1
Wade Chien, MD 1,2 , Neil Bhattacharyya, MD 2,3 Palate Length, Tonsil Size, and Sleep Apnea Severity 1 Department of Otolaryngology, Massachusetts Eye & Ear Infirmary, Boston, MA 2 Department of Otology and Laryngology, Harvard Medical School, Boston, MA 3 Department of Otolaryngology, Brigham and Women’s Hospital, Boston, MA Objectives: Determine relationship between soft tissue oropharyngeal measurements and sleep apnea severity. Methods: A prospective series of adult patients undergoing surgical therapy for obstructive sleep apnea (OSA) was studied. Tonsil size (graded 0 to 4+) and median (palatal spine-uvula tip) and lateral (posterior hard palate-free edge soft palate) dimensions of the soft palate were measured transorally at the time of surgery. From the preoperative polysomnographic and the medical record, respiratory disturbance index (RDI), lowest oxygen saturation (LSat) and body mass index (BMI) were determined. The relationship between both RDI and LSat and tonsil size and palatal dimensions was determined with multivariate linear regression adjusted for BMI. Results: 88 patients were enrolled. The mean±95% confidence interval values for the median and lateral soft palate lengths were 4.71±0.14 cm and 3.73±0.12 cm, respectively; the mean tonsil size was 1.8±0.3. The mean RDI and LSat were 44.0±5.6 events/hour and 84.7±2.4%, respectively. On multivariate regression, only BMI significantly predicted RDI (p=0.003); median (p=0.210) and lateral (p=0.507) palate lengths and tonsil size (p=0.860) did not. For LSat, both BMI and tonsil size were significant predictors (p<0.001 and p=0.017, respectively); median and lateral lengths did not (p=0.251 and p=0.376, respec- tively). Conclusion: Adjusted for BMI, soft palate length does not predict sleep apnea severity well. Adjusted for BMI, tonsil size predicts the LSat but not the RDI. These results highlight difficulties associated with correlating soft tissue structure with sleep apnea severity. Abstract: Obstructive sleep apnea (OSA) is a common sleep disorder that is thought to result from collapse of the upper airway during sleep. It has long been suspected that facial dysmorphosis plays an important role in the pathophysiology of OSA, and multiple cephalometric studies have been done to examine the relationship between craniofacial features and OSA. In this study, we tried to determine the relationship between oropharyngeal soft tissue measurements and sleep apnea severity by comparing measurements of tonsil and soft palate sizes with the respiratory disturbance index (RDI) and lowest oxygen saturation (LSat) in pa- tients with OSA. Introduction: 1. A prospective series of adult patients undergoing surgical therapy for obstructive sleep apnea (OSA) was studied. 2. Tonsil size (graded 0 to 4+) and median (palatal spine-uvula tip) and lateral (posterior hard palate-free edge soft palate) dimensions of the soft palate were measured transorally at the time of surgery (Figures 1A and 1B). 3. From the preoperative polysomnographic and the medical record, respiratory disturbance index (RDI), lowest oxygen saturation (LSat) and body mass index (BMI) were determined. 4. The relationship between both RDI and LSat and tonsil size and palatal dimensions was determined with univariate and multivariate linear regressions adjusted for BMI. Methods: 1. A total of 88 patients with OSA were enrolled. The mean BMI was 30.4. The mean ± 95% confi- dence interval values for the median and lateral soft palate lengths were 4.71±0.14 cm and 3.73±0.12 cm, respectively; the mean tonsil size was 1.8±0.3. The mean RDI and LSat were 44.0±5.6 events/hour and 84.7±2.4%, respectively. 2. On univariate regression anaylsis, only BMI significantly correlated with RDI and LSat. Median and lateral palate lengths, as well as tonsil size, did not correlate significantly (Table 1). 3. On multivariate regression analysis, only BMI significantly correlated with RDI and LSat. The tonsil size correlated significantly with LSat, but not RDI. Median and lateral palate lengths did not correlate significantly with either RDI or LSat (Table 2). Results: Obstructive sleep apnea (OSA) is a common type of sleep-disordered breathing which affects 2-4% of the middle-aged population (Young et al., 1993). Some of the common signs and symptoms of OSA include restless sleep, loud snoring, apneic episodes, excessive daytime sleepiness, and decreased cognitive function. Obstructive sleep apnea has been shown to increase the risk of diurnal hypertension, nocturnal dysrhythmias, pulmonary hypertension, ventricular failure, myocardial infarction, and stroke (Yamashiro and Kryger, 1994). In addition, patients with OSA have a 7-fold increase in the risk for motor vehicle acci- dent (Findley et al., 1988). The size of the upper airway is thought to play an important role in the pathophysiology of OSA (Davies and Stradling, 1990). This is determined by both the soft tissue and skeletal components of the upper airway. In this study, we focused on the soft tissue component of the upper airway, where the tonsil size and the soft palate dimensions were measured intraoperatively in OSA patients. We compared these measurements with the sleep study results (RDI, BMI, LSat) from the same patient to see if there are correlations between these different variables. In our study, we found that BMI significantly predicted the RDI on multivariate regression analysis. This is consistent with the findings from a study by Young et al. (1993), where the occurrence of sleep- disordered breathing in over 600 middle-aged adults was examined. Thy found that obesity significantly predicted an AHI score of 5 or higher (p<0.001), and is an important risk factor for OSA. We did not find any significant correlations between the median and lateral soft palate lengths and the RDI. This is similar to the finding from a study by Min et al. (1997), where the length of the uvula is not correlated with the RDI. However, the lack of correlation between the soft palate lengths and LSat in our study is contradictory to the study by Min et al. (1997), where a significant correlation was found be- tween these 2 variables (p=0.02). Interestingly, these authors also measured the horizontal width of the uvula, and found that the uvular width correlated significantly with the RDI and LSat. When the tonsil size was examined in our study, it did not correlate significantly with RDI. This is contradictory to a study by Erdamar et al. (2001), where 85 patients with OSA were studied. These au- thors found that the size of the tonsil significantly correlated with RDI (p=0.004). They also found a large difference in the mean RDI between patients with grade 1 tonsils (22.06) and grade 2 tonsils (43.53). Similarly, in a study by Friedman et al. (1999) looking at 172 patients with OSA, tonsil size was also found to be significantly correlated with RDI (p=0.008). It is unclear why our result differed from the other 2 studies, since the study populations are similar. It is conceivable that our study population is too small to realize a significant correlation between the tonsil size and RDI, though the study by Erdamar et al. (2001) has a similar population size to our study (85 vs. 88 patients, respectively). Even though it is commonly believed that the size of tonsils plays an important role in OSA and would correlate significantly with LSat on the polysomnography, no study to our knowledge has examined this relationship formally. In our study, we found that the size of tonsils did correlate significantly with LSat in OSA patients (p=0.017). This is consistent with the finding that patients with OSA and tonsillar hy- pertrophy have improved LSat after tonsillectomy (Shine et al., 2006). Discussion: 1. Median and lateral soft palate lengths did not predict the severity of obstructive sleep apnea well. 2. Tonsil size predicted the LSat, but not the RDI. 3. These results highlight the difficulties associated with correlating soft tissue structures with sleep apnea severity. Conclusions: 1. Brodsky L. Modern assessment of tonsils and adenoids. Pediatr Clinic N Am 1989; 36: 1551-69 2. Dell'AringaI AR, Juares AJC, de Melo C, et al., Histological analysis of tonsillectomy and adenoidectomy specimens - January 2001 to May 2003. Rev. Bras. Otorrinolaringol. vol.71 no.1 São Paulo Jan./Feb. 2005 3. Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep disordered breathing among middle-aged adults. N Engl J Med. 1993 Apr 29;328(17):1230-5 4. Yamashiro Y, Kryger MH. Why should sleep apnea be diagnosed and treated? Clin pulm Med 1994;1:250-9 5. Findley LJ, Unverzagt ME, Suratt PM. Automobile accidents involving patients with obstructive sleep apnea. Am Rev Resp Dis 1988;138:337-40 6. Davis RJO, Stradling JR. The relationship between neck circumference, radiographic pharyngeal anatomy, and the obstructive sleep apnea syndrome. Eur Respir J 1990;3:509-14 7. Min YG, Jang YJ, Rhee CK, Kim CN, Hong SK. Correlation between anthropometric measurements of the oropharyngeal area and severity of apnea in patients with snoring and obstructive sleep apnea. Auris Nasus Larynx, 1997 Oct;24(4):399-403 8. Erdamar B, Suoglu Y, Cuhadaroglu C, Katircioglu S, Guven M. Evaluation of clinical parameters in patients with obstructive sleep apnea and possible correlation with the severity of the disease. Eur Arch Otorhinolaryngol. 2001 Nov;258(9):492-5 9. Friedman M, Tanyeri H, La Rosa M, Landsberg R, Vaidyanathan K, Pieri S, Caldarelli D. Clinical predictors of obstructive sleep apnea. Laryngoscope. 1999 Dec;109(12):1901-7 10. Shine NP, Lannigan FJ, Coates, HL, Wilson A. Adenotonsillectomy for obstructive sleep apnea in obese children. Arch Otolaryngol Head Neck Surg. 2006 Oct;132(10):1123-7 References: M Figure 1B: Diagrammatic representa- tion of median (M) and lateral (L) soft palatal length measurement locations. Figure 1A: Grading system for tonsillar hypertrophy proposed by Brodsky (1989, figure taken from Dell’Aringa et al., 2005) Variable Pearson's r p-value RDI Median palate length 0.168 0.059 Lateral palate length 0.088 0.208 Tonsil size 0.119 0.136 BMI 0.352 <0.001 LSat Median palate length -0.119 0.134 Lateral palate length -0.062 0.282 Tonsil size 0.136 0.102 BMI -0.336 0.001 Table 1: Univariate Pearson correlation for RDI and LSat with anatomic predictor variables Table 2: Multivariate regression for RDI and LSat for anatomic predictor variables Variable Regression coeeficient p-value RDI Median palate length 8.6 0.210 Lateral palate length -5.5 0.507 Tonsil size 0.4 0.860 BMI 0.003 LSat Median palate length -3.0 0.251 Lateral palate length 2.9 0.376 Tonsil size 1.8 0.017 BMI -0.6 <0.001 1.4
