I thank Dr. Dahan for his interest in my article.1He presents data from three subjects (he acknowledges that this is a very small sample) that show some “within-day variation” in the acute hypoxic ventilatory response (AHVR). The figure he presents for one subject indicates that minute ventilation in euoxia is lowest in the morning, highest at midday, and in-between in the afternoon. I do not know whether this same pattern is the same in all subjects. The minute ventilation in euoxia was as high as approximately 20 l/min, and this implies either that the end-tidal partial pressure of carbon dioxide (Pco2) of Dahan's subject varied widely during the day or that, if end-tidal Pco 2was constant, the metabolic CO2output of the lung must have varied widely during the day. The AHVR variation must therefore have correlated with end-tidal Pco2and/or CO2output in Dahan's small series. However, neither Sahn et al. .2nor Zhang and Robbins3were able to find such correlations.
Nonetheless, the within-day variation reported by Dahan may indeed be a “circadian” influence, and I agree that the observation needs further study. Variation in the AHVR within individuals on repeat testing is well established,4but it seems that between-day variation is greater than within-day (circadian) variation.2,3The study of Zhang and Robbins3sheds important light on the issue of variation. They found that the method of inducing hypoxia (they studied square wave hypoxic input, incremental hypoxic steps, and simulated rebreathing) did not influence the AHVR measured. So it seems that the methodologic influences to which Dahan refers may not be as influential as we all (intuitively) might think them to be. This is perhaps further supported by my analysis that there is no actual difference in the result of studies using rebreathing as compared with those using step hypoxia in assessing the reduction in AHVR by low-dose anesthetic.5
Zhang and Robbins3also found that between-day variation was greater than within-day (circadian) variation, so although the within-day variation found by Dahan is of interest, it is possible that between-day variation is more important.
There are two practical questions: (1) Does this variation in AHVR impact upon the results of the relevant studies, and (2) how can we account for it or control for it experimentally?
With regard to the first question, if the variation in AHVR influenced the outcome of studies, it might prove difficult (because of the large confidence intervals) to show, for example, that anesthetics blunt AHVR. However, this is not the case, and many different studies consistently find that halothane, enflurane, and sevoflurane reduce AHVR. The only agent for which there is less consistency in different studies' results is isoflurane. Although this may be a result of inherent variation in AHVR, it is also possible (and perhaps more likely) to be due to some property of isoflurane that gives it a more variable effect than other agents.
With regard to the second question, there seem to be two general ways to control for variation. One is to conduct each experimental period in a study at precisely the same time of day. However, this does not control between-day variability (which seems more important). The second way is to conduct experimental periods at random times of day in a suitable number of different subjects, ideally using repeated experimental periods in the same subject, and then average the results to reduce any systematic variation. One problem is that repeated exposure is often (ethically) undesirable in anesthetic studies, but where possible, I prefer this second approach.
John Radcliffe Hospital, Oxford, United Kingdom. firstname.lastname@example.org