The cycle Wingate test (CWT) has been the traditional method used to estimate anaerobic power and capacity (2,16). However, from the 1970s to the present, the reliability and validity of the CWT have been criticized regarding its test loads (6,12), test duration (8), calculation of power outcomes (4), and inertial effect of the flywheel (13,14,17,22). In recent years, the Wingate Institution recommended the use of higher test loads than was originally suggested. Additionally, new software has allowed the use of different starting procedures, test durations, second by second data analyses, and inertia correction.
The CWT continues to be examined by correlating its noninvasive anaerobic outcomes with histochemical values, metabolic contributions, and physiological responses. Current questions about the CWT include the following: “Why does the Wingate test not lead to a considerable depletion of anaerobic energy reserves as expected?” and “How anaerobic is the CWT anaerobic test (5)?”
Work production capacity is closely related to active muscle mass involvement. With the cycle ergometer, the lack of activity pattern specificity may be the most important factor limiting the proportion of activated muscle mass during Wingate testing. To overcome this limitation, elliptical trainers have been tested regarding active muscle mass; exercise on elliptical trainers was found to result in greater muscle activity than was exercise on classic cycle ergometers (7). Elliptical trainers have also been found to be effective devices for estimating metabolic responses (3,15,23). Laboratory tests, able to predict aerobic power noninvasively by using submaximal (1,10) and maximal (9) incremental graded tests, and a military physical fitness test (21) have been developed on elliptical trainers.
Ozkaya et al. (19,20) modified an electromagnetically braked elliptical trainer to evaluate anaerobic performance. Using the same Wingate protocol, they demonstrated the disadvantages of cycle testing (20). Greater power outcomes were obtained with the modified elliptical test compared with those of the CWT (19), and reliable retest results were observed with this modification (20). Electronic system configurations are generally more complicated than their counterpart mechanic system configurations (11) and may have standardization problems in the future, as technology advances.
The main purpose of this study was to modify an elliptical trainer with a mechanic brake system (EWT) instead of an electromagnetically braked system, making it a more valid and reliable research tool in the laboratory. The secondary aims of the study were to determine the optimum test load for the EWT and measure the retest reliability of this load to best measure anaerobic outcomes.
Experimental Approach to the Problem
This study consisted of 3 phases: (a) an engineering study to design a new test ergometer for testing anaerobic performance, (b) a pilot study to narrow down (restrict) the load interval for the primary study, and (c) a primary study to determine a proper test load to reveal the most efficient anaerobic test outcomes and to evaluate the retest reliability of the selected test load.
The University Ethics Committee approved the study protocol. All the participants were informed about the possible effects of malnutrition, inadequate food or fluid intake, and other risks potentially related to the study, and written informed consent was obtained. Thirty healthy male university athletes were allowed to participate in the study with regard to their anaerobic performance (Table 1). All the volunteers underwent an average of 4–6 training sessions per week in their particular sports disciplines and had an average sports participation background of 8.8 6 2.1 years.
A mechanically braked elliptical trainer (Precor Experience series EFX 576i, Precor, Inc., Woodinville, WA, USA) was modified using the original interface and software of a traditional CWT (Monark 834E, Varberg, Sweden) to assess anaerobic power and capacity outcomes such as peak power (PP), average power (AP), minimum power (MP), power drop (PD), and fatigue index ratio (FI%) (Figure 1).
The large cogwheel was placed at the rear of the original elliptical trainer (6). Movements on the elliptical trainer using parts 1–4 were transferred to the large cogwheel (6) via the original crank (5). This movement was then transferred by the chain (8) to the small cogwheel (9), which in turn caused the 1.5-m flywheel (10) to revolve. The braking system was provided using a rope running around the circumference of the flywheel, attached to a pulley (12) and weight system (14–16). A starting signal was provided via the signaler (13) when the weights dropped. The rpm (revolutions per minute) calculator (7) provided data regarding revolutions of the crank. All data were collected by an interface (17) and transferred to the computer (18). At the end of the test, absolute and relative power values were automatically calculated by the original software system.
A pilot study was conducted using a large load range between 12% (L-12) and 24% (L-24) body mass workload. The purpose of the pilot study was to narrow (restrict) load intervals to determine a proper load range for the primary study. Five well-trained volunteer sprinters (mean age 21 ± 2 years) were selected to participate in the pilot study. An error detection procedure was administered using data attained from illogical (improper) interrelations among 5-second segments of 30-second all-out tests (20). For this purpose, these 5-second segments of 30-second test duration were numbered from 1 to 6. Expected segmental order (Figure 2) was examined by using error detection criteria as follows:
The segment number of PP has to be lower than the segment number of MP. If so, it is expected (zero points), if not, it is an error (one point).
- E.1. If PP Segment No < MP Segment No ⇒ Expected ‘0 points’; if not ⇒ Error ‘1 point.’
The segment number of PP has to be lower than that of the third segment. If so, it is expected (zero points), if not, it is an error (one point).
- E.2. If PP Segment No < 3⇒ Expected ‘0 points’; if not ⇒ Error ‘1 point.’
The segment number of MP has to be higher than that of the fifth segment. If so, it is expected (zero points), if not, it is an error (one point).
- E.3. If MP Segment No > 5 ⇒ Expected ‘0 points’; if not ⇒ Error ‘1 point.’
The power value of the second segment has to be higher than that of the power value of the third segment. If so, it is expected (zero points), if not, it is an error (one point).
- E.4. If the value of the second segment > value of the third segment ⇒ Expected ‘0 points’; if not ⇒ Error ‘1 point.’
The power value of the third segment has to be higher than the power value of the fourth segment. If so, it is expected (zero points), if not, it is an error (one point).
- E.5. If the value of the third segment > value of the fourth segment ⇒ Expected ‘0 points’; if not ⇒ Error ‘1 point.’
The power value of the fourth segment has to be higher than the power value of the fifth segment. If so, it is expected (zero points), if not, it is an error (one point).
- E.6. If the value of the fourth segment > value of the fifth segment ⇒ Expected ‘0 points’; if not ⇒ Error ‘1 point.’
Following the error detection procedure, loads between L-16 and L-20 were selected to be used for testing with the 30 well-trained athletes in the primary study.
Elliptical Wingate Testing
All loads were randomly tested using EWT for each volunteer. Warm-ups were performed at 20% of each load. The velocity during the warm-up period was 80–90 rpm, and the heart rates (HRs) were maintained at approximately 120 b·min−1. Three acceleration bursts during the third, fourth, and fifth minutes, lasting 2–3 seconds each, with verbal cues were performed as warm-up. At the end of the warm-up period, the subjects performed dynamic stretching exercises using relevant muscle groups during a 5-minute resting period. An unloaded time period, lasting 2–3 seconds, was implemented to enable the subjects to reach maximal revolution. By the time the workload was administered, the 30-second all-out test duration had been started. The participants received verbal motivation to facilitate maximum anaerobic power and capacity results during the tests.
Blood Lactate Analysis
Capillary blood samples were drawn from participants' fingertips at rest and at 5 minutes postexercise. Total blood lactate concentrations were analyzed using an electroenzymatic lactate analyzer (YSI 1500S, Yellow Springs Instruments, Inc., Yellow Springs, OH, USA). Delta lactate (ΔLa) was calculated as postexercise minus preexercise lactate concentration.
At the end of the tests, HRs were recorded as beats per minute by using a telemetric system (Polar RS400, Polar Electro Oy, Kempele, Finland).
Elliptical Wingate Retesting
A retest study was performed using the same standards on subsequent days at the same time of the day. The PP, AP, MP, PD, FI%, and ΔLa responses were compared with the initial test results.
Results were evaluated using SPSS Version 15.0 software (SPSS Inc., Chicago, IL, USA). A repeated measures design was used to estimate the most effective workload. Power outcomes of all workloads were used as factor levels. According to results of Mauchly's test of sphericity, the variance homogeneity assumption was not provided. The Greenhouse-Geisser test was therefore used to conduct a one-way analysis of variance (ANOVA). Bonferroni pairwise comparisons were used, as post hoc analysis, to investigate which group was different in a positive direction. Pearson correlation coefficient (r) was used to determine the linear correlation between tested variables. Findings with a p ≤ 0.05 were considered statistically significant.
The purpose of the pilot study was to narrow down load intervals to ascertain the proper load range for the primary study. According to the results of the pilot study, workloads lower than L-13 led to a plateau in power graphs from the beginning to the end of the test (Figure 3A). The FI% and PD were zero or negative. All-out tests with workloads greater than L-23 could not be completed because of exhaustion (Figure 3B). These values were not consistent with test outcomes that appear at the end of a correct Wingate application. As expected, problems occurred with very low (L-13, L-14, and L-15) and heavy workloads (L-23, L-22, and L-21). Plateaus occurred in power graphs when L-13, L-14, or L-15 loads were applied to the load scale (Figure 3C). Because inertial effects stem from extreme velocities, the subjects were not able to apply optimal force to the system. Undulations occurred in power graphs when L-23, L-22, or L-21 loads were applied because of extreme resistance administrations (Figure 3D); thus, the subjects were not able to reach the desired rpm values. Error scores of very low and heavy workloads are displayed in Table 2.
Workloads between L-16 and L-20 revealed expected (logical) power graph specificities as shown in Figure 2A,B. The PP appeared in the first (A) or second (B) segment. The MP was evident in the last segment. Because of satisfying the requirements, PD and FI% could be calculated in a valid fashion at the end of the test.
Descriptive statistics for the mean of PP, AP, FI%, and ΔLa in L-16, L-17, L-18, L-19, and L-20 are shown in Table 3.
According to results of Mauchly's test of sphericity, the assumption of variance homogeneity was not met (Table 4). The Greenhouse-Geisser test was therefore used to conduct a one-way ANOVA (Table 5).
Tested hypotheses were as follows:
Ha: At least one of the μ is different.
After the ANOVA test, it was concluded that at least one mean value of PP, AP, FI%, and ΔLa was significantly different from the remaining mean values (p < 0.05). The Bonferroni pairwise comparisons test was used as a post hoc analysis test to investigate group differences.
According to the results of the Bonferroni pairwise comparisons test, PP and AP outcomes using L-18 were significantly different from those of L-16, L-17, L-19, and L-20 in a positive direction. These findings showed that L-18 was the most effective load for PP and AP values (p < 0.05, Tables 6 and 7).
Also, the largest number of participants (70%) was able to reach their best anaerobic power and capacity outcomes with the L-18 test load.
The FI% values using both L-16 and L-18 were significantly different from those of the remaining loads (L-17, L-19, and L-20, p < 0.05). L-16 could not be selected as the optimal load because of very diminished power values; thus, L-18 was the most effective load for FI% (Table 8).
The highest ΔLa values were observed after testing with L-18 and L-19 loads (15.4 ± 1.7 and 15.3 ± 2.2 mM, p > 0.05). However, between these 2 loads, only L-18 yielded significantly higher ΔLa values than the others did (p < 0.05; Table 9).
Overall, L-18 yielded the greatest anaerobic power and capacity outcomes. The PP, AP, FI%, and ΔLa outcomes are shown in Figures 4 and 5.
No significant differences were found between test and retest results for the EWT (Ho was not rejected, p > 0.05).
The tested hypotheses were as follows:
- Ho: μtest − μ retest = 0,
- Ha: μtest − μ retest ≠ 0.
Test and retest variables were highly correlated (Figure 6). Pearson correlation coefficients are given in Table 10.
The popularity of elliptical trainers has been increasing in recent years because of effective activation of both upper and lower body muscle groups (7). Laboratory tests to predict V[Combining Dot Above]O2max noninvasively are now being performed using elliptical trainers (1,9,10). Ozkaya et al. (19,20) previously showed that an electromagnetically modified elliptical trainer could be used as a reliable device for performing all-out anaerobic testing because it provided greater power outputs than did traditional devices (19,20). Although cycle ergometers, treadmills, and nonmotorized treadmills are commonly used in exercise laboratories, elliptical trainers are gaining acceptance as alternative test devices in present day research.
Combining the advantages of mechanically braked cycle ergometers, such as cost effectiveness, ease of calibration, and reliability of result comparison (11,18), with the advantages of modified elliptical trainers, such as higher anaerobic performance production based on increased muscular activity, optimal leg cycling pattern, and other factors (7,19), may provide the overall best anaerobic testing environment. For these reasons, we set out to implement these ideas by combining the original brake system of a cycle ergometer with the main frame of an elliptical trainer.
Regarding our secondary aim of determining the most effective test load for well-trained adult male athletes, workloads lower than L-13 led to a plateau in the power graph from the beginning to the end of the test in our pilot study. Because of excessively low resistance, the subjects reached velocities that were too high resulting in an extreme inertial effect on the system. In contrast, it was not possible to complete the all-out test with workloads greater than L-23 because of exhaustion. The subjects could not reach optimal velocity grades against this extremely high resistance, so FI% and PD could not be calculated correctly. In the proper workload range, estimated with the error detection procedure before the primary study, L-18 was determined to be the most effective load for reaching maximum anaerobic output for whole-body tests on the EWT. Test and retest analyses of power outcomes were also highly correlated with L-18. Although our load optimization analysis determined 18% of body mass to be the ideal test load, load optimization of the traditional Wingate test is still being questioned. Taking into account widely accepted recommendations for traditional cycle Wingate testing, the same load for well-trained female athletes might also be appropriate; 25% smaller workloads might be more suitable for younger, less massive, or untrained persons (24).
Although the widely accepted workload for traditional cycle ergometers is 6–10%, the optimal workload for the elliptical trainer in our study was 18% of body mass. The estimated optimal workload is closely related to body movement patterns based on the specific configuration of the ergometers. Although cycle ergometers principally activate either the lower or upper body, both arms and legs can be used simultaneously on an elliptical trainer. In the study by Browder et al. (7), electromyography results indicate that movement patterns on the elliptical trainers involved greater muscle activity compared with that of cycle ergometers because of differences in movement patterns. This may explain why the optimal workload for an elliptical trainer was higher than the workload widely used for cycle ergometers.
Additional problems occurred when heavier loads (i.e., L-24; ∼24 kg for a 100-kg weighted subject) were tested on the EWT: Some of the modifications we made were damaged. In future studies, special ropes, chains, and larger load scales should be used to make the EWT system stronger so that heavier loads can be used without resulting in damage.
The mechanically braked elliptical modification is a reliable and consistent device for measuring the maximum anaerobic performance for the whole body. Anaerobic testing using this modification may be more useful to athletes and coaches than cycle ergometers because a greater proportion of muscle groups are involved during exercise on an elliptical trainer.
The authors would like to give special thanks to Professor Cem Seref Bediz, M.D., Dokuz Eylul University Medical Faculty, Department of Physiology, for his editorial assistance. The authors would also like to thank the Acibadem Health Group for the material support they provided. This study was sponsored by the Acibadem Health Group, Istanbul, Turkey.
1. Armour A, Michael T, Zabik R, Liu Y, Dawson M, Carl D. Development of a submaximal exercise test to predict V[Combining Dot Above]O2
max using an elliptical trainer
. Med Sci Sports Exerc 35: 310, 2003.
2. Bar-Or O. The Wingate anaerobic test: An update on methodology, reliability and validity. Sports Med 4: 381–394, 1987.
3. Batte AL, Darling J, Evans J, Lance LM. Physiologic response to a prescribed rating of perceived exertion on an elliptical fitness cross-trainer. J Sports Med Phys Fitness 43: 300–307, 2003.
4. Bell W, Cobner DV. Effect to individual time to peak power output on the expression of peak power output in the 30-s Wingate anaerobic test. Int J Sport Med 28: 135–139, 2007.
5. Beneke R, Pullman C, Bleif I, Leithaufer RM, Hütler M. How anaerobic is the Wingate anaerobic test for human? Eur J Appl Physiol 87: 388–392, 2002.
6. Bradley AL, Ball TE. The Wingate test: Effect of load on the power outputs of female athletes and non-athletes. J Appl Sport Sci Res 4: 193–199, 1992.
7. Browder KD, Dolny D, Cowin B, Hadley M, Jasper C, McAllister T, Stewart C, Terrel B. Muscle activation during elliptical trainer
and recumbent bike exercise. Med Sci Sports Exerc 37: 106, 2005.
8. Calbet JAL, Chavarren J, Dorado C. Fractional use of anaerobic capacity during a 30- and a 45-s Wingate test. Eur J Appl Physiol 76: 308–313, 1997.
9. Dalleck LC, Kravitz L, Robergs RA. Maximal exercise testing using the elliptical cross-trainer and treadmill. J Exerc Physiol Online 7: 94–101, 2004.
10. Dalleck LC, Kravitz L, Robergs RA. Development of a submaximal test to predict elliptical cross-trainer V[Combining Dot Above]O2
max. J Strength Cond Res 20: 278–283, 2006.
11. Dotan R. The Wingate anaerobic test's past and future and the compatibility of mechanically versus electro-magnetically braked cycle-ergometers. Eur J Appl Physiol 98: 113–116, 2006.
12. Dotan R, Bar-Or O. Load optimization for the Wingate anaerobic test. Eur J Appl Physiol 51: 409–417, 1983.
13. Franklin KL, Gordon RS, Baker JS, Davies B. Accurate assessment of work done and power during a Wingate anaerobic test. Appl Physiol Nutr Metab 32: 225–232, 2007.
14. Franklin KL, Gordon RS, Davie B, Baker JS. Assessing accuracy of measurements for a Wingate test using the Taguchi method. Sport Med 16: 1–14, 2008.
15. Hughes NJ, Dolny DG, Browder KD, Cowin B, Hedley M, Jasper C, McAllister T, Steward C, Terell B. Ratings of perceived exertion (RPE) during elliptical trainer
, treadmill bike and recumbent bike exercise. Med Sci Sports Exerc 37: 103, 2005.
16. Inbar O, Bar-Or O, Skinner JS. The Wingate Anaerobic Test. Champaign, IL: Human Kinetics, 1996.
17. MacIntosh BR, Bryan SN, Rishaug P, Norris SR. Evaluation of the Monark Wingate ergometer by direct measurement of resistance and velocity. Can J Appl Physiol 26: 543–558, 2001.
18. Micklewright D, Alkhatib A, Beneke R. Mechanically versus electromagnetically braked cycle ergometer
: performance and energy cost of the Wingate anaerobic test. Eur J Appl Physiol 96: 748–751, 2006.
19. Ozkaya O, Colakoglu M, Fowler D, Colakoglu S, Kuzucu OE. Wingate anaerobic testing with a modified electro-magnetically braked elliptical trainer
, part II: Physiological considerations. Isokinetic Exerc Sci 17: 115–119, 2009.
20. Ozkaya O, Colakoglu M, Ozgonenel O, Fowler D, Colakoglu S, Tekat A. Wingate anaerobic testing with a modified electro-magnetically braked elliptical trainer
, part I: Methodological considerations. Isokinetic Exerc Sci 17: 107–113, 2009.
21. Parcer SB, Griswold L, Vickers RR. Development of an elliptical trainer
physical fitness test (final report: NHRC-06-06, XBNMRC/MD). San Diego, CA: Naval Health Research Center, 2006.
22. Reiser RF II, Broker JP, Peterson ML. Inertial effects on mechanically braked Wingate power calculations. Med Sci Sport Exerc 32: 1660–1664, 2000.
23. Sweitzer ML, Kravitz L, Weingart HM, Dalleck LC, Chitwood LF, Dahl E. The cardiopulmonary responses of elliptical cross-trainer versus treadmill walking in CAD patients. J Exerc Physiol Online 5: 11–15, 2002.
24. Vandewalle H, Peres G, Heller J, Monod H. All-out anaerobic capacity tests on cycle ergometers: A comparative study on men and women. Eur J Appl Physiol 54: 222–229, 1985.