Previous investigations have established the physiological determinants of performance in a number of endurance sports, such as bicycling and running (6,7–9,11). These studies have identified maximal oxygen consumption (V̇O2max), oxygen consumption at lactate threshold (LT V̇O2), velocity at lactate threshold (LT), submaximal economy (ECON), and percentage of slow-twitch muscle fibers as the strongest predictors of performance (6,7–9,11). Due to the nature of running and bicycling, it has been possible to identify predictors of performance through sport-specific testing in the laboratory. However, this has proven more difficult with cross-country skiing, where ski-specific laboratory testing has only recently become possible. Previously, investigations assessing predictors of cross-country ski racing performance have primarily been limited to utilizing a nonspecific modality, such as treadmill running (1,4,14,21,25,29). As a result, a majority of the research has simply examined the relationship between running V̇O2max or LT and skiing performance (1,2,14,29,32).
Recently, it has been shown that ski-specific laboratory testing, such as treadmill roller skiing or upper-body ergometry, is better able to predict skiing performance when compared with treadmill running (4,10,21,25–27,32). The ability of ski-specific testing to more accurately predict performance is attributed to the extensive use of the upper-body musculature in cross-country ski racing, a variable not assessed during treadmill running (18,30). Consequently, in laboratory studies, upper-body aerobic and anaerobic power, assessed through treadmill roller skiing or upper-body ergometry, have been shown to better predict success in cross-country skiing than running V̇O2max or LT tests (21,25–27,32).
Investigations of cross-country skiing performed in the field have been limited to: 1) identifying physiological differences between skiing techniques, 2) describing the physiology of cross-country ski racing, and 3) testing V̇O2max(3,13–15,17,20,22,31). To our knowledge, no study to date has attempted to comprehensively identify determinants of cross-country skiing performance in the field. Thus, the purposes of the present investigation were, first, to extend research conducted in the laboratory into a ski-specific setting in the field and, second, to see whether determinants of performance are consistent for athletes of differing competitive abilities.
Thirteen male cross-country skiers, all members of an NCAA Division I Ski Team, gave their written informed consent to participate in this study. The subjects were well-trained collegiate athletes competitive in both regional and national races. Physical characteristics (mean ± SEM) of the subjects were as follows: age, 19.5 ± 0.7 yr; weight, 70.9 ± 1.5 kg; height, 180.5 ± 1.4 cm; and body fat, 8.6 ± 0.7%. Body density was estimated by the hydrostatic weighing method (16). Percent body fat was then calculated according to the formula developed by Brozek et al. (5).
The subjects performed three field-testing sessions on roller skis to determine LT, skiing ECON, V̇O2max, and peak upper-body oxygen consumption (UB V̇O2peak); they also performed a 10-km skating time-trial (TT) in random fashion. All testing sessions and the TT were completed within 3–4 wk during preseason training, approximately 2 months before the start of the competitive season. The testing sessions were integrated into the subject’s normal training routine in a manner that no high-intensity/volume training was performed 48 h before a testing session. To control for variations in rolling resistance between different brands of roller skis, the subjects used the same type of roller ski (R2 Roadskater, Pro-Ski Inc., Nordic Equipment Inc., Park City, UT) during each field-testing session.
During the three field-testing sessions, oxygen consumption was continuously measured and averaged every 20 s via a two-way pneumotach attached to a portable gas analysis system (KB1-C; AeroSport, Inc., Ann Arbor, MI). Heart rate (HR) was continuously measured during each test using an HR monitor (Polar Electro, Inc., Port Washington, NY) and recorded by the gas analysis system software. The gas analysis system was manually calibrated in the field before each trial using standard gases of a known composition. This system has been demonstrated to provide a valid measure of oxygen consumption against standard laboratory methods (16,24).
LT and ECON.
After a 5- to 10-min warm-up, the subjects completed a six-stage graded protocol on roller skis using the V-2 skating technique to determine LT and ECON. The subjects were instructed to complete the entire trial using the V-2 technique. The V-2 technique is used on flat terrain or gradual inclines and consists of a symmetrical double-pole plant for each leg stride. All of the subjects were familiar with the V-2 technique and routinely used this technique while roller skiing and on snow. The subjects completed six laps of a level 1.6-km loop at a specified velocity beginning at 14 km·h−1 and increasing by 2 km·h−1 each loop. A technician pacing the subject on a bicycle with a calibrated electronic speedometer controlled the subject’s velocity (13,19). The velocity for each loop varied on average by less than 0.3 km·h−1 from the desired speed. The subjects rested for approximately 1–2 min between each stage, during which time blood samples were taken via a finger stick. Blood samples were stored in a preservative solution of EDTA, Triton X-100, and sodium fluoride (12) for later analysis of lactate concentration (1500 Sport Lactate Analyzer; Yellow Springs, Inc., Yellow Springs, OH). The LT was defined as the stage at which blood lactate increased by 1 mmol·L−1 above baseline values, followed by further increases (6,8).
The first three stages of the LT test (14, 16, and 18 km·h−1) were used to assess skiing ECON. These three speeds were chosen because they represented a range of submaximal velocities for all subjects (∼35–60% V̇O2max). ECON was defined as the average V̇O2 (mL·kg−1·min−1), over the final 2 min of each stage, for all three velocities.
Maximal oxygen consumption test.
Skiing V̇O2max was determined through a protocol similar to that described by Mygind et al. (20) and Mahood et al. (18). The trial was completed using the skating technique on a 3-km course, with the first 2 km being conducted on relatively flat terrain (∼4% grade) and the final kilometer, on a steep uphill (10–15% grade). The subjects were not restricted to a specified technique during the trial; however, most used the V-2 technique until the final uphill section, when they switched to the V-1 technique that is predominantly used on steep inclines.
The subjects were instructed to ski the first kilometer at slightly below-race pace, then increase to race pace for the next kilometer, finishing with an all-out effort to volitional exhaustion. None of the subjects completed the entire 3-km course. The trial was considered to be a maximal effort if the following criteria were met: 1) a plateau in V̇O2 with increasing exercise intensity, 2) an R value > 1.10, and 3) a peak lactate > 8 mmol·L−1(9). All of the subjects met the above criteria. The average time to exhaustion was 8:44 ± 0:24 min:s. A blood sample was taken 1–2 min after completion of the trial for later analysis of peak lactate concentration.
A 1-km TT was used to determine upper-body peak V̇O2 (UB V̇O2peak). This trial was performed using the double-pole technique (which uses only the arms and trunk muscles for propulsion) on a gradual uphill (∼6–10% grade). UB V̇O2peak was defined as the highest V̇O2 (mL·kg−1·min−1) during the final minutes of the trial. Time to complete the trial was recorded (UB time) and a blood sample was taken 1–2 min after exercise for later determination of peak blood lactate concentration.
The subjects completed a 10-km roller-skiing TT, using the skating technique, as a measure of performance. To provide a further quantitative assessment of performance beyond the TT, subjects were rank-ordered (RANK) on a point basis from race results during the competitive season following the field-testing sessions.
Additionally, to determine whether predictors of performance were consistent for athletes of differing competitive abilities, the subjects were divided into two distinct groups of greater (group 1, N = 6) and lesser (group 2, N = 7) performance ability. This was based on the coach’s subjective assessment of the skier’s abilities and their performance during the competitive season.
The dependent physiological variables (SK V̇O2max, LT V̇O2, UB V̇O2, UB time, and ECON) from the three trials, TT time, and RANK were subjected to a Pearson’s product-moment correlation coefficient test. A forward stepwise multiple regression analysis was then used to determine the best predictor of TT time and RANK.
An independent t-test was performed to determine whether differences existed between groups for each of the dependent variables. All data are presented as mean ± SEM. Statistical significance was set at P < 0.05 for all analyses.
Table 1 summarizes the results from the three field-testing sessions. The subjects attained a peak V̇O2 of 52.0 ± 1.1 mL·kg−1·min−1 and a peak HR of 186.3 ± 1.5 beats·min−1 during the UBTT, which was equivalent to 77.2 ± 0.0% of and 97.0 ± 0.0% of SK V̇O2max and maximal HR, respectively. LT occurred at a V̇O2 of 44.1 ± 1.3 mL·kg−1·min−1 and an HR of 165.0 ± 2.2 beats·min−1, equivalent to 66.1 ± 1.4% of SK V̇O2max and 85.5 ± 0.8% of maximal HR, respectively.
Correlation coefficients between selected variables from the three trials to TT time and RANK are presented in Table 2. All of the dependent variables were significantly related to both of the measures of performance. Time to complete the UBTT (UB time) demonstrated the highest correlation to both TT time (Fig. 1) and RANK (Fig. 2).
Separate multiple regression analyses were performed with RANK and TT time as the dependent variables and SK V̇O2max, UB V̇O2peak, LT V̇O2, ECON (mL·kg−1·min−1), and UB time(s) as the independent variables. These analyses revealed that UB V̇O2peak and UB time accounted for 91% of the variance in RANK (F2,10 = 63.9), whereas 88% of the variance in TT time was explained by ECON and UB time (F4,8 = 44.6). The best predictor of both RANK and TT time was UB time, as demonstrated by the significant β values (0.77, P < 0.001, and 0.79, P < 0.001, respectively).
Analyses between groups revealed significant (P < 0.05) differences between groups for SK V̇O2max, UB V̇O2peak, HR at LT, ECON, and LT V̇O2 (Table 3) as well as percent SK V̇O2max at LT (69.3 ± 2.4 vs 63.3 ± 0.9). Additionally, the group 1 skiers were able to complete the UBTT 28.6 s faster (P < 0.05) that the group 2 skiers (212.6 ± 4.2 vs 241.2 ± 4.8 s).
Treadmill running has historically been used to assess the endurance capacity of cross-country skiers (1,3,4,14,25,28,31). Recent work, however, has demonstrated the importance of ski-specific testing for evaluating cross-country skiers (26,27). The unique aspect of the current investigation was that all tests were conducted in the field on roller skis. Although previous work in the field has looked at differences between skiing techniques, the physiological responses to ski racing, and V̇O2max of cross-country skiers, this represents the first study, to our knowledge, that extensively evaluated predictors of cross-country skiing performance in the field (2,3,13,15,17,18,20,22,31).
The major finding of this study was that ski-specific upper-body fitness, assessed through a 1-km UBTT, was the most important determinant of performance in this group of highly trained collegiate cross-country skiers. Time to complete the UBTT showed the highest correlation to both RANK (Fig. 1) and 10-km TT time (Fig. 2) and was the best predictor of either of these measures of performance. A similar negative correlation (r = −0.83) between a 1-km UBTT on snow and seasonal rank in male biathletes has been reported by Rundell and Bacharach (25).
The division of the subjects into two groups further demonstrated the importance of the upper-body component to cross-country skiing performance. Multiple regression analysis showed UB time to be the best predictor of both RANK (Fig. 2, A and B) and 10-km TT time (Fig. 1, A and B) within the two groups of athletes. Although the small number of subjects per group potentially limited the statistical power of the regression analyses, our results are consistent with previous investigations (10,21,30). Thus it would seem that regardless of skiing ability, upper-body endurance and power limit cross-country skiing performance.
The ability of the UBTT to predict performance over other ski-specific tests is most likely related to the extensive use of the upper body in cross-country skiing. It has been suggested that depending on terrain and technique used, greater than 50% of forward propulsion is generated by the upper body during ski skating (29). Additionally, it is estimated that to climb at race pace, the energy required greatly exceeds that which can be supplied by aerobic metabolism alone (23); therefore, a high level of upper-body aerobic and anaerobic power is necessary to succeed in cross-country skiing. This finding is supported by Staib et al. (30), who recently examined the relative contribution of upper-body aerobic and anaerobic power to performance in cross-country skiing. The authors found a strong negative correlation between upper-body anaerobic power (r = −0.68), upper-body peak V̇O2 (r = −0.74), and double-pole time to exhaustion (r = −0.80) to season rank. The major finding of that study was that upper-body anaerobic power and peak upper-body V̇O2 accounted for 75% of the variance in rank. This may explain why in our investigation UB time was a better predictor of performance than UB V̇O2peak; due to the short duration of the UBTT (3:49 ± 0:05 min:s), energy must be supplied by both aerobic and anaerobic metabolism. Although cross-country skiing is predominantly an aerobic sport, at times during a race (i.e., steep climbs and sprints), a high level of anaerobic power is required. Therefore, a test that taxes both energy pathways would likely better predict performance over a measure of just one of the variables.
The contribution of peak upper-body aerobic power to performance is nonetheless important. The strong negative correlation between UB V̇O2peak and RANK (r = −0.79) and 10-km TT time (r = −0.74) is in agreement with data from others (4,21,30) who found a similar relationship between skiing performance and peak upper-body V̇O2. The UB V̇O2peak of 52.0 ± 1.1 mL·kg−1·min−1 attained by the skiers in this study was equivalent to 77.2% of their skiing V̇O2max. Sharkey and Heidel (28) and Bilodeau et al. (4) reported similar upper- and lower-body ratios of 76.5% and 77.8%, respectively, for junior skiers—athletes comparable in age and competitive ability to those in the present investigation. These upper/lower-body ratios are lower than those reported by Rundell and Bacharach (25) of 89.3%, by Mygind et al. (21) of 89–95%, and by Staib et al. (30) of 84.3%. However, their subjects were elite international- and national-level athletes, unlike the subjects in the present investigation. Our findings confirm those of the previously cited studies in that peak upper-body aerobic power in skiers appears to be a reflection of both age and training status. Further, they confirm that a high upper/lower-body ratio is a prerequisite for success and that all cross-country skiers, particularly junior athletes, need to focus on the development of upper-body aerobic power through a consistent, structured upper-body training program.
The importance of upper-body aerobic power to performance can further be seen in the significantly higher (P < 0.05) peak upper-body V̇O2 for group 1 (Table 3). As a consequence of their higher upper-body aerobic power, and most likely anaerobic power, the group 1 skiers were able to complete the UBTT 13% faster than the group 2 skiers. This finding is in accordance with data from Staib et al. (30), who reported similar differences in UB V̇O2peak, endurance, and upper-body anaerobic power between two groups of skiers differing in competitive ability.
In addition to the necessity of upper-body aerobic and anaerobic power to cross-country skiing performance, the significant relationships between SK V̇O2max, LT V̇O2, UB V̇O2peak, and ECON to both of the performance variables demonstrates the need for a high level of ski-specific conditioning. Although V̇O2max was not the best predictor of performance within this group of endurance athletes, the importance of V̇O2max, however, is not to be discounted. An athlete’s V̇O2max sets the upper limit of oxidative energy production by the body and, consequently, regulates the extent to which UB V̇O2peak and LT V̇O2 can improve with training. The significantly (P < 0.05) higher V̇O2max of the group 1 skiers—and, accordingly, UB V̇O2peak and LT V̇O2—was undoubtedly one of the factors, along with better skiing ECON, that allowed them to perform at a higher level than the group 2 skiers.
It is important to note that several authors have suggested oxygen consumption be expressed as mL·min−1·kg−2/3, because this unit may more accurately reflect V̇O2 during cross-country skiing and thus better predict performance (2,14). In the present study, the results were not different when the analyses were repeated with V̇O2 expressed as mL·min−1·kg−2/3.
In summary this study confirms the importance of upper-body aerobic and anaerobic power to cross-country skiing performance. These results suggest that cross-country skiers should focus on developing the aerobic power and specific strength of the upper-body within a well-rounded endurance-training program. Lastly, this investigation demonstrated the potential of the 1-km UBTT as a simple field test to predict cross-country ski performance over more sophisticated and costly laboratory and field testing.
The authors thank Cory Schwatrz, Head Nordic Coach, and the members of the University of New Hampshire ski team for their participation and cooperation in this study. Additionally, thanks are extended to Mark Smith at AeroSport, Inc., who provided the metabolic equipment used to conduct this project. The results of this investigation do not constitute an endorsement by the authors or the American College of Sports Medicine of the AeroSport portable gas analysis system used in the study.
Address for correspondence: Robert W. Kenefick, Ph.D., Department of Kinesiology, University of New Hampshire, New Hampshire Hall, Durham, NH 03824; E-mail: firstname.lastname@example.org.
1. Bergh, U. Physiology of Cross-Country Ski Racing. Champaign, IL: Human Kinetics Publishers, 1982,pp. 16–23.
2. Bergh, U. The influence of body mass in cross-country skiing. Med. Sci. Sports Exerc. 19: 324–331, 1987.
3. Bilodeau, B., B. Roy, and M. R. Boulay. A comparison of three skating techniques and the diagonal stride on heart rate response and speed in cross-country skiing. Int. J. Sports Med. 12: 71–76, 1991.
4. Bilodeau, B., B. Roy, and M. R. Boulay. Upper-body testing of cross-country skiers. Med. Sci. Sports Exerc. 27: 1557–1562, 1995.
5. Brozek, J., F. Grande, T. Anderson, and A. Keys. Densitometric analysis of body composition, revision of some quantitative assumptions. Ann. N. Y. Acad. Sci. 110: 113–119, 1963.
6. Coyle, E. F. Integration of the physiological factors determining endurance performance ability. Exerc. Sport Sci. Rev. 23: 25–64, 1995.
7. Coyle, E. F., A. R. Coggan, M. K. Hopper, and T. J. Walters. Determinants of endurance in well trained cyclists. J. Appl. Physiol. 64: 2622–2630, 1988.
8. Coyle, E. F., W. H. Martin, A. A. Ehsani, et al. Blood lactate threshold
in some well-trained ischemic heart disease patients. J. Appl. Physiol. 54: 18–23, 1983.
9. Farrell, P. A., J. H. Wilmore, E. F. Coyle, J. E. Billing, and D. L. Costill. Plasma lactate accumulation and distance running performance. Med. Sci. Sports Exerc. 11: 338–344, 1979.
10. Gaskill, S. E., R. C. Serfass, and K. W. Rundell. Upper body power comparison between groups of cross-country skiers and runners. Int. J. Sports Med. 20: 290–294, 1999.
11. Hagberg, J. M., and E. F. Coyle. Physiological determinants of endurance performance as studied in competitive race walkers. Med. Sci. Sports Exerc. 15: 287–289, 1983.
12. Hill, D. W. Effect of storage on measured blood lactate concentration. Int. J. Sports Med. 16: 88–90, 1995.
13. Hoffman, M. D., P. S. Clifford, P. J. Foley, and A. G. Brice. Physiological responses to different roller skiing techniques. Med. Sci. Sports Exerc. 22: 391–396, 1989.
14. Ingjer, F. Maximal oxygen uptake as a predictor of performance ability in women and men elite cross-country skiers. Scand. J. Med. Sci. Sports 1: 25–30, 1991.
15. Katch, F. I., and V. L. Katch. Measurement and prediction errors in body composition assessment and the search for the perfect prediction equation. Res. Quart. 51: 249–253, 1980.
16. King, G. A., J. E. Mclaughlin, E. T. Howley, D. R. Bassett, Jr., and B. E. Ainsworth. Validation of AeroSport KB1-C portable metabolic system. Int. J. Sports Med. 20: 304–308, 1999.
17. Macdougall, J. D., R. Hughson, J. R. Sutton, and J. R Moroz. The energy cost of cross-country skiing among elite competitors. Med. Sci. Sports Exerc. 11: 270–273, 1979.
18. Mahood, N. V., R. W. Kenefick, J. Im, and K. W. Rundell. Validity of peak VO2
field testing for Nordic skiers. Med. Sci. Sports Exerc. 31: S335 1999.
19. Millet, G. Y., M. D. Hoffman, R. B. Candau, and P. S. Clifford. Poling forces during roller skiing: effects of technique and speed. Med. Sci. Sports Exerc. 30: 1645–1653, 1998.
20. Mygind, E., L. B. Andersen, and B. Rasmussen. Blood lactate and respiratory variables in elite cross-country skiing at racing speeds. Scand. J. Med. Sci. Sports 4: 243–251, 1994.
21. Mygind, E., B. Larsson, and T. Klausen. Evaluation of a specific test in cross-country skiing. J. Sports Sci. 9: 249–257, 1991.
22. Niinimaa, V., M. Dyon, and R. J. Shepard. Performance and efficiency of intercollegiate cross-country skiers. Med. Sci. Sports Exerc. 10: 91–93, 1978.
23. Norman, R. W., and P. V. Komi. Mechanical energetics of world class cross-country skiing. Int. J. Sports Biomech. 3: 353–369, 1987.
24. O’Neil, D. S., D. E. Devoe, and R. W. Gotshall. Validity of the AeroSport KB1-C gas analysis system. Med. Sci. Sports Exerc. 30: S125, 1998.
25. Rundell, K. W., and D. W. Bacharach. Physiological characteristics and performance of top U.S. biathletes. Med. Sci. Sports Exerc. 27: 1302–1310, 1995.
26. Rundell, K. W. Differences between treadmill running and treadmill roller skiing. J. Strength Cond. Res. 10: 167–172, 1996.
27. Rundell, K. W. Treadmill roller ski test predicts biathlon roller ski race results of elite US Biathlon women. Med. Sci. Sports Exerc. 27: 1677–1685, 1995.
28. Sharkey, B. J., and B. Heidel. Physiological tests of cross-country skiers. J. U. S. Ski Coaches Assoc. 5: 1–6, 1986.
29. Smith, G. A., R. C. Nelson, A. Feldman, and J. L. Rankinen. Analysis of V1 skating technique of Olympic cross-country skiers. Int. J. Sports Biomech. 5: 185–207, 1989.
30. Staib, J. L., J. Im, Z. Caldwell, and K. W. Rundell. Cross-country ski racing performance predicted by aerobic and anaerobic double poling power. J. Strength Cond. Res. 14: 282–288, 2000.
31. Stromme, S. B., F. Ingjer, and H. D. Meen. Assessment of maximal aerobic power in specifically trained athletes. J. Appl. Physiol. 42: 833–837, 1977.
32. Wisløff, U., and J. Helgerud. Methods for evaluating peak oxygen uptake and anaerobic threshold in upper-body of cross-country skiers. Med. Sci. Sports Exerc. 30: 963–970,1998.