Daniel A. Keir

So, YLI, Faricier, R, Guluzade, NA, Belfry, GR, Murias, JM, and Keir, DA. The respiratory compensation point predicts 19.2-km cycling performance in trained male cyclists. J Strength Cond Res XX(X): 000–000, 2026—This study tested the hypothesis that the step-ramp-step (SRS)-corrected cycling power output (PO) at the respiratory compensation point (RCP) predicts cycling time-trial performance and explored the independent and collective contributions to trial time of 4 physiological variables derived from single or repeated SRS protocols: peak oxygen uptake (V̇ o ₂ peak), cycling economy, RCP, and fatigue resistance. Twenty-eight trained male cyclists completed an SRS protocol before and after a 19.2-km virtual time trial (∼30 minutes) on their personal road bikes using a smart trainer. V̇ o ₂ peak, cycling economy, and SRS-corrected PO at RCP were derived from the pre-time-trial SRS protocol. Fatigue resistance was quantified as the change in peak ramp PO from pre- to post-time trial. Independent and multilinear regression evaluated the predictive strength of each variable. Relative SRS-corrected RCP (W·kg ⁻¹ ) showed the strongest independent correlation with time-trial completion time ( r = −0.61, p < 0.001). Relative V̇ o ₂ peak (mL·kg ⁻¹ ·min ⁻¹ ) and cycling economy (%) were also moderately associated with time-trial performance ( r = −0.42, p = 0.03; and r = −0.55, p < 0.001, respectively), whereas fatigue resistance was not significantly correlated. In the multilinear regression analysis, SRS-corrected RCP uniquely explained 30.9% of the model's total predictive variance ( r = 0.68, p < 0.01) followed by cycling economy (21.3%), ΔPO _peak (2.1%), and absolute V̇ o ₂ peak (3.3%). Among trained male cyclists, relative PO at SRS-corrected RCP is a strong and independent predictor of ∼30-minute time-trial performance. The SRS protocol provides a practical, single-session method to assess key physiological determinants of endurance performance, with valuable applications for training prescription and athlete monitoring.

Stable breathing depends on the integrated action of central and peripheral chemoreceptors, but the nature of this interaction in humans and its physiological importance remain controversial. Previous studies reported additive, hypo‐additive and hyper‐additive responses across species, preparations and protocols. Here, we used a novel method to quantify peripheral chemoreflex sensitivity across graded levels of central chemoreceptor activation to test whether central–peripheral chemoreceptor interaction in awake humans is fixed or state dependent. At low altitude, the form of interaction varied across individuals, with most showing additive responses (12/17) and a substantial subset showing hypo‐additivity (5/17). After 7–10 days of acclimatization to 3500 m, all five hypo‐additive individuals became additive, and four additive participants became hyper‐additive. This shift in interaction towards hyper‐additivity provides the first evidence that central‐peripheral chemoreceptor integration is plastic. Plasticity during wakefulness was not associated with individual differences in central sleep apnoea severity at high altitude. Thus, our findings confirm that central–peripheral chemoreceptor interaction differs among individuals and suggest that the form of interaction can be reshaped by hypobaric hypoxic stress. image Key points: How central and peripheral chemoreceptor drives to breathe interact in humans and whether this process is plastic remains uncertain. We quantified peripheral chemoreflex sensitivity to hypoxia (PChS) across graded levels of central chemoreceptor activation at low altitude and after 7–10 days of high‐altitude exposure (3500 m). Seventeen participants performed repeated modified rebreathing tests with end‐tidal fixed at 150 (hyperoxia), 60 and 45 mmHg (hypoxia). PChS was computed at 1 mmHg intervals of end‐tidal () from linear regression of the difference in ventilation between hypoxic and hyperoxic conditions vs . calculated arterial oxygen saturation (ScO ₂ ) (L·min ⁻¹ ·mmHg ⁻¹ ·%ScO ₂ ⁻¹ ). At low altitude, PChS increased with in all individuals; the relationship was linear (no interaction) in 12 participants, and 5 showed hypo‐additivity. After acclimatization, central and peripheral chemoreflex sensitivities increased. All hypo‐additive individuals became additive; four additive individuals became hyper‐additive. Central–peripheral chemoreceptor interaction is plastic in humans and can shift towards hyper‐additivity with sustained hypobaric hypoxia.

How ventilation () matches gas exchange () while maintaining isocapnia during exercise remains unresolved. We measured , , arterialized‐venous (PāCO ₂ ) and hypercapnic ventilatory responsiveness (HCVR = ∆/∆) during four steady‐state exercise intensities below the respiratory compensation point (RCP) and used a mathematical model to characterize feedforward contributions to exercise hyperpnoea and test the hypothesis that CO ₂ chemofeedback is inhibitory. Model assumptions were that the ventilatory response to exercise‐intensity changes (G _SYS = ∆/∆) equals the sum of a feedforward exercise stimulus proportional to (G _EX ) modified by the CO ₂ chemoreflex (HCVR). Using experimental data G _SYS was quantified as ∆/∆ between exercise steps and G _EX equalled G _SYS minus chemoreflex feedback (i.e. HCVRx∆PāCO ₂ /∆). Twenty healthy young participants completed a maximal cycling test and three submaximal step‐incremental protocols of two 8 min stages below the estimated lactate threshold (θ _LT ; MOD ₁  and MOD ₂ ) and two 12 min stages between θ _LT and RCP (HVY ₁ and HVY ₂ ). During visits 2 and 3, three rebreathing tests performed at each intensity provided HCVR (average –end‐tidal slope). During visit 4, heated hand vein blood provided PāCO ₂ at each intensity, and and were measured by metabolic cart. With higher steady‐state , one‐way repeated‐measures ANOVA showed that HCVR rose from 2.4 ± 0.9 at MOD ₁ to 2.8 ± 0.8, 3.2 ± 0.8 and 3.6 ± 1.0 L · min ⁻¹ · mmHg ⁻¹ at MOD ₂ , HVY ₁ and HVY ₂ , respectively; ∆PāCO ₂ /∆ decreased ( P < 0.001) from 0.2 ± 4.3 at MOD ₁ to −1.9 ± 5.0, −5.5 ± 4.5 and −8.0 ± 4.6 mmHg · L · min ⁻¹ ; and G _EX increased ( P < 0.001) from 26.9 ± 15.1 at MOD ₁ to 34.5 ± 20.3, 53.9 ± 19.1 and 73.4 ± 27.5 L · min ⁻¹ . Exercise hyperpnoea may reflect the balance of intensity‐dependent feedforward and inhibitory CO ₂ chemofeedback drive. image Key points: We aimed to quantify exercise intensity‐dependent changes in CO ₂ chemoreflexes and interpret their role in exercise hyperpnoea. Ventilatory, gas exchange, arterialized‐venous (PāCO ₂ ) and hypercapnic ventilatory responses (HCVR) were measured during successive work‐to‐work cycling exercise transitions from low‐to‐moderate to heavy‐intensity work rates. The measured HCVR at each intensity and ∆/∆ and ∆PāCO ₂ /∆ from one work rate to the next were included in a mathematical model to characterize feedforward contributions to exercise hyperpnoea at four intensities. With increasing steady‐state , ∆/∆ and HCVR progressively increased, whereas ∆PāCO ₂ /∆ progressively became more negative. Assuming ∆ is the sum of a feedforward gain plus linear CO ₂ chemofeedback gain responding to changes in and from rest to exercise, respectively, we estimated that feedforward gain rises with increasing exercise intensities, attaining a drive that exceeds requirements for regulation at its resting level. Exercise hyperpnoea, therefore, requires a feedforward exercise stimulus constrained by inhibitory chemofeedback from rest to exercise to minimize hypocapnia .

Objective. We tested the hypothesis that the oxygen uptake ( V ˙ O ₂ ) associated with systematic changes in skin temperature ( T _sk ) during ramp-incremental exercise would coincide with the respiratory compensation point (RCP) and not the gas exchange threshold (GET). Approach. Thirty-eight healthy adults (16 females; 24 ± 5 years, body mass 73.0 ± 11.4 kg, height 174.8 ± 8.5 cm) completed a ramp-incremental exercise test on a cycle ergometer (20–45 W min ⁻¹ depending on self-reported aerobic fitness; selected to induce task disengagement within 10–15 min) in a climate-controlled laboratory. Throughout the protocol, ventilatory and gas exchange variables were acquired breath-by-breath using a metabolic cart and T _sk of the posterior calf was recorded revolution-by-revolution using an infrared camera. Main results. The T _sk -time relationship exhibited either an abrupt or gradual deceleration as exercise time progressed. The V ˙ O ₂ associated with this change-point was identified visually or by finding the local maximum of the second derivative of a 6th-order polynomial fit to the data. The V ˙ O ₂ at GET and RCP were determined visually. Overall, the V ˙ O ₂ at GET, RCP and T _sk breakpoint identified visually or modelled were 1884 ± 466, 2633 ± 747, 1969 ± 462, and 1769 ± 508 ml∙min ⁻¹ , respectively. Compared to the V ˙ O ₂ at GET, higher values were reported when T _sk breakpoint was identified visually ( p = 0.04, ESg = 0.3, 95%CI[1, 170 ml min ⁻¹ ]) and lower by the second derivative ( p < 0.01, ESg = 0.4, 95%CI[30, 199 ml min ⁻¹ ]). Larger effect sizes were observed for RCP (visually: p < 0.001, ESg = 1.3, 95%CI[524, 803 ml min ⁻¹ ]; second derivative: p < 0.001, ESg = 1.7, 95%CI[724, 1003 ml min ⁻¹ ]). Intraclass correlation of V ˙ O ₂ between T _sk and GET ranged from very good-to-excellent (0.67–0.92) but were not significant for RCP. For GET, mean absolute percentage error was 12.5% (95%CI: 9.4%–15.7%) and 14.5% (95%CI: 11.1%–17.9%) for the visual and second-derivative methods, respectively, and for RCP, 23.2% (95% CI: 18.6%–27.8%) and 31.4% (95% CI: 26.7%–36.2%). Significance. Findings suggest a statistical association between intensity-dependent changes in metabolism and the T _sk response of the calf.

Purpose : This study investigated the effect of treadmill incline on wearable-derived running power output (rPO) at the lactate threshold (LT) and lactate turnpoint (LTP). Methods : Thirteen trained runners (5 females; 23 [5] y) completed 3 rPO-matched, step-incremental treadmill tests to exhaustion at fixed inclines of 0%, 5%, and 10%. Running speed was adjusted to maintain consistent rPO across conditions using standardized calculations. During each test, breath-by-breath oxygen uptake ( ) and rPO (via Stryd power meter) were continuously measured. Blood lactate concentration was measured during the final 30 seconds of each 3-minute stage. The LT and LTP were identified from the blood lactate concentration versus rPO using 3-segment and modified-poly-D _max fitting models, respectively. A repeated-measures analysis of variance assessed between-condition differences in LT and LTP ( P < .05). Results : The –rPO relationship was not different across incline conditions ( P = .147). The rPO at LT did not differ between inclines ( P = .454): 259 (49) W, 252 (43) W, and 249 (56) W for 0%, 5%, and 10%, respectively, nor did rPO at LTP ( P = .604): 309 (58) W, 316 (59) W, and 312 (60) W for 0%, 5%, and 10%, respectively. The at LT ( P = .727) and LTP ( P = .470) also showed no significant differences across conditions. Conclusion : The rPO at LT and LTP in highly trained runners are consistent across treadmill inclines between 0% and 10%. These findings support the use of rPO as a metric for monitoring exercise intensity across different gradient conditions in both training and performance settings.

Peripheral or central chemoreceptor sensitization is a well-documented cause of reflex sympathetic excitation in patients with heart failure and reduced ejection fraction (HFrEF).¹ In one series, augmented ventilatory responsiveness to hypercapnia and hypoxia, present in ∼half the patient cohort, predicted a 4-year survival of 49%.² The carotid body has been identified as a target for surgical or percutaneous intervention with the aim of benefiting such individuals by restoring autonomic equilibrium.^3,4 Importantly, this strategy is predicated on the assumption that transient intermittent hypoxic tests can accurately characterize a sensitized peripheral chemoreflex, with the conjecture that the slope or sensitivity of the efferent sympathetic response then parallels that of ventilation. However, chemoreceptor-mediated regulation of breathing and autonomic outflow engage different brainstem circuits.⁵ To date, selection of patients for peripheral chemoreflex intervention has involved quantification of ventilatory responses to intermittent hypoxia³: 100% nitrogen is inhaled over a short duration (∼2–8 breaths) to achieve a range of oxygen saturations (70%–100% S_aO₂). The peak of ventilation (V̇_E) elicited by each exposure is inversely related to the prior nadir in S_aO₂, generating a V̇_E–S_aO₂ slope for peripheral chemoreflex sensitivity with units of L·min⁻¹·%S_aO₂⁻¹. This measure is often called the ‘hypoxic ventilatory response’ (HVR).⁶ In patients with HFrEF, a value ≥0.75 L·min⁻¹·%S_aO₂⁻¹ is considered an ‘exaggerated’ HVR and used as an indication for carotid body intervention.⁷

We assessed peripheral chemoreflex (PChR) contributions to steady-state breathing at rest and at four exercise intensities below the respiratory compensation point by quantifying the fall in ventilation in response to surreptitious oxygen breathing. The magnitude by which ventilation fell increased progressively from rest to the highest intensity, indicating a rising PChR drive. Proton concentration from arterialized-venous blood also increased with intensity. Intensity-dependent increases in PChR drive may reflect a reflexive response to rising proton concentration.

Leg muscle sympathetic nerve activity (MSNA) diminishes in healthy (HC) individuals during mild dynamic exercise but not in age‐matched patients with heart failure due to reduced ejection fraction (HFrEF). To elucidate the neural mechanisms responsible for such sympathetic excitation, we studied 20 stable HFrEF patients (6F; mean age 62 ± 8 SD years) and 15 age‐matched HC (6F; mean age 59 ± 7). We quantified peak oxygen uptake () and separately, fibular MSNA (microneurography) at rest and during one‐leg cycling (2 min each, mild (unloaded) and moderate intensity (loaded = 30–40% )) throughout three interventions: (1) post‐exercise circulatory occlusion (PECO), which isolates the leg muscle metaboreflex (MMR); (2) supine posture, which stimulates cardiopulmonary baroreceptors (CPB); and (3) 32% inspired oxygen, to supress the peripheral chemoreflex (PC). One‐leg cycling increased MSNA and activated the leg MMR in patients with HFrEF but not HC. MSNA at rest and during mild exercise was lower when supine than seated in both cohorts. Breathing 32% oxygen lowered the MSNA of HC but not HFrEF. In both groups, hyperoxia decreased burst frequency during low‐intensity cycling. Hyperoxia abolished the ‘paradoxical’ sympatho‐excitation of HFrEF. Thirteen participants with HFrEF were reassessed after 4 months of conventional cardiopulmonary rehabilitation. Exercise training improved by 17% and attenuated the leg MMR response without altering CPB activation or PC suppression. We conclude that in HFrEF, all three autonomic reflexes are engaged to a varying degree by one‐leg cycling. Patient training attenuates the leg MMR without affecting CPB or PC modulation of MSNA during exercise. image Key points: In HFrEF patients, an exaggerated leg MMR is the dominant sympatho‐excitatory reflex during one‐leg cycling at moderate work rates; with their MSNA response relating inversely to . Activation of the cardiopulmonary baroreflex and peripheral chemoreflex by exercise also contribute, suggesting that exercising supine or while breathing 32% O₂ may complement conventional training protocols. An exercise‐based cardiac rehabilitation programme lowers sympathetic discharge at rest and during mild intensity cycling by abolishing specifically the leg MMR response.