Fig 1.
Time synchronized force and EMG data in CP-tests.
Results are shown for MLSSw - 10%, MLSSw, and MLSSw + 10% (A-D). (A) Correlation of mechanical parameters to MLSSw and MLSSw/kg (n = 14). Except for cadence (RPM), all parameters differ between workloads. Negative force is lower in group A. (B) Correlation of muscle activity to MLSSw and MLSSw/kg (n = 14). Only RF activity consistently differs between workloads and correlates to MLSSw and MLSSw/kg. The only exception is the VL at MLSSw + 10%. Note the negative correlation for ST and TA to MLSSw/kg at the same workload. (C) RF activity and correlation to MLSSw and MLSSw/kg at different workloads for group A and B (both groups n = 7). RF activity but also mean EMGs of all 4 muscles were consistently increased especially in the posterior part of the pedal cycle in group A compared to group B. (D) Correlation of frequency parameters (MPF) to IFEt, P%max, Pneg and EMG RF (n = 14). Note the pronounced MPF of RF in group A, which likely increases the quantitative amount of neural drive and, thus, the contractile force. P%max and Pneg are shown as mean of both cranks. IFE(t,a,p) is shown as mean of the right crank. For detailed description of data analysis and parameter definitions see Methods and S1 Fig. Black asterisks (*): two-tailed paired t-tests (A,B), yellow asterisks (*): two-tailed unpaired t-tests (A,C,D), red asterisks (*): correlation to MLSSw/kg (A,B,D), green asterisks (*): correlation to MLSSw (A,B), grey asterisks (*): correlation of MPF RF Burst to mechanical parameters and quantitative RF activity (D). Shown is mean. *P < 0.05, **P < 0.01, ***P < 0.001.
Fig 2.
Group comparisons of recruitment patterns of lower limbs muscles at the MLSSw in CP-test.
(A, B) Averaged bi-pedal crank force and EMG data across two pedal cycles (0–720°) at MLSSw–10% (yellow), MLSSw (red) and MLSSw + 10% (green) for group A (A) and group B (B) (both n = 7). Group comparisons reveal differences in force development, negative force level and recruitment patterns for RF, VL and TA (see gray shaded areas). Note also the vertical crossing points of the tangential force (Ft) and the delay at the TDC (Delay Ft crossing = DFC) at 180°/360°/540° in marked circles.
Fig 3.
Group comparisons of recruitment patterns and propulsive force around MLSSw in CP- test.
(A) Averaged bi-pedal crank force and EMG data across two pedal cycles (0–720°) of group A (green), B (black) and subject 9 (red). In CP-tests we recorded an earlier rise of EMG activity in Q3/Q4 and force development in Q1 in athletes with a higher MLSSw/kgbw. Note the more pronounced increase of RF and VL activity in group A and subject 9 at the transition from MLSSw to MLSSw +10%. (B) Averaged total propulsive force (Ft r+l) and EMG data across two pedal cycles (0–720°) at the MLSSw for group A (green, n = 7), group B (black, n = 7) and subject 9 (red). (B, second line) depicts curves for the absolute propulsive force. The early and strong rise of RF/VL activity prevents the loss of propulsive force (Ft r+l) at the TDC and increases the force development in Q1 (shaded red and green areas). Ftint of group A is 18% larger compared to group B. Subject 9 even had a 49% higher Ftint than group B (B, top row).
Fig 4.
Examples of RF/VL recruitment pattern and force development at the TDC for subjects at different performance levels in CP-tests.
(A) Averaged bi-pedal crank force and EMG data across two pedal cycles (0–720°) for athletes with a similar, high MLSSw (subject 9, red) and subject 11, black) and a lower MLSSw (subject 4, green). Note the vertical crossing points of the tangential force (Ft) and the delay at the TDC (Delay Ft crossing = DFC) at 180°/360°/540°. Although the DFCs of subjects 9 and 11 hardly differ (1.4°/2.4°), force development of subject 11 is clearly delayed in Q1. The high DFC of subject 4 (8.6°) results in an even more pronounced delay of force development. Note the double burst of the RF in subject 4. (B) Averaged total propulsive force (Ft r+l) and EMG data across two pedal cycles (0–720°) for subject 9 (red), 11 (black) and 4 (green). The relative force integral (Ftint) of subject 9 is larger than that of subject 11 (37%) and 4 (45%). Thus, with this pattern of recruitment, 37% and 45% more relative power is available for propulsion in subject 9, respectively (B, top row). (B, second line) depicts curves for the absolute propulsive force. At similar workloads (384W/366W), subject 9 has the highest force development in Q1 (red shaded areas). In contrast, subject 11 has the highest force development in Q2, during RF activity that causes hip flexion at the contralateral side of the pedal cycle (247°-354°, black shaded areas). Therefore, force distribution over a pedal revolution is more uniform in subject 9. Note also the different electromechanical delays (EMDs) and innervation onsets of the RF/VL of subjects 9, 11 and 4. Subject 9 innervates the RF from hip flexion to knee extension earlier and more intensely. Furthermore, the onset of innervation of the VL is later in VP 11 and, to an even higher extent, also in VP4. This interruption of muscle activity clearly shifts the onset of recruitment of knee extensors towards Q1 (see also Fig 6).
Fig 5.
Correlations of force and EMG parameters relevant for coordination and efficient force production in CP-tests (n=14).
These data show that the retrievable power at the MLSSw (P%max), the negative power (Pneg) and the mechanical effectiveness (IFEt) is determined by a strong and early activation of the RF in Q3/Q4. DFC, IFEt, Pneg and P%max are correlated to MLSSw/kgbw and are highly influenced by RF activity. Our data suggests that TA activation in the first half of the pedal cycle (Q1, see Figs 2 and 3) impedes cycling performance. Note that DFC is significantly reduced in group A and is inversely correlated to EMG and force/power parameters. Data were calculated for each pedal cycle. P%max and Pneg are shown as mean of both cranks. IFEt is shown as mean of the right crank. For detailed description of data analysis and parameter definitions see Methods and S1 Fig. Black asterisks (*): correlation to DFC, grey asterisks (*): correlation to P%max, yellow asterisks (*): correlation to IFEt, green asterisks (*): correlation to Pneg. Shown is mean. *P < 0.05, **P < 0.01, ***P < 0.001.
Fig 6.
Recruitment pattern of RF/VL at the transition from hip flexion to knee extension in CP-tests.
(A, B) Bi-pedal crank force (Ft right and Ft left), total propulsive force (Ft r+l), rate of force development (ΔFt), and averaged EMG across two pedal cycles at the MLSSw for subject 9 (red) (A) and for subject 9 vs highly trained (group A) and trained (group B) athletes (B). Vertical lines and circles (red = subject 9, green = group A, grey/black = group B) show reversal points of ΔFt (begin of force development during knee extension). Broken circles (red = subject 9, green = group A, black = group B) indicate the beginning of VL activation. Note the delayed VL activation (297°-320°) after termination of RF activity for hip flexion (234°-297°) in Q4 (A). Here, the RF likely operates together with the VL as a knee extensor. After onset of activation, the VL/RF-induced change in ΔFt (electromechanical delay, EMD) starts after 42 ms in subject 9, after 79 ms in group A and after 85ms in group B (B). A major difference between group A/subject 9 and group B is the more intense RF activation and the rapid onset of VL activity. Here, the pretension of the RF during hip flexion and subsequent strong and fast knee extension together with the other thigh extensors is the characteristic activation pattern of highly endurance trained cyclists (B). For detailed description of data analysis and parameter definitions see Methods and S1 Fig.
Fig 7.
Summary of the force/EMG pattern during MP-tests.
(A-D) Average tangential force and EMG data across two pedal cycles (0–720°) at different workloads: (i) for all subjects (subjects) (n = 14) (A, B), (ii) for all subjects vs subject 9 (C) and (iii) for all subjects vs group A/B (both n = 7) (D). The RF force/EMG curve differs from VL, TA and ST and reaches comparable activity only at high workloads (B). Simultaneously, the negative tangential force (Fneg) decreases with increasing workload (A) (see also Fig 1A). EMG activity of all subjects in the stages before the MLSSw is clearly higher than in subjects 9 (C). Between 39% max and 78% max, RF activity of subject 9 steeply increases in parallel with the average activity of all muscles. In contrast, RF activity across all subjects lags behind the mean EMG of all muscles up to high workload levels (C). RF activity increases clearly at the MLSSw in group A, which is not the case in group B and the overall sample (D). The other three recorded muscles (EMG m3) also tend to be more active in group A than in group B and the overall sample (D). For detailed description of data analysis and parameter definitions see Methods and S1 Fig. Shown is mean.
Fig 8.
Comparison of CP- and ILT-tests.
(A) Average normalized bi-pedal crank force and EMG data across two pedal cycles (0–720°) at the MLSSw for ILT-tests (black) vs CP-tests (red) (n = 11). Considering the methodological problems of comparing EMG data recorded on different days, activation of the four recorded lower limb muscles appeared to be earlier and more intense in CP-test. (B) Average normalized bi-pedal crank force, total average tangential force and average EMG data for test subject 9 across two pedal cycles (0–720°) at the MLSSw during ILT- (black) and CP-tests (red) (B, top row). (B, second line) depicts curves of absolute propulsive force (Newton). Due to the early and strong RF/VL activity during the CP test, the force drops less at TDC and rises more steeply in Q1 than in the ILT-test. Note the different electromechanical delays (EMDs) for RF and VL (for further explanations see Fig 6).
Fig 9.
Comparison of parameters relevant for coordination and effective force development in CP- and ILT-tests.
Time-based results for ILT- vs CP-tests reveal differences for RF and partially for ST, and TA activity as well as for the average EMG of all 4 recorded muscles (EMGm all) in the second part of the pedal cycle for all examined subjects (n = 11) and for subject 9. The delay of force crossings (DFCs) is inversely correlated to activation behavior of the RF which parallels results from CP-tests (see also Fig 7). Note that the correlation between IFEa and RF-activity between 180°-360° is missing in the ILT-test. P%max and Pneg are shown as mean of both cranks. IFE(t,a,p) is shown as mean of the right crank. For detailed description of data analysis and parameter definitions see Methods and S1 Fig. Grey asterisks (*): correlation to DFC, green asterisks (*): correlation to IFEa, black asterisks (*): two-tailed paired t-tests, yellow asterisks (*): two-tailed unpaired t-tests. Shown is mean. *P < 0.05, **P < 0.01, ***P < 0.001.
Fig 10.
Force, EMG and lactate kinetics around the lactate threshold in ILT-tests.
(A) Summary of ILT-tests (n = 11). Blood lactate concentration (mM), Pergo (Watt), rate of perceived exertion (RPE, 1-10) and heart rate (beats/min) for the final 6 increments of ILT-tests. A step-like blood lactate accumulation occurs at the individual lactate threshold (ILT, Δ[La−] = 0.7 mM) after a slight workload increment (ΔP = 6 Watt). Immediately after a comparable slight workload reduction (ΔP = 5 Watt), lactate accumulation slows down, indicating the maximal lactate steady state workload (MLSSw) [24]. (B) Averaged bi-pedal crank force and EMG data across two pedal cycles (0–720°) (n = 11) during three increments before and during two increments after the ILT. There are no visible differences in force and recruitment patterns. (C, D) Relationship of capillary lactate concentration [La−] to MLSSw/kg, mechanical and EMG parameters (C) and differences in lower limb muscle activity between group A and B (D) for the pre-ILT stage, at the ILT and for the post-ILT stage, calculated for each revolution. P%max is shown as mean of both cranks. IFEa is shown as mean of the right crank. For detailed description of data analysis and parameter definitions see Methods and S1 Fig. Red asterisk (*): correlation to MLSSw/kg (C), green asterisks (*): correlation to [La−] at ILT (C), grey asterisks (*): correlation to Δ[La−] (C), yellow asterisks (*): two-tailed unpaired t-tests (D). Shown is mean. *P < 0.05, **P < 0.01.