Comments on point:counterpoint: humans do/do not demonstrate selective brain cooling during hyperthermia.

TO THE EDITOR: Middle cerebral artery blood velocity (MCAVmean) is reduced up to 30% in humans following a passive increase in internal temperature of 1–1.5°C (1). If the diameter of the MCA remains unchanged during heat stress, reductions in MCAVmean are proportional to reductions in cerebral blood flow. In support of selective brain cooling in hyperthermic humans, White et al. (7) suggest that vasodilation of the cerebral vasculature exists, which in turn increases cranial perfusion and maintains the arterial-venous temperature difference. The authors state, “it remains to be explained how MCA velocity, and presumably cranial perfusion, is reduced in hyperthermic humans if mean arterial blood pressure is maintained and MCA caliber remains constant” (4, 7). This argument ignores the potent effects of changes in carbon dioxide partial pressures on cerebral perfusion, with hypercapnia increasing and hypocapnia decreasing cerebral blood flow, respectively (6). During moderate to pronounced passive heat stress, arterial and end-tidal carbon dioxide partial pressures decrease upward to 8 Torr (1, 2). Importantly, an 8-Torr reduction in arterial carbon dioxide partial pressure is estimated to reduce cerebral blood flow by 24% (5) through increases in resistance of the cerebral arterioles “downstream” from the MCA (3). Thus the clear and robust reduction in MCAVmean during passive heat stress is likely due primarily to decreases in carbon dioxide partial pressures causing increases in vascular resistance of cerebral arterioles distal to the MCA.