Ventilatory drive in children with obstructive apnea
Carole Marcus, University of Pennsylvania
6 October 2004
Dear Editor,
Recently, Fregosi et al published an article describing the ventilatory drive during wakefulness in children in relation to their apnea hypopnea index (1). In this article, they compared their results to those of a study we published a number of years ago (2). We have, therefore, reanalyzed the raw data from our study in order to address their questions.
Fregosi et al studied the ventilatory drive in a sample of children obtained from the general community. They found a correlation between the P0.1 during hypoxia and the apnea hypopnea index (AHI), although they found no relationship between steady state ventilatory responses to hypoxia/hypercapnia and the AHI. In the discussion, they compare their results to our study, in which rebreathing ventilatory responses to hypoxia and hypercapnia were performed in children with the obstructive sleep apnea syndrome (who had presented clinically with symptoms of obstructive sleep apnea) compared to normal controls (asymptomatic children with normal baseline polysomnograms). We found no difference in the ventilatory responses between the obstructive sleep apnea syndrome (OSAS) group and controls. However, in the original paper, we did not look at the correlation between the degree of sleep-disordered breathing and the ventilatory response.
We therefore reevaluated our data. At the time when this study was conducted, hypopneas were typically not measured in children (3); thus we looked at the obstructive apnea index rather than the apnea hypopnea index. We found no significant correlation between the apnea index and the hypoxic ventilatory response (r = -0.02), or the apnea index and the hypercapnic ventilatory response (r = 0.00). Neither was there a correlation between gas exchange abnormalities during sleep and the ventilatory responses.
There are several potential reasons for the difference in results between the two studies. The studies evaluated very different populations. Fregosi et al recruited asymptomatic children from the community. As a result, the degree of sleep-disordered breathing was extremely mild, and most of the subjects would not have been considered to have clinical OSAS. Of the 50 subjects considered in their study to have OSAS, only 1 of the 50 children had any obstructive apneas at all. In contrast, the mean obstructive apnea index in our study group was 16/hr; our controls, by definition, had an apnea index < 1/hr. The remaining 49 subjects in the Fregosi study only had hypopneas, using a liberal definition of hypopnea (30% decrease in airflow from baseline, with no desaturation or arousal needed). The American Academy of Sleep Medicine considers hypopneas in adults to be significant if there is a decrease in airflow > 50%, or a smaller decrease if associated with desaturation or arousal (4); recent data suggest that the same definition can be applied to children (5). Therefore, by most definitions, the “OSAS” group studied by Fregosi may have actually been normal. However, even when we looked only at our normal control group, we found no significant correlation between the hypoxic (r = -0.14) or hypercapnic (r = -0.03) ventilatory response. The correlation between P0.1 and ventilatory responses in the current study is interesting, and further study in a group of more severely affected children would be of value.
Sincerely,
Carole L. Marcus, M.B.B.Ch.
David Gozal, M.D.
Raanan Arens, M.D.
Thomas G. Keens, M.D.
Sally L. Davidson Ward, M.D.
Reference List
(1) Fregosi RF, Quan SF, Jackson AC, Kaemingk KL, Morgan WJ, Goodwin JL et al. Ventilatory drive and the apnea-hypopnea index in six-to-twelve year old children. BMC Pulm Med 2004; 4(1):4.
(2) Marcus CL, Gozal D, Arens R, Basinski DJ, Omlin KJ, Keens TG et al. Ventilatory responses during wakefulness in children with obstructive sleep apnea. Am J Respir Crit Care Med 1994; 149(3 Pt 1):715-721.
(3) American Thoracic Society. Standards and indications for cardiopulmonary sleep studies in children. Am J Respir Crit Care Med 1996; 153:866-878.
(4) Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep 1999; 22(5):667-689.
(5) Tang JP, Rosen CL, Larkin EK, Difiore JM, Arnold JL, Surovec SA et al. Identification of sleep-disordered breathing in children: variation with event definition. Sleep 2002; 25(1):72-79.
Competing interests
None
Response to comment submitted by Dr. Marcus et al.
Ralph Fregosi, The University of Arizona
17 November 2004
Marcus and colleagues recently commented on our paper, published in BMC Pulmonary Medicine [1], where we report a significant correlation (R = 0.31) between the obstructive apnea-hypopnea index (OAHI) and hypoxic sensitivity in 50 children. In contrast, in an earlier study Marcus et al did not detect impaired hypoxic sensitivity in children with obstructive sleep apnea, prompting the comment [2]. Marcus et al used a cross-sectional approach, wherein a group of children with obstructive sleep apnea were compared to an age-matched healthy control group. But differences in experimental design (cross-sectional vs. correlation) cannot explain the different findings because we also found a significant difference in the average hypoxic ventilatory response slope when we divided the subjects into high and low OAHI groups (see Table 1 in [1]) (i.e., with a technique similar to that used by Marcus et al [2]). Nevertheless, Marcus et al. reanalyzed their data [2] but found no correlation between the apnea index and the ventilatory response to hypoxia or hypercapnia, nor between nocturnal gas exchange abnormalities and the apnea index. In an effort to explain the different findings they suggest that our subjects may have been normal because of our recruitment approach and “liberal” definition of hypopnea, while their subjects were judged to have moderate-to-severe obstructive sleep apnea based on obstructive events alone. In other words, our OAHI includes obstructive apneas plus hypopneas, while the index that they used includes only true obstructive events.
Although there is no doubt that their subjects had more severe disease than ours, it is difficult to envision why our apparently “normal” subjects would reveal a correlation between hypoxic sensitivity and the apnea-hypopnea index, but children with more severe sleep apnea would not show such a correlation. But several other factors may at least partially explain the different findings. First, we used the “P0.1 technique” wherein the mouth pressure (P) recorded 0.1 seconds after brief airway occlusions is taken as the index of respiratory drive; the measurements were made in normoxia and at two levels of steady state, isocapnic hypoxia. In contrast, Marcus et al. [2] measured the pulmonary ventilation rate in response to a hypoxic rebreathing protocol, wherein oxygen saturation falls progressively. The P0.1 technique is considered a more sensitive estimate of respiratory drive than pulmonary ventilation, as the former is not influenced by airway resistance whereas the latter index is [3]. In addition, steady state estimates of hypoxic sensitivity are less variable than rebreathing methods [4]. The second issue relates to sample size. We studied 50 subjects with a range of OAHI values, whereas Marcus et al studied 20 patients with relatively severe disease and 19 completely normal control subjects. Subjecting such a data set to correlation analysis is problematic due to “clustering” of the values at opposite extremes of the scale in the two different populations. Moreover, if the correlation analysis was done on only the 20 children with obstructive sleep apnea, sample size becomes an issue as well. For example, in our sample of 50 subjects a correlation coefficient of 0.29 was required for statistical significance, assuming a P value of 0.05 and a power of 0.7; we attained a value of 0.31. If we had used the same statistical boundaries with a sample size of 20, a correlation coefficient of 0.49 would be needed to achieve statistical significance.
Finally, we wish to comment on the suggestion that our subject population may have been “normal”. Although we used a “liberal” definition of hypopnea in our study, we would disagree that our population did not have sleep disordered breathing (SDB). First, as we discussed in our paper [1], the definition of hypopnea that we used is associated with symptoms of SDB [5] and impaired learning [6] in this population. Second, we studied children with a broad range of SDB. The OAHI ranged from 1.2 to 52 events per hour. Furthermore, those children in the high OAHI group had a mean index of 12 in comparison to 3.1 in the low OAHI group (Table 1 [1]). Although it is plausible that there might be some ascertainment error amongst those children close to the boundary of our OAHI cutoff (i.e., 5 events per hour), it is improbable that those with high values were “normal”. Furthermore, as mentioned in the discussion section of the paper, we have found significant correlations between our definition of SDB and more restrictive ones, implying that any assessment of relative disease severity would not change with a different definition. Third, we would still maintain that the definition used to identify children with SDB remains unclear. The paper cited by Marcus et al [7] merely demonstrates a substantial variability in event prevalence as a function of event definition in children. It did not propose or provide data suggesting a definition of disease. In closing, we are grateful that these investigators took the time to re-analyze their data on ventilatory drive in children with obstructive sleep apnea [2], and we certainly agree that studying chemoreceptor reflexes in children with sleep-disordered breathing is important, and worthy of further study.
References
1. RF Fregosi, SF Quan, AC Jackson, KL Kaemingk, WJ Morgan, JL Goodwin, JC Reeder, RK Cabrera, E Antonio: Ventilatory drive and the apnea-hypopnea index in six-to-twelve year old children. BMC Pulm Med 2004, 4:4.
2. CL Marcus, D Gozal, R Arens, DJ Basinski, KJ Omlin, TG Keens, SL Ward: Ventilatory responses during wakefulness in children with obstructive sleep apnea. Am J Respir Crit Care Med 1994, 149:715-21.
3. WA Whitelaw, JP Derenne, J Milic-Emili: Occlusion pressure as a measure of respiratory center output in conscious man. Respir Physiol 1975, 23:181-99.
4. S Zhang, PA Robbins: Methodological and physiological variability within the ventilatory response to hypoxia in humans. J Appl Physiol 2000, 88:1924-32.
5. JL Goodwin, KL Kaemingk, RF Fregosi, GM Rosen, WJ Morgan, DL Sherrill, SF Quan: Clinical outcomes associated with sleep-disordered breathing in Caucasian and Hispanic children--the Tucson Children's Assessment of Sleep Apnea study (TuCASA). Sleep 2003, 26:587-91.
6. KL Kaemingk, AE Pasvogel, JL Goodwin, SA Mulvaney, F Martinez, PL Enright, GM Rosen, WJ Morgan, RF Fregosi, SF Quan: Learning in children and sleep disordered breathing: findings of the Tucson Children's Assessment of Sleep Apnea (tuCASA) prospective cohort study. J Int Neuropsychol Soc 2003, 9:1016-26.
7. JP Tang, CL Rosen, EK Larkin, JM DiFiore, JL Arnold, SA Surovec, JM Youngblut, S Redline: Identification of sleep-disordered breathing in children: variation with event definition. Sleep 2002, 25:72-9.
Ventilatory drive in children with obstructive apnea
6 October 2004
Dear Editor,
Recently, Fregosi et al published an article describing the ventilatory drive during wakefulness in children in relation to their apnea hypopnea index (1). In this article, they compared their results to those of a study we published a number of years ago (2). We have, therefore, reanalyzed the raw data from our study in order to address their questions.
Fregosi et al studied the ventilatory drive in a sample of children obtained from the general community. They found a correlation between the P0.1 during hypoxia and the apnea hypopnea index (AHI), although they found no relationship between steady state ventilatory responses to hypoxia/hypercapnia and the AHI. In the discussion, they compare their results to our study, in which rebreathing ventilatory responses to hypoxia and hypercapnia were performed in children with the obstructive sleep apnea syndrome (who had presented clinically with symptoms of obstructive sleep apnea) compared to normal controls (asymptomatic children with normal baseline polysomnograms). We found no difference in the ventilatory responses between the obstructive sleep apnea syndrome (OSAS) group and controls. However, in the original paper, we did not look at the correlation between the degree of sleep-disordered breathing and the ventilatory response.
We therefore reevaluated our data. At the time when this study was conducted, hypopneas were typically not measured in children (3); thus we looked at the obstructive apnea index rather than the apnea hypopnea index. We found no significant correlation between the apnea index and the hypoxic ventilatory response (r = -0.02), or the apnea index and the hypercapnic ventilatory response (r = 0.00). Neither was there a correlation between gas exchange abnormalities during sleep and the ventilatory responses.
There are several potential reasons for the difference in results between the two studies. The studies evaluated very different populations. Fregosi et al recruited asymptomatic children from the community. As a result, the degree of sleep-disordered breathing was extremely mild, and most of the subjects would not have been considered to have clinical OSAS. Of the 50 subjects considered in their study to have OSAS, only 1 of the 50 children had any obstructive apneas at all. In contrast, the mean obstructive apnea index in our study group was 16/hr; our controls, by definition, had an apnea index < 1/hr. The remaining 49 subjects in the Fregosi study only had hypopneas, using a liberal definition of hypopnea (30% decrease in airflow from baseline, with no desaturation or arousal needed). The American Academy of Sleep Medicine considers hypopneas in adults to be significant if there is a decrease in airflow > 50%, or a smaller decrease if associated with desaturation or arousal (4); recent data suggest that the same definition can be applied to children (5). Therefore, by most definitions, the “OSAS” group studied by Fregosi may have actually been normal. However, even when we looked only at our normal control group, we found no significant correlation between the hypoxic (r = -0.14) or hypercapnic (r = -0.03) ventilatory response. The correlation between P0.1 and ventilatory responses in the current study is interesting, and further study in a group of more severely affected children would be of value.
Sincerely,
Carole L. Marcus, M.B.B.Ch.
David Gozal, M.D.
Raanan Arens, M.D.
Thomas G. Keens, M.D.
Sally L. Davidson Ward, M.D.
Reference List
(1) Fregosi RF, Quan SF, Jackson AC, Kaemingk KL, Morgan WJ, Goodwin JL et al. Ventilatory drive and the apnea-hypopnea index in six-to-twelve year old children. BMC Pulm Med 2004; 4(1):4.
(2) Marcus CL, Gozal D, Arens R, Basinski DJ, Omlin KJ, Keens TG et al. Ventilatory responses during wakefulness in children with obstructive sleep apnea. Am J Respir Crit Care Med 1994; 149(3 Pt 1):715-721.
(3) American Thoracic Society. Standards and indications for cardiopulmonary sleep studies in children. Am J Respir Crit Care Med 1996; 153:866-878.
(4) Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep 1999; 22(5):667-689.
(5) Tang JP, Rosen CL, Larkin EK, Difiore JM, Arnold JL, Surovec SA et al. Identification of sleep-disordered breathing in children: variation with event definition. Sleep 2002; 25(1):72-79.
Competing interests
None
Response to comment submitted by Dr. Marcus et al.
17 November 2004
Marcus and colleagues recently commented on our paper, published in BMC Pulmonary Medicine [1], where we report a significant correlation (R = 0.31) between the obstructive apnea-hypopnea index (OAHI) and hypoxic sensitivity in 50 children. In contrast, in an earlier study Marcus et al did not detect impaired hypoxic sensitivity in children with obstructive sleep apnea, prompting the comment [2]. Marcus et al used a cross-sectional approach, wherein a group of children with obstructive sleep apnea were compared to an age-matched healthy control group. But differences in experimental design (cross-sectional vs. correlation) cannot explain the different findings because we also found a significant difference in the average hypoxic ventilatory response slope when we divided the subjects into high and low OAHI groups (see Table 1 in [1]) (i.e., with a technique similar to that used by Marcus et al [2]). Nevertheless, Marcus et al. reanalyzed their data [2] but found no correlation between the apnea index and the ventilatory response to hypoxia or hypercapnia, nor between nocturnal gas exchange abnormalities and the apnea index. In an effort to explain the different findings they suggest that our subjects may have been normal because of our recruitment approach and “liberal” definition of hypopnea, while their subjects were judged to have moderate-to-severe obstructive sleep apnea based on obstructive events alone. In other words, our OAHI includes obstructive apneas plus hypopneas, while the index that they used includes only true obstructive events.
Although there is no doubt that their subjects had more severe disease than ours, it is difficult to envision why our apparently “normal” subjects would reveal a correlation between hypoxic sensitivity and the apnea-hypopnea index, but children with more severe sleep apnea would not show such a correlation. But several other factors may at least partially explain the different findings. First, we used the “P0.1 technique” wherein the mouth pressure (P) recorded 0.1 seconds after brief airway occlusions is taken as the index of respiratory drive; the measurements were made in normoxia and at two levels of steady state, isocapnic hypoxia. In contrast, Marcus et al. [2] measured the pulmonary ventilation rate in response to a hypoxic rebreathing protocol, wherein oxygen saturation falls progressively. The P0.1 technique is considered a more sensitive estimate of respiratory drive than pulmonary ventilation, as the former is not influenced by airway resistance whereas the latter index is [3]. In addition, steady state estimates of hypoxic sensitivity are less variable than rebreathing methods [4]. The second issue relates to sample size. We studied 50 subjects with a range of OAHI values, whereas Marcus et al studied 20 patients with relatively severe disease and 19 completely normal control subjects. Subjecting such a data set to correlation analysis is problematic due to “clustering” of the values at opposite extremes of the scale in the two different populations. Moreover, if the correlation analysis was done on only the 20 children with obstructive sleep apnea, sample size becomes an issue as well. For example, in our sample of 50 subjects a correlation coefficient of 0.29 was required for statistical significance, assuming a P value of 0.05 and a power of 0.7; we attained a value of 0.31. If we had used the same statistical boundaries with a sample size of 20, a correlation coefficient of 0.49 would be needed to achieve statistical significance.
Finally, we wish to comment on the suggestion that our subject population may have been “normal”. Although we used a “liberal” definition of hypopnea in our study, we would disagree that our population did not have sleep disordered breathing (SDB). First, as we discussed in our paper [1], the definition of hypopnea that we used is associated with symptoms of SDB [5] and impaired learning [6] in this population. Second, we studied children with a broad range of SDB. The OAHI ranged from 1.2 to 52 events per hour. Furthermore, those children in the high OAHI group had a mean index of 12 in comparison to 3.1 in the low OAHI group (Table 1 [1]). Although it is plausible that there might be some ascertainment error amongst those children close to the boundary of our OAHI cutoff (i.e., 5 events per hour), it is improbable that those with high values were “normal”. Furthermore, as mentioned in the discussion section of the paper, we have found significant correlations between our definition of SDB and more restrictive ones, implying that any assessment of relative disease severity would not change with a different definition. Third, we would still maintain that the definition used to identify children with SDB remains unclear. The paper cited by Marcus et al [7] merely demonstrates a substantial variability in event prevalence as a function of event definition in children. It did not propose or provide data suggesting a definition of disease. In closing, we are grateful that these investigators took the time to re-analyze their data on ventilatory drive in children with obstructive sleep apnea [2], and we certainly agree that studying chemoreceptor reflexes in children with sleep-disordered breathing is important, and worthy of further study.
References
1. RF Fregosi, SF Quan, AC Jackson, KL Kaemingk, WJ Morgan, JL Goodwin, JC Reeder, RK Cabrera, E Antonio: Ventilatory drive and the apnea-hypopnea index in six-to-twelve year old children. BMC Pulm Med 2004, 4:4.
2. CL Marcus, D Gozal, R Arens, DJ Basinski, KJ Omlin, TG Keens, SL Ward: Ventilatory responses during wakefulness in children with obstructive sleep apnea. Am J Respir Crit Care Med 1994, 149:715-21.
3. WA Whitelaw, JP Derenne, J Milic-Emili: Occlusion pressure as a measure of respiratory center output in conscious man. Respir Physiol 1975, 23:181-99.
4. S Zhang, PA Robbins: Methodological and physiological variability within the ventilatory response to hypoxia in humans. J Appl Physiol 2000, 88:1924-32.
5. JL Goodwin, KL Kaemingk, RF Fregosi, GM Rosen, WJ Morgan, DL Sherrill, SF Quan: Clinical outcomes associated with sleep-disordered breathing in Caucasian and Hispanic children--the Tucson Children's Assessment of Sleep Apnea study (TuCASA). Sleep 2003, 26:587-91.
6. KL Kaemingk, AE Pasvogel, JL Goodwin, SA Mulvaney, F Martinez, PL Enright, GM Rosen, WJ Morgan, RF Fregosi, SF Quan: Learning in children and sleep disordered breathing: findings of the Tucson Children's Assessment of Sleep Apnea (tuCASA) prospective cohort study. J Int Neuropsychol Soc 2003, 9:1016-26.
7. JP Tang, CL Rosen, EK Larkin, JM DiFiore, JL Arnold, SA Surovec, JM Youngblut, S Redline: Identification of sleep-disordered breathing in children: variation with event definition. Sleep 2002, 25:72-9.
Competing interests
None declared