalent subtypes. Nearly two thirds of the patients belong to two of the five subtypes collectively, benign paroxysmal vertigo and acephalgic migraines. Coincidentally, these are the two pediatric migraine equivalents explicitly recognized by the International Headache Society (IHS) classification scheme [22]. Within our series, acute confusional migraines were a relatively rare occurrence. Typically the confusion and agitation that is associated with this particular migraine equivalent prompt an emergency room visit by concerned parents and caregivers. This presentation may partly explain the relative infrequency of this particular subtype encountered in a practice profile that was predominantly outpatient and ambulatory in its orientation. The design of our study does not permit the derivation of precise prevalence or incidence figures for the subtypes in general pediatric patients as a whole. This will need to be the object of future prospective study that is population-based. The female preponderance in our sample overall parallels that which is observed in migraines as a whole [1]. This finding was especially obvious in the three most common subtypes encountered: benign paroxysmal vertigo, acephalgic migraines, and abdominal migraines/ cyclical vomiting. Among all subtypes there was a strong family history of migraines relatively easily elicited. This finding once again places these disorders within the migrainous spectrum [2]. It suggests that there is a hereditary (i.e., genetic) predisposition to the occurrence of the subtypes. Specifically, as previously suggested, migraine equivalents can be considered as age-dependent,
Al-Twaijri and Shevell: Pediatric Migraine Equivalents 367
developmental expressions of the inherited migrainous predisposition [3]. Factors that determine the expression of one subtype or another at a specific age have not yet been elucidated. A wide range in age of onset for each subtype was observed with the exception of benign paroxysmal torticollis, which is limited to the infant and toddler age groups. General trends can be noted and are consistent with that already reported in the literature [3]. Specifically, benign paroxysmal vertigo and abdominal migraine/cyclical vomiting are largely disorders of toddlers and children, whereas acephalgic migraines and acute confusional migraines are largely disorders of mid-childhood to adolescence. An ascending order in terms of the average age of onset of each subtype is discernible according to the following trend: benign paroxysmal torticollis, benign paroxysmal vertigo, abdominal migraine/cyclical vomiting, acephalgic migraine, and acute confusional migraine. The average age of onset for both benign paroxysmal torticollis and benign paroxysmal vertigo was higher than expected. For benign paroxysmal torticollis, the average age was skewed upwards by three children who developed their symptoms initially at 2, 3, and 3 years of age, respectively. For benign paroxysmal vertigo, the average was skewed upwards by six children who initially became symptomatic beyond 8 years of age. Coexisting typical migraine syndromes (i.e., migraines with or without aura [common or classic migraines]) was evident at the time of diagnosis of the migraine equivalent or during the course of follow-up (that was variable in length) more often than not in children with either abdominal migraine/cyclical vomiting or acephalgic migraines. A co-existing typical migraine syndrome occurred infrequently in children with either of the two younger age-ofonset subtypes, specifically benign paroxysmal torticollis and benign paroxysmal vertigo. This finding was expected because it is relatively rare to reliably diagnose typical migraine syndromes in the child less than 5 years of age [23]. It was also not unexpected that more typical migraine syndromes were not observed frequently in children presenting with acute confusional migraines because this particular periodic migraine syndrome complex may be the initial presentation of a migrainous tendency [8,20]. It should be noted that the above comments refer to the occurrence of co-existing typical migraine syndromes together with these migraine equivalent subtypes and are not a determination of the eventual lifetime occurrence of typical migraine syndromes in this patient population. From the foregoing, it is readily apparent that migraine equivalents are a relatively frequent occurrence, likely higher than previously thought by the practitioner within the pediatric population that is referred for neurologic evaluation. They represent a subset of pediatric migraineurs and occupy a clear place among the migrainous spectrum. Initial clues to understanding their pathogenesis will likely be derived from our understanding of the more typical and frequent migraine syndromes. A particular
368
PEDIATRIC NEUROLOGY
Vol. 26 No. 5
challenge will be framing this understanding within a context that accounts for the apparent age-dependent expression of migraine equivalents and their interindividual variability in expression among those with a discernible migrainous predisposition. The authors would like to acknowledge the secretarial assistance of Alba Rinaldi in the preparation of this manuscript. M.I.S. is a Chercheur Boursier Clinicien (Clinical Research Scholar) of the Fonds de Recherche en Sante du Quebec. M.I.S. is grateful for the support of the MCH Foundation during the writing of this manuscript. Presented in part at the thirtieth Annual Meeting of the Child Neurology Society, Victoria, British Columbia. References [1] Ziegler DK. Headache. Public health problem. Neurol Clin 1980;8:781-790. [2] Barlow CF. The expression of childhood migraine. Headaches and migraine in childhood. Philadelphia: JB Lippincott, 1984:46-75. [3] Shevell M. A guide to migraine equivalents. Contemp Pediatr 1998;15:71-79. [4] Deonna T, Martin D. Benign paroxysmal torticollis in infancy. Arch Dis Child 1981;56:956-959. [5] Lanzi G, Balottin U, Fazzi E, Tagliasacchi M, Manfrin M, Mira E. Benign paroxysmal vertigo of childhood: A long-term follow-up. Cephalgia 1994;6:458-460. [6] Symon DNK, Russell G. The relationship between cyclic vomiting syndrome and abdominal migraine. J Pediatr Gastroenterol Nutr 1995;21:S42-S43. [7] Shevell MI. Acephalgic migraines of childhood. Pediatr Neurol 1996;14:211-215. [8] Sheth RD, Riggs JE, Bodensteiner JB. Acute confusional migraine: Variant of transient global amnesia. Pediatr Neurol 1995;12:129131. [9] Cohen HA, Nussinovitch M, Ashkenasi A, Straussberg R, Kauschanksy A, Frydnan M. Benign paroxysmal torticollis in infancy. Pediatr Neurol 1993;9:488-490. [10] Roulet E, Deonna T. Benign paroxysmal torticollis in infancy. Dev Med Child Neurol 1988;30:409-410. [11] Drigo P, Carli G, Laverda AM. Benign paroxysmal torticollis of infancy. Brain Dev 2000;22:169-172. [12] Fenichel GM. Migraine as a cause of benign paroxysmal vertigo in childhood. J Pediatr 1967;71:114-115. [13] Koehler B. Benign paroxysmal vertigo of childhood: A migraine equivalent. Eur J Pediatr 1980;134:149-151. [14] Curatolo P, Sciarretta A. Benign paroxysmal vertigo and migraine. Dev Med Child Neurol 1987;29:405-406. [15] Drigo P, Carli G, Laverda AM. Benign paroxysmal vertigo of childhood. Brain Dev 2001;23:38-41. [16] Li BK. Cyclic vomiting: New understanding of an old disorder. Contemp Pediatr 1996;13:48-52. [17] Mortimer MJ, Kay J, Jaron A. Clinical epidemiology of childhood abdominal migraine in urban general practice. Dev Med Child Neurol 1993;35:243-248. [18] Bentley D, Kehely A, Al-Bayaty M, Michie CA. Abdominal migraine as a cause of vomiting in children: A clinician’s view. J Pediatr Gastroenterol Nutr 1995;21:S49-S51. [19] Golden GS. The Alice in Wonderland syndrome in juvenile migraine. Pediatrics 1979;63:517-519. [20] Barlow CF. Confusional migraine. Headaches and migraine in childhood. Philadelphia: JB Lippincott, 1984:109-115. [21] Shaabat A. Confusional migraine in childhood. Pediatr Neurol 1996;15:23-25. [22] Olesen J. The classification and diagnosis of headache disorders. Neurol Clin 1990;8:793-799. [23] Barlow CF. Migraine in the infant and toddler. J Child Neurol 1994;9:92-94.
Pediatric Migraine Equivalents: Occurrence and Clinical Features in Practice Waleed A. Al-Twaijri, MD* and Michael I. Shevell, MD, CM*† Migraine equivalents of infancy, childhood, and adolescence are recognized periodic, paroxysmal syndromes without associated headache that are thought to be migrainous in etiology. Five such equivalents are presently recognized. Their clinical features and relative frequency in ambulatory pediatric neurology practice have not been well documented. Utilizing a comprehensive, standardized computer database, the occurrence of these migraine equivalents in a single pediatric neurology practice together with their observed clinical features were documented over an 8-year period. Of a total of 5,848 patients in the database, of whom 1,106 were migraineurs, 108 patients (1.8% of total, 9.8% of migraineurs) were identified to have migraine equivalents. The following distribution among migraine equivalents was observed: benign paroxysmal torticollis 11 (10.2% of patients with migraine equivalents), benign paroxysmal vertigo 41 (38%), abdominal migraine/cyclical vomiting 20 (18.5%), acephalgic migraine 31 (28.7%), and acute confusional migraine 5 (4.6%). In each type, with the exception of benign paroxysmal torticollis and acute confusional migraine, females clearly predominated, and in all types a strong positive family history of migraine was elicited (68%–100%). There was variation in the age of onset of a particular equivalent with considerable overlap observed. Coexisting more typical migraines were observed in from 10% (benign paroxysmal torticollis) to 70% (abdominal migraines/cyclical vomiting) of the cases. In conclusion, pediatric migraine equivalents occur with relative frequency in ambulatory practice, possessing discrete clinical features that have a clear relationship to more typical migrainous phenomena. © 2002 by Elsevier Science Inc. All rights reserved.
Al-Twaijri WA, Shevell MI. Pediatric migraine equivalents: occurrence and clinical features in practice. Pediatr Neurol 2002;26:365-368.
From the Departments of *Neurology/Neurosurgery and †Pediatrics; McGill University; Division of Pediatric Neurology; Montreal Children’s Hospital; Montreal, Quebec, Canada.
Communications should be addressed to: Dr. Shevell; Room A-514; Montreal Children’s Hospital; 2300 Tupper Street; Montreal, Quebec, H3H 1P3 Canada. Received September 27, 2001; accepted December 5, 2001.
© 2002 by Elsevier Science Inc. All rights reserved. PII S0887-8994(01)00416-7 ● 0887-8994/02/$—see front matter
Introduction Migraine and its variants are frequent causes for referral for neurologic assessment in pediatric patients [1]. Over the latter half of the twentieth century, clinicians have recognized discrete migraine syndromes occurring in infancy, childhood, and adolescence that lack a prominent headache component [2,3]. These specific syndrome complexes have been labeled migraine equivalents, migraine precursors, or periodic syndromes of childhood [3]. They are thought to bear a pathogenetic and causal relationship to more typical childhood migraines by their paroxysmal or periodic nature, similarity to more well-established adult migraine syndromes, backdrop of a strong family history of migraines in the affected child, and the later evolution in the individual child to more typical migraines [3]. At present, five such migraine equivalent syndromes have been generally recognized in pediatric patients: benign paroxysmal torticollis [4], benign paroxysmal vertigo [5], abdominal migraine/cyclical vomiting [6], acephalgic migraine [7], and acute confusional migraine [8]. The literature on these migraine equivalent syndromes largely consists of case reports and series limited to a description of each of the recognized subtypes on its own. There are no reports yet on their relative frequency of occurrence within the context of ambulatory pediatric neurology practice and their frequency within the context of migraines encountered as a whole. In addition, summarizing, comparing, and contrasting the clinical features observed for each subtype in a single series will both
Al-Twaijri and Shevell: Pediatric Migraine Equivalents 365
Table 1. Frequency distribution of migraine equivalent subtypes identified (N ⴝ 108) Number of Patients Identified (% of Total Number of Migraine Equivalents)
Migraine Equivalent Subtype Benign paroxysmal vertigo Acephalgic migraines Abdominal migraines/cyclical vomiting Benign paroxysmal torticollis Acute confusional migraines
41 (38%) 31 (28.7%) 20 (18.5%) 11 (10.2%) 5 (4.6%)
provide refinement of their individual clinical recognition and illustrate their individual distinctiveness along the migraine spectrum. Materials and Methods Patients were identified through systematic review of a comprehensive standardized computer database of the ambulatory general pediatric neurology practice of a single pediatric neurologist (M.S.). The database includes demographic, diagnostic (neurologic and nonneurologic), and treatment information on all patients observed in the context of this practice. Information is entered into the database by the neurologist at the time of initial assessment and modified, as necessary, depending on clinical evolution and the results of laboratory testing at each subsequent follow-up visit. As part of this general ambulatory practice, children were observed in three separate locales: (1) hospital-based private office practice, (2) hospital-based neurology outpatient clinic in conjunction with rotating house staff (neurology and pediatric), and (3) a suburban private office practice situated in a large pediatric group clinic. This practice profile does not include a specialty “headache” clinic or solicit in particular referrals of patients with migraine. The above database in its entirety was searched for all records containing the diagnosis of one of the five possible migraine equivalents (benign paroxysmal torticollis, benign paroxysmal vertigo, abdominal migraine/cyclical vomiting, acephalgic migraine, or acute confusional migraine). The search was limited to those children observed for their initial neurologic assessment between July 1991 and June 1999 (8 years inclusive). All such records identified in this comprehensive search had their case files retrieved and reviewed systematically to yield the results presented below. The diagnosis of a migraine equivalent and its specific type was made by the initial assessing neurologist. In the absence of a biomarker for any of these migraine equivalents, diagnosis was based on the presence of the typical clinical features specific to a migraine equivalent subtype thus far noted and the absence of any positive laboratory testing that may suggest another possible etiology for the child’s symptomatology [3].
Results A total of 5,848 patients were entered into the database for the 8-year inclusive period that was the time frame of Table 2.
the data search. A total of 1,106 migraineurs were identified, representing 18.9% of the patient population seen in the ambulatory pediatric neurology practice. A total of 108 children with a migraine equivalent were identified, representing 1.8% of all patients entered into the database and 9.8% of those identified as migraineurs. Thus approximately one in 10 of all migraineurs were suffering from a recognized migraine equivalent subtype. The frequency distribution of migraine equivalent subtypes identified is presented in Table 1. Forty-one children had benign paroxysmal vertigo, 31 had acephalgic migraines, 20 had abdominal migraines/cyclical vomiting, 11 had benign paroxysmal torticollis, and only five had an acute confusional migraine. From a frequency perspective, benign paroxysmal vertigo represented 38% of all migraine equivalents; acephalgic migraine, 28.7%; abdominal migraine/cyclical vomiting, 18.5%; benign paroxysmal torticollis, 10.2%; and acute confusional migraine, 4.6%. Thus two thirds of all migraine equivalents diagnosed were accounted for by the subtypes of benign paroxysmal vertigo and acephalgic migraines taken together. For the 20 patients with abdominal migraine/cyclical vomiting, six had presented initially with recurrent vomiting, whereas 14 had presented with episodes of abdominal pain. Table 2 contains the distribution among migraine equivalent subtypes of sex, age of onset (range and mean), frequency of family history of migraines, and recognition of coexisting more typical migraine syndromes (i.e., migraines without aura [common migraines] and migraines with aura [classical migraines]). Among migraine equivalents as a whole, there was a preponderance of females (63 females [58.3%] and 45 males [41.7%)]. In all subtypes, females exceeded males with the exception of a relatively even distribution among the benign paroxysmal torticollis and acute confusional migraine subtypes. There existed a strong family history of migraine in all subtypes ranging from a low of 65% (abdominal migraine/cyclical vomiting) to a high of 100% (acute confusional migraine). With the exception of benign paroxysmal torticollis, there was a wide range in the age of onset. Three subtypes demonstrated the potential for onset in infancy (benign paroxysmal torticollis, benign paroxysmal vertigo, and abdominal migraine/cyclical vomiting). The other two subtypes (acephalgic migraine/acute confusional migraine) began in either mid-childhood or adolescence. The average age of onset for each of these subtypes in ascending order was as follows: benign paroxysmal torti-
Frequency distribution of pertinent clinical features among the various migraine subtypes
Migraine Equivalent Subtype
No. of Patients
Gender (M/F)
Mean Age of Onset (range)
Family History of Migraine (%)
Coexisting Typical Migraine Syndrome (%)
Benign paroxysmal vertigo Acephalgic migraines Abdominal migraines/cyclical vomiting Benign paroxysmal torticollis Acute confusional migraines
41 31 20 11 5
17/24 12/19 8/12 6/5 2/3
5 yr (1-13 yr) 6.8 yr (5-14 yr) 6.5 yr (9 mo-13 yr) 1.17 yr (2 mo-3 yr) 10.9 yr (6.5-15 yr)
28 (68%) 22 (70%) 13 (65%) 8 (73%) 5 (100%)
4 (10%) 15 (48%) 14 (70%) 2 (18%) 1 (20%)
366
PEDIATRIC NEUROLOGY
Vol. 26 No. 5
collis, benign paroxysmal vertigo, abdominal migraine/ cyclical vomiting, acephalgic migraine, and acute confusional migraine. There was a wide range of overlap in terms of the age of onset between the different migraine equivalent subtypes. For children with abdominal migraine/cyclical vomiting, those who presented initially with cyclical vomiting had an average age of onset of symptoms of 3.6 years (range ⫽ 1-7 years), whereas those whose initial symptom consisted of abdominal pain had an average age of onset of 7.6 years (range ⫽ 9 months to 13 years of age). Coexisting more typical migraine syndromes (migraines with or without aura) were recognized relatively rarely (i.e., less than one fourth of cases) in three of the subtypes: benign paroxysmal vertigo, benign paroxysmal torticollis, and acute confusional migraine. In the other two subtypes (abdominal migraine/cyclical vomiting and acephalgic migraine), coexisting typical migraine syndromes were observed relatively commonly, that is in more than one half of the cases of these subtypes noted in this series. Discussion Five discrete migraine equivalent syndromes have thus far been recognized to occur within the pediatric patients. Benign paroxysmal torticollis [4,9,10,11] typically is thought to begin during infancy and presents with torticollis, with or without associated pallor, vomiting, or behavioral changes. The torticollis is episodic, may occur to either side, and usually lasts between 4 hours and up to 4 days in duration. Typically the frequency and duration of these episodes decline as the child gets older, and by early to mid-childhood the episodes have resolved in their entirety. Benign paroxysmal vertigo [5,12-15] typically has a later onset of between 1 and 3 years of age. The episodes are brief, usually lasting between 1 and 5 minutes, and are characterized by the appearance of unsteadiness, pallor, and sometimes evident fear. The verbal child is able to describe a sensation of spinning or “vertigo.” Typically these episodes resolve within 2 years after their onset. Abdominal migraine and cyclical vomiting [6,16-18] typically begin in mid-childhood between 4 and 10 years of age. Often these episodes have recognized specific precipitants and are characterized by crampy, periumbilical abdominal pain or pronounced nausea and vomiting with a high frequency of emeses of small volume. Frequently there is associated pallor, and the episodes typically are longer than 1 hour. There is usual resolution of this migraine equivalent within 2 years after onset. Acephalgic migraines [7,19] also begin typically in mid- to late childhood between 5 and 12 years of age. This migraine equivalent is characterized by the occurrence of typical migrainous auras, predominantly visual in type, featuring sometimes micropsia or metamorphopsia without associated headache. These auras typically last less than 10 minutes. The final recognized migraine equivalent subtype is that of acute confusional migraine [8,20,21].
This condition typically begins in late childhood to adolescence. It may be the initial presentation of a migrainous predisposition and is often preceded by minor head trauma. The youngster appears confused and agitated and may have pronounced memory disturbances. Headaches are only rarely recognized within the context of the acute symptomatology that can last up to approximately 8 hours in duration. Our series highlights the relative frequency of migraine equivalents. These discrete syndrome complexes taken together as a whole represented close to 10% of all migraineurs identified in an ambulatory general pediatric neurology practice. Given the relative paucity of literature reports on these individual syndromes, this is larger than expected. Furthermore, these entities are likely underdiagnosed because the symptoms associated with them acutely may be underrecognized by primary practice providers as being migrainous in origin [3]. The frequency of these equivalents observed in our series has implications with respect to the training and education of both future and current practice providers. This occurrence is essential for the accurate recognition of these discrete syndromes because their proper diagnosis is both reassuring given their typical benign evolution and also effectively limits unnecessary testing when classic features are present [3]. It is interesting to note the relative frequency within the practice profile of each of these migraine equivalent subtypes. Nearly two thirds of the patients belong to two of the five subtypes collectively, benign paroxysmal vertigo and acephalgic migraines. Coincidentally, these are the two pediatric migraine equivalents explicitly recognized by the International Headache Society (IHS) classification scheme [22]. Within our series, acute confusional migraines were a relatively rare occurrence. Typically the confusion and agitation that is associated with this particular migraine equivalent prompt an emergency room visit by concerned parents and caregivers. This presentation may partly explain the relative infrequency of this particular subtype encountered in a practice profile that was predominantly outpatient and ambulatory in its orientation. The design of our study does not permit the derivation of precise prevalence or incidence figures for the subtypes in general pediatric patients as a whole. This will need to be the object of future prospective study that is population-based. The female preponderance in our sample overall parallels that which is observed in migraines as a whole [1]. This finding was especially obvious in the three most common subtypes encountered: benign paroxysmal vertigo, acephalgic migraines, and abdominal migraines/ cyclical vomiting. Among all subtypes there was a strong family history of migraines relatively easily elicited. This finding once again places these disorders within the migrainous spectrum [2]. It suggests that there is a hereditary (i.e., genetic) predisposition to the occurrence of the subtypes. Specifically, as previously suggested, migraine equivalents can be considered as age-dependent,
Al-Twaijri and Shevell: Pediatric Migraine Equivalents 367
developmental expressions of the inherited migrainous predisposition [3]. Factors that determine the expression of one subtype or another at a specific age have not yet been elucidated. A wide range in age of onset for each subtype was observed with the exception of benign paroxysmal torticollis, which is limited to the infant and toddler age groups. General trends can be noted and are consistent with that already reported in the literature [3]. Specifically, benign paroxysmal vertigo and abdominal migraine/cyclical vomiting are largely disorders of toddlers and children, whereas acephalgic migraines and acute confusional migraines are largely disorders of mid-childhood to adolescence. An ascending order in terms of the average age of onset of each subtype is discernible according to the following trend: benign paroxysmal torticollis, benign paroxysmal vertigo, abdominal migraine/cyclical vomiting, acephalgic migraine, and acute confusional migraine. The average age of onset for both benign paroxysmal torticollis and benign paroxysmal vertigo was higher than expected. For benign paroxysmal torticollis, the average age was skewed upwards by three children who developed their symptoms initially at 2, 3, and 3 years of age, respectively. For benign paroxysmal vertigo, the average was skewed upwards by six children who initially became symptomatic beyond 8 years of age. Coexisting typical migraine syndromes (i.e., migraines with or without aura [common or classic migraines]) was evident at the time of diagnosis of the migraine equivalent or during the course of follow-up (that was variable in length) more often than not in children with either abdominal migraine/cyclical vomiting or acephalgic migraines. A co-existing typical migraine syndrome occurred infrequently in children with either of the two younger age-ofonset subtypes, specifically benign paroxysmal torticollis and benign paroxysmal vertigo. This finding was expected because it is relatively rare to reliably diagnose typical migraine syndromes in the child less than 5 years of age [23]. It was also not unexpected that more typical migraine syndromes were not observed frequently in children presenting with acute confusional migraines because this particular periodic migraine syndrome complex may be the initial presentation of a migrainous tendency [8,20]. It should be noted that the above comments refer to the occurrence of co-existing typical migraine syndromes together with these migraine equivalent subtypes and are not a determination of the eventual lifetime occurrence of typical migraine syndromes in this patient population. From the foregoing, it is readily apparent that migraine equivalents are a relatively frequent occurrence, likely higher than previously thought by the practitioner within the pediatric population that is referred for neurologic evaluation. They represent a subset of pediatric migraineurs and occupy a clear place among the migrainous spectrum. Initial clues to understanding their pathogenesis will likely be derived from our understanding of the more typical and frequent migraine syndromes. A particular
368
PEDIATRIC NEUROLOGY
Vol. 26 No. 5
challenge will be framing this understanding within a context that accounts for the apparent age-dependent expression of migraine equivalents and their interindividual variability in expression among those with a discernible migrainous predisposition. The authors would like to acknowledge the secretarial assistance of Alba Rinaldi in the preparation of this manuscript. M.I.S. is a Chercheur Boursier Clinicien (Clinical Research Scholar) of the Fonds de Recherche en Sante du Quebec. M.I.S. is grateful for the support of the MCH Foundation during the writing of this manuscript. Presented in part at the thirtieth Annual Meeting of the Child Neurology Society, Victoria, British Columbia. References [1] Ziegler DK. Headache. Public health problem. Neurol Clin 1980;8:781-790. [2] Barlow CF. The expression of childhood migraine. Headaches and migraine in childhood. Philadelphia: JB Lippincott, 1984:46-75. [3] Shevell M. A guide to migraine equivalents. Contemp Pediatr 1998;15:71-79. [4] Deonna T, Martin D. Benign paroxysmal torticollis in infancy. Arch Dis Child 1981;56:956-959. [5] Lanzi G, Balottin U, Fazzi E, Tagliasacchi M, Manfrin M, Mira E. Benign paroxysmal vertigo of childhood: A long-term follow-up. Cephalgia 1994;6:458-460. [6] Symon DNK, Russell G. The relationship between cyclic vomiting syndrome and abdominal migraine. J Pediatr Gastroenterol Nutr 1995;21:S42-S43. [7] Shevell MI. Acephalgic migraines of childhood. Pediatr Neurol 1996;14:211-215. [8] Sheth RD, Riggs JE, Bodensteiner JB. Acute confusional migraine: Variant of transient global amnesia. Pediatr Neurol 1995;12:129131. [9] Cohen HA, Nussinovitch M, Ashkenasi A, Straussberg R, Kauschanksy A, Frydnan M. Benign paroxysmal torticollis in infancy. Pediatr Neurol 1993;9:488-490. [10] Roulet E, Deonna T. Benign paroxysmal torticollis in infancy. Dev Med Child Neurol 1988;30:409-410. [11] Drigo P, Carli G, Laverda AM. Benign paroxysmal torticollis of infancy. Brain Dev 2000;22:169-172. [12] Fenichel GM. Migraine as a cause of benign paroxysmal vertigo in childhood. J Pediatr 1967;71:114-115. [13] Koehler B. Benign paroxysmal vertigo of childhood: A migraine equivalent. Eur J Pediatr 1980;134:149-151. [14] Curatolo P, Sciarretta A. Benign paroxysmal vertigo and migraine. Dev Med Child Neurol 1987;29:405-406. [15] Drigo P, Carli G, Laverda AM. Benign paroxysmal vertigo of childhood. Brain Dev 2001;23:38-41. [16] Li BK. Cyclic vomiting: New understanding of an old disorder. Contemp Pediatr 1996;13:48-52. [17] Mortimer MJ, Kay J, Jaron A. Clinical epidemiology of childhood abdominal migraine in urban general practice. Dev Med Child Neurol 1993;35:243-248. [18] Bentley D, Kehely A, Al-Bayaty M, Michie CA. Abdominal migraine as a cause of vomiting in children: A clinician’s view. J Pediatr Gastroenterol Nutr 1995;21:S49-S51. [19] Golden GS. The Alice in Wonderland syndrome in juvenile migraine. Pediatrics 1979;63:517-519. [20] Barlow CF. Confusional migraine. Headaches and migraine in childhood. Philadelphia: JB Lippincott, 1984:109-115. [21] Shaabat A. Confusional migraine in childhood. Pediatr Neurol 1996;15:23-25. [22] Olesen J. The classification and diagnosis of headache disorders. Neurol Clin 1990;8:793-799. [23] Barlow CF. Migraine in the infant and toddler. J Child Neurol 1994;9:92-94.