Article Text

other Versions

Download PDFPDF

Isolated mild sleep-associated hypoventilation in children with Down syndrome
  1. Wai Wong,
  2. Dennis Rosen
  1. Division of Respiratory Diseases, Boston Children's Hospital, Boston, Massachusetts, USA
  1. Correspondence to Dr Dennis Rosen, Division of Respiratory Diseases, Boston Children’s Hospital, 300 Longwood Avenue, Boston, MA 02115, USA; dennis.rosen{at}


Introduction Children with Down syndrome (DS) have a high incidence of obstructive sleep apnea (OSA) that is often associated with hypoventilation. Little is known, however, about the prevalence of sleep-associated hypoventilation independent of OSA in these children.

Methods Retrospective chart review of all children with DS under 18 years of age undergoing polysomnography at a tertiary care paediatric hospital during a 2-year period. Exclusion criteria were as follow: those requiring oxygen or positive-pressure ventilation; with tracheostomy, baseline hypoxia, unrepaired cyanotic heart disease, pulmonary hypertension, and cerebral palsy; or OSA with >5 obstructions/hour.

Results 86 children met inclusion criteria. 68 (79%) had ETCO2values >50 mm Hg during sleep. 37 (43%) ranged 50–55 mm Hg, and 12 (14%) met American Academy of Sleep Medicine criteria for hypoventilation of ETCO2 >50 mm Hg for >25% of total sleep time (TST). Average pulse-oximetry saturation (SpO2) values during sleep were 97.8% (SD ±1; range: 95.1–99.9). Average percentage of TST with SpO2 >92% was 99.89%.

Conclusion Mildly elevated ETCO2 in the absence of OSA is common in children with DS. This may reflect underlying differences in autonomic control of ventilation in these children and may be considered a normal variant not necessitating intervention other than close monitoring for pulmonary hypertension.

  • down syndrome
  • sleep disordered breathing
  • hypoventilation
  • autonomicnervous system
View Full Text

Statistics from

What is already known?

Children with Down syndrome often have hypoventilation associated with obstructive sleep apnea. However, it is unclear whether they may have a predisposition to hypoventilation separate from the obstruction due to differences in autonomic control of breathing.

What this study adds?

Children with Down syndrome have a higher incidence of mild hypoventilation in the absence of other sleep-disordered breathing. This may be a normal variant secondary to their condition that does not require aggressive treatment.


Down syndrome (DS), or trisomy 21, is the most common chromosomal variance in humans with an incidence of 1 in every 691 live births.1 Among the many medical problems common to people with DS is a high incidence of obstructive sleep apnea (OSA) relative to the general paediatric population.2–4 Many children with DS who have OSA often have coexisting hypoventilation, which in children is defined by the American Academy of Sleep Medicine (AASM) as arterial, transcutaneous or end-tidal carbon dioxide levels >50 mm Hg for >25% of total sleep time (TST).5 In one study by Schott et al,6 OSA and hypoventilation were found to coexist in 39% of children with DS undergoing polysomnography.

Hypoventilation is often viewed as indicating a more serious degree of obstruction, in which minute ventilation is compromised, and may lead clinicians to pursue more aggressive treatment for children in whom the number of discrete obstructive events may be low but who have coexisting mild hypoventilation. However, it may also be that children with DS are predisposed to maintaining higher than normal ETCO2 values in comparison with typical children because of innate differences in their autonomic control of breathing. These may essentially be a part of having DS itself, similar to other differences in autonomic activity that have been observed and described previously in children with DS.7–10

Because we are unaware of previous studies looking at the prevalence of isolated mild hypoventilation in children with DS, we undertook this study to characterise gas exchange in children with DS with either minimal or no OSA. Our hypothesis was that mildly elevated ETCO2 values in children with DS may constitute a normal variant that, if recognised as such, could be expectantly followed instead of being treated.


Under an Institutional Review Board-approved protocol, a retrospective chart review of all standard overnight polysomnography performed in the outpatient setting at an academic, tertiary care, paediatric hospital sleep laboratory on patients with DS aged ≤18 years between January 2012 and December 2013 was conducted. All studies were performed without sleep deprivation or sedation. Polysomnography was conducted and scored using the 2007 AASM guidelines.11 Parameters measured included: 10-lead electroencephalogram; electrooculogram; submental electromyogram and anterior tibial electromyography; airflow measurement with oronasal thermistor and nasal pressure transducer; thoracoabdominal movement measured via impedance plethysmography; pulse oximetry and capnography and video and microphone recording to assess snoring, breathing patterns and movement. PSG was performed using digital polysomnographic equipment (Natus Sleepworks, Natus Medical Incorporated, San Carlos, California, USA), and oximetry was done using Masimo Rad 9 with SET Technology, using a 2-s window. Sleep stages, TST, leg movements and arousal index were routinely scored. Obstruction was scored per the 2007 AASM manual for scoring of sleep and associated events as follows. Events were scored as obstructive apneas when lasting for two breath cycles or greater; associated with a >90% reduction in airflow measured by thermistor compared with baseline for ≥90% of the discrete respiratory event and were associated with ongoing respiratory effort for the duration of the decreased flow. Events were scored as obstructive hypopnea when lasting for two breath cycles or greater; associated with a >50% reduction in airflow measured by nasal-pressure signal amplitude compared with baseline for ≥90% of the discrete respiratory event; were associated with ongoing respiratory effort for the duration of the decreased flow and associated with arousal, awakening or desaturation of >3%. Sleep was scored in 30-s epochs. Only epochs in which definable electroencephalogram and respiratory signals were present for >50% of the 30-s period were counted. A minimum of 480 scored epochs (240 min) were necessary for a polysomnogram to be included in this study. Studies with <4 hours of TST, time in stage R of <30 min, time in non-R of <90 min and unreliable ETCO2 recording (in which no ETCO2 was recorded or the majority of ETCO2 values were consistently recorded as <30 mm Hg within a given 30-s epoch) were excluded.

A total of 258 patients were identified. Exclusion criteria included the need for supplemental oxygen or noninvasive positive-pressure ventilation (NPPV) during sleep; presence of tracheostomy, baseline hypoxia with oxygen saturation of <97% in room air during wake; unrepaired cyanotic cardiac disease; chronic lung disease; pulmonary hypertension; neuromuscular disease; severe or profound cerebral palsy or a history of hypoxic-ischaemic encephalopathy. Patients with partial-continuous upper airway obstruction, or obstructive or central sleep apnea with an Apnea–Hypopnea Index (AHI) ≥5/hour were also excluded.


Of the 258 patients, 172 patients were excluded. About 86 children with DS met the inclusion criteria. Seventy-one were excluded due to inadequate ETCO2 collection, rapid eye movement/ non-rapid eye movement (REM/NREM) sleep data, oxygen saturation data or <4 hours of TST. Sixty patients had an AHI of >5/hour. Ten patients used either oxygen or NPPV regularly, and one patient had a tracheostomy. Two patients had severe or profound cerebral palsy, one had a progressive neuromuscular disorder and one patient had pulmonary hypertension. Twenty-five patients had average wake baseline oxygen saturation of <97% in room air. One patient was excluded due to being inpatient status.

Subject demographics are presented in table 1. The average age was 5.9 years (SD: ±3.7; range: 0.25–16). Males comprised 52% of the study, whereas females comprised 48% of the study. The average body mass index was 19.8 (SD: ±3.4; range: 13.8–29.8). The primary symptoms for ordering a polysomnography were snoring or other symptoms of OSA (79%), adenotonsillar hypertrophy (14%) and screening (7%). As demonstrated in table 2, the average TST was 439 min (SD: ±46.6; range: 278–521). Obstructive AHI average was 1.13/hour (SD: ±1.2; range: 0–4.7). Central AHI average was 0.85/hour (SD: ±1.1; range: 0–7.2).

Table 1

Demographics of children with Down syndrome

Table 2

Total sleep time, AHI and ETCO2 data in children with Down syndrome.


During wakefulness, 92.5% (SD: ±0.2; range: 0–100) of the time was spent with ETCO2 ranging 30–50 mm Hg. 6.3% (SD: ±0.1; range: 0–81) of the time was spent with ETCO2>50 mm Hg, with 5.6% (SD: ±0.1; range: 0–56) of the time ranging from 50 to 55 mm Hg, and 0.7% (SD: ±0; range: 0–44) of the range >55 mm Hg. During sleep, 89.3% (SD:±0.2; range 1–100) of the time was spent with ETCO2 ranging from 30 to 50 mm Hg. 10.7% (SD: ±0.2; range: 0–99) of the time was spent with ETCO2>50 mm Hg, with 9.7% (SD: ±0.2; range: 0–82) of the time ranging from 50 to 55 mm Hg, and 1% (SD: ±0.1; range: 0–63) of the range >55 mm Hg. Table 2 illustrates further details for ETCO2 values and percentage of TST spent in NREM and REM sleep. Figure 1 demonstrates the distribution of ETCO2 values.

Figure 1

ETCO2 and %TST during awake and sleep. TST, total sleep time.

Of 86 children, 68 (79%) children had ETCO2 values >50 mm Hg at a point during sleep. Forty-four (51%) children also had ETCO2 values >50 mm Hg while awake. Thirty-seven (43%) children had mildly elevated ETCO2 values between 50 and 55 mm Hg during sleep, including 19 (22%) children with mildly elevated ETCO2 values while awake as well. The remaining 31 (36%) children had ETCO2 values ≥55 mm Hg during sleep, among whom 12 (14%) had similarly elevated ETCO2 values while awake. Ten (12%) children with elevated ETCO2 measurement met the AASM criteria for hypoventilation (ETCO2 values >50 mm Hg for >25% of TST).5 One child (1%) had ETCO2 values >55 mm Hg for >25% of TST. Of these 10 patients who met AASM criteria, figure 2 demonstrates the distribution of percentage of TST spent greater than ETCO50 mm Hg. None of the eight children <2 years of age met AASM criteria for hypoventilation.

Figure 2

Distribution of %TST > ETCO2 50 mm Hg. TST, total sleep time.

Oxygen saturation

As demonstrated in table 3, the average SpO2 while awake was 98.3% (SD:±0.8; range: 97–100). The average SpO2 during sleep was 97.8% (SD: ±1; range: 95.1–99.9), with average SpO2 during NREM being 97.7% (SD: ±1; range: 95.3–99.9) and SpO2 during REM being 97.9% (SD: ±1; range: 94.9–99.9), as illustrated in figure 3. During sleep, the average percentage of TST with SpO>96% was 87.32 (SD: ±18.46; range: 12.85–100), whereas the average percentage of TST with SpO2 >92%–96% was 12.57 (SD: ±18.34; range: 0–86.81). The average percentage of TST with SpO2 >88%–92% was 0.09 (SD: ±0.02; range: 0–3.39) and for SpO2 <88% was 0.01 (SD: ±0.04; range: 0–0.37). The average percentage of TST with SpO2 >92% was 99.89%.

Table 3

Oxygen saturation by average and percentage of TST ranges

Figure 3

SpO2 and percentage of total sleep time (TST) during sleep.


It is useful to contrast the degree of hypoventilation found in these otherwise healthy children with DS with that which has been described elsewhere in otherwise healthy typical children and adolescents. Uliel et al 12 found ETCO2 values >50 mm Hg for 0.29%±0.24% of TST in healthy typical children and adolescents between the ages of 1 and 15 years. Marcus et al 13 found that healthy typical children and adolescents from ages 1.1 to 17.4 years had similar values of ETCO2 >50 mm Hg for 0.5%+4% of TST; 18% of those children had minimal OSA with an AHI <3.1.13 In younger, otherwise healthy typical children, Montgomery-Downs et al found values of ETCO2 >50 mm Hg in 3.2–5.9 year olds for 4.0%±15.3% of TST and in 6–8.6 year olds of 2%±7.1% of TST. Collectively, the ETCO2 values for the entire group were >50 mm Hg for 2.8%±11.3% of TST.14

In contrast to the studies cited above, we found that the percentage of TST spent with ETCO2>50 and even 55 mm Hg was much higher in this group of otherwise healthy children with DS. We propose that this may be the result of underlying differences in autonomic control of breathing in children with DS, similar to other differences in their autonomic nervous systems that have been described previously, such as a reduced cardiac response to sympathetic stimulation.8

It is certainly possible that other factors were involved in the mild hypoventilation seen. While 7% of the children with DS were referred for polysomnography in accordance with the recommendations of the American Academy of Pediatrics Paediatrics,15 and not because of overt signs of OSA, most (79%) of the children did, in fact snore, and some had mild OSA. Thus, it is possible that some of the hypoventilation noted resulted from partial-continuous upper airway obstruction. A prospective case–control study in which children with DS with and without clinical signs or symptoms of OSA may help to better answer this question. Likewise, hypotonia, which is common to children with DS, may also play a role in preventing those in whom it is severe from increasing their minute ventilation sufficiently. With that said, none of the eight children <2 years of age, in whom one would expect this to be most pronounced, had hypoventilation.

Because of the apparently inherent differences in autonomic control of ventilation between otherwise healthy children with and without DS, it may be acceptable to adopt a more conservative and expectant approach towards isolated mild hypoventilation with ETCO2 <55 mm Hg in otherwise healthy children with DS and without additional types of sleep-disordered breathing or cardiopulmonary disease. One important caveat is that these children would need to be carefully monitored for evidence of pulmonary hypertension. Although the pulmonary hypertension that develops in the setting of alveolar hypoventilation is usually attributed to hypoxic pulmonary vasoconstriction, pulmonary vasoconstriction can occur with isolated acidosis and hypercarbia as well.16–18


Otherwise healthy children with DS and no other evidence of SDB besides mildly elevated ETCO2 may be managed conservatively with observation and close monitoring for the development of pulmonary hypertension.


View Abstract


  • Contributors WW and DR carried out the initial analyses, reviewed and revised the manuscript, and approved the final manuscript as submitted.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

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.