Assessment of inspiratory flow limitation invasively and noninvasively during sleep

Abstract

To define the standard of airway flow limitation, pharyngeal pressure and flow rate were measured during wakefulness and sleep in seven habitual snorers with widely varying degrees of sleep-induced increases in upper airway resistance. Inspiratory pressure:flow relationships were used to group breaths into four categories of flow limitation, including linear (Level 1), mildly alinear (Level 2), constant flow rate with no pressure dependence (Level 3), and decreasing flow rate throughout significant portions of inspiration, i.e., negative pressure dependence (Level 4). These pressure:flow rate gold standards of flow limitation were used to evaluate a flow limitation index derived from the time profile (or "shape") of three noninvasive estimates of flow rate: (1) pneumotach flow rate, (2) differentiated sum respiratory inductance plethysmography (RIP), and (3) nasal pressure. A nonflow limited template for each of these noninvasive measurements was taken from awake breaths and the difference in area determined between the template breath and each of the noninvasive signals measured during nonrapid eye movement (NREM) sleep. The noninvasive flow limitation indices were found to be effective in differentiating severe types of inspiratory flow limitation, i.e., Level 1 versus Level 3 or Level 4 (sensitivity/specificity > 80%). On the other hand, these indirect indices were not able to consistently detect mild levels of flow limitation (Level 1 versus Level 2; sensitivity/specificity = 62 to 72%); nor were these noninvasive estimates of flow rate "shape" sensitive to breaths with a high but fixed resistance throughout inspiration. The area index derived from measurements of pressure at the nares (Pn) was the most sensitive, nonperturbing, noninvasive measure of flow rate and flow limitation, and we recommend its use for recognizing most of the common types of moderate to severe levels of airway flow limitation in sleeping subjects.