Ventilatory response to exercise in adolescents with cystic fibrosis and mild-to-moderate airway obstruction

Data regarding the ventilatory response to exercise in adolescents with mild-to-moderate cystic fibrosis (CF) are equivocal. This study aimed to describe the ventilatory response during a progressive cardiopulmonary exercise test (CPET) up to maximal exertion, as well as to assess the adequacy of the ventilatory response for carbon dioxide (CO2) exhalation. Twenty-two adolescents with CF (12 boys and 10 girls; mean ± SD age: 14.3 ± 1.3 years; FEV1: 78.6 ± 17.3% of predicted) performed a maximal CPET. For each patient, data of a sex- and age matched healthy control was included (12 boys and 10 girls; mean ± SD age: 14.3 ± 1.4 years). At different relative exercise intensities of 25%, 50%, 75%, and 100% of peak oxygen uptake (VO2peak), breathing pattern, estimated ventilatory dead space ventilation (VD/VT ratio), minute ventilation (VE) to CO2 production relationship (VE/VCO2-slope), partial end-tidal CO2 tension (PETCO2), and the VE to the work rate (VE/WR) ratio were examined. VO2peak was significantly reduced in CF patients (P = 0.01). We found no differences in breathing pattern between both groups, except for a significantly higher VE at rest and a trend towards a lower VE at peak exercise in patients with CF. Significantly higher values were found for the estimated VD/VT ratio throughout the CPET in CF patients (P < 0.01). VE/VCO2-slope and PETCO2 values differed not between the two groups throughout the CPET. VE/WR ratio values were significantly higher in CF during the entire range of the CPET (P < 0.01). This study found an exaggerated ventilatory response (high VE/WR ratio values), which was adequate for CO2 exhalation (normal VE/VCO2-slope and PETCO2 values) during progressive exercise up to maximal exhaustion in CF patients with mild-to-moderate airway obstruction.


Background
Peak oxygen uptake (VO 2peak ) is reported to be limited in patients with cystic fibrosis (CF) (Bongers et al. 2012;Hjeltnes et al. 1984;Keochkerian et al. 2008;Shah et al. 1998;Wideman et al. 2009). This reduction seems to have a multifactorial cause (Selvadurai et al. 2002). Respiratory, cardiovascular, as well as peripheral muscle function are reported as potential exercise limiting mechanisms (Almajed and Lands 2012). In patients with mild to moderate pulmonary disease, non-pulmonary factors, such as low muscle mass, impaired skeletal muscle function and centrally mediated oxygen delivery, seem to predominate in limiting exercise capacity (Regnis et al. 1996;Moorcroft et al. 2005;Saynor et al. 2014). In more severe patients with CF (forced expiratory volume in one second (FEV 1 ) <60% of predicted), ventilatory constraints and impaired gas exchange become more important determinants.
Due to continuous airflow obstruction, as reflected by a decreased FEV 1 and dynamic hyperinflation, adolescents with CF have been described to develop a rapid shallow breathing pattern at rest (Hart et al. 2002) and during exercise (Keochkerian et al. 2005). This can be accompanied with a decreased ventilatory capacity and concomitant reduced VO 2peak (Keochkerian et al. 2008). Children and adolescents with CF with static hyperinflation at rest (residual volume to total lung capacity ratio (RV/TLC) >30%) seem to be more prone to a ventilatory limitation during exercise, which appears to be associated with decreased exercise performance (Sovtic et al. 2013;Werkman et al. 2010).
A recent study found that exercise limitation in adult patients with CF is multifactorial and that it was dominantly correlated with FEV 1 and nutritional and inflammatory status, but also with the magnitude of the overall ventilatory response during exercise (Pastré et al. 2014). In patients with severe airway obstruction (FEV 1 < 50% of predicted), multivariate analysis revealed the FEV 1 to be a significant independent predictor of exercise capacity, whereas the ratio between minute ventilation and carbon dioxide exhalation (VE/VCO 2 ratio) at peak exercise was the major determinant of exercise limitation in patients with mild-to-moderate disease (FEV 1 > 50% of predicted) (Pastré et al. 2014). However, knowledge about the ventilatory response to exercise in adolescent patients with mild to moderate CF is ambiguous and inconclusive. Several studies in mild-to-moderate adolescents with CF describe an exaggerated ventilatory response with a rapid shallow breathing pattern at rest (Hart et al. 2002) and during exercise (Keochkerian et al. 2008). On the contrary, a recent study in adolescents with mild CF did not find any evidence for a different ventilatory response and/or rapid shallow breathing pattern during exercise (Borel et al. 2014). As the two studies demonstrating exaggerated ventilatory responses were performed in adolescents with CF with lower FEV 1 values, this suggests that the ventilatory response to exercise is at least partially affected by the degree of airway obstruction.
Moreover, questions can be raised whether the adopted rapid shallow breathing pattern is beneficial as higher breathing frequencies seem to increase ventilatory dead space ventilation (VD/VT ratio), as has been reported in patients with CF waiting for lung transplantation (Thin et al. 2004). However, adding additional dead space volume during exercise in patients with mild CF lung disease had no influence on VO 2peak and the duration of the exercise test, and it even increased ventilation which was attributed to an increased tidal volume with no change in respiratory rate (Dodd et al. 2006). The increased dead space volume was accompanied by higher VE/VCO 2 ratios during exercise (Dodd et al. 2006). This finding in patients with mild CF suggests that the ventilatory response during exercise in adolescents with CF might differ with healthy controls and that this might alter during the course of the disease. Unfortunately the recent study of Borel et al. (2014), which focussed on ventilation during the entire range of exercise, only mentioned mechanical constraints influencing ventilation and did not discuss metabolic issues related to ventilation. Moreover, they included small and unequal groups of only prepubertal children.
As lung function decreases over time in most patients with CF, exercise capacity eventually becomes limited by the lungs reaching their mechanical limits to expand (approximately at FEV 1 values ≤60%pred) (Almajed and Lands 2012). Insight in the breathing pattern during progressive exercise in adolescents with CF in a broad spectrum of lung function deterioration, as well as the adequacy of this ventilatory response for carbon dioxide (CO 2 ) exhalation, is clinically relevant for future therapeutic interventions. Therefore, the aim of the current study was 1) to describe the ventilatory response during a progressive cardiopulmonary exercise test (CPET) up to maximal exertion in adolescents with mild-to-moderate CF and 2) to assess the adequacy of the ventilatory response for CO 2 exhalation (determined from partial end-tidal carbon dioxide tension (P ET CO 2 )) throughout the CPET in mild-to-moderate patients with CF. Our hypothesis is that adolescents with CF develop an obstructive breathing pattern, combined with a relatively large VD/VT ratio, which limit CO 2 wash out.

Participants
Exercise data of twenty-two adolescents (12 boys and 10 girls from 12 to 17 years of age, mean ± SD age: 14.3 ± 1.3 years) with mild-to-moderate CF were randomly selected from the exercise database from the CF Center at the University Medical Center Utrecht. The database contained anonymous patient data of anthropometry, lung function and exercise capacity which was measured as part of usual care at the routine annual check-up. Therefore, all patients were free from acute pulmonary or gastrointestinal exacerbation at the time of testing. Testing procedures used in this study met the assumptions for standard of practice for the routine care of patients with CF. For each patient with CF, an age, sex and anthropometrically matched healthy control who performed a maximal CPET in our hospital retrospectively was selected (untrained and normal physical activity level). All participants and their guardians provided approval for inclusion of the data in research studies. After evaluation, the medical ethical committee of the University Medical Center Utrecht determined that inclusion of the data conformed to the regulations of the Dutch CF Registration and that inclusion of the data in this study met the ethical polices of the University Medical Center Utrecht, as well as the regulations of the Dutch government.

Anthropometric measures
Body mass and body height were determined using an electronic scale (Seca, Hamburg, Germany) and a stadiometer (Ulmer Stadiometer, Ulm, Germany) respectively. Body mass index (BMI) was calculated as the body mass in kilograms divided by the square of the body height in meters. Standard deviation (SD) scores were calculated for body height for age, body mass for age, and BMI for age using Dutch normative values (Fredriks et al. 2000). The equation of Haycock et al. (1978), validated in infants, children, and adults, was used to obtain the participants' body surface area (BSA).

Spirometry and plethysmography
In the patients with CF, spirometry and plethysmography were performed by qualified lung function analysts of the CF Center at the University Medical Center Utrecht. Since the healthy controls were not known with any disease, spirometry and plethysmography was not performed in this group. In order to prevent the potential influence of bronchial reactivity during exercise, spirometry and plethysmography were performed after bronchodilation with salbutamol (800 μg). FEV 1 was obtained from flow volume curves (Masterscreen, Jaeger, Würzburg, Germany). Residual volume (RV) and total lung volume (TLC) were determined in a body plethysmograph (Master Lab system, Jaeger, Würzburg, Germany). The RV was expressed as a percentage of TLC (RV/TLC ratio) as well. In order to improve comparative possibilities with the reports of other studies in CF, we used the commonly used reference values of Zapletal et al. (1987) to express lung function values as percentage of predicted values.

Cardiopulmonary exercise test
All participants completed a maximal CPET using an electronically braked cycle ergometer (Jaeger Physis, Carefusion, Houten, The Netherlands) after bronchodilation with salbutamol. After three minutes of rest measurements, cycling started at a workload of 0 W. Then, the work rate (WR) was linearly incremented with a 15 W•min −1 ramp protocol (Godfrey 1974) until the patient stopped due to volitional exhaustion, despite strong verbal encouragement. The test effort was considered maximal when the participant showed objective (heart rate (HR) at peak exercise (HR peak ) >180 beats · min −1 and/or a respiratory exchange ratio (RER) at peak exercise (RER peak ) >1.0) (Armstrong and Welsman 2008) and subjective (unsteady biking, sweating, facial flushing, and clear unwillingness to continue despite encouragement) signs of maximal effort.
Throughout the CPET, participants breathed through a facemask (dead space volume 63 or 72 mL, dependent on size) (Hans Rudolph Inc, Kansas City, MO) that was connected to a calibrated metabolic cart (Oxycon Pro, Carefusion, Houten, the Netherlands). Gas analyzers were calibrated using gases of known concentration, whereas the flow meter was calibrated using a three-liter syringe (Hans Rudolph Inc, Kansas City, MO). Expired gas passed through a flow meter and a gas analyzer connected to a computer, which calculated breath-by-breath minute ventilation (VE), oxygen uptake (VO 2 ), VCO 2 and RER from conventional equations. Output from the flow meter and gas analyzers were averaged over ten-second intervals and stored for further use. Relative VO 2peak (VO 2peak /kg) was calculated by dividing VO 2peak by body mass. HR was monitored continuously by a threelead electrocardiogram (Hewlett-Packard, Amstelveen, Netherlands) and transcutaneous O 2 saturation at the index finger was measured by pulse oximetry (Nellcor 200 E, Nellcor, Breda, the Netherlands). Peak exercise parameters were defined as the highest values achieved within the last 30 seconds prior to exhaustion.

Data analysis
The ventilatory threshold was defined as the point at which the ventilatory equivalent for oxygen and the partial end-tidal oxygen tension reached a minimum and thereafter began to rise in a consistent manner, coinciding with an unchanged ventilatory equivalent for carbon dioxide and a peak in the P ET CO 2 course (American Thoracic Society, American College of Chest Physicians 2003; Ohuchi et al. 1996). When this ventilatory equivalents method appeared to provide uncertain results for a participant's ventilatory threshold (n = 4, 18%, in the CF patients, n = 0 in the healthy controls), the point at which the linear slope of the relation between the VCO 2 and VO 2 changed was taken as the ventilatory threshold, according to the Vslope method (Beaver et al. 1986). The ventilatory threshold (VO 2 ) was expressed as an absolute value, relative value (VO 2 normalized for body mass) and as a percentage of the attained VO 2peak . The estimated VD/VT ratio was calculated by using the P ET CO 2 . The graphical presentation of VE as a function of VCO 2 during the progressive CPET was used to determine the point at which VE increased out of proportion to VCO 2 , the respiratory compensation point.
Next to resting values, tidal volume, breathing frequency, VE, VD/VT ratio, and P ET CO 2 were determined as an average of 30 seconds at different exercise intensities of 25%, 50%, 75% and 100% of VO 2peak . Both VE and tidal volume were adjusted for body mass as well. To allow for fair comparison, resting VO 2 was subtracted from VO 2peak . Subsequently, 25%, 50%, 75% and 100% of this delta VO 2 was calculated and summed with the resting VO 2 for each participant (CF patients and healthy controls). Furthermore, we calculated the VE to VO 2 relationship (VE/VO 2 -slope) and the VE to VCO 2 relationship (VE/VCO 2 -slope) at the same different exercise intensities by linear regression of the exercise data up to 25%, 50%, 75%, and 100% of VO 2peak using the least squares Values are presented as mean ± SD. Abbreviations: BSA body surface area, CF cystic fibrosis, HR peak peak heart rate, NS not statistically significant, P ET CO 2 partial end-tidal carbon dioxide tension, RER peak peak respiratory exchange ratio, VCO 2peak peak carbon dioxide output, VCO 2peak /kg peak carbon dioxide output normalized for body mass, VD/VT ratio physiological dead space ventilation, VE minute ventilation, VE/kg minute ventilation normalized for body mass, VE/VCO 2 -slope slope of the relationship between minute ventilation and carbon dioxide output, VE/VO 2 -slope slope of the relationship between minute ventilation and oxygen uptake, VE/WR minute ventilation to work rate ratio, VO 2peak peak oxygen uptake, VO 2peak /kg peak oxygen uptake normalized for body mass, WR peak peak work rate, WR peak /kg peak work rate normalized for body mass. *P < 0.05; **P < 0.01; ***P < 0.001.
approach. We examined the participants' response of the VE to the WR (VE/WR ratio) at similar intensities. Data from the first minute of exercise were excluded since the breathing pattern during the first minute of exercise frequently appears to be unstable.

Statistical analysis
The Statistical Package for the Social Sciences (SPSS, version 15.0; SPSS Inc., Chicago, IL) was used for the dataanalysis. Shapiro-Wilk tests for normality were performed in order to evaluate the data distribution of each variable. Differences between adolescents with CF and their healthy counterparts in anthropometry and in exercise data at different exercise intensities were examined with Mann-Whitney U tests. Data are presented as mean values ± SD. A P-value <0.05 was considered as statistically significant.

Results
Anthropometric data for the patients with CF and the healthy controls are presented in Table 1, with no significant anthropometric between-group differences. Lung function characteristics of the adolescents with CF are also shown in Table 1. Patients had mild-to-moderate airway obstruction (mean FEV 1 expressed as a percentage of predicted of 79 ± 17%) and a mild-to-moderate degree of static hyperinflation (mean RV/TLC ratio of 34 ± 9%). More specifically, 14 patients (5 boys and 9 girls) had an absolute RV/TLC ratio greater than 30%, which suggests static hyperinflation (Eid et al. 2000). All participants performed a maximal effort during the CPET (mean test duration of 623 ± 139 and 659 ± 194 seconds for the CF patients and healthy controls respectively), and the results are presented in Table 2. Compared with the healthy controls, adolescents with CF attained significantly lower values for HR peak , peak WR (WR peak ), WR peak /kg, VO 2peak , VO 2peak /kg, VO 2peak /m 2 (normalized for BSA), peak VCO 2 (VCO 2peak ), VCO 2peak /kg, and peak VE (VE peak ) normalized for body mass (VE peak /kg), whereas they attained significantly higher values for the estimated VD/VT ratio at peak exercise, the VE/VO 2slope up to the ventilatory threshold, P ET CO 2 at rest, and the VE/WR ratio at peak exercise.
Breathing pattern components, tidal volume and breathing frequency, adopted during exercise did not differ significantly between patients with CF and healthy adolescents, except for a significantly higher breathing frequency at rest (19 ± 3 versus 22 ± 5 breaths · min −1 ; P = 0.02) and a trend for lower absolute tidal volume values at or near maximal exercise in patients with CF (1.74 ± 0.34 versus 1.62 ± 0.37 L; P = 0.32). Consequently, VE at rest was significantly higher (11.5 ± 2.1 versus 13.8 ± 3.4 L · min −1 ; P = 0.02), whereas VE peak (75.8 ± 18.8 versus 89.5 ± 27.2 L . min −1 ; P = 0.08) tended to be lower in patients with CF. VE normalized for body mass (VE/kg) was significantly higher at rest (0.2 ± 0.05 versus 0.3 ± 0.06 L · kg −1 · min −1 ; P < 0.01) and at 25% of VO 2peak (0.4 ± 0.07 versus 0.5 ± 0.08 L · kg −1 · min −1 ; P = 0.03) in the patients with CF. Values for breathing frequency divided by the tidal volume (rapid shallow breathing index) tended to be higher in patients with CF at rest (32 ± 13 versus 38 ± 13; P = 0.14) and at 25% of VO 2peak (27 ± 10 versus 32 ± 15; P = 0.31). Furthermore, estimated VD/VT ratio values were significantly higher during all Figure 1 Changes in the estimated VD/VT ratio during exercise at similar percentages of VO 2peak in the healthy adolescents and the adolescents with CF. Dashed lines correspond to the ventilatory threshold in both groups. Abbreviations: CF = cystic fibrosis; VD/VT ratio = physiological dead space ventilation; VO 2peak = peak oxygen uptake. **P < 0.01; ***P < 0.001. exercise intensities in patients with CF (see Figure 1, P < 0.01, P < 0.01, P < 0.01, and P < 0.001 at 25%, 50%, 75%, and 100% of VO 2peak respectively). Figure 2 shows significantly higher VE/VO 2 -slope values at 25% (P = 0.01) and 50% (P < 0.01) of VO 2peak in patients with CF, whereas VE/VCO 2 -slope values differed not significantly between the two groups throughout the entire range of the CPET. Furthermore, Figure 2 demonstrates that, except for resting values (P = 0.03), P ET CO 2 values differed not significantly between the two groups during the CPET. Figure 3 represents the RER at rest and throughout progressive exercise in order to elucidate the higher VE/ VO 2 ratios during sub-maximal exercise. At rest and during sub-maximal exercise, patients with CF attained significantly higher RER values (P < 0.001, P < 0.001, P < 0.001, P < 0.01, and P = 0.26 at rest, 25%, 50%, 75%, and 100% of VO 2peak respectively). The VE/WR ratio depicted in Figure 4 was significantly higher in CF during the entire range of the CPET (P < 0.01, P < 0.001, P < 0.001, and P < 0.001 at 25%, 50%, 75%, and 100% of VO 2peak respectively).

Discussion
The present study aimed to 1) describe the ventilatory response during a progressive CPET and 2) assess the adequacy of the ventilatory response for CO 2 exhalation during exercise in mild-moderate adolescents with CF. First, we found an exaggerated ventilatory response during exercise. Second, this ventilatory response to exercise seems to be adequate for CO 2 exhalation in patients with mild-to-moderate CF. The latter is illustrated by a similar course of the VE/VCO 2 -slope throughout exercise, higher RER values during sub-maximal exercise, and the ability to maintain P ET CO 2 values within normal limits throughout the entire range of the CPET.
The basic physiological factors that determine and modify the ventilatory response to exercise are: 1) CO 2 output of the exercising muscles, 2) the arterial CO 2 setpoint, 3) the VD/VT ratio, and 4) the change in the arterial pressure of CO 2 (PaCO 2 ) during exercise (Wasserman et al. 1996). Several of these concepts are addressed to in this study. First, muscular CO 2 output during (sub-)maximal exercise seems to be increased in patients with CF (Bongers et al. 2012;Hebestreit et al. 2005;Nguyen et al. 2014), which is illustrated in the current study by a higher RER. The higher RER in patients with CF is suggested to reflect a higher reliance on glucose oxidation to meet energy demands during exercise (Hebestreit et al. 2005;Nguyen et al. 2014). Altered substrate utilization in CF (de Meer et al. 1995;Moser et al. 2000;Selvadurai et al. 2003) might explain the increased RER at rest and for (See figure on previous page.) Figure 2 Changes in VE/VO 2 -slope, VE/VCO 2 -slope, and P ET CO 2 during exercise at similar percentages of VO 2peak in the healthy adolescents and the adolescents with CF. Dashed lines represent the ventilatory threshold in both groups. Abbreviations: CF = cystic fibrosis; P ET CO 2 = partial end-tidal carbon dioxide tension; VE/VCO 2 -slope = slope of the relationship between minute ventilation and carbon dioxide output; VE/VO 2 -slope = slope of the relationship between minute ventilation and oxygen uptake; VO 2peak = peak oxygen uptake. *P < 0.05; **P < 0.01. Figure 3 RER during exercise at similar percentages of VO 2peak in healthy adolescents and patients with CF. Vertical dashed lines correspond to the ventilatory threshold in both groups. Abbreviations: CF = cystic fibrosis; RER = respiratory exchange ratio; VO 2peak = peak oxygen uptake. **P < 0.01; ***P < 0.001. the lower exercise intensities. In addition, slowed VO 2 kinetics during exercise and recovery in patients with CF might also be a possible explanation for an increased RER during sub-maximal exercise (Kusenbach et al. 1999;Massin et al. 2000;Pouliou et al. 2001;Hebestreit et al. 2005;Stevens et al. 2009;Saynor et al. 2014). This results in a greater dependency on glycolytic energy systems at lower exercise intensities resulting in an increased VCO 2 . Higher aerobic or anaerobic glycolytic energy expenditure during exercise increases ventilation as supported by a study in patients with familial mitochondrial myopathy (Heinicke et al. 2011). The current study showed an exaggerated VE relative to metabolic rate, indicated by high VE/VO 2 -slope values and VE/ WR ratios, as well as an elevated RER, with no apparent signs of pulmonary insufficiency.
Second, the VD/VT was significantly higher at all exercise intensities in patients with CF, which seems to be exaggerated with disease progression (post-hoc analysis, data not shown). High VD/VT values during exercise have previously been reported in patients with CF (Cerny et al. 1982;Coates et al. 1988;Thin et al. 2004;Wilkens et al. 2010). This, in combination with the trend of higher P ET CO 2 in patients with CF, indicates an abnormal alveolar dead space ventilation (Wilkens et al. 2010), which might explain the relatively high VE values at rest and during sub-maximal exercise in patients with CF. Unfortunately, the exaggerated ventilatory response we found and the shift from oxidative to glycolytic energy metabolism, as shown by the attenuation of the RER slope at exercise intensities above 50% of VO 2peak when compared to healthy controls, seems not to be able to totally compensate the limited exercise capacity in patients with CF (significantly reduced VO 2peak /kg and WR peak /kg). Our results are in agreement with the study of Borel et al. (2014) who found no effect of mild CF on breathing pattern and breathing strategy during an incremental CPET. Although breathing pattern and breathing strategy of adolescents with CF were comparable with healthy controls, most studies still report a reduced exercise capacity in adolescents with CF (Almajed and Lands 2012; Rand and Prasad 2012;Saynor et al. 2014). Saynor et al. (2014) recently suggested that centrally mediated oxygen delivery might be the principally limiting the aerobic function of pediatric CF patients with mild-tomoderate airway obstruction during ramp incremental cycling exercise. However, based on the results of the study of Keochkerian et al. (2008) on the ventilatory response in children with CF, one would still expect that CF could alter the ventilatory response during a progressive CPET. The current study demonstrated no significant effect of CF on the ventilatory response for CO 2 exhalation (VE/VCO 2 -slope and P ET CO 2 ) to exercise in patients with CF with mild-to-moderate airway obstruction.
As described above, we found higher values for RER, VE/WR, and VE/VO 2 -slope values during sub-maximal exercise. These results suggest a higher ventilatory demand rather than a higher ventilatory response during sub-maximal exercise in patients with mild CF. An explanation for the absence of an impact of CF on the ventilatory response to exercise is the mild-to-moderate severity of the lung disease in our population. Keochkerian et al. (2008) reported that the more severe the airway obstruction, the more rapid and shallow the breathing pattern.
The current study has some limitations. Firstly, by categorizing exercise intensity by %VO 2peak , there is no Figure 4 Changes in the VE/WR ratio during exercise at similar percentages of VO 2peak in healthy adolescents and patients with CF. Dashed lines correspond to the ventilatory threshold in both groups. Abbreviations: CF = cystic fibrosis; VE/WR = minute ventilation to work rate ratio; VO 2peak = peak oxygen uptake. **P < 0.01; ***P < 0.001. standardization of exercise intensity relative to the ventilatory response for each participant. However, standardizing for exercise intensity relative to VO 2 (%VO 2peak ) was also done by other studies (Borel et al. 2014;Keochkerian et al. 2008). Since the ventilatory threshold occurred at a similar percentage of VO 2peak in the pediatric CF patients and the healthy controls (58 ± 9% versus 55 ± 9%; P = 0.32), we believe that we did compare groups within the same physiological exercise intensity domain. Secondly, the sample size was relatively small and included mainly patients with CF with mild to moderate airflow obstruction. For this reason, these findings cannot be generalized to patients with severe airflow obstruction. Nevertheless, the current study sample is representative for the CF population in a tertiary CF Center. Thirdly, the estimated VD/VT ratio cannot be accurately predicted from the P ET CO 2 in patients with an increased VD/VT ratio due to lung disease (Wasserman et al. 2005), so caution must be taken with the interpretation of these results. Fourth, unfortunately we were not able to correct VO 2peak for fat free mass as this variable was not routineously measured in patients with CF in the CF Center at the University Medical Center Utrecht. Finally, the used criteria for a maximal effort are subject to debate, especially in patients with CF. It has previously been demonstrated that traditional testing protocols and verification criteria significantly underestimate VO 2peak in both healthy (Barker et al. 2011) and children with CF (Saynor et al. 2013). We did not verify the attainment of a true VO 2peak with a supramaximal exercise testing procedure.

Implications and future research
As a higher ventilatory demand seems to be present during submaximal exercise in mild-moderate patients with CF, a small decline in ventilatory capacity might hamper the precarious balance between ventilation and homeostasis with further disease progression in patients with CF. The main findings presented in this study highlight the importance for the clinician to aim for attenuation of lung function decline even in patients with CF with a relatively preserved lung function ("normal" FEV 1 ). For future research it would be interesting to evaluate the latter hypothesis in patients with CF with more severe airway obstruction. Moreover, the differences we found during the course of sub-maximal exercise highlight the importance to evaluate the submaximal exercise response as well when interpreting a CPET, and not just focus on peak exercise parameters.