- Email: [email protected]
Body mass index as an indicator of obstructive sleep apnea in pediatric Down syndrome
International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Body mass index as an indicator of obstructive sleep apnea in pediatric Down syndrome§ Courtney B. Shires a,*, Sandra L. Anold b,2, Robert A. Schoumacher b,2, George W. Dehoff c,3, Sreekrishna K. Donepudi a,1, Rose Mary Stocks a,1 a
Department of Otolaryngology-Head & Neck Surgery, University of Tennessee Health Science Center, 910 Madison Avenue, Room 428, Memphis, TN 38163, USA Department of Pediatrics, University of Tennessee Health Science Center, 50 N Dunlap Street, Memphis, TN 38103, USA c Department of Internal Medicine, University of Tennessee Health Science Center, 956 Court Avenue, Room H314, Memphis, TN, USA b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 19 September 2009 Received in revised form 20 March 2010 Accepted 23 March 2010 Available online 7 May 2010
Objective: Our objective was to determine if higher body mass index (BMI) increases the likelihood of, obstructive sleep apnea (OSA) in pediatric Down syndrome (DS) patients. Methods: We performed a, retrospective chart review of 63 DS patients evaluated by overnight polysomnography from December 1995 to February 2005. Patients aged less than 2 years were excluded. Remaining patients were grouped, according to presence (n = 19) or absence (n = 33) of OSA based on apnea hypopnea index (AHI). OSA, and non-OSA DS groups were age matched while blinded to patient attributes other than age and OSA, status. Patients without appropriate age matches were excluded. We recorded various patient information, including age, sex, height, weight, number of apneas, number of hypopneas, respiratory distress index (RDI), apnea–hypopnea index (AHI), lowest oxygen saturation during sleep, mean oxygen saturation, number of arousals per hour, and mean time spent in REM sleep. We calculated BMI using the, standard kg/m2 formula and converted this into a Z-score. Results: Fifty-two DS patients were analyzed with average age of 9.3 4.5 years (10.2 4.2 in 33 OSA patients, 7.8 4.3 in 19 non-OSA patients). There were 28 males and 24 females. The OSA group mean BMI Zscore was 2.09 0.94, and the non-OSA group Z-score was 1.4 1.40. The Z-scores for BMI were statistically significant between OSA and non-OSA patients with p = 0.03 by t-test. Conclusions: When age and sex adjusted, BMI has a statistically significant association with the presence of OSA in Down syndrome patients. The incidence of OSA also increases with increasing age in this population. ß 2010 Elsevier Ireland Ltd. All rights reserved.
Keywords: BMI Obstructive sleep apnea Down syndrome
1. Objective Our objective was to determine the contribution of body mass index (BMI) to the incidence of obstructive sleep apnea (OSA) in pediatric Down syndrome (DS) patients. 2. Introduction Children with DS are known to have a wide variety of medical problems associated with the disorder. The DS population has a
§ Presented at the American Society of Pediatric Otolaryngologists 2007 Spring Meeting at the Combined Otolaryngology Spring Meetings in San Diego, CA, USA on April 27, 2007. * Corresponding author. Tel.: +1 901 448 5886; fax: +1 901 448 5120. E-mail address: [email protected] (C.B. Shires). 1 Tel.: +1 901 448 5886; fax: +1 901 448 5120. 2 Tel.: +1 9014485500; fax: +1 901 448 7836. 3 Tel.: +1 901 448 5814; fax: +1 901 448 7836.
0165-5876/$ – see front matter ß 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2010.03.050
higher incidence of OSA (30–50%) than the general population (0.7–2.0%) [7]. Children and adolescents with DS have a tendency to be overweight beginning in infancy and continuing throughout the growing years [12]. High BMI is a known OSA risk factor among children and adults without DS, but the role it plays in DS children has been a subject of debate. OSA, if untreated, can cause exacerbation of problems seen in DS as well as cause problems directly related to OSA. Untreated OSA can have serious pulmonary sequelae such as hypercarbia, acidosis, and pulmonary hypertension with possible cor pulmonale. The possibility of these health problems in a population at higher health risk is a serious concern. Untreated OSA can also lead to sleep fragmentation, sleep deprivation, daytime somnolence, decreased sleep time, and decreased REM time, all of which can aggravate learning and behavioral problems already encountered in DS patients. Due to the possible serious health complications of OSA in these children, we believe it is important to learn as much as possible about the etiology of this disease.
C.B. Shires et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
3. Methods Our study was designed as a retrospective chart review. We obtained institutional review board approval from both the University of Tennessee Health Science Center and Methodist LeBonheur Healthcare. First, we identified 63 consecutive children with DS that had been evaluated by overnight polysomnography (PSG). The identification of DS was made based both on clinical history and photographs. The PSG were performed between the years of 1995 and 2005 in the LeBonheur Pediatric & Adolescent Sleep Disorders Center at LeBonheur Children’s Hospital in Memphis, TN. Data were recorded from the date of the PSG, and alternate data were used from clinic visits or surgery if necessary. A standardized night-time polysomnography at the sleep laboratory included 16-channel strip recorder on a computerized system. All studies involved a minimum period of 5 h and a maximum of 10 h of sleep in a time and temperature controlled room, accompanied by one of the parents without previous sedation or sleep deprivation. The children were continuously observed by trained technicians through an infra-red video camera. The following parameters were measured: electroencephalogram (C3/A2, C4/A1, O1/A2), electromyogram (submental and tibial), electro-oculogram (right and left), oxygen saturation (SaO2), oro-nasal airflow (thermistor), abdominal and chest wall movement, and electrocardiography. We recorded various patient information including age, sex, height, weight, number of apneas, number of hypopneas, respiratory distress index (RDI), apnea–hypopnea index (AHI), lowest oxygen saturation during sleep, mean oxygen saturation, number of arousals per hour, and mean time spent in REM sleep. Sleep architecture was scored using standard criteria for our sleep laboratory. Certified polysomnography technicians scored the sleep studies, which were later interpreted by board-certified sleep medicine specialists. Respiratory events were significant if they lasted two or more respiratory cycles and were associated with a >3% fall in SaO2 and/or were terminated by an arousal. Obstructive apneas were defined as a reduction in airflow of >80% from baseline amplitude with continuing or increasing effort as reflected by the thoracic or abdominal plethysmographic bands. Hypopneas were defined as a decrease in airflow to between 20% and 50% of the baseline amplitude and >2% reduction in SaO2. The AHI was defined as the number of apneas or hypopneas occurring per hour of sleep time (total and for NREM and REM sleep) as a marker of severity of obstructive sleep apnea (OSA). We also recorded any surgery related to OSA and existence of GERD. BMI was defined as kg/m2 and was translated into a BMI Zscore. BMI Z-scores, also known as BMI standard deviation scores, are measures of relative weight adjusted for child age and sex. These values correspond to growth chart percentiles, and are particularly useful to monitor changes in patients with a BMI score above the 99th percentile or below the first percentile. BMI is not calculated for children less than 2 years of age, so we excluded 11 children on this basis. We divided the remaining 52 patients into a group with OSA and a group without OSA. The determination of whether a patient had OSA was based on interpretation of the sleep studies by a pulmonologist certified in sleep medicine. In our laboratory, a diagnosis of OSA requires an AHI > 1.0, which is consistent with the criteria used by Dryken et al. [12]. Mild OSA was defined as having an AHI between 1 and 5, moderate OSA was defined as having an AHI between 5 and 10, and severe OSA was defined as having an AHI greater than 10. Linear regression analysis was used to evaluate the association of AHI with BMI Z-score, tonsil size, and age. DS and non-DS OSA data were compared using the independent samples t-test. A p-value of <0.05 was considered significant.
769
Table 1 Population characteristics.
Total Male Female BMI Z-score Mean age (years)
OSA
Non-OSA
33 18 (55%) 15 (45%) 2.09 + 0.94 10.2 + 4.2
19 10 (54%) 9 (46%) 1.4 + 1.4 7.8 + 4.3
4. Results Sixty-three pediatric DS patients with overnight PSG were identified. Eleven patients were excluded due to age less than 2 years. Fifty-two patients were included in the study and these were divided into two groups based on OSA status (Table 1). Thirty-three patients (63.5%) were found to have OSA on PSG and 19 patients (36.5%) were found not to have OSA. Of the 52 patients included in the study, 24 were females (46%) and 28 were males (54%). The overall mean age of participants at the time of sleep study was 9.3 years 4.5 (9.8 years for males, 8.7 years for females). The mean age of patients found to have OSA was 10.2 years 4.2, and the mean age of non-OSA patients was 7.8 4.3. We compared the difference between the mean age of the two groups by t-test with a results of p < 0.001. The mean BMI in all participants was 24.7 + 12.4 (24.1 for females, 25.3 for males). The mean BMI Z-score of all participants was 1.84 (standard deviation 1.12), 1.85 for females (SD 0.99) and 1.83 for males (SD 1.24). The mean BMI Z-score in the OSA group was 2.09 0.84, and the mean BMI Z-score for the non-OSA group was 1.4 1.40. We compared the difference between the mean BMI Z-scores for the OSA and the non-OSA groups by t-test with a result of p = 0.03. Five children had previously undergone tonsillectomy and adenoidectomy. The average BMI in these children was 36.4, and the average AHI was 8.72. Three of these children continued to have OSA postoperatively. Of the 14 children with OSA who had not undergone tonsillectomy and adenoidectomy, the average tonsil size was 3, and the average AHI was 28. The positive correlation between tonsil size and AHI reached statistical significance (r2 = 0.75, p = 0.002) (Fig. 5). 5. Discussion In previous studies, OSA has been found to be prevalent in young patients with DS. In a pilot study, Dyken evaluated 19 patients (nine boys, 10 girls) in a consecutively encountered, nonselected population of young patients with Down syndrome using standard overnight polysomnography [12]. He found OSA in 79% of the subjects. Higher BMI was significantly associated with a higher apnea index and a lower SaO2 level. In our study, 52 patients were evaluated and 33 patients (63.5%) were found to have OSA. This is a considerably lower percentage of patients with DS found to have OSA than in Dyken’s study, but still a significant fraction. One established risk factor for OSA in the general adult and pediatric population is a high BMI, but the contribution of high BMI to development of OSA in DS patients is unknown. Since there are no BMI charts specific to DS children in the United States, the CDC BMI-for-age charts that include all children were utilized to plot our data. The BMI charts that are available are inadequate for children with this disorder as illustrated by the number of children with BMI values beyond the upper extreme of the scale (Fig. 1). 94% of our children with DS fall above the 50th percentile and nearly 88% of children in our study were above the 85th percentile of the CDC chart. This is a much higher value than that reported by Cronk. She reported that 15% of children with DS were overweight (weight
770
C.B. Shires et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
Fig. 1. (a) Plots of male BMI on standard BMI-for-age chart. (b) Plots of female BMI on standard BMI-for-age chart.
Fig. 2. (a) Plots of BMI of male DS patients on Swedish DS BMI chart. (b) Plots of BMI of female DS patients on Swedish DS BMI chart.
C.B. Shires et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
greater than 85th percentile) during early infancy and nearly 50% of girls age 3 years and boys in early childhood were overweight. She found that the percentage of overweight DS children varied across the remaining age intervals but was always greater than 30% [12]. We would consider the development of disorder specific charts to be a positive step in the care and management of these patients. Syndrome specific growth charts have been developed for multiple disorders, including Prader–Willi syndrome [1,9], Turner syndrome [10], Noonan syndrome [11], and Down syndrome [12]. In 2002, Myrelid from Sweden evaluated 354 DS children and developed a BMI chart based on these patients. He noted that the Swedish male DS patients were taller than male patients from the United States, and that the females were lighter in weight but of similar height to U.S. females [8]. Both factors support the higher BMI values we observed in our patients. Most of our DS patients (88%) fall above the mean of the Swedish DS BMI charts, which illustrates that the DS population in Memphis, TN is generally heavier than the Swedish DS population. Therefore, syndrome specific charts may also need to be adjusted for different locations. When plotting our DS patients on both the CDC BMI chart and the Swedish DS BMI chart, we can see a slight difference in the number of children that would be considered overweight. The same percentage of male patients (93%) was above the 50th percentile on both the CDC BMI chart and DS BMI chart. However, 96% of DS females were above the 50th percentile on the CDC BMI chart, and 83% of females were above the 50th percentile on the DS BMI chart (p = 002). Thus, the syndrome specific BMI chart appeared to be more applicable in the female gender. Our data plotted in Fig. 2 demonstrate visually the difference we found between our OSA and non-OSA BMI values plotted on a syndrome specific BMI chart. 100% of DS males with OSA in our study were above the 50th percentile for BMI on the Swedish BMI chart, while 80% of our males DS without OSA were above the 50th percentile for BMI on the Swedish BMI chart. 87% of our DS females with OSA were above the 50th percentile for BMI on the Swedish BMI chart, while 78% of our DS females without OSA were above the 50th percentile for BMI on the Swedish BMI chart. There appears to be a difference in the percentage of obesity between the OSA and non-OSA groups; however the majority of DS children in our study battle being overweight regardless of OSA status. Using linear regression, there was no statistically significant correlation between AHI and BMI (r2 = 0.2, p = 0.09) (Fig. 3). However, when the BMI Z-scores (BMI adjusted for child age and sex) were used, statistical statistical significance was reached (p = 0.03). In the Swedish study, the average AHI was 12.9 episodes per hour, and in our study, the average AHI was 18.7 The children of
Fig. 3. Scatter plot comparing AHI to BMI.
771
Fig. 4. Scatter plot comparing AHI to age.
our study showed higher BMI’s and higher AHI’s. Fitzgerald found no associated between BMI and degree of OSA; however, BMI and not adjusted BMI values were used in that study [6]. The importance of the relationship between OSA and DS is that BMI is a modifiable risk factor. Maintaining a lower BMI could be a positive addition to the management of OSA in these children. Interestingly, several DS patients with an extremely high BMI had a normal sleep study. The converse was also observed with patients who had both a low BMI and OSA. Based on this fact, OSA is likely a multifactorial disease with various potential contributing factors [5]. Other factors that we consider as contributors are the hypothyroidism, relative midfacial hypoplasia, mandibular hypoplasia, hypertrophy of tonsils and adenoids, and hypotonia seen in DS children. Some of these risk factors are fixed, and some require invasive procedures to adjust. In contrast to our study, Fitzgerald reviewed polysomnograms of 33 children with DS, 91% of whom were not obese. Interestingly, 97% had obstructive sleep apnea, with an average AHI of 12.9 and an average oxygen desaturation of 4%. This supported the recommendation of conducting polysomnograms in all children with DS who snore regardless of BMI [6]. We also wanted to compare age and BMI to evaluate if Down syndrome children become more obese with age. Linear regression showed a positive correlation between age and BMI (r2 = 0.39, p < 0.001) (Fig. 4), supporting an increase in obesity as the DS children age. This is in contrast to Berger’s investigations, in which he found that in patients with primary snoring, mild and moderate OSA had a similar increase in AHI over time which depended mainly on weight gain and to a lesser degree on time. However, he found that age and high BMI, but not AHI, were significant risk factors for developing hypertension and/or cardiovascular disease [2]. While tonsil size and AHI do have a significant positive correlation (p = 0.002) in our study, others have shown that T&A in DS children improves some parameters of OSA, but not as markedly as in non-DS children [10]. T&A is a common initial surgical treatment for OSA in DS children; however, 30–50% develop persistent and recurrent OSA which eventually require CPAP, BiPAP, or tracheostomy [11]. Eight of the 19 children with Down syndrome and OSA in our study also carried the diagnosis of gastroesophageal reflux disease. The average AHI in these children was 9.282, while the average AHI in children without GERD was 13.56. Therefore, GERD does not significantly worsen OSA in this population. There was no statistical difference in the number of patients with DS and OSA among genders. 46% of our patients with DS and
C.B. Shires et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
772
patient population. While there are many possible factors in the development of OSA, a high BMI is one that can be modified. In DS children at risk for developing OSA, a recommendation to maintain a lower BMI could positively affect their clinical course. OSA remains a complex disorder, and we encourage further investigation into the relationship between OSA and BMI in this patient population. We will endeavor to further investigate BMI, as well as other possible contributors to the development of OSA in these patients in the future.
References
Fig. 5. Scatter plot comparing AHI to tonsil size.
OSA were female, and 54% were male. This is in contrast to Plywaczewski, who found that obesity and male gender were the main risk factors for the development of OSA [3]. There are limitations to our study, as this was a retrospective review. We did not record the incidence of glossoptosis in these children. We were unable to document snoring in some of the subjects, and therefore chose to not include this data point in our study. However, Fitzgerald has recommended routine polysomnography in Down syndrome children independent of a history of snoring [4,6]. We also did not include cephalometric measures, the incidence of midface/mandibular hypoplasia, or the incidence of hypothyroidism in these children. 6. Conclusion We found a positive relationship between BMI and OSA in our pediatric DS patient population. This relationship mirrors the established relationship in non-DS children and adults. We also found a significant relationship between age and OSA in our
[1] H. Heussler, M. Harris, D. Cooper, C. Dakin, S. Suresh, G. Williams, Hypersomolence in Prader Willi syndrome, J. Intellect. Disabil. Res. 52 (10) (2008 Oct) 814. [2] G. Berger, R. Berger, A. Oksenberg, Progression of snoring and obstructive sleep apnoea: the role of increasing weight and time, Eur. Respir. J. 33 (February(2)) (2009) 338–345. [3] R. Plywaczewski, P. Bielen, M. Bednarek, L. Jonczak, D. Gorecka, P. Sliqinski, Influence of neck circumference and body mass index on obstructive sleep apnoea severity in males, Pneumonol. Alerfol. Pol. 76 (5) (2008) 313–320. [4] L.G. Morris, A. Kleinberger, K.C. Lee, L.A. Liberatore, O. Burschtin, Rapid risk stratification for obstructive sleep apnea, based on snoring severity and body mass index, Otolaryngol. Head Neck Surg. 139 (November (5)) (2008) 615–618. [5] A. Ursavas, M. Karadag, Y.O. Ilcol, I. Ercan, B. Burgazlioglu, F. Coskun, R.O. Gozu, Low level of IGF-1 in obesity may be related to obstructive sleep apnea syndrome, Lung 185 (September–October (5)) (2007) 309–314. [6] D.A. Fitzgerald, A. Paul, C. Richmond, Severity of obstructive apnoea in children with Down syndrome who snore, Arch. Dis. Child. 92 (May (5)) (2007) 423–425. [7] O. Resta, M.P. Barbaro, T. Giliberti, G. Caratozollo, M.G. Cagnazzo, F. Scarpelli, M.C. Nocerino, Sleep related breathing disorders in adults with Down syndrome, Downs Syndr. Res. Pract. 8 (August (3)) (2003) 115–119. [8] A. Myrelid, J. Gustafsson, B. Ollars, G. Anneren, Growth charts for Down’s syndrome from birth to 18 years of age, Arch. Dis. Child. 87 (August (2)) (2002) 97– 103. [9] M.G. Butler, F.J. Meany, An anthropometric study of 38 individuals with Prader– Labhart–WIlli syndrome, Am. J. Med. Genet. 26 (1987) 44–55. [10] M.M. Shete, R.M. Stocks, M. Sebelik, R. Schoumacher, Effects of adenotonsillectomy on polysomnographic patterns in Down syndrome children with obstructive sleep apnea: a comparative study with children without Down syndrome, Int. J. Pediatr. Otorhinolaryngol. 74 (March (3)) (2010) 241–244. [11] M. Strome, Down’s syndrome: a modern otorhinolaryngological perspective, Laryngoscope 96 (December) (1986) 1340–1343. [12] M. Dryken, D. Lin-Dykin, S. Poulton, M. Zimmerman, Prospective polysomnographic analysis of obstruction sleep apnea in Down syndrome, Arch. Pediatr. Adolesc. Med. 157 (July) (2003) 655–660.
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Body mass index as an indicator of obstructive sleep apnea in pediatric Down syndrome§ Courtney B. Shires a,*, Sandra L. Anold b,2, Robert A. Schoumacher b,2, George W. Dehoff c,3, Sreekrishna K. Donepudi a,1, Rose Mary Stocks a,1 a
Department of Otolaryngology-Head & Neck Surgery, University of Tennessee Health Science Center, 910 Madison Avenue, Room 428, Memphis, TN 38163, USA Department of Pediatrics, University of Tennessee Health Science Center, 50 N Dunlap Street, Memphis, TN 38103, USA c Department of Internal Medicine, University of Tennessee Health Science Center, 956 Court Avenue, Room H314, Memphis, TN, USA b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 19 September 2009 Received in revised form 20 March 2010 Accepted 23 March 2010 Available online 7 May 2010
Objective: Our objective was to determine if higher body mass index (BMI) increases the likelihood of, obstructive sleep apnea (OSA) in pediatric Down syndrome (DS) patients. Methods: We performed a, retrospective chart review of 63 DS patients evaluated by overnight polysomnography from December 1995 to February 2005. Patients aged less than 2 years were excluded. Remaining patients were grouped, according to presence (n = 19) or absence (n = 33) of OSA based on apnea hypopnea index (AHI). OSA, and non-OSA DS groups were age matched while blinded to patient attributes other than age and OSA, status. Patients without appropriate age matches were excluded. We recorded various patient information, including age, sex, height, weight, number of apneas, number of hypopneas, respiratory distress index (RDI), apnea–hypopnea index (AHI), lowest oxygen saturation during sleep, mean oxygen saturation, number of arousals per hour, and mean time spent in REM sleep. We calculated BMI using the, standard kg/m2 formula and converted this into a Z-score. Results: Fifty-two DS patients were analyzed with average age of 9.3 4.5 years (10.2 4.2 in 33 OSA patients, 7.8 4.3 in 19 non-OSA patients). There were 28 males and 24 females. The OSA group mean BMI Zscore was 2.09 0.94, and the non-OSA group Z-score was 1.4 1.40. The Z-scores for BMI were statistically significant between OSA and non-OSA patients with p = 0.03 by t-test. Conclusions: When age and sex adjusted, BMI has a statistically significant association with the presence of OSA in Down syndrome patients. The incidence of OSA also increases with increasing age in this population. ß 2010 Elsevier Ireland Ltd. All rights reserved.
Keywords: BMI Obstructive sleep apnea Down syndrome
1. Objective Our objective was to determine the contribution of body mass index (BMI) to the incidence of obstructive sleep apnea (OSA) in pediatric Down syndrome (DS) patients. 2. Introduction Children with DS are known to have a wide variety of medical problems associated with the disorder. The DS population has a
§ Presented at the American Society of Pediatric Otolaryngologists 2007 Spring Meeting at the Combined Otolaryngology Spring Meetings in San Diego, CA, USA on April 27, 2007. * Corresponding author. Tel.: +1 901 448 5886; fax: +1 901 448 5120. E-mail address: [email protected] (C.B. Shires). 1 Tel.: +1 901 448 5886; fax: +1 901 448 5120. 2 Tel.: +1 9014485500; fax: +1 901 448 7836. 3 Tel.: +1 901 448 5814; fax: +1 901 448 7836.
0165-5876/$ – see front matter ß 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2010.03.050
higher incidence of OSA (30–50%) than the general population (0.7–2.0%) [7]. Children and adolescents with DS have a tendency to be overweight beginning in infancy and continuing throughout the growing years [12]. High BMI is a known OSA risk factor among children and adults without DS, but the role it plays in DS children has been a subject of debate. OSA, if untreated, can cause exacerbation of problems seen in DS as well as cause problems directly related to OSA. Untreated OSA can have serious pulmonary sequelae such as hypercarbia, acidosis, and pulmonary hypertension with possible cor pulmonale. The possibility of these health problems in a population at higher health risk is a serious concern. Untreated OSA can also lead to sleep fragmentation, sleep deprivation, daytime somnolence, decreased sleep time, and decreased REM time, all of which can aggravate learning and behavioral problems already encountered in DS patients. Due to the possible serious health complications of OSA in these children, we believe it is important to learn as much as possible about the etiology of this disease.
C.B. Shires et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
3. Methods Our study was designed as a retrospective chart review. We obtained institutional review board approval from both the University of Tennessee Health Science Center and Methodist LeBonheur Healthcare. First, we identified 63 consecutive children with DS that had been evaluated by overnight polysomnography (PSG). The identification of DS was made based both on clinical history and photographs. The PSG were performed between the years of 1995 and 2005 in the LeBonheur Pediatric & Adolescent Sleep Disorders Center at LeBonheur Children’s Hospital in Memphis, TN. Data were recorded from the date of the PSG, and alternate data were used from clinic visits or surgery if necessary. A standardized night-time polysomnography at the sleep laboratory included 16-channel strip recorder on a computerized system. All studies involved a minimum period of 5 h and a maximum of 10 h of sleep in a time and temperature controlled room, accompanied by one of the parents without previous sedation or sleep deprivation. The children were continuously observed by trained technicians through an infra-red video camera. The following parameters were measured: electroencephalogram (C3/A2, C4/A1, O1/A2), electromyogram (submental and tibial), electro-oculogram (right and left), oxygen saturation (SaO2), oro-nasal airflow (thermistor), abdominal and chest wall movement, and electrocardiography. We recorded various patient information including age, sex, height, weight, number of apneas, number of hypopneas, respiratory distress index (RDI), apnea–hypopnea index (AHI), lowest oxygen saturation during sleep, mean oxygen saturation, number of arousals per hour, and mean time spent in REM sleep. Sleep architecture was scored using standard criteria for our sleep laboratory. Certified polysomnography technicians scored the sleep studies, which were later interpreted by board-certified sleep medicine specialists. Respiratory events were significant if they lasted two or more respiratory cycles and were associated with a >3% fall in SaO2 and/or were terminated by an arousal. Obstructive apneas were defined as a reduction in airflow of >80% from baseline amplitude with continuing or increasing effort as reflected by the thoracic or abdominal plethysmographic bands. Hypopneas were defined as a decrease in airflow to between 20% and 50% of the baseline amplitude and >2% reduction in SaO2. The AHI was defined as the number of apneas or hypopneas occurring per hour of sleep time (total and for NREM and REM sleep) as a marker of severity of obstructive sleep apnea (OSA). We also recorded any surgery related to OSA and existence of GERD. BMI was defined as kg/m2 and was translated into a BMI Zscore. BMI Z-scores, also known as BMI standard deviation scores, are measures of relative weight adjusted for child age and sex. These values correspond to growth chart percentiles, and are particularly useful to monitor changes in patients with a BMI score above the 99th percentile or below the first percentile. BMI is not calculated for children less than 2 years of age, so we excluded 11 children on this basis. We divided the remaining 52 patients into a group with OSA and a group without OSA. The determination of whether a patient had OSA was based on interpretation of the sleep studies by a pulmonologist certified in sleep medicine. In our laboratory, a diagnosis of OSA requires an AHI > 1.0, which is consistent with the criteria used by Dryken et al. [12]. Mild OSA was defined as having an AHI between 1 and 5, moderate OSA was defined as having an AHI between 5 and 10, and severe OSA was defined as having an AHI greater than 10. Linear regression analysis was used to evaluate the association of AHI with BMI Z-score, tonsil size, and age. DS and non-DS OSA data were compared using the independent samples t-test. A p-value of <0.05 was considered significant.
769
Table 1 Population characteristics.
Total Male Female BMI Z-score Mean age (years)
OSA
Non-OSA
33 18 (55%) 15 (45%) 2.09 + 0.94 10.2 + 4.2
19 10 (54%) 9 (46%) 1.4 + 1.4 7.8 + 4.3
4. Results Sixty-three pediatric DS patients with overnight PSG were identified. Eleven patients were excluded due to age less than 2 years. Fifty-two patients were included in the study and these were divided into two groups based on OSA status (Table 1). Thirty-three patients (63.5%) were found to have OSA on PSG and 19 patients (36.5%) were found not to have OSA. Of the 52 patients included in the study, 24 were females (46%) and 28 were males (54%). The overall mean age of participants at the time of sleep study was 9.3 years 4.5 (9.8 years for males, 8.7 years for females). The mean age of patients found to have OSA was 10.2 years 4.2, and the mean age of non-OSA patients was 7.8 4.3. We compared the difference between the mean age of the two groups by t-test with a results of p < 0.001. The mean BMI in all participants was 24.7 + 12.4 (24.1 for females, 25.3 for males). The mean BMI Z-score of all participants was 1.84 (standard deviation 1.12), 1.85 for females (SD 0.99) and 1.83 for males (SD 1.24). The mean BMI Z-score in the OSA group was 2.09 0.84, and the mean BMI Z-score for the non-OSA group was 1.4 1.40. We compared the difference between the mean BMI Z-scores for the OSA and the non-OSA groups by t-test with a result of p = 0.03. Five children had previously undergone tonsillectomy and adenoidectomy. The average BMI in these children was 36.4, and the average AHI was 8.72. Three of these children continued to have OSA postoperatively. Of the 14 children with OSA who had not undergone tonsillectomy and adenoidectomy, the average tonsil size was 3, and the average AHI was 28. The positive correlation between tonsil size and AHI reached statistical significance (r2 = 0.75, p = 0.002) (Fig. 5). 5. Discussion In previous studies, OSA has been found to be prevalent in young patients with DS. In a pilot study, Dyken evaluated 19 patients (nine boys, 10 girls) in a consecutively encountered, nonselected population of young patients with Down syndrome using standard overnight polysomnography [12]. He found OSA in 79% of the subjects. Higher BMI was significantly associated with a higher apnea index and a lower SaO2 level. In our study, 52 patients were evaluated and 33 patients (63.5%) were found to have OSA. This is a considerably lower percentage of patients with DS found to have OSA than in Dyken’s study, but still a significant fraction. One established risk factor for OSA in the general adult and pediatric population is a high BMI, but the contribution of high BMI to development of OSA in DS patients is unknown. Since there are no BMI charts specific to DS children in the United States, the CDC BMI-for-age charts that include all children were utilized to plot our data. The BMI charts that are available are inadequate for children with this disorder as illustrated by the number of children with BMI values beyond the upper extreme of the scale (Fig. 1). 94% of our children with DS fall above the 50th percentile and nearly 88% of children in our study were above the 85th percentile of the CDC chart. This is a much higher value than that reported by Cronk. She reported that 15% of children with DS were overweight (weight
770
C.B. Shires et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
Fig. 1. (a) Plots of male BMI on standard BMI-for-age chart. (b) Plots of female BMI on standard BMI-for-age chart.
Fig. 2. (a) Plots of BMI of male DS patients on Swedish DS BMI chart. (b) Plots of BMI of female DS patients on Swedish DS BMI chart.
C.B. Shires et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
greater than 85th percentile) during early infancy and nearly 50% of girls age 3 years and boys in early childhood were overweight. She found that the percentage of overweight DS children varied across the remaining age intervals but was always greater than 30% [12]. We would consider the development of disorder specific charts to be a positive step in the care and management of these patients. Syndrome specific growth charts have been developed for multiple disorders, including Prader–Willi syndrome [1,9], Turner syndrome [10], Noonan syndrome [11], and Down syndrome [12]. In 2002, Myrelid from Sweden evaluated 354 DS children and developed a BMI chart based on these patients. He noted that the Swedish male DS patients were taller than male patients from the United States, and that the females were lighter in weight but of similar height to U.S. females [8]. Both factors support the higher BMI values we observed in our patients. Most of our DS patients (88%) fall above the mean of the Swedish DS BMI charts, which illustrates that the DS population in Memphis, TN is generally heavier than the Swedish DS population. Therefore, syndrome specific charts may also need to be adjusted for different locations. When plotting our DS patients on both the CDC BMI chart and the Swedish DS BMI chart, we can see a slight difference in the number of children that would be considered overweight. The same percentage of male patients (93%) was above the 50th percentile on both the CDC BMI chart and DS BMI chart. However, 96% of DS females were above the 50th percentile on the CDC BMI chart, and 83% of females were above the 50th percentile on the DS BMI chart (p = 002). Thus, the syndrome specific BMI chart appeared to be more applicable in the female gender. Our data plotted in Fig. 2 demonstrate visually the difference we found between our OSA and non-OSA BMI values plotted on a syndrome specific BMI chart. 100% of DS males with OSA in our study were above the 50th percentile for BMI on the Swedish BMI chart, while 80% of our males DS without OSA were above the 50th percentile for BMI on the Swedish BMI chart. 87% of our DS females with OSA were above the 50th percentile for BMI on the Swedish BMI chart, while 78% of our DS females without OSA were above the 50th percentile for BMI on the Swedish BMI chart. There appears to be a difference in the percentage of obesity between the OSA and non-OSA groups; however the majority of DS children in our study battle being overweight regardless of OSA status. Using linear regression, there was no statistically significant correlation between AHI and BMI (r2 = 0.2, p = 0.09) (Fig. 3). However, when the BMI Z-scores (BMI adjusted for child age and sex) were used, statistical statistical significance was reached (p = 0.03). In the Swedish study, the average AHI was 12.9 episodes per hour, and in our study, the average AHI was 18.7 The children of
Fig. 3. Scatter plot comparing AHI to BMI.
771
Fig. 4. Scatter plot comparing AHI to age.
our study showed higher BMI’s and higher AHI’s. Fitzgerald found no associated between BMI and degree of OSA; however, BMI and not adjusted BMI values were used in that study [6]. The importance of the relationship between OSA and DS is that BMI is a modifiable risk factor. Maintaining a lower BMI could be a positive addition to the management of OSA in these children. Interestingly, several DS patients with an extremely high BMI had a normal sleep study. The converse was also observed with patients who had both a low BMI and OSA. Based on this fact, OSA is likely a multifactorial disease with various potential contributing factors [5]. Other factors that we consider as contributors are the hypothyroidism, relative midfacial hypoplasia, mandibular hypoplasia, hypertrophy of tonsils and adenoids, and hypotonia seen in DS children. Some of these risk factors are fixed, and some require invasive procedures to adjust. In contrast to our study, Fitzgerald reviewed polysomnograms of 33 children with DS, 91% of whom were not obese. Interestingly, 97% had obstructive sleep apnea, with an average AHI of 12.9 and an average oxygen desaturation of 4%. This supported the recommendation of conducting polysomnograms in all children with DS who snore regardless of BMI [6]. We also wanted to compare age and BMI to evaluate if Down syndrome children become more obese with age. Linear regression showed a positive correlation between age and BMI (r2 = 0.39, p < 0.001) (Fig. 4), supporting an increase in obesity as the DS children age. This is in contrast to Berger’s investigations, in which he found that in patients with primary snoring, mild and moderate OSA had a similar increase in AHI over time which depended mainly on weight gain and to a lesser degree on time. However, he found that age and high BMI, but not AHI, were significant risk factors for developing hypertension and/or cardiovascular disease [2]. While tonsil size and AHI do have a significant positive correlation (p = 0.002) in our study, others have shown that T&A in DS children improves some parameters of OSA, but not as markedly as in non-DS children [10]. T&A is a common initial surgical treatment for OSA in DS children; however, 30–50% develop persistent and recurrent OSA which eventually require CPAP, BiPAP, or tracheostomy [11]. Eight of the 19 children with Down syndrome and OSA in our study also carried the diagnosis of gastroesophageal reflux disease. The average AHI in these children was 9.282, while the average AHI in children without GERD was 13.56. Therefore, GERD does not significantly worsen OSA in this population. There was no statistical difference in the number of patients with DS and OSA among genders. 46% of our patients with DS and
C.B. Shires et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 768–772
772
patient population. While there are many possible factors in the development of OSA, a high BMI is one that can be modified. In DS children at risk for developing OSA, a recommendation to maintain a lower BMI could positively affect their clinical course. OSA remains a complex disorder, and we encourage further investigation into the relationship between OSA and BMI in this patient population. We will endeavor to further investigate BMI, as well as other possible contributors to the development of OSA in these patients in the future.
References
Fig. 5. Scatter plot comparing AHI to tonsil size.
OSA were female, and 54% were male. This is in contrast to Plywaczewski, who found that obesity and male gender were the main risk factors for the development of OSA [3]. There are limitations to our study, as this was a retrospective review. We did not record the incidence of glossoptosis in these children. We were unable to document snoring in some of the subjects, and therefore chose to not include this data point in our study. However, Fitzgerald has recommended routine polysomnography in Down syndrome children independent of a history of snoring [4,6]. We also did not include cephalometric measures, the incidence of midface/mandibular hypoplasia, or the incidence of hypothyroidism in these children. 6. Conclusion We found a positive relationship between BMI and OSA in our pediatric DS patient population. This relationship mirrors the established relationship in non-DS children and adults. We also found a significant relationship between age and OSA in our
[1] H. Heussler, M. Harris, D. Cooper, C. Dakin, S. Suresh, G. Williams, Hypersomolence in Prader Willi syndrome, J. Intellect. Disabil. Res. 52 (10) (2008 Oct) 814. [2] G. Berger, R. Berger, A. Oksenberg, Progression of snoring and obstructive sleep apnoea: the role of increasing weight and time, Eur. Respir. J. 33 (February(2)) (2009) 338–345. [3] R. Plywaczewski, P. Bielen, M. Bednarek, L. Jonczak, D. Gorecka, P. Sliqinski, Influence of neck circumference and body mass index on obstructive sleep apnoea severity in males, Pneumonol. Alerfol. Pol. 76 (5) (2008) 313–320. [4] L.G. Morris, A. Kleinberger, K.C. Lee, L.A. Liberatore, O. Burschtin, Rapid risk stratification for obstructive sleep apnea, based on snoring severity and body mass index, Otolaryngol. Head Neck Surg. 139 (November (5)) (2008) 615–618. [5] A. Ursavas, M. Karadag, Y.O. Ilcol, I. Ercan, B. Burgazlioglu, F. Coskun, R.O. Gozu, Low level of IGF-1 in obesity may be related to obstructive sleep apnea syndrome, Lung 185 (September–October (5)) (2007) 309–314. [6] D.A. Fitzgerald, A. Paul, C. Richmond, Severity of obstructive apnoea in children with Down syndrome who snore, Arch. Dis. Child. 92 (May (5)) (2007) 423–425. [7] O. Resta, M.P. Barbaro, T. Giliberti, G. Caratozollo, M.G. Cagnazzo, F. Scarpelli, M.C. Nocerino, Sleep related breathing disorders in adults with Down syndrome, Downs Syndr. Res. Pract. 8 (August (3)) (2003) 115–119. [8] A. Myrelid, J. Gustafsson, B. Ollars, G. Anneren, Growth charts for Down’s syndrome from birth to 18 years of age, Arch. Dis. Child. 87 (August (2)) (2002) 97– 103. [9] M.G. Butler, F.J. Meany, An anthropometric study of 38 individuals with Prader– Labhart–WIlli syndrome, Am. J. Med. Genet. 26 (1987) 44–55. [10] M.M. Shete, R.M. Stocks, M. Sebelik, R. Schoumacher, Effects of adenotonsillectomy on polysomnographic patterns in Down syndrome children with obstructive sleep apnea: a comparative study with children without Down syndrome, Int. J. Pediatr. Otorhinolaryngol. 74 (March (3)) (2010) 241–244. [11] M. Strome, Down’s syndrome: a modern otorhinolaryngological perspective, Laryngoscope 96 (December) (1986) 1340–1343. [12] M. Dryken, D. Lin-Dykin, S. Poulton, M. Zimmerman, Prospective polysomnographic analysis of obstruction sleep apnea in Down syndrome, Arch. Pediatr. Adolesc. Med. 157 (July) (2003) 655–660.