Response.

Abstract

Dear Editor-in-Chief We appreciate the interest of such influential colleagues in our recent meta-analysis (5). Point-by-point responses to their comments are presented below: In our view, experts in meta-analysis concur that “The Cochrane Risk of Bias Assessment” tool is designed to evaluate randomized controlled trials, which have yet to be performed on the topic of the current meta-analysis. Likewise, PRISMA, but not MOOSE, guidelines focus on randomized controlled trials. An identical comment concerning arteriovenous oxygen difference (a-vO2diff) is acknowledged and fully addressed in the manuscript. The training study of Roca et al. (8) reported neither maximal cardiac output (max) nor a-vO2diff (i.e., the outcomes required for the present meta-analysis); therefore, that study cannot be included. Of note, the process of article selection is straightforward and thoroughly described (5).We concur with the prevailing notion that any variable potentially influenced by exercise training (ExT) may be modified more with ExT in untrained versus trained individuals. If, however, such a variable (e.g., a-vO2diff at maximal exercise) is close to optimal levels in the untrained state, ExT will induce a less marked effect. This conjecture is firmly opposed by “the null heterogeneity among studies and the associations between the standardized mean differences (SMDs) in max and V˙O2max (P = 0.006), and the SMD in a-vO2diff and duration of training (P = 0.0002)” ([5], p. 2031). To relieve our colleagues’ main concern, the meta-analysis reveals that a-vO2diff is indeed increased with ExT, when this lasts 12 or more weeks (5). This finding has also been observed in healthy age individuals (4). With shorter (5–11 wk) ExT duration, max, but not a-vO2diff, is increased irrespective of age, according to 18 studies comprising nearly 200 individuals (4,5). Moreover, Poole et al. (7) seem to ignore solid human experimental proof demonstrating that the increase in V˙O2max with short-term (6 wk) ExT is abolished when the increase in max is manipulated to pretraining values—despite the presence of skeletal muscle adaptations, including increases in capillarization and mitochondrial content (3). Poole et al. (7) appear to perceive a-vO2diff as an equivalent of muscle O2 diffusing capacity (DO2m), which it is not. DO2m is computed as V˙O2 divided by the muscle O2 pressure gradient (9); thus, it is intrinsically determined by strict muscle O2 diffusing capacity and convective O2 delivery, which in turn is primarily a function of muscle blood flow. Accordingly, it cannot be discarded that DO2m increases with ExT are essentially the consequence of enhanced max leading to augmented muscle blood flow. In addition, our colleagues seem to rely on a theoretical model (9) based on untenable assumptions (6) to speculate that any potential increase in V˙O2max should, at least in part, be limited by skeletal muscle attributes determining DO2m. The collapse of their model is strikingly evident, considering the aforementioned unequivocal evidence (3) and experimental data gathered for a century in humans and other vertebrates (2). Furthermore, recent (again apparently unnoticed) catheterization studies demonstrate a twofold functional reserve in DO2m at V˙O2max in healthy untrained individuals (1). Ultimately, caution should be exercised before proposing therapeutic interventions only targeting skeletal muscle adaptations if attempting to improve V˙O2max substantially. David Montero Zurich Center for Integrative Human Physiology Institute of Physiology, University of Zurich Zurich, SWITZERLAND Candela Diaz-Cañestro Center for Molecular Cardiology University of Zurich Schlieren, SWITZERLAND Carsten Lundby Zurich Center for Integrative Human Physiology Institute of Physiology, University of Zurich Zurich, SWITZERLAND