Article Text

Download PDFPDF
Presentation and outcome of water-related events in children with long QT syndrome
  1. Louise Albertella,
  2. Jackie Crawford,
  3. Jonathan R Skinner
  1. Greenlane Paediatric and Congenital Cardiac Services, Starship Children's Hospital, Auckland, New Zealand
  1. Correspondence to Dr JR Skinner, Greenlane Paediatric and Congenital Cardiac Services, Starship Children's Hospital, Park Road, Grafton, Auckland, New Zealand; jskinner{at}

Statistics from

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.


Long QT syndrome (LQTS) occurs in about 1 in 2500 of the general population.1 Of the 12 genotypes, LQTS type 1 is the most common type and is classically associated with exercise-related events, especially swimming,2 and the QT interval prolongs with Epinephrine infusion3 and after exercise.4 Mutations are in KCNQ1 and the cardiac potassium channel IKs is defective, such that the action potential is prolonged.5 6 Of the first 388 index cases with LQTS at the Mayo clinic, 28 (11%) had a history of swimming triggered events; 85% of these were due to LQTS type 1, 6% to LQTS type 2 and 3% were genotype negative.7

The relationship of drowning and near drowning to LQTS is well established. However, the nature and outcome of these events are not well described in children. It is essential to have a high index of suspicion when faced with these cases in order to make a prompt diagnosis and prevent unnecessary mortality and morbidity in these children, and take the opportunity to find and protect undiagnosed family members. We recently found that one third of a group of unrelated probands diagnosed with LQTS were initially misdiagnosed as having a seizure disorder, resulting in years of inappropriate treatments.8 During these years of delay (a mean of 9 years in those treated as epileptic), there were a number of sudden deaths in family members which might have been prevented. With this in mind we aimed to review the details of presentation and outcome, and time to final diagnosis, of children with LQTS who presented with a syncopal event related to water.


Data were retrieved from LQTS probands (the first to be identified in a family) registered with the Cardiac Inherited Diseases Registry in New Zealand. This consent-based voluntary patient and family registry permits the storage of detailed clinical and genetic information.9,,11 Detailed review of the records from children with water-related syncope was undertaken, and who were subsequently proven to have LQTS. In particular, we report the exact mode of presentation, as reported by first-hand witness accounts, if there was any delay in the diagnosis being made, and the outcome for the proband and other family members. The corrected QT interval (QTc) reported is the maximal QT interval, measured in V5 and lead II, corrected for heart rate using the Bazett formula (QT interval divided by the square root of the preceding R-R interval).


Data from 10 genotyped LQTS subjects with a documented history of water-related syncope in childhood were retrieved. Nine had LQTS type 1 (KCNQ1), and in one case the genotype could not be identified (see table 1). The male:female ratio was 1:1. The age of the subject at the time of the water-related event ranged from 3 to 13 years. The water-related event led to a diagnosis in six subjects, but three subjects were diagnosed 1–17 years later due to another subsequent event, one of which was fatal. Initial QTc ranged from 450 ms to 600 ms. Minimal QTc on follow-up ranged from 340 ms to 460 ms. All subjects were started on β-blocker medication, three also had an intracardiac defibrillator (ICD) inserted, two of whom have also had left cardiac sympathectomy. The probands were subsequently found to have between two and seven affected family members each, through family screening. In the family of one subject (case number 7), one parent was gene negative and the other declined testing.

Table 1

Genotype and mutation characteristics among 10 children with long QT syndrome and water-triggered cardiac events

Nature of presentation and outcome

Subject 1 was 12 years old when he had an episode of underwater syncope in a friend's swimming pool (table 2). He felt something tugging at his foot and dived down to investigate. He then describes waking up on the bottom of the pool, before swimming up to the surface. Prior to the event, he recalls immersing his face and looking at the bottom of the pool. He had had an episode of syncope 2 weeks prior at home in the bathroom. He remains well on β-blocker medication 5 months since presentation. He recently had a left cardiac sympathectomy. Three affected family members have been identified to date. His paternal grandfather was reported to have had seizures while swimming as a youngster.

Table 2

Clinical details and outcome for 10 children with long QT syndrome having water-triggered cardiac events

Subject 2, who is now 29 years old, was 7 years old when she had an episode of underwater syncope in a public swimming pool. She was racing and was pulled out at the end of the pool having gone stiff. No resuscitation was required and she regained consciousness spontaneously. She was diagnosed with epilepsy and started on antiepileptic medication. But it was not until 7 years later that she was formally diagnosed with LQTS. This was following an episode of documented ventricular fibrillation (VF) requiring cardio version while watching cricket. Aside from this episode her seizures were always related to exercise. She remains well on β-blocker medication but has had one episode of torsade de pointes recorded on her ICD. Three affected family members have been identified.

Subject 3, who is now 23 years old, was 13 years old when she had an episode of syncope followed by a seizure after she had been swimming in a lake. She felt weak while swimming, having dived into the water, but managed to get herself out of the lake before she lost consciousness. Her QTc was so long (>600 ms) that an ICD was implanted the next month. She remains well on β-blocker medication and has had one episode of torsade de pointes recorded on her ICD. She has seven affected family members.

Subject 4, who is now 14 years old, was also 13 years old when she had an episode of underwater syncope while racing in a public swimming pool. She was noted by her coach to stop swimming suddenly and appeared to be looking at her hand but not moving. She was pulled from the water by fellow swimmers who were struck by her ‘floppiness’. Her coach commented that her ‘heart was racing’. She did not require any resuscitation and remains well on β-blocker medication. She has three affected family members and had an Aunt who died while swimming, who had a past medical history of developmental delay and epilepsy.

Subject 5, now 24 years old, was 10 years old when he also had an episode of underwater syncope while racing in a public swimming pool. He had been diagnosed with LQTS 2 years prior to the event when he presented with syncope while running. He remains well on β-blocker medication 16 years after the diagnosis. He has two affected family members and there is a history of a sibling who died from sudden infant death syndrome.

Subject 6, now 32 years old, was 11 years old when he presented with an episode of underwater syncope. He was pulled out of the water at a public swimming pool, having swum most of the length of the pool. He did not require resuscitation. He had previous symptoms of syncope while running and playing rugby. He remains well 21 years after the initial event on β-blocker medication, with reports of paroxysmal atrial fibrillation only. He has two affected siblings.

Subject 7, now 19 years old, was 8 years old when he had a near drowning at a private pool. He had been performing somersaults under water but did not surface. He was found on the bottom of the pool, and it was estimated that he had been submersed for 2–5 min. Cardiopulmonary resuscitation was initiated and the ambulance crew arrived 10 min later. VF was documented, and he was cardioverted with a Direct Current shock. Following the event and resuscitation, he is severely intellectually impaired and has spastic quadriplegia. He has gross QT prolongation (see figure 1). One parent was gene negative. We have been unable to screen the other side of this family.

Figure 1

12 lead ECG from subject 7, after a near drowning. ECG was taken several days after recovery from cardiac arrest, with normal serum electrolytes and at normothermia. He has a mutation in KCNQ1. The T wave is grossly abnormal, with a broad base, typical of long QT type 1. In lead V5, QT is 0.64 s, the preceding R–R interval is 1.13 s giving a QTc of 0.60 s.

Subject 8 was 10 years old when he had a syncopal episode after swimming underwater at the school pool. He had become weak while swimming and was helped out of the pool. He proceeded to have a brief seizure before regaining consciousness spontaneously. Three years earlier he had been diagnosed with epilepsy and was receiving antiepileptic medications. At the age of 13 he collapsed while warming up for a school hockey game. He could not be resuscitated. The autopsy was negative. It was noted that his EEG's in life had always been normal. This case has been reported previously.12 Molecular genetic testing on his newborn screening card revealed a mutation in KCNQ1, the gene linked to LQTS type 1. The mutation was also present in his mother and she was placed on β-blocker medication, the sibling was negative. The wider family has declined genetic testing.

Subject 9, now 42 years old, was 5 years old when she had an episode of underwater syncope. Prior to final diagnosis at the age of 22 she had had recurrent episodes of syncope and seizures. After diagnosis she had a left cardiac sympathectomy. After many years with no symptoms, she had a further episode of syncope at the age of 36, relaxing at a hot pool while off β-blocker medications, and another on β blockers while swimming, when she was pulled from the water and had a brief seizure. She has since had three episodes of torsade de pointes documented on her ICD. Three family members are also affected. Her mother died while dancing at the age of 27.

Subject 10, now 13 years old, was 3 years old when she had an episode of syncope and simply slipped through a flotation device in a pool. There had been no facial immersion. The diagnosis of LQTS was raised at the time but a formal diagnosis was not reached until a year later. She had a past history of syncope preceded by activity and had been diagnosed with breath holding episodes. Five gene-carrying family members have been identified.


This case series reinforces the association between swimming-related cardiac events and children with LQTS. The condition was found to be familial in all nine cases where this was possible. Other family members have been diagnosed, counseled and offered preventive therapies.

Failure to consider LQTS after the event was common. It is worthy of note that a previous family history of seizures or sudden death associated with exercise or swimming in three families had not led to a diagnosis. In addition, one child had classical water-related collapse and seizure diagnosed erroneously as an epileptic seizure, delaying his diagnosis, such that he was only ultimately diagnosed with LQTS posthumously.

Seven out of ten children were swimming, three of whom were competing in a race at the time of the event. The mechanism behind such arrhythmias occurring in water is presumed to be multifactorial. It has been observed that facial immersion in cold water prolongs the QT interval in those with hereditary LQTS, but not in controls.13 Hypothermia (less than 35°C centigrade) has also been shown to prolong the QT interval.14 15

The dive reflex16 causing a fall in heart rate, competes with the release of Epinephrine due to excitement and exercise. Such release of Epinephrine is presumably also behind the association of water-triggered syncope due to catecholaminergic polymorphic ventricular tachycardia, a rarer but highly malignant cardiac ion channelopathy also linked to sudden death in the young.17 18 This condition is diagnosed by detection of ventricular extrasystoles or ventricular tachycardia on exercise test; the resting ECG is normal.18

However, two of the subjects in the present report presented with a water-related cardiac event that cannot be explained by the above mechanisms described. Subject 9 already had a diagnosis of LQTS type 1 and at the time was not taking β-blocker medication. She was simply sitting in a hot pool when the sudden episode of syncope occurred. Since we do not have a rhythm strip from that event, it is possible she had a common faint given that this was warm water. However, subject 10 was within a flotation ring when she had a sudden unheralded syncope, and slipped through, limp, under the water, to be pulled out by her mother.

The association with water might also be partly a phenomenon of age and chance. Data from the international long QT registry have shown us that the risk of death in LQTS varies by age, history of syncope, gender, QT interval and genotype.19 20 Risk of sudden death is highest among males aged 5–15 years during exercise. A transient loss of consciousness underwater may prove fatal, and would rarely escape notice, whereas on land recovery usually occurs and is perhaps more likely to be passed over as a common faint.

How should the QT interval be measured?

The QTc used should be the longest recorded (it is worth taking more than one good quality 12 lead ECG, on separate days) in lead II or V5, at a time remote from a cardiac arrest, allowing time for cerebral recovery if necessary, with normal serum electrolytes and at normothermia. Best repeatability is reported by measuring the QT interval from the onset of the Q wave to the end of the T wave as defined by extending the slope of the rapid descent of the T wave down to the isoelectric line (‘Teach the Tangent’).21

Assessment of risk and management of patients with LQTS

Heart rate QTc over 550 ms indicate particularly high risk. During childhood, boys are at highest risk; females are at highest risk during adult life. Recent syncopal episodes indicate elevated risk at any age.19 20

Those affected must avoid swimming and diving. All gene carriers need to avoid certain medications which prolong the QT interval, which usually act by a direct effect on the cardiac potassium ion channels ( β blockers, taken daily without fail, reduce the risk of sudden death in LQTS by 75% in LQTS type 1 and 50% in LQTS type 2.19 Left cardiac sympathectomy is highly effective in all groups and this can now be performed by thoroscopy.22 Defibrillators are usually reserved for those who have had a cardiac arrest or who are unresponsive to, or unable to take, β blockers.


Syncope, cardiac arrest or seizure in children secondary to LQTS can occur during or after swimming, especially racing, but facial immersion is not essential. Failure to recognise arrhythmic syncope as a cause of water-triggered syncope or seizures results in many years of delayed diagnosis, which may result in a fatal outcome for the child, and potentially other, hitherto unidentified, affected family members.

Syncope or seizure after swimming as well as while swimming or playing in water should be considered to be arrhythmogenic until proven otherwise and should be investigated at least by 12 lead ECG, exercise testing and a thorough family history.


The Cardiac Inherited Disease Registry in New Zealand is supported by Cure Kids.



  • Provenance and peer review Commissioned; externally peer reviewed.