Cyclic vomiting syndrome and benign paroxysmal torticollis are associated with a high risk of developing primary headache: A longitudinal study

Abstract

Background and aim

Episodic syndromes that may be associated with migraine are a group of disorders affecting patients with migraine or with an increased risk of presenting it, and likely represent an early life expression of migraine. Cyclic vomiting syndrome and benign paroxysmal torticollis are well characterized and represent a frequent cause of request for specialist consultations. The aim of this study is to longitudinally assess the rate of headache in patients presenting with cyclic vomiting syndrome and benign paroxysmal torticollis during infancy, and to define the main clinical features of the disorder.

Methods

We administered a questionnaire to the parents of all our pediatric patients with previous diagnosis of cyclic vomiting syndrome and/or benign paroxysmal torticollis according to ICHD-3; questions were focused on the main clinical features of the disorder as well as the prognosis, with particular emphasis on the development of headache.

Results

For the final analysis we considered 82 patients with cyclic vomiting syndrome and 33 with benign paroxysmal torticollis. Seventy-nine percent of patients with cyclic vomiting syndrome presented with headache during the follow-up, with a mean age at onset of 6 years; 67% of patients with benign paroxysmal torticollis suffered from headache during the follow-up, with a mean age at onset of 5 years.

Discussion

Cyclic vomiting syndrome and benign paroxysmal torticollis are associated with a very high risk of developing headache, mostly migraine, later in life. In both groups of patients, the vast majority presented with different episodic syndromes that may be associated with migraine at different ages, thus suggesting an age-dependent evolution of migraine-like symptoms before the onset of clear migrainous headache.

Introduction

Episodic syndromes that may be associated with migraine are defined as a group of disorders affecting patients with migraine or with an increased risk of presenting it (1). They are considered as an early life expression of migraine (2). The new ICHD-3 classification includes recurrent gastrointestinal disturbance (previously known as chronic abdominal pain), cyclic vomiting syndrome (CVS), abdominal migraine, benign paroxysmal vertigo, benign paroxysmal torticollis (BPT), infantile colics, alternating hemiplegia of childhood, and vestibular migraine (3). Additional conditions are motion sickness and periodic sleep disorders (such as sleep walking, sleep talking, night terrors and bruxism) (3). Common clinical features of these conditions are a periodic or paroxysmal pattern, an association with subsequent diagnosis of migraine (26%) and a high rate of positive familial history for migraine (1,4,5).

Although episodic syndromes that may be associated with migraine were initially reported almost one century ago, there is a paucity of longitudinal studies investigating the prevalence of primary headaches and their characteristics in this group of individuals (2). More data are available on the incidence of previous episodic syndromes in patients actually suffering from migraine. A recent study of our group on a large sample of children with tension-type headache and/or migraine found a previous history of these disorders in 70.3% of patients (6).

Among episodic syndromes that may be associated with migraine, CVS and BPT are well characterized and represent a frequent cause of request for specialist consultations. CVS is characterized by repetitive attacks of severe nausea and vomiting affecting about 2% of school-age children (7). Attacks have a widely variable duration and may be accompanied by pallor and lethargy (1). Episodes are usually highly stereotyped in the same individual, and have a cyclic pattern, with a complete resolution of symptoms between attacks (1). CVS onset is usually reported between 4 and 7 years of age, with a significant delay in diagnosis (8). The frequency of episodes is highly variable, even reaching three episodes per month (9). Migraine is diagnosed in 40–46% of cases (1,10). Although this is mainly an exclusion diagnosis, clinical diagnostic criteria are available (3). BPT is a benign condition characterized by repetitive episodes of head tilt, usually with alternating side, with complete and spontaneous remission (3). The onset is usually in the first year of life, with episodes recurring every 45–75 days and lasting 4.5–6 days (4). Migraine was reported in 12% of cases (4).

The aim of this study is to longitudinally assess the rate of headache in patients presenting with cyclic vomiting and benign paroxysmal torticollis during infancy, and to define the main clinical features of the disorder.

Material and methods

For this longitudinal retrospective study, we collected the data of all pediatric patients referring to our Hospital (Bambino Gesù Children's Hospital, IRCCS, a tertiary center for the diagnosis and cure of pediatric headache) between January 2000 and January 2018 and complaining of recurrent episodes of vomiting and/or paroxysmal torticollis. In all of them, other possible organic diagnoses were excluded. We searched the electronic hospital database by using ICD codes and text diagnosis of these two conditions. The diagnosis of CVS and BPT respectively was confirmed applying ICHD-3 diagnostic criteria. At least one parent of all selected patients was contacted by phone and/or email and administered a questionnaire aimed at assessing the main clinical features of the disorder as well as the prognosis, with particular emphasis on the development of headache. Headache clinical features (age at onset, frequency of attacks, responsiveness to symptomatic therapy) were evaluated according to ICHD-3 diagnostic criteria.

Local Ethical Board approval was obtained.

Results

Patients' demographics

We identified a total of 150 patients with a diagnosis of CVS. Sixty-three were excluded due to lack of data and unavailability of parents/caregivers to give information on follow-up (such as modified phone number/contact details, unwillingness of parents to participate, lack of response to phone calls/e-mails), and five were excluded due to coexistent medical conditions that could cause vomiting. Therefore, the final analysis included 82 patients (41 male, 41 female). Mean age at the time of interview was 12 years (range 4–24 years). As for BPT, we identified 38 patients, but five were excluded due to unavailability of parents/caregivers to give information on follow-up. The final analysis therefore included 33 patients (10 male, 23 female) with a mean age at the time of interview of 7 years (range 4–17 years). The main information is summarized in Figure 1.

Figure 1.

Flow-chart summarizing inclusion and exclusion patients with CVS (a) and BPT (b) with their final outcome.

CVS

Mean age at CVS onset was 3 years 9 months (range 1 month to 14 years), and the majority of patients (67%, n = 55) presented this disturbance for 3 years or more. Fifty-seven percent of patients (n = 47) presented more than five episodes of vomiting for each attack. Although the exact frequency of vomiting per hour was difficult to obtain in all patients, parents usually referred to different vomiting occurring in one hour as a single episode. The frequency of episodes was widely variable, ranging from sporadic episodes to daily attacks (for at least a brief period) in eight patients. The vast majority of patients (85%, n = 70) also presented with other episodic syndromes that may be associated with migraine. In particular, recurrent gastrointestinal disturbance was present in 45% of patients (n = 37), infantile colic in 39% (n = 32), motion sickness in 30% (n = 25), recurrent limb pain in 27% (n = 22), BPT in 15% (n = 12), and benign paroxysmal vertigo in 13% (n = 11). Nearly half of all patients (48%, n = 39) presented with more than one episodic syndrome during infancy and childhood. Seventy-eight percent (n = 64) of patients presented a positive familial history of migraine, while a familial history of episodic syndromes associated with migraine was difficult to assess due to missing/forgotten information. Seventy-nine percent (n = 65) of patients presented with headache during the follow-up. Mean age at headache onset was 6 years, but 29 out of 65 patients (45%) started between 3 and 6 years of age. In 63% of patients developing headache, there was at least a monthly attack frequency, with almost 11% of patients complaining of four or more episodes per month. Phonophobia was present in 34% of patients, photophobia in 46%, nausea in 31% and vomiting in 26%. Sixty-five percent of patients used symptomatic therapy, which was effective in 88% of them. The application of ICHD-3 criteria allowed a diagnosis of migraine in 71% of patients (n = 46) and of tension-type headache in the remaining 29% (n = 19).

BPT

Mean age at BPT onset was 5 months, with 91% of patients (n = 30) presenting the first episodes in the first year of life and 49% of patients (n = 16) presenting this disturbance in the first 3 months of life. In 24% of patients, each episode of BPT lasted for some hours; in 36% some days, and in 27% it even lasted several weeks. The vast majority of patients (73%) presented more than five episodes throughout their life, with 46% presenting more than 10. Thirty out of 33 patients (91%) also presented with other episodic syndromes. In particular, 79% presented recurrent gastrointestinal disturbance, 61% recurrent limb pain, 45% benign paroxysmal vertigo, 15% motion sickness and 9% CVS. Eighty-eight percent (n = 29) of patients presented a positive familial history of migraine, while familial history for episodic syndromes was difficult to assess due to missing/forgotten information. Sixty-seven percent (n = 21) of patients suffered from headache during the follow-up, with a mean age at onset of 5 years. In 67% of cases, headache attacks started at or before 5 years of age. In 60% of patients developing headache, attacks lasted for several hours, and in 41% of cases there was at least a monthly attack frequency, with 18% of subjects complaining of four or more episodes per month. Photophobia, phonophobia and nausea were present in 52% of patients, and in 19% there was a worsening of pain with physical activity. Sixty-eight percent of patients used symptomatic therapy, which was effective in 60% of them. The application of ICHD-3 criteria allowed a diagnosis of migraine in 81% of patients (n = 17) and of tension-type headache in the remaining 19% (n = 4).

Discussion

In our sample, CVS was associated with a very high risk of developing subsequent migraine, with a prevalence of 79% versus the expected 7.7% of the general pediatric population (11), and versus a previously reported 40-46% (1). In this subgroup of patients, the age at onset of migraine was slightly lower than expected (6 years vs. 7 years). Compared to patients who had been suffering from CVS, in patients presenting with BPT during early infancy the rate of subsequent migraine was slightly lower, but always higher than in the general population and as previously reported (67% vs. 12%) (4). It is also worth noting that 13 subjects in the CVS sample and three in the BPT group are 7 years old or younger, thus making it possible that migraine will develop in the coming years. This means that the real prevalence of migraine in subjects over 7 years of age with a history of CVS and/or BPT is probably even higher than that reported in the present study. Among children developing headache during the follow-up in both groups, migraine was present in 73% of patients, and tension-type headache in the remaining 27%. Previous literature data on the association between episodic syndromes that may be associated with migraine and subsequent migraine and/or tension-type headache found contradictory results, with an apparent prevalence of migraine in some papers (14,15), and no significant differences in others (6). However, it is to be underlined that previous studies examined the prevalence of episodic syndromes in two distinct populations of patients with migraine and tension-type headache. On the other hand, in the present work, we longitudinally followed up children with episodic syndromes that may be associated with migraine to explore the symptoms appearing over time. Furthermore, this distinction should be regarded with particular caution, as it is well known that these disorders share many clinical features at pediatric age, raising the hypothesis of a common pathogenetic mechanism (16).

In both groups of patients, the vast majority presented with different episodic syndromes at different ages, thus suggesting an age-dependent evolution of migraine-like symptoms before the onset of clear migraine headache. However, it should be underlined that for different conditions there are many diagnostic limitations, due to the lack of inclusion in the current ICHD-3 classification (i.e. recurrent limb pain) or for the inclusion as “additional conditions” lacking diagnostic criteria (i.e. motion sickness and recurrent sleep disturbances).

Limitations of the study

Our study shows a limitation that must be recognized. Contacting patients' parents via phone calls and/or e-mails can indeed hamper the verification of the clinical diagnostic criteria. This increases the known difficulty in the applicability of ICHD criteria to children, particularly up to 6 years of age (12,13). However, even if this point can bias the correct diagnosis of patients' headache, it does not interfere with the percentage of patients with previous diagnosis of CVS or BPT who have developed headache.

Conclusions

The present study shows that CVS and BPT are associated with a high risk of subsequent development of primary headache (migraine and tension-type headache). Indeed, the prevalence of primary headache in our patients was definitely higher than that observed in the general pediatric population. Application of ICHD-3 clinical diagnostic criteria is crucial to correctly classify patients with episodic syndromes that may be associated with migraine; however, not all conditions are included in this classification. Indeed, recurrent limb pain is commonly considered as a migraine equivalent in clinical practice, but it is not considered among the episodic syndromes that may be associated with migraine in ICHD-3. Furthermore, motion sickness and recurrent sleep disturbances, though listed in the ICHD-3, are considered as “additional conditions” without explicit diagnostic criteria. Therefore, these symptoms are very difficult to explore rigorously. More longitudinal studies on larger samples of patients are warranted to confirm the existing relationship between episodic syndromes and subsequent primary headaches and unravel the underlying physiopathological mechanisms linking these conditions.

Clinical implications

Episodic syndromes that may be associated with migraine are a common complaint in the pediatric population.

There is a very high risk of developing migraine/tension-type headache after CVS and/or BPT.

Most patients present more than one episodic syndrome before developing primary headache, suggesting an age-dependent expression of migraine and migraine-like symptoms.