In Response to regional peak mucosal cooling predicts the perception of nasal patency

The letter by Garcia et al. raised a concern that the reported correlation between computational fluid dynamics (CFD)-simulated and rhinomanometry-measured nasal resistance (NR) in our recent article is smaller than expected, although is still statistically significant (r 5 0.41, P< 0.01). This concern appears unwarranted given that this is the first published comparison between CFD simulations and rhinomanometric measurements in a cohort of live human subjects, and that there are no prior data in the literature to indicate what such a correlation should be. In addition, Garcia et al. are mistaken in their contention that our prior retrospective study in human subject included rhinomanometric measurements (no, it did not). Garcia et al.’s expectation of a higher correlation is purely based on experimental measurements performed in plastic replica nasal models in which resistances were measured under fluid mechanics lab settings. Measuring nasal resistance in live human subjects via commercial portable rhinomanometry is undoubtedly associated with higher variance; and the measurement process requires considerable subject cooperation, for example, holding the flow tube to ensure a tight seal to the external naris, etc. Because the reported resistances in our article are averages over 7 to 8 consecutive breaths, we reexamined the raw data and found that the standard deviation of resistance within these 7 to 8 breaths can range from 5% to 15%. The test–retest reliability of rhinomanometry over longer durations (1 hour) has been reported with a similar variation of 7% 17%, supporting the integrity of our data; however, that study applied a nasal decongestant that removed any effect of the nasal cycle. The nasal cycle may in fact be an additional contributor to the relatively low correlation. The time delay between rhinomanometry and the computed tomography scan in our protocol was roughly half an hour (comprising a subway ride between two sites, paperwork, and waiting at a busy outpatient radiology center). For some subjects, this half-hour delay may result in a phase change in their nasal cycles. Combining the NR of the two sides could mitigate some of the nasal cycle fluctuations because they may cancel out each other. So, we recalculated bilateral NR in this same subject cohort based on the unilateral NRs of the two sides (1/Rtotal 5 1/Rleft 1 1/Rright) and indeed found a better correlation between the CFD and rhinomanometry (r 5 0.53, P<0.01). We concur with Garcial et al. that other sources of variation may arise from the rigid nasal wall assumption in both the CFD and plastic nasal models, the sitting versus supine positions, or the different pressure drops between the CFD and rhinomanometry in obtaining the NR. Any or all of these conditions may account for a significant portion of the remaining variability. Nevertheless, the bigger picture from our article and the extant literature is that we should look beyond nasal resistance to account for individual perceptions of nasal obstruction. Regardless of nasal resistance, human beings must inspire an adequate tidal volume per breath to survive, and this tidal volume must affect a certain degree of nasal cooling. Therefore, nasal cooling may be partially decoupled from nasal resistance. Our previous study has shown that if we maintain the nasal resistance but modulate the nasal cooling, the perception of nasal patency actually follows the nasal cooling changes. We believe that this represents the mechanism whereby we perceive whether or not we are receiving a sufficient tidal volume, independent of significant fluctuations in nasal resistance during the nasal cycle or during restful breathing versus heavy exercise. The effort involved in breathing to overcome nasal resistance varies substantially across these conditions; however, our perception of adequate airflow (subjective nasal patency) remains remarkably stable.