In this issue of The Journal of Physiology, Dempsey and colleagues (Sheel et al. 2001) contribute another key chapter in their ongoing series of elegant investigations on novel interactions involving the respiratory muscles, autonomic nervous system and cardiovascular regulation in humans. Earlier, they demonstrated that manipulation of the work of breathing during maximal exercise resulted in marked changes in locomotor muscle blood flow, cardiac output and both whole-body and active limb oxygen uptake (Harms et al. 1997, 1998). They also established the remarkable metabolic costs of supporting respiratory muscle function during maximal exercise, requiring up to 16 % of the cardiac output (Harms et al. 1998). Importantly, the reduced locomotor muscle blood flow and vascular conductance in the elevated work of breathing condition was associated with augmented noradrenaline (norepinephrine) spillover from the active limbs, suggesting enhanced sympathetic vasoconstriction (Harms et al. 1997). These physiological effects of the work of breathing have important functional consequences, as demonstrated by an ~15 % improvement in endurance performance with respiratory muscle unloading (Harms et al. 2000).
The next generation of experiments attempted to establish the mechanisms underlying these fascinating physiological connections. In a paper recently published in this journal (St Croix et al. 2000), high-resistance, prolonged duty cycle breathing at rest, resulting in respiratory muscle fatigue, evoked an increase in leg muscle sympathetic nerve activity (MSNA) that was independent of central respiratory motor output, indicating a reflex origin. Moreover, the temporal nature of the response (MSNA was unchanged during the initial 1–2 min of the fatiguing task but increased progressively thereafter) was characteristic of a slower-developing muscle metaboreflex (chemoreflex), rather than a mechanoreflex stimulated by force development (which would be expected to evoke sympathoexcitation at the start of contractions).
The present article by Sheel et al. (2001) represents a critical extension of this work by establishing that this presumed respiratory muscle-limb reflex has the ability, at least under resting conditions, to reduce significantly limb blood flow and vascular conductance. Thus, together with previous observations (St Croix et al. 2000), the present contribution provides compelling evidence for the existence of a metaboreflex, with its origin in the respiratory muscles, that can modulate limb perfusion via stimulation of sympathetic nervous system vasoconstrictor neurones (