We wanted to study the heart rate response to chemoreceptor input. 17 healthy term infants born to daily smokers (median 10 cig./d) and 15 controls matched by sex and maternal age were studied 1, 3, and 10 days and 10 wk postnatally. During natural sleep, inspiratory gas was supplied in a lightweight face mask. Repeated 15-second challenges with 100% O2 followed by 60 s of air were presented. Beat-by-beat heart rate, breath-by-breath ventilation computed from strain-gauge respiratory tracing and EOG were recorded continuously.

Selecting quiet sleep only, we obtained 105 experimental runs with a median of 6 hyperoxic challenges. The coherently averaged heart rate and ventilatory response to hyperoxia were calculated for each infant at each age. Timing of responses and maximal change in RR interval and ventilation were analysed with ANOVA.

Short-term variation in RR interval and ventilation were closely correlated. Hyperoxia depressed ventilation and increased RR interval. The RR interval change lagged 4 s behind the ventilatory response. Median increase in RR interval during hyperoxia was 12 ms. The absolute value of the RR interval change was positively correlated with prestimulus RR interval (p<0.001), prestimulus respiratory frequency (p=0.02), and magnitude of ventilatory depression (p=0.02). Postnatal age and smoke exposure introduced no additional effects. Thus, a dynamic interplay of cardiac and respiratory reflexes decides the heart rate response to chemoreceptor stimuli in neonates.