Article Text

Download PDFPDF

The prognosis of cyclical vomiting syndrome
  1. F Dignan,
  2. D N K Symon,
  3. I AbuArafeh,
  4. G Russell
  1. Department of Child Health, University of Aberdeen, Aberdeen AB25 2ZD, UK
  1. Dr Russellconnie.durno{at}


AIMS The medium term prognosis of cyclical vomiting syndrome (CVS) was studied to determine the proportion of affected individuals who had gone on to develop headaches fulfilling the International Headache Society criteria for migraine.

METHODS Twenty six (76%) of 34 CVS sufferers identified from the authors' clinical records were traced, and all agreed to participate. Each child was matched to a control, and telephone interviews were conducted using a standardised questionnaire.

RESULTS Thirteen (50%) of the subjects had continuing CVS and/or migraine headaches while the remainder were currently asymptomatic. The prevalence of past or present migraine headaches in subjects (46%) was significantly higher than in the control population (12%).

CONCLUSION Results support the concept that CVS is closely related to migraine.

  • migraine
  • cyclical vomiting syndrome
  • headache
View Full Text

Statistics from

Cyclical vomiting syndrome (CVS), first described by Gee,1 is a childhood condition with an estimated prevalence of 1.9%.2 It is characterised by recurrent, sudden, and self limiting episodes of nausea, vomiting, and lethargy during which the child looks pale, ill, and miserable.3-5Each attack lasts from a few hours to a few days, and there is complete resolution of symptoms between attacks.

The aetiology of the condition is uncertain. It has been linked to epilepsy,6 7 psychological problems,8-10metabolic disorders,11 and autonomic dysfunction12 but in the great majority of children no abnormality is found. Many authors consider that CVS is a “migraine equivalent”.13 In a previous study,14 we found a similar pattern of associated symptoms in children with CVS and those with abdominal migraine, also considered a migraine equivalent. We have also noted relations between CVS and migraine in an epidemiological study; in particular, children with CVS have an excess of migraineurs among their first degree relatives.2

Also supportive of a relation to migraine is the response of CVS to antimigraine drugs. It may respond to pizotifen,8 15 a serotonin antagonist, to propanolol,16 and to cyproheptadine,17 an antihistamine with serotonin antagonist and calcium blocking properties, all of which are used in the prophylaxis of migraine. A positive response has also been noted in a case treated with sumatriptan,18 a 5-HT1agonist used in the acute management of migraine.

Further support for a relation between CVS and migraine comes from electrophysiological studies. Quantitative electroencephalogram (EEG) changes in CVS are similar to patterns recorded in migraine.19 Good20 performed a detailed analysis of visual evoked responses, and identified similarities between CVS and migraine, with or without aura.

Although the links between CVS and migraine are widely (although not universally21) accepted, it is unclear if children with CVS continue to have attacks of vomiting as they get older, or if the condition transforms into more typical migraine. We have therefore examined the medium term prognosis of CVS to determine the proportion who have gone on to develop headaches that fulfil the International Headache Society22 criteria for migraine.



The study was approved by the Joint Ethical Committee of the University of Aberdeen and Grampian Health Board.


Subjects were identified from clinical records maintained by DNKS, IAA, and GR. We identified 34 individuals in whom a diagnosis of CVS had been made, based on previously published criteria2(see table 1). We identified controls, matched for sex and week of birth, by using the hospital birth records; three potential controls were found for each case and were contacted in turn until one responded.

Table 1

Criteria for the diagnosis of cyclic vomiting


An initial letter was sent to the home address of each child, detailing the aims of the study and explaining that they would subsequently be contacted by telephone. At telephone interview, we then enquired about headache and vomiting in the past year. We used a previously validated headache questionnaire23; table 2shows additional questions on cyclical vomiting. In addition to questioning the child, one parent, usually the mother, was consulted to confirm the accuracy of the information.

Table 2

Questions relating to cyclic vomiting attacks


The proportions of symptomatic subjects were compared using the χ2 test.



Of the 34 children fulfilling our criteria for CVS, current addresses were available for 26 (76%). These comprised 15 females and 11 males, and all agreed to participate in the study. The mean age of respondents was 17 years. There were no significant age or sex differences between those who could and could not be traced.


Table 3 presents the results. Of the CVS subjects, eight (31%) were still experiencing vomiting attacks. Twelve (46%) had a current or previous history of headaches that fulfilled the International Headache Society criteria for migraine, compared to only three (12%) of the controls, a difference that was statistically significant (using Fisher's exact test, two tailed, p = 0.0128).

Table 3

The prognosis of cyclic vomiting syndrome


This investigation is the first case control study to follow up a population of children previously diagnosed with CVS. There is no laboratory test to confirm the diagnosis of CVS, but we have found our clinical criteria easy to apply in both clinical24 and epidemiological2 settings. Although confusion might arise in psychological disorders such as anorexia nervosa and bulimia, in metabolic disorders such as urea cycle disorders, organic acidaemias, and mitochondrial myopathy, encephalopathy with lactic acidosis and stroke (MELAS), raised intracranial pressure, and infections, especially of the urinary tract, there are usually other clues to these disorders.

We chose to approach our patients for telephone rather than personal interview because our hospital service covers a wide geographic area, and clinical interviews in the hospital were unlikely to find favour with families living at a distance. In addition, commitments at school, college, and work make attendance problematical for this age group. The diagnosis of migraine is based on symptoms, and there are no diagnostic clinical or laboratory findings, so that clinical examination will only rarely influence the diagnosis. We experienced no problems in applying the questionnaire by telephone, and found it easy to interview the teenage subjects and to obtain confirmatory details from a parent.

This follow up study of patients with CVS indicates a strong predisposition on the part of sufferers to develop migraine headache. Almost half had experienced typical migraine headaches, 27% currently and 19% previously. In contrast, only 12% of the control population had ever had migraine, a figure similar to that found in larger populations.25 In a previous study,12 we found that 38% of children also suffered from migraine headaches at the time they consulted for CVS. The present study suggests that this proportion increases with the passage of time. Generally, the onset of CVS precedes that of migraine headaches,26 being at the ages of 5.3 and 8.3 years respectively in our local school population.2 23 It may well be that our findings reflect the tendency for migraine to present with abdominal symptoms in early childhood, with some but not all going on to develop migraine headache as they get older.

About one third of our subjects continued to suffer from cyclical vomiting attacks well into their teens, a finding in keeping with previous reports.27 28 It is therefore inappropriate to regard CVS as a purely paediatric condition that disappears at puberty or during adolescence. Not only do sufferers tend to develop migraine headaches, but many also continue to have episodic vomiting.

Our results confirm that there is an association between CVS and migraine headaches and that in a significant proportion of children, the condition transforms into typical migraine headache.


View Abstract

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.