Mild obstructive sleep apnea: beyond the AHI

Abstract

A common conundrum faced by sleep medicine practitioners is how to manage the large group of patients with mild sleep apnea. Many patients are referred for sleep evaluation, with symptoms thought to be due to obstructive sleep apnea (OSA). Often polysomnography demonstrates only mild sleep apnea, and the clinician and patient are faced with the dilemma of whether to use continuous positive airway pressure (CPAP) therapy or an oral appliance. In making this important decision the clinician incorporates the commonly used definition of mild sleep apnea as an apnea-hypopnea index of between 5 and 14 apneas or hypopneas per hour of sleep. Moderate sleep apnea is defined as 15-29 events per hour, and severe is 30 and above events per hour. These arbitrary thresholds originated in the early 1980s when knowledge of this condition was in its infancy and little was known about the long term health effects. The definition was based on the finding of apneas, defined by the complete cessation of airflow for at least 10 seconds. The concept of hypopnea and respiratory-effort related arousal (RERA) came later and with frequently changing definitions that have been the subject of significant controversy throughout the last 30 years. Many sleep centers include these RERA’s in the definition of respiratory disturbance index, which is incorrectly used interchangeably with AHI. While the sleep literature has demonstrated the untoward effects of moderate to severe sleep apnea, there has been considerable debate about the clinical significance of mild sleep apnea, that is, an AHI between 5 and 15. The current paper by Quan, et al (1) is a significant contribution to the literature in sleep medicine addressing this important clinical question. This paper reports data drawn from the APPLES study, a large multi-center, well-conducted study designed to determine if CPAP therapy improves sleepiness, mood disorder, or cognitive function in patients with OSA, that has subsequently produced several important publications (2-6). As part of the study, extensive data was obtained on each of these neurocognitive parameters including the Epworth Sleepiness Scale, Stanford Sleepiness Scale, Hamilton Rating Scale for Depression, Profile of Mood States, and Sleep Apnea Quality of Life Index, all validated questionnaires used frequently in the sleep literature. In this part of the study, 199 patients with an AHI>5 but <15 were compared to 40 patients enrolled in the study, but with and AHI<5. The mean AHI was 10 per hour in the mild OSA group, and was 3 per hour in the non-OSA group. Size of the study was statistically large enough to determine significant differences. Remarkably, there