Comparison of the hyperoxic test and the alternate breath test in infants

Peripheral chemoreceptor function has been tested using either the hyperoxic test (HT), which decreases minute ventilation ((V)over dot(E)) by causing physiologic chemodenervation, or the alternate breath test (ABT), which induces (V)over dot(E) alternations by delivering rapid hypoxic stimuli through breath-by-breath alternations in fractional inspired 02 between normoxia (0.21) and hypoxia (0.15). No previous studies have compared ventilatory responses to both tests in the same infants. We hypothesized that the (V)over dot(E) decrease during HT would be significantly related to (V)over dot(E) alternations during ABT. Eighteen infants (postnatal age 21 +/- 14 d) underwent two 30-s HTs and two ABTs (quiet sleep, face mask, and pneumotachograph; mass spectrometry measurement of inspired and expired O-2 and CO2 fractions; and breath-by-breath analysis). The tests were done in random order. Decreases in (V)over dot(E) and mean inspiratory flow (tidal volume over inspiratory time, V-T/T-I) during HTs were significantly correlated to their respective percentage coefficients of alternation during ABTs (r = 0.69 and 0.70, respectively, p < 0.01). Principal components analysis showed that the (V)over dot(E) and V-T/T-I decreases during HTs were due chiefly to a fall in V-T, whereas (V)over dot(E) and V-T/T-I alternations were ascribable to alternations in both V-T and T-I. Intraindividual coefficients of variation of (V)over dot(E) changes were significantly lower during HTs than during ABTs. We conclude that (1) ventilatory responses to HT and ABT are significantly correlated despite differences in the mechanisms of the VE changes; (2) the better reproducibility of the (V)over dot(E) response to HT as compared with ABT may be an advantage in clinical practice.