Flow-volume indices in snorers with and without-OSAS.

In a paper on obstructive sleep apnoea recently published in the Journal, BOGAARD et al. [1] found correlations between maximal sleep O2 desaturation on the one hand and peak inspiratory flow and forced inspiratory volume % predicted on the other. In their figures 2 and 3, they illustrated this by means of linear regression lines. However, for several reasons the regressions are not linear. One reason is that flowvolume indices above a certain value will not influence the arterial oxygen tension. The other reason is the curvilinear relationship between oxygen tension and oxygen saturation. Therefore, I believe that it is misleading to draw these straight regression lines, even though it is convenient (too convenient) to have the computer perform this manoeuvre.