Article Text

Download PDFPDF
Increased interbreath variability of gas exchange during exercise in children with cardiomyopathy
  1. Tony Reybrouck1,
  2. Steven Vangesselen2,
  3. Luc Mertens3,
  4. Marc Gewillig3
  1. 1Department of Cardiac Rehabilitation, University Hospital Gasthuisberg, Leuven, Belgium
  2. 2Department of Rehabilitation Sciences, University of Leuven (K U Leuven), Leuven, Belgium
  3. 3Department of Paediatric Cardiology, University Hospital Gasthuisberg, Leuven, Belgium
  1. Correspondence to:
    Professor T Reybrouck
    Department of Cardiovascular Rehabilitation, University Hospital Gasthuisberg, Herestraat, 3000 Leuven, Belgium; tony.reybrouck{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.

During exercise testing in patients with chronic heart failure and severely depressed left ventricular function, an increased variability of gas exchange with a phenomenon of periodic breathing with an oscillatory pattern in oxygen uptake (VO2) has been recognised.1 Previous studies were performed in adult patients with chronic heart failure related to ischaemic heart disease. The aim of the present study was to analyse whether an increased interbreath variability for gas exchange with an oscillatory pattern could also be identified during exercise in children with cardiomyopathy. Moreover, we also analysed whether this new parameter correlates with cardiac function.


In all, 49 patients were used for this analysis. Twenty patients were studied with dilated cardiomyopathy (DCMP; mean (SD) age 11.7 (3.7) years). At a regular outpatient visit the patients were clinically investigated, and underwent an echocardiographic and electrocardiographic examination. Myocardial function was assessed by determination of the fractional shortening on the echocardiogram. The patients were compared with 29 age-matched normal controls (Nl) (10.7 (2.8) years).

Maximal exercise test was performed on a treadmill. Speed was set at 5.6 km/h, and the inclination was increased by 2% every minute until exhaustion. Gas exchange was measured breath-by-breath. During exercise, the interbreath variability for VO2 was calculated as the mean value of all absolute differences between the VO2 of all single breaths during 1 min, and the mean of all these breaths and was expressed as a percentage of the mean value for VO2 during that minute.2

Exercise capacity was assessed by determination of the maximal VO2 or symptom-limited VO2 and was expressed as a percentage of the Nl mean value.2


The left ventricular end diastolic diameter averaged 55.3 (9.3) mm in DCMP versus 41 (4.1) mm in Nl (p<0.001). The fractional shortening averaged 22.5 (7.4%) in DCMP versus 37.3 (4.2%) in Nl (p<0.001).

For submaximal exercise at 2% inclination, the variability for VO2 amounted to 9.6 (5.4) for DCMP versus 7.1 (2.3%) for Nl (p = 0.006) and at 4% inclination 8.6 (3.9) for DCMP versus 6.2 (2) for Nl (p = 0.006).

In patients with DCMP an increased variability of VO2 was found for four levels of exercise in half of the group, which exceeded the 95% confidence limit (CL) of Nl (variability >8%). In these patients an oscillatory pattern in VO2 could also be observed. For the total group (patients with DCMP and Nl), at 2% inclination on the treadmill, significant negative correlations were found between the magnitude of the variability for the VO2 and the VO2 max (r = −0.47, p = 0.05) and the fractional shortening (r = −0.48; p = 0.05) as an estimate of myocardial function.

Maximal oxygen uptake in DCMP averaged 79.8 (17.3%) of Nl (p<0.005, patients v Nl).


This study shows a high-interbreath variability of VO2 during graded exercise testing in patients with DCMP with a significantly increased variability of VO2 in half of the patients. This increased variability with an oscillatory pattern for VO2 correlates with a poor left ventricular function. In the present study, the prevalence of increased interbreath variability with an oscillatory pattern is higher than values reported in literature (10–20%).3,4 This may be explained by differences in methods such as duration of oscillations3,5 or comparing different time windows.4 In the present study, we defined an increased breath-by-breath variability as exceeding 8% (upper limit of 95% CL). In subjects with normal left ventricular function no oscillatory pattern of VO2 was observed. A similar observation has been made in adult patients with DCMP and systolic dysfunction.1 The underlying mechanism for an oscillatory pattern of VO2 remains to be determined. Previous studies have shown that increased interbreath variability for VO2 during exercise correlates with poor left ventricular function.5 Similarly, in the present study an increased interbreath variability of VO2 was observed in the subjects with the lowest values for fractional shortening. Oscillatory changes of VO2 during exercise correlate with an oscillatory pattern of pulmonary blood flow. In patients with reduced left ventricular function, Yajima3 showed during simultaneous breath-by-breath gas exchange measurements and beat-to-beat radionuclide recordings of ejection fraction oscillatory fluctuation for both parameters.

Increased interbreath variability with oscillatory changes of VO2 may significantly influence the determination of the maximal value for VO2. With modern gas analysis equipment data are often reported with a time window of 10–15 s. Therefore, if a patient has an increased interbreath variability of VO2 with a periodic oscillatory breathing, the peak of this value may exceed the true VO2max. The overestimation will depend upon the size of the oscillation and can amount up to 5 ml O2/min/kg or up to 35% of the average VO2.4 This can be equal to the effect of some interventions—for example, physical training, pharmacological treatment; such difference can therefore be important for managing patients during a work up for heart transplantation. To obviate these problems, a time interval of 60 s or minimally 30 s should be recommended.



  • Competing interests: None declared.