Association of Obstructive Sleep Apnea With Calvarial and Skull Base Thinning

• Published in: JAMA otolaryngology-- head & neck surgery Vol. 144; no. 6; p. 513
• Main Authors: Rabbani, Cyrus, Saltagi, Mohamad Z, Ye, Michael J, Patel, Janaki M, Manchanda, Shalini, Nelson, Rick F
• Format: Journal Article
• Language: English
• Published: United States 01.06.2018
• ISSN: 2168-619X
• DOI: 10.1001/jamaoto.2018.0347

Spontaneous cerebrospinal fluid leaks (sCSF-L) of the temporal bone are associated with obesity, calvarial thinning, and obstructive sleep apnea (OSA), and the incidence has doubled in the past decade. It is currently unknown if OSA is independently associated with skull thinning. To determine if patients with OSA have thinner skulls than patients without OSA. A retrospective cohort study of patients who underwent a level 1 polysomnogram (PSG) and also had high-resolution computed tomographic (CT) imaging of the head from January 2010 to March 2017 at Indiana University was carried out. Patients with moderate to severe OSA (apnea-hypopnea index [AHI]≥25/h) and without OSA (AHI<5/h) were matched for age and body mass index (BMI, calculated as weight in kilograms divided by height in meters squared). Measurement of calvarial thickness, extracranial zygoma thickness, skull base height and tegmen dehiscence (>4 mm) when blinded to OSA status. Primary outcomes were calvarial, skull base, and zygoma thickness differences between patients with OSA vs those without OSA. A total of 22 933 patients had a PSG and 1012 also had head CT imaging. Of the 1012 patients with both PSG and CT, the mean (SD) age was 50.8 (16.2) years and 624 (61.7%) were women. Those patients with moderate to severe OSA (56) and without OSA (58) were matched for mean (SD) age (50.3 [6.5] vs 49.8 [6.1] years]) and BMI (37.4 [8.1] vs 38.6 [6.8]). Patients with OSA had thinner mean (SD) calvaria (2.73 [0.67] vs 2.47 [0.52] mm; difference, -0.26 mm; 95% CI, -0.49 to -0.04; Cohen d, 0.44) and thinner skull bases (5.03 [1.40] vs 4.32 [1.28] mm; difference, -0.71; 95% CI, -1.23 to -0.19; Cohen d, 0.53). The mean (SD) extracranial zygoma thickness was the same (4.92 [0.87] vs 4.84 [0.84] mm; difference, -0.07 mm; 95% CI, -0.39 to 0.24). The tegmen mastoideum was dehiscent in nearly twice as many patients with OSA as those without (37% vs 20%; difference, 17%; 95% CI, 0.4-32). Obstructive sleep apnea was independently associated with intracranial bone (calvaria and skull base) thinning and not with extracranial (zygoma) thinning. These findings support a possible role of OSA in the pathophysiologic development of sCSF-L.