Article Text

Download PDFPDF

The Young Everest Study: effects of hypoxia at high altitude on cardiorespiratory function and general well-being in healthy children
  1. E Scrase1,
  2. A Laverty1,
  3. J C D Gavlak1,
  4. S Sonnappa2,
  5. D Z H Levett3,
  6. D Martin3,
  7. M P W Grocott3,
  8. J Stocks2
  1. 1
    Department of Paediatric Respiratory Medicine, Great Ormond Street Hospital for Children NHS Trust, London, UK
  2. 2
    Portex Anaesthesia, Intensive Therapy and Respiratory Medicine Unit, UCL, Institute of Child Health, London, UK
  3. 3
    UCL Centre for Altitude Space and Extreme Environment Medicine, Institute of Human Health and Performance, UCL, London;
  1. Dr E Scrase, Department of Respiratory Medicine, Level 6 Cardiac Wing, Great Ormond Street Hospital for Children, Great Ormond Street, London WC1N 3JH, UK; E.Scrase{at}


Objectives: To assess the effect of altitude and acclimatisation on cardiorespiratory function and well-being in healthy children.

Methods: A daily symptom diary, serial measurements of spirometry, end-tidal carbon dioxide (etCO2) and daytime and overnight pulse oximetry (SpO2), were undertaken at sea level and altitudes up to 3500 m in healthy children during a trekking holiday. SpO2 at altitude was compared with that in flight and during acute hypoxic challenge (breathing 15% oxygen) at sea level.

Results: Measurements were obtained in nine children aged 6–13 years (median 8). SpO2 decreased significantly during the hypoxic challenge (difference −5%, 95% CI −6 to −3%, p<0.01) but remained above 90% in all children. There was a significant fall in daytime and overnight SpO2 (95% CI −11.9 to −7.5% and −12 to −8, respectively) and etCO2 (−8.5 to −4.5 mm Hg) as the children ascended to 3500 m. There was a significant increase in SpO2 (95% CI 1.1 to 4.9%) and a further drop in etCO2 (−5.9 to −0.8 mm Hg) after a week at altitude, etCO2 being negatively correlated with SpO2. There was no correlation between SpO2 during hypoxic challenge, in flight or at altitude. Lung function remained within 7% of baseline in all but two children, in whom reductions of up to 23% in FVC and 16% FEV1 were observed at altitude. The children generally remained well, but the Lake Louise scoring system was unreliable in this age group.

Conclusions: A wide range of physiological responses to altitude are evident in healthy children. This study should inform future larger studies in children to improve understanding of responses to hypoxia in health and disease.

View Full Text

Statistics from

Arterial hypoxaemia is a common feature of cardiopulmonary disease and sleep disorders, which are among the commonest causes of childhood morbidity.1 Children exposed to reductions in inspired oxygen levels as a result of air travel25 or altitude6 may be vulnerable to the pathological consequences of hypoxaemia.7 With increasing altitude, barometric pressure falls, ambient air at 3500 m having an oxygen partial pressure equivalent to 13% oxygen at sea level. Modern aircraft are pressurised to ∼8000 ft/2438 m, the equivalent to breathing 15.1% oxygen, and a hypoxic challenge is recommended in children with lung disease to assess fitness to fly with or without supplemental oxygen.4 8 There is a lack of data on the range of normal desaturation in children during air travel, and studies on healthy individuals have been recommended.5 9 10

Increasing numbers of healthy, lowland children are holidaying at high altitude (above 2500 m)7 11 and as such may experience some of the hypoxaemic-related symptoms common to children with lung disease. Although rapid exposure to low oxygen leads to hypoxaemia and, potentially, impaired physical function, including acute mountain sickness (AMS), high altitude pulmonary oedema and/or cerebral oedema6 1114 slow progressive exposure to hypoxia allows the development of adaptive mechanisms, which serve to normalise oxygen content and protect the tissues from these effects.13 14 Nevertheless, the increased ventilation that occurs on exposure to altitude is accompanied by marked hypocapnoea and respiratory alkalosis, which may in turn lead to disturbed sleep and periodic breathing.14

The commonly accepted definition of AMS uses the Lake Louise classification15 to rank the presence and severity of symptoms (see appendix).16 Although a preverbal version has been developed for very young children,7 17 18 there is minimal information regarding the validity of Lake Louise definitions in school-age children.19 This lack of information about the effects of altitude in young children led Pollard et al6 to suggest that until further evidence is obtained, children under 10 years of age should not sleep above 3000 m.

Caudwell Xtreme Everest (CXE) is one of the largest field studies to investigate mechanisms underlying adaptation to hypoxia in healthy adults.20 21 The infrastructure for that study provided the opportunity to perform similar measurements in healthy children who were undertaking a family trek in the region at the same time.


The aims of this study were to observe healthy lowland children during a trek at altitude in order to: (1) ascertain whether pulse oximetry during acute hypoxic challenge at sea level reflects that at similar oxygen levels in flight and at an equivalent altitude of 2600 m; (2) describe the cardiorespiratory and symptom responses to altitude; (3) explore the relationship between changes in cardiorespiratory variables and reported symptoms at altitude.



Children 6 years and over, without previous exposure to high altitude, who were already booked to trek with their parents in Nepal were recruited. The study was approved by UCL Research Ethics Committee and informed written assent/consent obtained from children and their parents. No financial compensation was provided for participation. Medical screening was undertaken to ascertain suitability for enrolment. Before departure, arrangements were discussed with the trek organisers to maximise the children’s safety.


The itinerary and testing schedule are summarised in table 1. The trek was planned to minimise AMS by slow graded ascent.7 Measurements detailed below were undertaken in the Respiratory Laboratory (London, UK) and during a weekend trek at sea level and repeated in Nepal at 1300 m, on arrival at 3500 m and after 9 days acclimatisation at 3500 m. Measurements marked by an asterisk were repeated daily during the trek:

Table 1 Effect of altitude on vital signs, lung function and symptom scores in children
  • The Lake Louise symptom score* (LLSS)11 13 14 16 (appendix): completed by the children, with parental help when necessary.

  • Vital signs: peripheral oxygen saturation* (SpO2) (Minolta Pulsox300, Stowood Scientific, Oxford, UK), heart rate* (HR), end-tidal carbon dioxide (etCO2) via nasal catheter (Tidal Wave Sp, Novametrix-Respironics, USA), respiratory rate and blood pressure.

  • Spirometric lung function, according to international guidelines22 adapted for children,23 using an ultrasonic flowmeter (EasyONE; NDD, Zurich, Switzerland), independent of gas density24 and validated for use at altitude (C Buess, personal communication).

  • Hypoxic challenge (fitness-to-fly, London only);3 4 SpO2 and HR monitored by pulse oximetry while breathing 15% oxygen in a modified body plethysmograph for 20 minutes.

  • In-flight monitoring of SpO2 approximately 4 h into flight, cabin pressurised to approximately 2430 m.

  • SpO2 and HR at 2600 m were recorded for direct comparison at an equivalent inspired oxygen with those in flight and during fitness-to-fly assessments.

Statistical analysis

The Wilcoxon non-parametric t-test was used to assess within-subject changes. Spearman Correlation was used to assess relationships between variables (SPSS 15.0 for Windows).


Serial measurements at sea level, in flight and at altitude were obtained in nine healthy children (five boys) aged 6–13 years (median 8).

Hypoxic challenge

There was a significant drop in SpO2 (difference −5.0%, 95% CI −6 to −3%, p<0.01) with marked intersubject variability in response to the laboratory hypoxic challenge (fig 1). No child desaturated below 90%, the cut-off for fitness to fly without supplemental oxygen.8 Mean SpO2 and HR during the test (93.4% (SD 1.7) and 86 beats per minute (bpm) (SD 10), respectively) were similar to those recorded at equivalent levels of inspired oxygen in flight and at 2600 m (fig 1, table 1), but there was marked individual variability on these three occasions.

Figure 1

Baseline peripheral oxygen saturation (SpO2) and heart rate compared with that recorded during an acute hypoxic fitness-to-fly (FTF) challenge at sea level, and at equivalent inspired levels of oxygen (PiO2) in flight and at altitude. Bars are mean with SD, dark bars SpO2% and pale bars heart rate beats per minute.

General well-being and LLSS (at altitude)

Despite occasional fatigue and minor injuries even the youngest children coped extremely well with the trekking regime, confirming the value of gradual ascent at altitude. When compared with either the objective physiological measures (see below) or observations by parents and investigators, the children appeared both to under and overreport symptoms on the LLSS; average values at sea level being similar to those at altitude (table 1).

Effect of altitude on cardiorespiratory function

On ascent to 3500 m there was a small increase in resting daytime respiratory rate (table 1). Resting HR increased gradually with increasing altitude, being on average 24 bpm faster at 3500 m than at sea level (95% CI 9 to 26 bpm, p<0.01). There was, on average, a 9% fall in daytime SpO2 (95%CI −11.9 to −7.5%, p<0.01), an 11% fall in overnight SpO2 (95% CI −12 to −8%, p<0.01) and a 7 mm Hg reduction in etCO2 (95% CI −8.5 to −4.5 mmHg, p<0.01), when compared with baseline, with marked between-subject variability (fig 2). At 3500 m, overnight SpO2 was significantly correlated with daytime values (R2  =  0.51, p = 0.03), overnight values being, on average, 3% (95% CI 1.4 to 5.2%, p = 0.011) lower than those recorded during the day.

Figure 2

Alterations in (A) end-tidal carbon dioxide (etCO2) and (B) peripheral oxygen saturation (SpO2) on ascent to altitude in healthy children and following acclimatisation. Mean is indicated in bold. Child C and child H represent the two extremes of response. These data show that during the transition from sea level to 3500 m (day 5), a 7 mm Hg reduction in etCO2 in child C was accompanied by a 12% fall in SpO2, whereas in child H, a 9 mm Hg reduction in etCO2 was only associated with a 9% fall in SpO2. After acclimatisation (day 14) at altitude, there was no further change in child H, but increased hyperventilation in child C resulted in a 12 mm Hg reduction in etCO2 compared with that recorded at sea level and an improvement in SpO2 to only 6% lower than that at sea level. See text for further discussion.

Repeat measures of SpO2 and etCO2 at 3500 m after a week of trekking and sleeping at up to 3860 m demonstrated an average 3% recovery in SpO2 (95% CI 1.1 to 4.9%, p<0.01) and a 3 mm Hg further drop in etCO2 (95% CI −5.9 to −0.8 mm Hg, p = 0.01) when compared with values on arrival at this altitude. Individual changes in two children who represented the extremes of response are highlighted in fig 2. Child H, who had the highest SpO2 and among the lowest etCO2 at sea-level, maintained this pattern during ascent to 3500 m but showed no further changes 9 days later, suggesting rapid acclimatisation. By contrast, child C had relatively low SpO2 and high etCO2 on all occasions. The reciprocal relationship between SpO2 and etCO2 can be seen in fig 3. There was a marked left shift in the carbon dioxide/oxygen relationship at altitude, with further adaptation following acclimatisation. One child showed a drop in diastolic pressure at 1300 m and a further decrease at 3500 m, but there were no significant group changes in blood pressure.

Figure 3

Relationship between change in end-tidal carbon dioxide (etCO2) and peripheral oxygen saturation (SpO2) on ascent to altitude in healthy children. × sea-level, ▾ 3500 m, ▵ 3500 m + 9 days. These data illustrate the reciprocal relationship between etCO2 and SpO2, which was significant on all measurement occasions: (R2  =  0.31, p = 0.12 at sea level; R2  =  0.67, p = 0.007 on arrival at 3500 m; R2  =  0.53, p = 0.026 after 9 days at 3500 m).

Lung function

On ascent to 3500 m there was a fall in forced expiratory volume in 1 s (FEV1) (difference −5%, 95% CI −12 to 0.7, p = 0.04) and in forced vital capacity (FVC) (−3%, 95% CI −19.6 to 1.0, p = 0.07). FEV1 and FVC remained within 7% of baseline in all but two children, who showed decreased FEV1 (12% and 16%) and FVC (20% and 23%) at altitude (fig 4).

Figure 4

Effect of altitude on lung function. Percentage change in forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC) (3500 m− baseline) is shown for each child, together with the fall in daytime and overnight peripheral oxygen saturation (SpO2) on arrival at 3500 m compared with sea level. The two children with the largest reductions in lung function also had some of the most marked drops in both day and night time SpO2 at altitude.


These results suggest that healthy children demonstrate a cardiorespiratory response to high altitude similar to that described in adults, with a fall in SpO2 and etCO2, increase in HR and a small reduction in FVC.13 The children responded to continued exposure to the reduced inspired oxygen at altitude with further increases in ventilation, indicated by further reductions in etCO2, and some recovery of oxygenation levels. There was, however, marked individual variability in hypoxic response, which warrants further investigation.

There were minimal changes in lung function in all but two children. Although these changes were not accompanied by respiratory symptoms, these two children displayed the most marked falls in SpO2 and higher etCO2 levels on arrival at 3500 m, suggesting a slower hypoxic ventilatory response (HVR). To our knowledge, the effect of altitude on spirometric lung function has not been reported previously in children, but similar results have been found in adults. Senn et al25 reported an average fall in FVC and FEV1 of 6% and 5%, respectively, with marked individual variation. These authors concluded that changes in pulmonary function after rapid ascent to high altitude (4559 m) were consistent with subclinical interstitial fluid accumulation, although respiratory muscle weakness and fatigue could not be excluded.

The greatest potential weakness of this study was the limited number of children studied. Given the lack of previous evidence regarding safety and the most relevant tests to undertake, it was considered unethical to recruit young children purely for scientific research. As these children had already booked a trekking holiday, there were no ethical concerns regarding either financial remuneration or exposure to unnecessary risks.6 26 Furthermore, by utilising the existing CXE infrastructure we had a unique opportunity to perform a wide range of tests on a limited number of children.

Fitness-to-fly assessments

There are limited published data with which to compare these results, although the mean SpO2 recorded in flight is similar to that previously reported in healthy adults and children.27 28 A recent review of the risks of air travel in children confirmed the poor relationship between desaturation levels during fitness-to-fly tests and those during flights or at altitude,4 and recent studies have suggested that the cut-off for fitness to fly in young children should be adjusted to SpO2 of 85% or greater during hypoxic challenge.29 30 Although the acute hypoxia test is frequently used to assess HVR in climbers before going to altitude,13 it is possible that partial pressure carbon dioxide remains normal over such a short exposure, and that variability in response may reflect not only differences in HVR but also those in the oxygen–haemoglobin dissociation curve. Interpretation of fitness to fly would have been facilitated by simultaneous measurements of transcutaneous carbon dioxide, which we would recommend for future studies. It has also been suggested that adaptation of this test to ascertain the extent to which desaturation is related to ventilation–perfusion mismatch rather than shunt, may be more predictive.31

Hypoxic ventilatory response

The HVR is mediated by the carotid body, with wide intersubject variability in response.12 13 Adults respond to altitude-induced hypoxia by increasing minute volume to improve oxygenation, a similar response recently being noted in children.32 Increases in haemoglobin concentration can compensate for the fall in SpO2 such that the oxygen content of the blood can be maintained up to 7000 m.21 Our findings that neither the fitness-to-fly test nor a low SpO2 at altitude predicted symptoms of altitude illness in these children is consistent with data from adults, which suggest that, despite a weak negative correlation between SpO2 and symptoms, this is rarely helpful clinically.33 Similar nocturnal reductions in SpO2 and etCO2 have been reported in prepubertal children and adults.32 The major value of SpO2 appears to be monitoring sudden changes within individuals, especially if accompanied by an increase in symptoms. The marked intersubject variability in response to hypoxaemia at altitude, which has also been reported in adults,21 may reflect both genetically and environmentally determined differences.34 In addition to known variations in the HVR,21 differences in the shape of the oxyhaemoglobin dissociation curve and the balance of conflicting effects of shifts to the left with alkalosis versus the right shift with increased 2, 3-diphosphoglycerate might contribute to observed variability. Assessment of urine pH in future studies may help to clarify such contributions.


The incidence of AMS is unclear in younger subjects, because children may report similar symptoms when travelling at sea level due to travel or disruption to daily routine.35 The LLSS has been adapted for preverbal children,17 18 but such adaptations have yet to be applied in older children. The combination of apparent under and overreporting, with no relationship to either physiological assessments or subjective observations, rendered the LLSS relatively meaningless in this study. These findings are in keeping with a report published after the completion of our study.36 The need for further adaptation and validation of the LLSS in children has been highlighted as an urgent priority, as it can be challenging to differentiate behavioural changes in children from potentially more serious underlying medical problems.6 26 It is therefore essential that parents are acquainted with the symptoms of altitude illness and its management before altitude travel and are aware of their child’s reactions during travel, irrespective of altitude.7

What is already known on this topic

  • There is a lack of data on the range of normal desaturation in children during air travel and studies on healthy subjects have been recommended.

  • There are limited published data about the effects of altitude in young children.

What this study adds

  • This study suggests that an acute hypoxic challenge, as administered during the fitness-to-fly test, does not predict individual responses either during flight or at altitudes with an equivalent inspired oxygen.

  • Cardiorespiratory responses during gradual ascent to 3500 m and subsequent acclimatisation to altitude appear similar in children and adults, although there is much individual variability.

  • This study helps to clarify the extent to which results from adults can be extrapolated to younger age ranges.

Clinical relevance

One of the aims of this study was to relate changes in cardiorespiratory responses to altitude-induced symptoms, but the lack of overt symptoms in these children, even in those with quite marked changes in SpO2 and lung function, precluded such conclusions. This lack of symptoms probably reflects the slow ascent deliberately adopted for the trek, as more marked symptoms of altitude sickness have been reported during rapid ascents.32 37 Although full interpretation of these data will not be possible until all results from the CXE project are available,20 38 the data reported here contribute to the sparse physiological evidence available from children at altitude. These results also describe the normal acute and adaptive responses to hypoxaemia in children without the confounding elements of disease process that occur in subjects with cardiorespiratory disease or those undergoing intensive care.38 As tissue hypoxia is a universal phenomenon among children who are critically ill and is frequently due to arterial hypoxaemia, useful insights may be gained by examining the biophysiological responses of healthy children exposed to low levels of environmental oxygen.


Results from this study suggest that, with sufficient preparation and vigilance, healthy children as young as 6 years of age can be taken to altitudes of 3500 m without major adverse effects, and that such children are willing and able to undertake a wide range of physiological assessments. Whereas the hypoxic challenge did not predict the degree of desaturation in flight or at altitude among this small group of healthy children, it correctly classified all as “fit to fly”. Cardiorespiratory responses to both acute exposure to altitude and subsequent acclimatisation appear similar in children and adults. A more reliable method of monitoring symptoms of AMS is required for children. Evidence from this study will help the design of future larger studies of children at altitude, as well as clarifying the extent to which results from adults can be extrapolated to younger age ranges.


The authors would like to thank: the children and families who participated in this study, all of whom generously contributed their time for the study without any financial compensation; the members and sponsors of the Caudwell Xtreme Everest team (; Kalsang Sherpa and Susie Sherpa Baer (The Walking and Climbing Company) and all the Sherpas who accompanied them on the trek and who ensured they had a safe and enjoyable trek, while carefully transporting medical equipment between locations; Smiths Medical for financial support; NDD for lending the Easyone Spirometers without charge; Dr Donald Urquhart for undertaking physical examination and medical screening of the children before departure and Professor Monty Mythen for professional support and scientific advice.

Appendix: Lake Louise scoring of AMS

AMS is classified as a total score of 3 or more following recent ascent to altitude, but only if headache and at least one other symptom is present (ie, poor sleep, fatigue and reduced appetite scores 3 but is not AMS; likewise, a severe headache alone scores 3 and is also not AMS).


View Abstract


  • Competing interests: None.

  • Funding: The Young Everest Study received funding from Smiths Medical. Research at the Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust benefits from R&D funding received from the NHS Executive.

  • Ethics approval: The study was approved by UCL Research Ethics Committee.

  • Patient consent: Obtained.

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.