Resistance machines are used abundantly by exercisers as a means to improve neuromuscular function. Whereas the majority of manufacturers produce constant resistance machines, some machines provide variable resistance. Machines providing constant resistance maintain the lever arm distance throughout the range of motion, whereas those providing variable resistance manipulate the lever arm distance to provide greater resistance during certain phases of the range of motion. For example, the use of a CAM system has been used to manipulate the distance of the lever arm in resistance machines (16). Variable resistance devices aim to emulate the varying force-joint angle curves of human muscle and/or groups of muscles (6,24) and by doing so intensively activate the muscle(s) throughout the range of motion (16).
During a multijoint leg press exercise, a linearly increasing strength curve is observed from a low knee angle to a maximum force close to full knee extension (6,24). Constant resistance during a dynamic action (e.g., concentric leg extension) would not fully stress the capabilities of the neuromuscular system during this exercise because the resistance is only near maximal at low knee angles. Consequently, if the resistance machines could match the human torque capabilities demonstrated by the force-joint angle curve, perhaps a more optimal training stimulus could be achieved compared with constant resistance.
Several studies have investigated the effectiveness of different resistance machines in matching the force-joint angle curves of humans during various exercises (2,9,19,28). These studies have highlighted the difficulty in matching the varying resistance provided by the machines with human torque capabilities. This inability to maximally activate the neuromuscular system throughout the range of motion has led to criticism of training using variable resistance. However, as long as the resistance provided by the machine does not exceed that capable by the human (20), greater neuromuscular activation during variable resistance exercise is observed (16). Furthermore, during acute loading, greater neuromuscular activation seems to lead to earlier onset of fatigue (17).
Given that previous studies have shown greater neuromuscular and hormonal responses because of greater intensity (26) or total volume/work (8,26,30), respectively, it is surprising that no study has directly investigated the effect of variable resistance loading. Endocrine hormones respond to loading and act as acute signals, which are part of several systems that interact, to cause muscle growth (31). Three of the most studied hormones are testosterone (TES; affects protein synthesis via androgen receptors and satellite cells), 22-kDa growth hormone (GH; affects protein synthesis via JAK2 signaling), and cortisol (COR; a primary stress hormone and anti-inflammatory agent) (31). Acute increases in these hormones may be important in modulating long-term adaptations to resistance exercise (31). However, loading protocols differ regarding the source of fatigue (26) and acute hormonal responses (14,26,30) and thereby cause adaptations primarily within the neural system (maximal strength loading) or primarily within the muscles, that is, peripheral fatigue (hypertrophic loading) (3). Consequently, it is important to distinguish between these two different resistance loading methods. Therefore, the purpose of this study was to investigate the acute effects of constant and variable resistance exercise on neuromuscular and endocrine responses during maximal strength and hypertrophic loadings. It may be hypothesized that through greater neuromuscular activation throughout the range of motion, a greater acute neuromuscular and endocrine response will be observed during variable resistance loadings.
Thirteen healthy, young men (mean ± SD: age = 28.4 ± 3.7 yr, height = 180.3 ± 3.9 cm, body mass = 78.7 ± 10.2 kg, muscle mass = 38.1 ± 5.2 kg, fat percentage = 15.4% ± 3.3%) provided informed consent to take part in the study. None of the subjects had previously taken part in systematic strength training at a frequency greater than once per week; however, the subjects were physically active and regularly took part in recreational physical activity (endurance or ball games activity took place no more than three times per week). The study was conducted according to the Declaration of Helsinki (1964), and ethical approval was granted by the ethics committee of the University of Jyväskylä, Finland.
The subjects reported to the laboratory 2 wk before the first loading protocol to adjust the devices used during loading and to determine their maximum isometric and concentric leg extension strength capabilities. Upon arrival to the laboratory, the subjects' height and weight were measured, followed by body composition measurements (InBody 720 body composition analyzer; Biospace Co. Ltd., Seoul, South Korea). Once completed, the EMG electrode placements (vastus lateralis (VL), vastus medialis (VM), rectus femoris (RF), and biceps femoris (BF)) were measured according to the SENIAM guidelines and marked by indelible ink tattoos.
An electromechanical isometric leg extension device (Department of Biology of Physical Activity, University of Jyväskylä, Finland) was then set so that the subjects' knee angle was 107° and hip angle was 110°, and the position was recorded. The subject then performed at least three maximum voluntary contractions (for device and protocol details, see Häkkinen et al. ). Briefly, the subjects were instructed to push "as fast and as hard as possible" for approximately 3 s, and if the maximum force during the third trial was greater than 5% compared with the previous trials, a fourth trial was performed.
A David M16 leg press device (David Sports Ltd., Helsinki, Finland) was set at a starting (flexed knee) position of 58° ± 2.3° knee angle (approximate hip angle of 70°), with the finishing position, at full knee extension (180°), giving a hip angle of 120°. One-repetition maximum (1RM) testing was performed according to the protocol described in one of our previous studies (32). These 1RM loads were used to determine the loads used during set 1 of each loading protocol.
Venous blood collection and analysis.
Venous blood was collected from an antecubital vein using sterile techniques with the blood transferred into serum tubes (Venosafe, Terumo, Belgium). The samples were held for 15 min at room temperature before being centrifuged for 10 min at 3500 rpm (Megafuge 1.0 R, Heraeus, Germany). Once the serum had been separated from red blood cells, it was pipetted into tubes and stored in the refrigerator (−80°C) for future analysis.
Serum samples were analyzed for total TES, 22-kDa GH, and COR in our laboratory by chemical luminescence techniques (analytical sensitivity; total TES = 0.5 nmol·L−1, 22-kDa GH = 0.01 μg·L−1, COR = 5.5 nmol·L−1) using the Immulite 1000 and hormone-specific immunoassay kits (Immulite, Siemens, IL) (intra-assay reliability CV% (coefficient of variation); total TES = 5.7%, 22-kDa GH = 5.8%, COR = 7.9%). Hormone values presented are uncorrected for plasma volume changes.
Resistance loading protocols.
Individual subjects' tests took place at the same time of day (±1 h) throughout the study, and the average test time was 15:45. Each loading test was separated by at least 7 d of rest, but no more than 10 d passed between the loadings. When the subjects arrived, a preloading blood sample was taken after the procedures described earlier. Blood samples were also obtained immediately postloading, 15 min postloading, and 30 min postloading. Similarly, fingertip blood lactate was obtained preloading, postloading, and 15 min postloading and collected into capillary tubes (20 μL), which were placed in a 1-mL hemolyzing solution and analyzed automatically after the completion of testing (EKF diagnostic, Biosen, Germany). In addition to the loading blood samples, resting serum blood samples (control condition) were obtained at the same time of day as the loadings.
As protein ingestion can influence serum TES levels (11), the subjects were given a standardized fluid supplement (0.5 g supplement per 1 kg body mass), consisting of 68% carbohydrate, 30% protein, and 2% fat, to ingest 3 h preloading. The effect of hydration status on strength performance (21) was also controlled through the ingestion of 0.5 L of water 1 h preloading.
Loading pretest and posttest consisted of two isometric leg extension trials at 107° knee angle (a third was taken if the force during trial two was greater than the first by more than 5%). Short-term recovery was assessed by leg extension trials 15 min postloading and 30 min postloading. EMG was assessed during all recordings (for details, see the EMG recording and analysis section).
After the pretests, the subjects began the loading protocol, which was assigned via random selection. The maximal strength loading protocols consisted of 15 sets of 1RM (15 × 1 at 100% 1RM), and the hypertrophic loading protocols consisted of 5 sets of 10RM (5 × 10 starting at 80% 1RM). Three minutes of rest was given between maximal strength loading sets, while 2 min of rest was given during hypertrophic loading protocols (as in previous studies from our laboratory; e.g., see Häkkinen and Pakarinen ). Verbal encouragement was given throughout the loadings. The loads were adjusted during the loadings to enable completion of the required repetitions (reps). If the subject was not able to complete the required reps, assistance was provided. All subjects completed four loading protocols in total: two maximal strength and two hypertrophic loading protocols using the different device settings.
Concentric force, knee angle, and muscle activation were recorded during each rep of the maximal strength loading protocols and from reps 2, 5, and 8 during the hypertrophic loadings. Raw concentric leg press data were recorded and then filtered and analyzed by a customized script (Signal 2.16; CED, Cambridge, UK). A low-pass filter was used for force (20 Hz) and angle data (75 Hz). Concentric leg press force was analyzed for the whole range of motion (60°-180°) and in 20° segments (e.g., 60°-80°, etc.).
EMG recording and analysis.
Bipolar surface EMG electrodes were placed so that the tattoo was between the two electrodes (intertrial reliability CV% = 7.2). The VL, VM, RF, and BF of the right leg were measured during all loadings. SENIAM guidelines were followed for skin preparation, electrode placement, and orientation (interelectrode distance = 20 mm, input impedance < 10 kΩ, common mode rejection ratio = 100 dB, 500 gain). Raw signals passed from the transportable pack, around the subjects' waist, to the receiving box (Telemyo 2400R; Noraxon, Scottsdale, AZ) and were then relayed to the computer via an AD converter (Micro1401; CED).
Analysis of concentric and isometric EMG was performed using a customized script (Signal 2.16; CED) and filtered using a band pass of 20-350 Hz before analysis. The concentric EMG was analyzed using root mean square (RMS) for the whole range (60°-180° knee angle) and also for the 20° segments (e.g., 60°-80°, etc.). The RMS EMG of the 20° segments was also analyzed through the area-under-the-curve methods for 60°-120° and 120°-180°. Integrated EMG was used to analyze isometric actions, and the maximum isometric integrated EMG was determined from the period 500-1500 ms.
Conventional statistical methods were used to obtain means, SD, SE, and correlation coefficients. All loading variables were analyzed using ANOVA with repeated measures and post hoc tests with Bonferroni adjustments, except GH, which failed to satisfy the conditions for parametric tests. GH was therefore analyzed using Friedman's test for multiple comparisons, and then Wilcoxon matched pairs were used as post hoc tests. Area-under-the-curve analysis was analyzed using a paired t-test. Significance was defined as P ≤ 0.05.
Force and muscle activation.
The peak 1RM and 10RM loads were significantly higher during the constant resistance loadings compared with the variable resistance loadings (204 ± 29 and 178 ± 30 kg vs 197 ± 30 and 165 ± 26 kg, P < 0.05 and P < 0.001, respectively). Also, the volume load (reps × sets × load), corrected for assistance during failed reps, used during both constant resistance loadings was higher than during variable resistance loadings (maximal strength loading = 2825 ± 424 vs 2666 ± 478 kg, P < 0.01; hypertrophic loading = 8036 ± 1409 vs 7339 ± 1261 kg, P < 0.01) and during both hypertrophic loadings was significantly higher than during both maximal strength loadings (P < 0.001). When assessing total work on the basis of average concentric force (reps × sets × average concentric force), there was no difference between constant and variable resistance loadings (maximal strength loading = 25,625 ± 4041 vs 26,155 ± 4529 N; hypertrophic loading = 80,154 ± 12,882 vs 81,953 ± 11,666 N).
Concentric load increased from sets 1 to 3 during both constant and variable resistance hypertrophic loadings (P < 0.01), plateaued from sets 3 to 4, and then decreased from sets 4 to 5 (P < 0.05; Fig. 1). During constant resistance maximal strength loadings, the load increased from sets 1 to 3 (P < 0.01) and then decreased thereafter (P < 0.05; Table 1). Only decreases were observed from sets 5 to 15 during variable resistance maximal strength loadings (P < 0.01; Table 1).
Average knee extensor ((VL + VM + RF)/3) concentric EMG (60°-180° knee angle) increased significantly after rep 2 in each set during variable resistance hypertrophic loadings (P < 0.05; Fig. 1A). Furthermore, the magnitude of the EMG signal was greater for rep 2 (P < 0.05) and rep 8 (P = 0.06, effect size = 0.49) during set 5 compared with set 4 (Fig. 1A) despite decreased load (P < 0.01). Increases in average extensor concentric EMG were observed in sets 3 to 5 during constant resistance hypertrophic loadings (Fig. 1B). There was no change in average extensor concentric EMG during maximal strength loadings using either constant or variable resistance (Table 1).
Figure 2 shows additional force (120°-180° knee angle) generated during variable resistance hypertrophic loading and muscle activation (in this case VL activation) throughout the range of motion (Fig. 2A). During hypertrophic loadings, average knee extensor ((VL + VM + RF)/3) muscle activation (area-under-the-curve analysis) was greater from 120° to 180° knee angle for variable resistance compared with constant resistance (Fig. 2B). These differences were significant at set 1/rep 2 and set 3/rep 5, with a trend (P = 0.06) at set 3/rep 8. There were no differences observed for force or muscle activation between constant and variable resistance during maximal strength loadings.
FIGURE 2-Average con...Image Tools
Isometric force decreased throughout all loadings, although reductions due to hypertrophic loadings (constant resistance = −50% and variable resistance = −52%) were greater than due to maximal strength loadings (constant resistance = −27% and variable resistance = −29%) (P < 0.01; Table 2). After both hypertrophic loadings, force increased during 15 min of recovery. However, 15-30 min after constant resistance hypertrophic loading, maximum isometric force continued to recover, while there were no further increases after variable resistance loading (P < 0.05; Table 2). Isometric muscle activation during both maximal strength loadings reduced preloading to postloading (−15% during both loadings) and remained lower than preloading throughout recovery (P < 0.05; Table 2). During both hypertrophic loadings, isometric muscle activation decreased preloading to postloading (constant resistance = −14%, P < 0.05, and variable resistance = −19%, P < 0.01), increased during 15 min of recovery, and plateaued thereafter (Table 2).
Blood lactate and serum hormone concentrations.
Preloading to postloading increases in blood lactate concentration due to all loadings was significant (P < 0.01), and blood lactate levels reduced after 15 min of recovery (P < 0.05). After hypertrophic loadings, blood lactate levels were greater than after maximal strength loadings (postloading: constant resistance maximal strength loading = 3.36 ± 1.4 mmol·L−1; variable resistance maximal strength loading = 3.87 ± 2.3 mmol·L−1; constant resistance hypertrophic loading = 11.9 ± 2.2 mmol·L−1; variable resistance hypertrophic loading = 13.4 ± 2.3 mmol·L−1, P < 0.01). Furthermore, postloading blood lactate levels were significantly higher after variable resistance hypertrophic loading compared with constant resistance hypertrophic loading (P < 0.05).
During variable resistance hypertrophic loading, serum total TES concentration increased preloading to postloading (P < 0.01), and then a significant decrease was observed from postloading to 30 min postloading (Fig. 3A). The preloading to postloading change was significantly different compared with resting conditions (P < 0.05; Fig. 3B). Serum GH increased preloading to postloading during both hypertrophic loadings (constant resistance, P = 0.051) and remained elevated at both 15 and 30 min postloading (P < 0.05; Fig. 4A). A preloading to postloading increase in COR concentration was observed during variable resistance hypertrophic loading, and COR levels were significantly higher than preloading after 15 and 30 min of recovery during both hypertrophic loadings (P < 0.05; Fig. 4B). Serum COR levels were significantly higher than resting values 15 and 30 min postloading. The change in plasma volume before and after variable resistance hypertrophic loading was significantly greater than before and after constant resistance hypertrophic loading (−15% ± 3% vs −8% ± 3%, P < 0.01). At postloading, GH concentration was increased after variable resistance maximal strength loading (Fig. 4A). No changes were observed for TES or COR during either maximal strength loadings.
Relative changes in blood lactate and serum GH preloading to postloading were significantly related (r = 0.497, P < 0.05, n = 18) when both hypertrophic loading results were combined. Also, during variable resistance hypertrophy loading, relative changes in blood lactate and isometric force preloading to postloading were inversely related (r = −0.615, P < 0.05, n = 13).
We observed differences in acute blood lactate and serum hormone responses of TES and COR between constant and variable resistance hypertrophic loadings (5 × 10RM), although both loadings were performed with maximum effort. The impact of using variable loading on the neuromuscular system was observed during analysis of the hypertrophic loading's concentric actions and also indications of prolonged neuromuscular fatigue during isometric performance were recorded. These findings indicate greater fatigability during variable resistance loading and that this fatigability influences acute hormonal responses.
To ensure that each set was a true reflection of the subject's maximum strength capabilities, the load was increased if the subject completed the desired number of rep without assistance (Table 1 and Fig. 1). Therefore, we are confident that our subjects performed all loadings with maximum effort so that any differences observed were due to the independent variables. During hypertrophic loadings, there was a progressive increase in quadriceps EMG magnitude ((VL + VM + RF)/3) from rep 2 to rep 8 (Fig. 1). This may indicate a gradual recruitment of higher threshold motor units to compensate for the fatigue of the previously recruited motor units. Comparing single reps between constant and variable resistance hypertrophic loadings shows that, during each rep, activation of the quadriceps is higher at larger knee angles (120°-180°) because of variable resistance (Fig. 2). This greater activation of the quadriceps, combined with the high volume protocol (50 reps), may have produced greater peripheral fatigue, possibly through alterations within excitation-contraction coupling (1,27) or higher intracellular pH (indicated by higher blood lactate). Therefore, during loading, force is maintained through either a) greater discharge rate and/or b) recruitment of higher threshold motor units, following the size principle (5) of Hennemann et al. (10), causing an increase in the magnitude of the EMG signal. During variable resistance loading, this was observed as early as set 1 and despite a reduction of load (sets 4 and 5). This suggests that variable resistance loading influences motor unit recruitment strategies to a greater extent than constant resistance loading through the interaction of greater initial neuromuscular demand and greater subsequent peripheral fatigue.
The present isometric results agree with previous studies (12,13), in that hypertrophic loading caused a greater level of neuromuscular fatigue compared with maximal strength loadings. Both central and peripheral fatigue contribute to decreases in isometric force and EMG during hypertrophic loading, whereas central fatigue is likely to be the primary cause during maximal strength loading (5). Evidence for this was provided by the inverse relationship between isometric force and blood lactate (variable resistance hypertrophic loading) and larger increases in force and EMG during 0-15 min postloading recovery (both hypertrophic loadings). However, from 15 min postloading to 30 min postloading, the rate of recovery after variable resistance hypertrophic loading seemed to reduce (Table 2), with both force and quadriceps EMG remaining significantly lower than preloading values. This was not observed after constant resistance hypertrophic loading and may indicate greater/longer lasting central fatigue (13) or disruption to excitation-contraction coupling mechanisms (5). Interestingly, during the maximal strength loadings, there was no change in the magnitude of the concentric quadriceps EMG signal ((VL + VM + RF)/3). This may indicate an ability to maintain central drive throughout the concentric contractions despite observed fatigue in maximum isometric contractions (7,33). Nevertheless, it is probable that both maximal strength loadings impacted the neuromuscular system equally because we observed equivalent levels of neuromuscular fatigue and recovery during the isometric testing.
Previous studies have shown that hypertrophic loading causes minor but significant changes in serum TES and large increases in GH, COR, and blood lactate postloading (15,9,29). However, exercises involving several large muscle groups, such as squat loadings, leads to a greater magnitude of TES response than during leg press loading (14,23,26,30). Furthermore, increases in blood lactate have been associated with increases in serum GH after hypertrophic loading (14,21), as shown in this study (r = 0.497, P < 0.05, n = 18). In the present study, our results agreed with those observed using a similar variable resistance leg press device (15), and this is the first study to compare variable and constant resistance loading using a leg press device. In contrast to hypertrophic loading, only minor changes were expected to occur because of maximal strength loading, especially during single rep sets (14,29). This was true also in the present study because the only significant change was the increased GH immediately after variable resistance loading (Fig. 4A).
As blood lactate concentration was significantly higher after variable resistance hypertrophic loading compared with constant resistance loading, it is not surprising that greater acute increases were observed in serum hormone concentrations. Smilios et al. (30) observed greater acute increases in blood lactate, TES, GH, and COR when the number of sets was increased from 2 to 4 (although no further increases were observed after six sets). Furthermore, changes in plasma volume-shown to be the cause of acute increased total TES concentration (25,29)-are thought to be associated with accumulation of lactate and other metabolites in the muscles (5,29). The magnitude of plasma volume change in the present study may therefore explain the difference observed in significant versus nonsignificant increases in total TES concentrations during the present hypertrophic loadings (Fig. 3).
Given that distance remains the same in both loadings and that two different methods have been used to determine total work (volume load and average force), our data showed that the subjects did not perform more work using variable resistance loadings. A higher blood lactate concentration postloading, which may have influenced serum hormone concentrations (17,21), is likely to be due to greater energy demands from a greater number of motor units and/or larger motor units active during each set. Therefore, we propose that the greater acute responses observed in serum hormone concentrations are most likely due to the loading method-in this case, a greater neuromuscular activation required at larger knee angles.
Serum COR concentration was already observed immediately postloading during variable resistance loading. Although a catabolic hormone, COR is thought to play an important role in tissue remodeling and hypertrophy processes. A longer exposure to elevated COR levels may indicate larger muscle damage, and a greater level of remodeling may occur because acute COR increase has been highly correlated with serum creatine kinase concentration 24 h postloading (22). Furthermore, COR indicates the level of psychophysiological stress. Therefore, a greater increase in COR concentration may indicate that the variable resistance loading was more metabolically demanding (19) than the constant resistance loading, which supports our plasma volume and force and EMG data, indicating greater peripheral fatigue. In the present study, we have observed greater acute changes in serum hormone concentrations because of a greater stress placed on the anaerobic energy system, possibly through a greater challenge to the neuromuscular system.
In summary, we observed differences in neuromuscular and hormonal responses between constant and variable hypertrophic loadings. A greater level of neuromuscular fatigue and larger acute responses in serum hormone concentrations occurred after hypertrophic variable resistance loading. With the exception of GH, differences in acute neuromuscular or hormonal responses were not observed using the present study design for maximal strength loading (15 × 1 at 100% 1RM). Perhaps the potential advantages of variable resistance training are more suited to hypertrophic training practices. Future well-controlled training studies are needed to determine whether the larger acute responses observed in hypertrophic variable resistance loading in the present study will induce larger long-term adaptations concerning muscle hypertrophy and strength development.
This project was partly funded by the David Sports Ltd. and the Department of Biology of Physical Activity, University of Jyväskylä. The project did not receive funding from the National Institutes of Health, the Wellcome Trust, the Howard Hughes Medical Institute, etc.
The authors thank Mr. Risto Puurtinen for his excellent assistance during blood collection and hormone analysis and Ms. Lisa Davis and Mr. Heikki Peltonen for their assistance during measurements.
The results of the present study do not constitute endorsement by the American College of Sports and Medicine.
1. Allen DG, Lamb GD, Westerblad H. Impaired calcium release during fatigue. J Appl Physiol
2. Cabell L, Zebas CJ. Resistive torque validation of the Nautilus multi-biceps machine. J Strength Cond Res
3. Campos GER, Luecke TJ, Wendeln HK, et al. Muscular adaptations in response to three different resistance-training regimens: specificity of repetition maximum training zones. Eur J Appl Physiol
4. Deleted in proof.
5. de Ruiter CJ, Elzinga MJH, Verdijk PWL, van Mechelen W, de Haan A. Changes in force, surface and motor unit EMG during post-exercise development of low frequency fatigue in vastus lateralis muscle. Eur J Appl Physiol
6. Eloranta V, Komi PV. Function of the quadriceps femoris muscle under maximal concentric and eccentric contractions. Electromyogr Clin Neurophysiol
7. Fry AC, Kraemer WJ, Lynch MJ, Triplett TN, Koziris PL. Does short-term near-maximal intensity machine resistance training induce overtraining? J Strength Cond Res
8. Gotshalk LA, Loebel CC, Nindl BC et al. Hormonal responses of multiset versus single-set heavy-resistance exercise protocols. Can J Appl Physiol
9. Harman E. Resistive torque analysis of 5 Nautilus exercise machines. Med Sci Sports Exerc
10. Hennemann E, Somjen G, Carpenter DO. Functional significance of cell size in spinal motoneurons. J Neurophysiol
11. Hulmi JJ, Tannerstedt J, Selänne H, Kainulainen H, Kovanen V, Mero AA. Resistance exercise with whey protein ingestion affects mTOR signaling pathway and myostatin in men. J Appl Physiol
12. Häkkinen K. Neuromuscular fatigue in male and female athletes during heavy resistance exercise. Int J Sports Med
13. Häkkinen K. Neuromuscular fatigue in males and females during strenuous heavy resistance loading. Electromyogr Clin Neuropysiol
14. Häkkinen K, Pakarinen A. Acute hormonal responses to two different fatiguing heavy-resistance protocols in male athletes. J Appl Physiol
15. Häkkinen K, Pakarinen A. Acute hormonal responses to heavy resistance exercise in men and women at different ages. Int J Sports Med
16. Häkkinen K, Komi PV, Kauhanen H. Scientific evaluation of specific loading of the knee extensors with variable resistance, "isokinetic" and barbell exercises. Medicine Sport Sci
17. Häkkinen K, Kaunhanen H, Komi PV. Effects of fatiguing loading with a variable resistance equipment on neural activation and force production of the knee extensor muscles. Electromyogr Clin Neurophysiol
18. Häkkinen K, Kallinen M, Izquierdo M, et al. Changes in agonist-antagonist EMG, muscle CSA, and force during strength training in middle-aged and older people. J Appl Physiol
19. Izquierdo M, Ibanez J, Calbet JAL, et al. Cytokine and hormone responses to resistance training. Eur J Appl Physiol
20. Johnson JH, Colodny S, Jackson D. Human torque capability versus machine resistive torque for four Eagle resistance machines. J Appl Sport Sci Res
21. Judelson DA, Maresh CM, Farrell MJ, et al. Effect of hydration state on strength, power, and resistance exercise performance. Med Sci Sports Exerc
22. Kraemer WJ, Fleck SJ, Dziados JE. Changes in hormonal concentrations after different heavy-resistance exercise protocols in women. J Appl Physiol
23. Kraemer WJ, Häkkinen K, Newton RU, et al. Effects of heavy-resistance training on hormonal response patterns in younger vs. older men. J Appl Physiol
24. Kulig K, Andrews JG, Hay JG. Human strength curves. Exerc Sport Sci Rev
25. McCall GE, Byrnes WC, Fleck SJ, Dickinson A, Kraemer WJ. Acute and chronic hormonal responses to resistance training designed to promote muscle hypertrophy. Can J Appl Physiol
26. McCaulley GO, McBride JM, Cormie P, et al. Acute hormonal and neuromuscular responses to hypertrophy, strength and power type resistance exercise. Eur J Appl Physiol
27. Piitulainen H, Komi PV, Linnamo V, Avela J. Sarcolemmal exitability as investigated with M-waves after eccentric exercise in humans. J Electromyogr Kinesiol
28. Pizzimenti MA. Mechanical analysis of the Nautilus leg curl machine. Can J Sport Sci
29. Pullinen T, Mero A, Huttunen P, Pakarinen A, Komi PV. Resistance exercise-induced hormonal responses in men, women, and pubescent boys. Med Sci Sports Exerc
30. Smilios I, Pilianidis T, Karamouzis M, Tokmakidis SP. Hormonal responses after various resistance exercise protocols. Med Sci Sports Exerc
31. Spiering BA, Kraemer WJ, Anderson JM, et al. Resistance exercise biology: manipulation of resistance exercise programme variables determines the responses of cellular and molecular signalling pathways. Sports Med
32. Walker S, Peltonen J, Ahtiainen JP, Avela J, Häkkinen K. Neuromuscular fatigue induced by an isotonic heavy-resistance loading protocol in knee extensors. J Sports Sci
33. Walker S, Ahtiainen JP, Häkkinen K. Acute neuromuscular and hormonal responses during contrast loading: effect of 11 weeks of contrast training. Scand J Med Sci Sports