PEDIATRICS Vol. 107 No. 2 February 2001, pp. 265-269

Prospective Study of Children With Cyanotic and Pallid Breath-Holding Spells

Francis J. DiMario Jr, MD

From the Department of Pediatrics, University of Connecticut, Connecticut Children's Medical Center, Hartford, Connecticut.

	ABSTRACT

Top Abstract Results Discussion References

Objective.  In this investigation, we sought to prospectively document the natural history of severe breath-holding spells (BHS) among children with both cyanotic and pallid BHS who were referred for neurologic consultation.

Design.  Prospective cohort study.

Methods.  A structured interview was undertaken at the time of initial consultation and at subsequent 1-year intervals regarding type of BHS, frequency of spells, associated phenomenon, sequelae, family history, and age at termination of spells.

Results.  A total of 95 children (48 boys, 47 girls) with BHS were identified and followed over a 9-year interval. There were no significant differences between genders. Median onset age was between 6 and 12 months old with 15% presenting younger than 6 months. A median frequency of spells was weekly with 30% experiencing 1 or more spells per day. The median age at peak frequency was between 12 and 18 months old with a range extending from 6 months to 4 years of age. Of the patients whose BHS had remitted for >12 months' time (n = 67), the last spell occurred at a median age of 37 to 42 months. Of those children whose BHS were still occurring, the oldest age at time of latest spell was at 7 years old. Hypoxic convulsions were associated with BHS in 15% of all participants. A positive family history of BHS was identified in 34% of all families with equal frequency distributed between paternal and maternal sides.

Conclusions.  In this study, new data concerning the natural history of BHS have been determined. The information is important for family counseling and identifying intervention strategies, and serves as baseline data to evaluate the efficacy of future treatment approaches.  Key words:  breath-holding, pallid infantile syncope, cyanotic spells, anoxic convulsions.

Severe breath-holding spells (BHS) of childhood are a readily identified clinical entity. The diagnosis is based on a distinctive and stereotyped sequence of clinical events beginning with a provocation resulting in crying or emotional upset that leads to a noiseless state of expiration accompanied by color change and ultimately loss of consciousness and postural tone.^1,2 Two clinical types are recognized based on the child's coloration (cyanotic or pallid) during these events.² Children typically experience 1 predominate type; however, individual children may experience mixed types. BHS is an involuntary, nonvolitional, reflexic, nonepileptic paroxysmal phenomenon of childhood. The episodes occur during full expiration despite its misnomer. Current data supports underlying autonomic dysregulation as a common mediating pathway that results in loss of consciousness.^3-9 This is emphasized by the observations that some children have both types (cyanotic and pallid) of BHS with 1 type predominating, and that within families multiple children may have different types of BHS.¹⁰ There is a genetic basis identifiable in significant proportions of patients. An autosomal dominant trait with reduced penetrance has been demonstrated.¹⁰ Retrospective data accounts for much of the known epidemiology of BHS.^2,11-14 In this study, we attempted to obtain prospective data concerning the natural history of a cohort of children with BHS.

	PATIENTS AND METHODS

All children referred for evaluation by a pediatric neurologist (F.J.D.) and diagnosed with either cyanotic or pallid BHS were eligible. Patients identified between 1988 and 1996 were considered for analysis. A diagnosis of BHS was based on a history of multiple episodes (3 or more) defined by the following clinical sequence: provocation, followed by crying to a point of noiselessness and accompanying change of color (cyanotic or pallid), and ultimately a loss of consciousness with an associated alteration in body tone. Each child was examined and diagnosis made. All children with identified pallid spells underwent an electrocardiogram to exclude prolonged QT interval.

A structured interview was used to obtain specific data regarding the type of BHS, age at onset, age at peak frequency, frequency range, associated phenomenon, family pedigree, and age at termination defined as 1 year without spells. Yearly evaluation or telephone updates were obtained on the cohort until BHS had terminated or family was lost to follow-up. All data were tabulated and summarized.

	RESULTS

Top Abstract Results Discussion References

A total of 95 children was identified and included in the current study. These include 47 boys and 48 girls. No specific medication treatment interventions were used. A single child with pallid BHS was treated with atropine between the ages of 16 and 28 months during which time she experienced no spells. One additional child was treated with thyroid hormone replacement for hypothyroidism identified 2 years after onset BHS. All parents were given instructional material and underwent education as to the underlying pathophysiology of BHS. Follow-up contact was completed for all patients at 1 year after initial diagnosis. There were 84/95 patients who had continued spells during the first observation year. A total of 17 patients were lost to follow-up after at least 1 year follow-up contact and 11 had ongoing BHS. Sixty-seven have completed follow-up (ie, spells terminated for 1 year). Of all patients evaluated, 49/95 had cyanotic spells only, 27/95 had pallid spells only, and 19/95 had both types with one predominating over the other (see Fig 1). In BHS patients, 85% of the time patients had a color change associated with their spells whereas 15% of the time a clear color change was not apparent. There were no significant differences between breath-holders with predominant cyanotic or pallid spells in terms of any clinical follow-up data. All (46/46) electrocardiogram recordings were normal. Therefore, all subsequent results pertain to all breath-holders as a single group. An immediate family member was identified as having had BHS in 34% of families. This was true for children with cyanotic or pallid spells.

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Pie chart depicting percentages of patients with cyanotic, pallid, and both type BHS.

The vast majority of children had their age of onset between 6 and 12 months of age. However, 12% of patients experienced their first breath-holding spell within the first 6 months of life. A minority of 5% (4/95) began to experience BHS in the perinatal period (day 1 through 29 of life, and as early as within hours of birth in a single case). Importantly, only 3 patients had a late onset between 25 and 30 months of age (see Fig 2).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Bar graph depicting age of onset BHS.

The frequency of BHS was variable. By definition, patients included in the current study had at least 3 spells before inclusion. At the time of peak frequency, 24/95 had >1 episode per day. The median frequency was daily to weekly spells with an average being weekly (see Fig 3). The age at peak frequency was also variable but occurred between 13 and 24 months with a mean of 18 months of age (see Fig 4). Of the 67 patients in whom an age at termination of spells was identified, the mean and median age was 36 months. There were 15 children whose termination age was after age 4 years. These included 7 children with cyanotic BHS, 5 with pallid BHS, and 3 with both types pallid predominant.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Bar graph depicting the number of patients attaining each frequency of BHS at their maximum frequency.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Bar graph depicting age at peak frequency of BHS (given in months).

Of all children included in this study 15/95 had hypoxic convulsions associated with their BHS. These most often occurred at the terminal phase of the spell and were not amenable to anticonvulsant therapy where tried (n = 3). Antiepileptic drugs used included phenobarbital and phenytoin. Among the 67 patients who had had a termination of BHS, 12 have thus far had episodes of syncope (18%). The majority of patients who experienced subsequent syncopal events had them brought on by environmental factors (ie, minor trauma, heat, fright, or prolonged standing). Four children have been identified as learning disabled and are in special education classes.

	DISCUSSION

Top Abstract Results Discussion References

BHS have been reported to occur in approximately 0.1% to 4.6% of well children.^2,12 A retrospective review of self-referrals to the Harriet Lane home over a 22-year period yielded 83 patients who sought consultation for severe BHS from a total of about 95 000 patient visits.¹² In 1967, Lombroso² published results from a prospective study conducted over a 2-year interval as part of a larger collaborative project. Approximately 5000 patients were solicited by direct questionnaire resulting in some 225 cases of severe BHS amounting to a prevalence of approximately 4.6%. Data regarding the relative prevalence of simple BHS is not as well-documented. Bridge,¹² as part of a smaller prospective study, in the early 1940s compiled questionnaires regarding the occurrence of these simple spells in a well-child population of 188 patients that resulted in an approximate prevalence of 27%. As part of a prospective outpatient study screening for the prevalence of temper tantrums in 800 3- to 12-year-old school children in New Delhi, India, 182 were found to have temper tantrums.¹⁵ A comparable control group of an additional 182 children were further analyzed resulting in a total of 26 (7%) children from the combined groups who experienced BHS.¹⁵ No distinction between severe and simple was made in this sample. A prospective outpatient survey conducted in Denver, Colorado, over a 5-month period surveyed parents for the presence of BHS in a pediatric practice.¹⁶ Of the 697 children surveyed, mild BHS (recognizable cyanosis without loss of consciousness) were identified in 33 or 4.7% of the sample. Severe BHS (comprising children who lost consciousness with or without convulsions) were identified in 12 children or 1.7% of the sample.¹⁶

Data presented in the current report is consistent with that presented by previous authors. Important differences, however, do exist which may in part be accounted for by the nature in which data were accrued. Clinical types of BHS were originally delineated in the study published by Lombroso and Lerman.² In their group of 225 severe breath-holders, 62% were identified as becoming cyanotic; 19% remained uncertain or not classifiable. Later, Laxdal et al¹⁴ retrospectively evaluated 150 children with severe BHS of whom 129 could be categorized. They found 54% with cyanotic spells, 22% with pallid spells, 12% having both types, and 7% not classifiable.¹⁴ This is consistent with data we have provided in our cohort. The approximate ratio of cyanotic:pallid:both has emerged as 5:3:2 ratio fairly consistently.

The onset age of BHS has been identified within the first 12 months of life in most studies addressing this issue. Bridge¹² reported 66 of 83 patients having had their onset by age 18 months of age and Laxdal et al¹⁴ reported 87% of 150 patients having had their onset before age 18 months. Lombroso² reported data on 193 patients identifying onset age before 12 months in 66%, and virtually all presented by 36 months of age. Importantly, 7% of his patients had onset within 2 months of life compared with 12% within the first 6 months of life, and 5% within the first 29 days of life in the current study. However, BHS with onset at this early age may not be diagnosed with certainly until sufficient observation time and clinical investigation can be obtained.

The frequency of BHS varies from multiple episodes per day to as infrequent as yearly.^2,12-14 A majority of children, however, will experience multiple episodes per week and, as identified in the current study, as many as one third will experience >1 episode per day at peak frequency. In the current study, boys seem to have an earlier age to peak frequency (13-18 months) than do girls (19-24 months; see Fig 4). This has not been identified previously. Another important piece of new data are that the mean age at termination was found to be approximately 36 to 42 months old (see Fig 5). This is slightly earlier than figures previously published. Retrospective data on 51 of an original 83 patients reported by Bridge¹² after follow-up of 9 years suggested that approximately half had termination of spells by 4 years of age. Eighty of 83 patients followed up an average of 12 years by Lombroso² and 106 of 123 patients followed up an average of 9 years by Laxdal et al¹⁴also reported a similar termination age. These new data have particular importance when considering the effectiveness of therapeutic interventions because efficacy has to be judged against natural course of expected age at peak frequency and age at expected resolution. An important clinical observation emphasized in the current study is that there are some children who exhibited both cyanotic and pallid BHS. This occurs in families in whom 1, both, or neither parent had experienced BHS as children themselves. This implies that the common unifying pathophysiology, autonomic dysregulation, is sufficiently sensitive to allow either sympathetic overactivity in the case of cyanotic BHS or parasympathetic overactivity in the case of pallid BHS to predominate in most and vacillate between these alternatives in some children. As such, one would not expect significant differences in the natural history of either BHS type. This is precisely what our data shows. In this study there were 15/67 children with the oldest ages at termination (4 years). Of these 15 children, 7 had cyanotic BHS, 5 had pallid BHS, and 3 had both types with pallid predominant.

View larger version (37K): [in this window] [in a new window]   Fig. 5.   Bar graph depicting age at which the last episode of BHS had occurred after 1 year remission.

Several authors have identified a familial tendency for BHS spells within family members.^2,4,14,17,18 A range of 20% to 35% of all patients will have an identified family member who had suffered BHS at some time during their childhood.^2,4,14,17,18 Careful analyses of a portion (n = 57) of the families contained in the current study have been previously reported in detail.¹⁰ It was found that 27% of 114 proband parents and 21% of 43 proband siblings had current or previous BHS, and 85 members of extended family pedigrees had current or previous severe BHS.¹⁰ This was consistent with an autosomal dominant inheritance with reduced penetrance.¹⁰ There was an essentially equal male to female ratio.¹⁰ Daoud and colleagues in a prospective study of iron therapy in children with BHS found that nearly 50% of their patients had positive family histories of BHS.¹⁹ In that study sample, 70% were identified as being products of consanguineous marriages. Although direct genetic transmission does not explain the occurrence of BHS in all children, a substantial proportion of cases may have a genetic predisposition.

An important part of evaluating children with BHS is counseling their parents. Accurate knowledge of an expected natural history course can allay some parental anxiety as to what to expect. Data presented within this study identify an earlier age at peak frequency than previously reported with a slight gender difference. Similarly, there is an earlier mean age at resolution than previously reported. This is important when considering what parents should anticipate for their child at a given age. This should also prompt providers to investigate potential underlying complicating factors in children whose course seems to be deviant from expected natural history. Additionally, these facts have clear relevance when considering the efficacy of various treatment interventions.

	FOOTNOTES

Received for publication Nov 8, 1999; accepted Jun 13, 2000.

This work was presented in part at the Child Neurology Society Meetings; October 21-24, 1998; Montreal, Canada.

Reprint requests to (F.J.D.) Department of Pediatrics, Connecticut Children's Medical Center, 282 Washington St, Hartford, CT 06101.

	ABBREVIATIONS

BHS, breath-holding spells.

	REFERENCES

Top Abstract Results Discussion References

1. DiMario FJ Jr Breath-holding spells in childhood. Am J Dis Child 1992; 146:125-131 [Abstract]
2. Lombroso CT, Lerman P Breath-holding spells (cyanotic and pallid infantile syncope). Pediatrics. 1967; 39:563-581 [Abstract/Free Full Text]
3. DiMario FJ Jr, Burleson JA Autonomic nervous system function in severe breath-holding spells. Pediatr Neurol. 1993; 9:268-274 [CrossRef][Medline]
4. DiMario FJ Jr, Chee CM, Berman PH The evaluation of the autonomic nervous system in children with pallid breath-holding spells. Clin Pediatr. 1990; 29:17-24
5. Kahn A, Rebuffat E, Sottiaux M, Muller MF, Bochmer A, Grosswasser J Brief airway obstructions during sleep in infants with breath-holding spells. J Pediatr. 1990; 117:188-193 [CrossRef][Medline]
6. DiMario FJ Jr, Bauer L, Baxter D Respiratory sinus arrhythmia in children with severe cyanotic breath-holding spells. J Child Neurol. 1997; 12:260-262 [Medline]
7. DiMario FJ Jr, Bauer L, Baxter D Respiratory sinus arrhythmia in children with severe cyanotic and pallid breath-holding spells. J Child Neurol. 1998; 13:440-442 [Medline]
8. Hunt CE Relationship between breath-holding spells and cardiorespiratory control: a new perspective. Pediatrics. 1990; 117:245-247 [CrossRef]
9. Anas NG, McBride JT, Brooks JG Childhood cyanotic breath-holding spells and cardiorespiratory control. J Pediatr. 1991; 118:656 [Medline]
10. DiMario FJ Jr, Sarfarazi M Family pedigree analysis of children with severe breath-holding spells. J Pediatr. 1997; 130:646-651
11. Hinman A, Dickey LB Breath-holding spells. A review of the literature and eleven additional cases. Am J Dis Child 1956; 91:23-33
12. Bridge EM, Livingston S, Tietze C Breath-holding spells. Their relationship to syncope, convulsions, and other phenomena. J Pediatr. 1943; 23:539-561 [CrossRef]
13. Livingston S. Disorders simulating epilepsy. In: Diagnosis and Treatment of Convulsive Disorders in Children. Springfield IL: Charles C. Thomas Pub; 1954:71-75
14. Laxdal T, Gomez MR, Reiher J Cyanotic and pallid syncopal attacks in children (breath-holding spells). Dev Med Child Neurol. 1969; 11:755-763 [Medline]
15. Bhatia MS, Dhar NK, Singhal PK, Nigam VR, Malik SC, Mullick DN Temper tantrums: prevalence and etiology. Clin Pediatr. 1990; 29:311-315
16. Linder CW Breath-holding spells in children. Clin Pediatr. 1968; 7:88-90
17. Maulsby R, Kellaway P. Transient hypoxic crisis in children. In: Kellaway P, ed. Neurological and Electroencephalographic Correlative Studies in Infancy. New York, NY: Grune and Stratton; 1964:349-360
18. Sibert PL, Gubbay SS Familial cyanotic breath-holding spells. J Pediatr Child Health 1992; 28:254-256
19. Daoud A, Batieha A, Al-Sheyyab M, Abuekteish F, Hijazi S Effectiveness of iron therapy on breath-holding spells. J Pediatr. 1997; 130:547-550 [CrossRef][Medline]