Share this article on:

Validity of Electromyographic Fatigue Threshold as a Noninvasive Method for Tracking Changes in Ventilatory Threshold in College-Aged Men

Kendall, Kristina L1; Smith, Abbie E1; Graef, Jennifer L1; Walter, Ashley A2; Moon, Jordan R1; Lockwood, Christopher M1; Beck, Travis W2; Cramer, Joel T2; Stout, Jeffrey R1

The Journal of Strength & Conditioning Research: January 2010 - Volume 24 - Issue 1 - p 109-113
doi: 10.1519/JSC.0b013e31819b79bc
Original Research

Kendall, KL, Smith, AE, Graef, JL, Walter, AA, Moon, JR, Lockwood, CM, Beck, TW, and Stout, JR. Validity of electromyographic fatigue threshold as a non-invasive method for tracking changes in ventilatory threshold in college-aged men. J Strength Cond Res 24(1): 109-113, 2010-The submaximal electromyographic fatigue threshold test (EMGFT) has been shown to be highly correlated to ventilatory threshold (VT) as determined from maximal graded exercise tests (GXTs). Recently, a prediction equation was developed using the EMGFT value to predict VT. The aim of this study, therefore, was to determine if this new equation could accurately track changes in VT after high-intensity interval training (HIIT). Eighteen recreationally trained men (mean ± SD; age 22.4 ± 3.2 years) performed a GXT to determine maximal oxygen consumption rate (o2peak) and VT using breath-by-breath spirometry. Participants also completed a discontinuous incremental cycle ergometer test to determine their EMGFT value. A total of four 2-minute work bouts were completed to obtain 15-second averages of the electromyographic amplitude. The resulting slopes from each successive work bout were used to calculate EMGFT. The EMGFT value from each participant was used to estimate VT from the recently developed equation. All participants trained 3 days a week for 6 weeks. Training consisted of 5 sets of 2-minute work bouts with 1 minute of rest in between. Repeated-measures analysis of variance indicated no significant difference between actual and predicted VT values after 3 weeks of training. However, there was a significant difference between the actual and predicted VT values after 6 weeks of training. These findings suggest that the EMGFT may be useful when tracking changes in VT after 3 weeks of HIIT in recreationally trained individuals. However, the use of EMGFT to predict VT does not seem to be valid for tracking changes after 6 weeks of HIIT. At this time, it is not recommended that EMGFT be used to predict and track changes in VT.

1Metabolic and Body Composition Lab; and 2Biophysics Lab, Department of Health and Exercise Science, Huston Huffman Center, University of Oklahoma, Norman, Oklahoma

Address correspondence to Jeffrey R. Stout,

Back to Top | Article Outline


Several investigations have used surface electromyography (EMG) to characterize the fatigue-induced increase in EMG amplitude and to identify the power output associated with the onset of neuromuscular fatigue during cycle ergometry. Previously, Matsumoto et al. (20) and Moritani et al. (22) proposed an incremental cycle ergometer test using fatigue curves to identify the maximal power output an individual can maintain without evidence of fatigue, described as the electromyographic fatigue threshold (EMGFT). The EMGFT test is an adaptation to original physical working capacity of deVries et al. (4) at the fatigue threshold test, using a supramaximal protocol. The EMGFT involves determining the rate of rise in electrical activity from the vastus lateralis during four 2-minute work bouts on a cycle ergometer, with varying power outputs. It has been suggested that the rise in electrical activity is a result of progressive recruitment of additional motor units or an increase in the firing frequency of motor units that have already been recruited.

Moritani et al. (22) suggested a strong physiological link between myoelectrical changes at fatigue and anaerobic threshold. Furthermore, the EMGFT method has been reported as a valid and reliable technique for examining the transition from aerobic to anaerobic metabolism during exercise (10,17,19). Identifying a reliable, noninvasive, and submaximal way to estimate the onset of fatigue has potential use in clinical populations, as well as serving as a training tool for athletes.

Chronic adaptations from high-intensity interval training (HIIT) have been shown to increase intramuscular buffering capacity and improve energy substrate utilization and ventilatory threshold (VT), leading to a delay in the onset of muscle fatigue, resulting in improvements in performance (13,16,18). In particular, VT has been used to identify an individual's level of aerobic fitness (30,33,34) and is demonstrated to be strongly related to the EMGFT (8,17,19,24). Recently, Graef et al. (7) developed an equation using the EMGFT to predict the VT [(W) = 0.665 (EMGFT) + 41.53]; however, no studies have determined if the VT prediction equation can accurately assess changes in VT after HIIT. We hypothesize that a noninvasive submaximal test in conjunction with the prediction equation will be able to accurately track changes in VT. Therefore, the purpose of this study was to validate the equation by Graef et al. (7), for tracking changes in VT after HIIT.

Back to Top | Article Outline


Experimental Approach to the Problem

The EMGFT has been widely used as a noninvasive method for determining the onset of muscle fatigue. Furthermore, it has been shown to occur at a similar power output as VT (12,31,35). The present study used a previously developed equation to track changes in VT from the EMGFT after an HIIT intervention. The predicted values were then statistically compared with observed VT values from a graded exercise test (GXT) on the cycle ergometer.

Back to Top | Article Outline


Eighteen recreationally trained (1-5 h·w 1 in one or a combination of the following: aerobic exercise [i.e., jogging, cycling, walking], resistance training, and recreational sports) college-aged men (mean ± SD-age: 22.4 ± 3.2 years; height: 178.4 ± 6.2 cm; weight: 78.5 ± 11.3 kg; pre o2peak: 36.75 ± 8.10 mL·kg 1·min 1) volunteered to participate in this study. All procedures were approved by the University of Oklahoma Institutional Review Board for Human Subjects, and written informed consent was obtained from each participant before any testing.

Back to Top | Article Outline


Determination of o2peak and Observed Ventilatory Threshold

Participants performed a continuous GXT on an electronically braked cycle ergometer (Corival 400; Lode, Groningen, The Netherlands) to determine maximal oxygen consumption rate (o2peak) and VT. After a 5-minute warm-up at 50 W, the workload increased 25 W every 2 minutes until the participant was unable to maintain 70 rpm or until volitional fatigue.

Ventilatory threshold was previously described by Matsumoto et al. (20) as the relationship between minute ventilation (E) and oxygen consumption (o2). To determine VT, regressed lines were fit to the lower and upper portions of the E vs. o2 curve, before and after the break points, respectively. The intersection of these 2 lines was defined as VT. Test-retest reliability for the VT protocol, previously reported by Amann et al. (1), resulted in an intraclass correlation coefficient (ICC) of 0.95 (SEM: 14.4 W).

Back to Top | Article Outline

Gas Exchange Analysis

Open-circuit spirometry was used to analyze the gas exchange data using the ParvoMedics TrueOne 2400 Metabolic Measurement System (ParvoMedics, Sandy, UT). Oxygen and carbon dioxide were analyzed through a sampling line after the gases passed through a heated pneumotach and mixing chamber. The data were averaged over 15-second intervals. The highest average o2 value during the GXT was recorded as the o2peak if it coincided with at least 2 of the following criteria: (a) a plateau in heart rate or heart rate values within 10% of the age-predicted HRmax, (b) a plateau in o2 (defined by an increase of no more than 150 mL·min 1), or (c) a respiratory exchange rate value greater than 1.15 (22).

Back to Top | Article Outline


Pre-gelled bipolar (2.54 cm center to center) surface electrodes (Ag-Ag Cl, Quinton Quick Prep; Quinton Instruments Co., Bothell, WA) were placed over the lateral portion of the vastus lateralis muscle, midway between the greater trochanter and the lateral condyle of the femur. A reference electrode was placed over the seventh cervical vertebra. The raw EMG signals were pre-amplified (gain × 1,000) (EMG 100C; Biopac Systems, Inc., Santa Barbara, CA), sampled at 1,000 Hz, and band-pass filtered from 10 to 500 Hz (zero-lag eighth-order Butterworth filter). All EMG amplitude values were stored on a personal computer (Dell Inspiron 8200; Dell, Inc., Round Rock, TX) and analyzed off-line using custom-written software (LabVIEW v. 7.1; National Instruments, Austin, TX).

Back to Top | Article Outline

Determination of the EMGFT

After a 5-minute warm-up on an electronically braked cycle ergometer (Quinton Corival 400), participants completed four 2-minute cycling bouts at incrementally ascending workloads (75-300 W). The initial workload corresponded with the workload at which VT occurred, determined during the GXT. Adequate rest was given between bouts to allow for participants' heart rate to drop within 10 beats of their resting heart rate. The rates of rise in EMG amplitude values (EMG slope) from the 4 workloads were plotted over 120 seconds. The EMG slope values for each of the 4 power outputs were then plotted to determine EMGFT. The line of best fit was extrapolated to the y-axis, and the power output at which it intersected the y-axis was defined as the EMGFT. The participants completed the EMGFT test 4 times: familiarization trial, baseline, mid-, and post-training.

Test-retest reliability for the EMGFT test, previously determined at the University of Oklahoma, resulted in an ICC of 0.935 (SEM: 5.03 W). The ICC from this laboratory was higher than previously reported using the vastus lateralis (ICC = 0.65) (4).

Back to Top | Article Outline

High-Intensity Interval Training

Participants were required to visit the lab 3 nonconsecutive days per week for 6 weeks to perform the HIIT. To account for neural vs. metabolic adaptations, subjects trained for 3 weeks, retested, and then trained for an additional 3 weeks. Subjects trained 3 days a week, with each training session lasting approximately 30 minutes. After a 5-minute warm-up (50 W), participants completed 5 or six 2-minute work bouts at a predetermined percentage of their maximum o2peak power output, resting 1 minute in between exercise bouts (Figure 1). The dates and times of training were recorded for each subject to ensure training compliance. No injuries were recorded during the study, possibly due to the low-impact nature of the training protocol.

Figure 1

Figure 1

Back to Top | Article Outline

Statistical Analyses

A previously developed equation by Graef et al. (7) [(W) = 0.665 (EMGFT) + 41.53] was used to predict VT from the EMGFT test. A one-way repeated-measures analysis of variance (ANOVA) was used to compare the observed and predicted VT values after both 3 and 6 weeks of HIIT. SPSS software (Version 14.0; SPSS, Chicago, IL) was used for all statistical comparisons. The α-level was set at p ≤ 0.05 to determine statistical significance.

Back to Top | Article Outline


Table 1 presents the mean power outputs (W) and delta scores for both observed and predicted VTs after 3 (mid) and 6 (post) weeks of training. Repeated-measures ANOVA indicated no significant difference (p = 0.607) between observed and predicted power outputs at VT after 3 weeks of HIIT. However, there was a significant difference (p = 0.014) between the observed and predicted VT values after 6 weeks of training.

Table 1

Table 1

Back to Top | Article Outline


The use of the EMGFT has been accepted as a reliable and noninvasive method for identifying the onset of neuromuscular fatigue (3,8,10,17,19,26,27). Furthermore, several studies have demonstrated a significant relationship between power output values at EMGFT and VT (10,17,35). However, the present research is the first to use an equation (7), determined from individual EMGFT values, to assess changes in VT after a period of HIIT. The results from the study suggest the use of EMGFT to accurately predict changes in VT after 3 weeks of HIIT in recreationally trained individuals. However, the use of EMGFT to predict VT does not seem to be valid for tracking changes after 6 weeks of HIIT.

High-intensity interval training can broadly be defined as repeated sessions of brief intermittent exercise bouts at intensities close to o2peak (≥90% of o2peak) (6). The purpose of HIIT is to repeatedly stress the body, physiologically, resulting in chronic adaptations and improving metabolic efficiency. During a high-intensity bout of exercise, the buffering capacity of muscle fibers may be a factor in reducing the accumulation of protons. This accumulation can lead to elevated levels of muscle and blood lactate concentrations, which contributes to muscle fatigue (14,15). Lactate formation has been shown to have a cause-and-effect relationship with VT; an increase in lactate buffering by the bicarbonate system leads to an increase in carbon dioxide production and ultimately ventilation rate (2). In addition, HIIT has been used as a method to improve muscle buffering capacity by increasing the oxidative capacity of muscle fibers, a mechanism that may improve VT (16). Specifically, Laursen et al. reported that 4 weeks of HIIT significantly improved VT, suggesting that the improvement in performance was due to peripheral skeletal muscle adaptations enhancing muscle oxidative capacity. In agreement, Burke et al. (2) and Poole et al. (28) demonstrated significant increases in VT after an HIIT program of 7 and 8 weeks, respectively. Along with increases in ventilation, the accumulation of hydrogen ions (H+) and lactate during high-intensity exercise has been shown to impair muscle contractility, making it necessary to recruit additional motor units to maintain the same power output (21,23,32). One way to measure this change is through surface EMG, quantifying the level of activation of working muscles, and possibly detecting signs of muscle fatigue (12). Due to this relationship, EMGFT has been suggested as a noninvasive method for determining VT (10,11,17).

During a high-intensity exercise bout, as the workload increases, the accumulation of lactate and H+ may impair muscle contractility. It becomes necessary to recruit additional motor units to compensate for poor contractility, which may lead to an increase in EMG amplitude (5,31). The EMGFT test has previously been used as a reliable technique for identifying neuromuscular fatigue during submaximal cycle ergometry by characterizing an increase in EMG amplitude of the working muscle over time (10,17,19). The EMGFT theoretically represents an estimate of the highest power output that can be sustained without signs of fatigue (20). Furthermore, GXTs, which require subjects to exercise at increasing workloads until exhaustion, are often employed as a way to measure VT. However, several studies have demonstrated that the EMGFT and VT occur at the same power output (29,35), suggesting the use of EMGFT as an alternative method for estimating VT that does not require maximal effort or the use of gas exchange measurement equipment (12,20,24). Adding to this argument, Lucia et al. (17) observed no significant differences when testing elite cyclists, between the exercise intensity corresponding to the EMGFT and the intensity at VT.

Although it has been well documented that VT and EMGFT occur at similar power outputs during cycle ergometry (4,9,19,20,25), the results from the present study are the first to suggest that an equation used to predict VT from EMGFT, developed by Graef et al. (7), may accurately track changes in VT after short-term HIIT. When using the equation [VT (W) = 0.665 (EMGFT) + 41.53] with an HIIT intervention, no significant differences were found between VT values measured using a metabolic cart and VT values derived from the EMGFT equation after baseline testing and after 3 weeks of HIIT training. However, after 6 weeks of training, the equation proved to be less accurate, resulting in significantly different VT values (Table 1). This could be due to the fact that the equation was developed using the EMGFT method, which reflects a neuromuscular threshold of fatigue rather than a metabolic (VT) threshold of fatigue. Our data suggest that the estimated VT from the EMGFT may have detected a neuromuscular adaptation during the first 3 weeks of HIIT training rather than a metabolic change. Therefore, the EMGFT estimate of VT may be unable to accurately detect significant metabolic changes that occurred during the second 3 weeks of training. Future studies are warranted to determine whether using an aerobically trained population vs. untrained in the Graef et al. (7) study, to develop an equation may more accurately track changes in VT after training.

Back to Top | Article Outline

Practical Applications

In summary, the results of this study suggest that the EMGFT test may be an attractive alternative for estimating and tracking changes in VT after short-term HIIT training in previously untrained subjects. The use of the EMGFT may be a more advantageous tool for determining VT for a couple of reasons: (a) most fitness professionals do not have access to expensive equipment (i.e., metabolic cart) necessary for measuring VT and (b) the use of the submaximal EMGFT test may be more valuable in a clinical setting, where maximal exertion is often not recommended and leads to severe discomfort. However, the use of EMGFT was not accurate at tracking VT changes after 6 weeks of HIIT. Therefore, it is not recommended that EMGFT be used to track changes in VT after 6 weeks of intense interval training.

Back to Top | Article Outline


The authors declare that they have no competing interests. The results of the present study do not constitute endorsement of the product by the authors or the National Strength and Conditioning Association.

Back to Top | Article Outline


1. Amann, M, Subudhi, AW, Walker, J, Eisenman, P, Shultz, B, and Foster, C. An evaluation of the predictive validity and reliability of ventilatory threshold. Med Sci Sports Exerc 36: 1716-1722, 2004.
2. Burke, J, Thayer, R, and Belcamino, M. Comparison of effects of two interval-training programmes on lactate and ventilatory thresholds. Br J Sports Med 28: 18-21, 1994.
3. Day, JR, Rossiter, HB, Coats, EM, Skasick, A, and Whipp, BJ. The maximally attainable VO2 during exercise in humans: The peak vs. maximum issue. J Appl Physiol 95: 1901-1907, 2003.
4. deVries, HA, Moritani, T, Nagata, A, and Magnussen, K. The relation between critical power and neuromuscular fatigue as estimated from electromyographic data. Ergonomics 25: 783-791, 1982.
5. Enoka, RM and Stuart, DG. Neurobiology of muscle fatigue. J Appl Physiol 72: 1631-1648, 1992.
6. Gibala, MJ and McGee, SL. Metabolic adaptations to short-term high-intensity interval training: A little pain for a lot of gain? Exerc Sport Sci Rev 36: 58-63, 2008.
7. Graef, JL, Smith, AE, Kendall, KL, Walter, AA, Moon, JR, Lockwood, CM, Beck, TW, Cramer, JT, and Stout, JR. The relationships among endurance performance measures as estimated from VO2PEAK, ventilatory threshold, and electromyographic fatigue threshold: A relationship design. Dyn Med 7: 15, 2008.
8. Hanon, C, Thepaut-Mathieu, C, Hausswirth, C, and Le Chevalier, JM. Electromyogram as an indicator of neuromuscular fatigue during incremental exercise. Eur J Appl Physiol Occup Physiol 78: 315-323, 1998.
9. Helal, JN, Guezennec, CY, and Goubel, F. The aerobic-anaerobic transition: Re-examination of the threshold concept including an electromyographic approach. Eur J Appl Physiol Occup Physiol 56: 643-649, 1987.
10. Hug, F, Faucher, M, Kipson, N, and Jammes, Y. EMG signs of neuromuscular fatigue related to the ventilatory threshold during cycling exercise. Clin Physiol Funct Imaging 23: 208-214, 2003.
11. Hug, F, Laplaud, D, Lucia, A, and Grelot, L. EMG threshold determination in eight lower limb muscles during cycling exercise: A pilot study. Int J Sports Med 27: 456-462, 2006.
12. Hug, F, Laplaud, D, Savin, B, and Grelot, L. Occurrence of electromyographic and ventilatory thresholds in professional road cyclists. Eur J Appl Physiol 90: 643-646, 2003.
13. Jenkins, DG and Quigley, BM. The influence of high-intensity exercise training on the Wlim-Tlim relationship. Med Sci Sports Exerc 25: 275-282, 1993.
14. Karlsson, J, Bonde-Petersen, F, Henriksson, J, and Knuttgen, HG. Effects of previous exercise with arms or legs on metabolism and performance in exhaustive exercise. J Appl Physiol 38: 763-767, 1975.
15. Klausen, K, Knuttgen, HG, and Forster, HV. Effect of pre-existing high blood lactate concentration on maximal exercise performance. Scand J Clin Lab Invest 30: 415-419, 1972.
16. Laursen, PB, Shing, CM, Peake, JM, Coombes, JS, and Jenkins, DG. Influence of high-intensity interval training on adaptations in well-trained cyclists. J Strength Cond Res 19: 527-533, 2005.
17. Lucia, A, Sanchez, O, Carvajal, A, and Chicharro, JL. Analysis of the aerobic-anaerobic transition in elite cyclists during incremental exercise with the use of electromyography. Br J Sports Med 33: 178-185, 1999.
18. MacDougall, JD, Hicks, AL, MacDonald, JR, McKelvie, RS, Green, HJ, and Smith, KM. Muscle performance and enzymatic adaptations to sprint interval training. J Appl Physiol 84: 2138-2142, 1998.
19. Maestu, J, Cicchella, A, Purge, P, Ruosi, S, Jurimae, J, and Jurimae, T. Electromyographic and neuromuscular fatigue thresholds as concepts of fatigue. J Strength Cond Res 20: 824-828, 2006.
20. Matsumoto, T, Ito, K, and Moritani, T. The relationship between anaerobic threshold and electromyographic fatigue threshold in college women. Eur J Appl Physiol Occup Physiol 63: 1-5, 1991.
21. Moritani, T and deVries, HA. Reexamination of the relationship between the surface integrated electromyogram (IEMG) and force of isometric contraction. Am J Phys Med 57: 263-277, 1978.
22. Moritani, T, Takaishi, T, and Matsumoto, T. Determination of maximal power output at neuromuscular fatigue threshold. J Appl Physiol 74: 1729-1734,1993.
23. Mortimer, JT, Magnusson, R, and Petersen, I. Conduction velocity in ischemic muscle: Effect on EMG frequency spectrum. Am J Physiol 219: 1324-1329, 1970.
24. Nagata, A, Muro, M, and Moritani, T. Anaerobic threshold determination by blood lactate and myoelectric signals. Jpn J Physiol 31: 585-597, 1981.
25. Nagata, A, Muro, M, Moritani, T, and Yoshida, T. Anaerobic threshold determination by blood lactate and myoelectric signals. Jpn J Physiol 31:585-597, 1981.
26. Orr, GW, Green, HJ, Hughson, RL, and Bennett, GW. A computer linear regression model to determine ventilatory anaerobic threshold. J Appl Physiol 52: 1349-1352, 1982.
27. Pavlat, DJ, Housh, TJ, Johnson, GO, Schmidt, RJ, and Eckerson, JM. An examination of the electromyographic fatigue threshold test. Eur J Appl Physiol Occup Physiol 67: 305-308, 1993.
28. Poole, DC and Gaesser, GA. Response of ventilatory and lactate thresholds to continuous and interval training. J Appl Physiol 58: 1115-1121, 1985.
29. Sihvonen, T, Airaksinen, O, and Hanninen, O. Electrical behavior (rms-EMG) of frontalis muscle during incremental exercise. Acta Physiol Scand 134: 140, 1988.
30. Tanaka, K and Matsuura, Y. Marathon performance, anaerobic threshold, and onset of blood lactate accumulation. J Appl Physiol 57: 640-643, 1984.
31. Taylor, AD, Bronks, R, and Bryant, AL. The relationship between electromyography and work intensity revisited: A brief review with references to lacticacidosis and hyperammonia. Electromyogr Clin Neurophysiol 37: 387-398, 1997.
32. Tesch, PA, Komi, PV, Jacobs, I, Karlsson, J, and Viitasalo, JT. Influence of lactate accumulation of EMG frequency spectrum during repeated concentric contractions. Acta Physiol Scand 119: 61-67, 1983.
33. Tesiorowski, AM, Harris, M, Chan, KJ, Thompson, CR, and Montaner, JS. Anaerobic threshold and random venous lactate levels among HIV-positive patients on antiretroviral therapy. J Acquir Immune Defic Syndr 31: 250-251, 2002.
34. Thin, AG, Linnane, SJ, McKone, EF, Freaney, R, FitzGerald, MX, Gallagher, CG, and McLoughlin, P. Use of the gas exchange threshold to noninvasively determine the lactate threshold in patients with cystic fibrosis. Chest 121: 1761-1770,2002.
35. Viitasalo, JT, Luhtanen, P, Rahkila, P, and Rusko, H. Electromyographic activity related to aerobic and anaerobic threshold in ergometer bicycling. Acta Physiol Scand 124: 287-293, 1985.

cardiorespiratory fitness; muscle fatigue; interval training; cycling

© 2010 National Strength and Conditioning Association