^{1}

^{*}

^{1}

^{1}

^{2}

^{2}

^{3}

^{4}

^{1}

^{1}

^{2}

^{3}

^{4}

Edited by: Luca Paolo Ardigò, University of Verona, Italy

Reviewed by: Thomas Leonhard Stöggl, University of Salzburg, Austria; Thue Kvorning, Team Danmark, Denmark

This article was submitted to Exercise Physiology, a section of the journal Frontiers in Physiology

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

The purpose of the present study was to compare time results from a roller-skiing double poling (DP) time trial with different physiological variables, muscular strength variables, and DP characteristics in both male and female young competitive skiers with the same relative training background. In order to do this, 28 (16 women and 12 men) well-trained 16–25-year-old cross-country skiers from three Norwegian high schools for skiers, as well as local high performance competitive skiers from the South-East of Norway were recruited to participate in the study. All participants were tested for; maximal oxygen uptake in running, Peak oxygen uptake in DP, lactate threshold in DP, DP economy, time to voluntary exhaustion in DP, force analyses in DP, one repetition maximum and power output in pulldown, and leg press and a time trial during DP roller skiing. The results expressed strong correlations between roller skiing time trial performance and maximal strength in pull-down, both independent (_{xy} = −0.83, _{xy–z} = −0.50, _{xy} = 0.78, _{xy} = −0.71, _{xy} = −0.48,

Cross-country skiing is an aerobic endurance sport, and the contribution from the aerobic system is approximately 70 – 95% (_{2}_{max}), and a good work economy, which both contributes to a high velocity at lactate threshold (LT) (

Previous studies have focused on determining factors for performance in cross-country skiing and have found strong correlations between VO_{2}_{max} and performance (

Some previous studies have shown a relationship between maximal strength

Double poling (DP) is a high-speed cross-country skiing technique. Total racetime now contains a much larger percentage of DP than only a few years ago (

In previous studies on DP in cross-country skiers, some DP characteristics that are linked to maximal muscular strength have been identified (

Male and female athletes at the same relative performance level show sex differences in both VO_{2}_{max} (

The purpose of the present study was therefore to compare roller-skiing time trial (TT) performance with different physiological variables, muscular strength variables, and DP characteristics in both male and female young competitive skiers with the same relative training background. The hypothesis was that maximal upper body strength would significantly impact DP characteristics and performance.

The main objective of this cross-sectional study was to evaluate correlations between performance in DP cross-country roller skiing and different physiological variables, muscular strength variables, and DP characteristics in both male and female young competitive skiers with the same training background. Comparisons between male and female skiers as well as correlation analyses both independent of and corrected for sex, were thus performed.

A total of 28 (16 women and 12 men) well-trained 16–25-year-old cross-country skiers from three Norwegian high schools for skiers, as well as local high performance competitive skiers from the South-East of Norway participated in this study (

Characteristics of skiers (

BW (kg) | 70.1 ± 7.5 | 10.7 | 65.9 ± 5.2 | 7.9 | 73.2 ± 7.5^{∗∗} |
10.2 |

Age (years) | 18.5 ± 1.3 | 7.0 | 18.3 ± 1.2 | 6.6 | 18.6 ± 1.4 | 7.5 |

VO_{2max} running |
||||||

Ml⋅kg^{–1}⋅min^{–1} |
64.0 ± 10.4 | 16.3 | 53.8 ± 4.4 | 8.2 | 72.2 ± 5.1^{∗∗} |
7.1 |

_{2max}, maximal oxygen uptake; Ml⋅kg

^{–1}⋅min

^{–1}, milliliters per kilogram BW per minute.

^{∗∗}

In order to evaluate physiological and technical variables related to performance in DP, the following tests were carried out; VO_{2}_{max} running, VO_{2}_{peak} DP, LT in DP, DP economy (C_{DP}), time to voluntary exhaustion in DP in the ramp VO_{2}_{peak} test, force analyses in DP, one repetition maximum (1RM) and power output in pulldown and leg press, and performance during a DP roller skiing time trial (TT).

The skiers were tested over two consecutive days. Day one consisted of an incremental VO_{2}_{max} test in running and a DPTT test with 1-h rest in between. The subjects started at an intensity of 8–12 km⋅h^{–1} and a 6% inclination. Every 30 s the inclination increased by 1% until 8% inclination was reached. Then the speed was increased by 0.5 km⋅h^{–1} every 30 s. The test terminated at voluntary fatigue, and additionally heart rate (HR) ≥98% of HR_{max}, respiratory exchange ratio (RER) ≥1.05, as well as a plateau of the VO_{2} curve was used to evaluate if VO_{2}_{max} was obtained (_{2} measurements were made by the metabolic test system, Metalyzer II Cortex (Biophysic GmbH, Leipzig, Germany), with a mixing chamber. The treadmill used for running was a Woodway PPS 55 sport (Waukesha, WI, United States). All HR measurements were made by Polar s610 HR monitors (Kempele, Finland).

The double poling time trial test took place in a paved roller ski course track of 940 m with a height difference of 11 m. The subjects completed six laps, totaling 5640 m. This test was organized as an interval start with 30 s between each subjects. The subjects were told to use the DP technique throughout the whole test, and drafting was not allowed (using cycling TT rules). In this test, differences in temperature and humidity in between test days, may lead to differences in rolling resistance. Therefore, we performed a calibration test to calculate a correction factor. One of the test leaders conducted a 50 m roller-timing test with the same roller skis immediately after the time trial test every test day. The test was conducted in a tucked position, with the same test person every day, in a gentle slope, approximately 10%, and with time measured by use of photocell equipment (Musclelab system, Ergotest Inovation, Porsgrunn, Norway). Ten runs were performed for each test, ensuring proper warm up of the wheels, and the average time of the last three runs was used to calculate the correction factor.

The second day of testing consisted of a DP test on a cross-country skiing treadmill, (Rodby RL 2700E, Rodby Innovation, Vänge, Sweden) and two maximal strength tests with 1-h rest in between. The subjects were acquainted to the cross-country skiing treadmill by use of a 30-min workout ahead of the pretest. The first 15 – 20 min consisted of 3–5 four-minute submaximal work periods. Whole blood lactate concentration was measured with a Lactate Scout+ (SensLab GmbH, Leipzig, ray Inc., Kyoto, Japan). Then C_{DP}, force measurements and DP characteristics were evaluated. By use of a force transducer, measurements of force and DP characteristics, were possible. The force transducers were integrated in the poles and is a part of the Musclelab system (Ergotest Innovation, Porsgrunn, Norway). The dimension of the force transducer was 4 cm of length and 2 cm in diameter, placed 8 cm below the grip bar, as an integrated part of the pole. Outside the force transducer, a sender with the dimension 4 cm × 4.6 cm × 1 cm was placed. The total weight of the system added 100 g to the pole. The sender communicated by a Nordic semiconductor Gazell stack with a 2.4 GHz band (Nordic Semiconductor, Norway) with the Musclelab system, with a sampling rate of 200 Hz and a resolution of 14 bits. Over all accuracy was 0.9% of full scale. Test retest reliability was checked at our lab, exhibiting a standard error mean of <1%.

The system was calibrated by use of two different external weight loads on top of the pole placed in a vertical position, while the other end of the pole was placed on the force platform for a secondary control. The reading from the sensor of the pole unloaded was recorded and then the reading from the sensor of the pole with external load was recorded. Force was then computed using the formula F = (signal – offset) gain. The subjects started at a work intensity assumed to represent 50 – 70% of their VO_{2}_{peak} in DP, corresponding to 4% inclination and 11.5 km⋅h^{–1} for men and 6 or 7 km⋅h^{–1} for women. Every 4 min after the first step, the speed was increased by 1–3 km⋅h^{–1}, until the protocol terminated at a lactate level above the subjects’ LT. LT was defined as the warm up lactate value (i.e., the lowest measured lactate value) + 2.3 mmol L^{–1}. LT was expressed as the VO_{2} in% of VO_{2}_{peak} DP (%VO_{2}_{peak}), whereas the velocity at LT was expressed as km.h^{–1}. This is in accordance with the protocol proposed by ^{–1} with the YSI apparatus. As the constant difference in [La−]_{b} between whole blood and hemolyzed blood is 40%, the 1.5 mmol L^{–1} measured by YSI equals 2.3 mmol L^{–1} measured by Lactate Scout+. The advantage of using individual warm-up values compared with e.g., a fixed 4 mmol L^{–1}, is that this is less vulnerable to day-to-day variations in subjects [La−]_{b}, as previously discussed in _{DP}, were made between minute 3:00 and 03:20 in each work period. The force transducer measured force through the poles, DP frequency, and CT. C_{DP} was calculated as oxygen consumption at LT. All DP characteristic measurements were performed at the same relative intensity, i.e., at LT velocity.

Maximal aerobic speed (MAS) was calculated based on the oxygen consumptions measured in the submaximal work periods and the VO_{2}_{peak} in DP, and was defined as the velocity where the horizontal line representing VO_{2}_{peak} meets the extrapolated linear regression representing the sub maximal VO_{2} measured in the LT assessment. The same method was used for cycling in _{2}_{peak}DP/C_{DP}. Since _{2peak} DP/C_{DP} is expressed as a velocity (m min^{–1}).

One minute after the last submaximal work period, the subjects carried out an all-out test where time to exhaustion and VO_{2}_{peak} in DP were measured. This test was implemented as an incremental ramp protocol. The output speed was set to 11.5 and 6 km⋅h^{–1} for men and women, respectively. The inclination was set to 6%, and remained constant through the whole test. The speed was increased by 1 km⋅h^{–1} every 30 s until 18 km⋅h^{–1} (men) and 10 km⋅h^{–1} (women) were reached. The speed was then increased by 0.5 km⋅h^{–1} every 30 s until voluntary exhaustion. The subjects were encouraged to perform their best. Voluntary exhaustion was defined as the point where the subjects could no longer manage to keep the position at the treadmill, but slowly moved backward reaching a pre-defined mark 1 m behind their original position on the mill. The time to exhaustion was registered and the mean of the two subsequent highest registered VO_{2}-values, each representing 10 s intervals by the mixing chamber, was representing VO_{2}_{peak} DP.

After a rest period of minimum 60 min, the subjects were then tested for 1RM and power output in leg press (OPS161 Interchangeable leg press, Vertex, United States), and pulldown (Gym 2000, Vikersund, Norway). From pilot testing in _{2}_{max} and MAS testing in running and cycling, compared to 1RM without these prior tests. Leg press was chosen as a measure of lower body maximal strength for several reasons. More specialized DP related exercises such as hip flexion involves both the lower body, truncus and to some extent the upper body. The more specialized the exercises, the more practice is needed to perform valid and reliable 1RM tests. Only two maximal strength tests, one for the upper body and one for the lower body were chosen, due to the large total number of tests in this study.

Each lift was performed with a controlled slow eccentric phase, a complete stop of movement for approximately 1 s in the lowest position (leg press) or the highest position (pulldown), followed by a maximal mobilization of force in the concentric phase. The measurements of lifting time, distance of work, and thus power output were performed using the Muscle Lab system (Ergotest Innovation AS, Porsgrunn, Norway). Sensors were placed vertically below the center of the weight loads in both leg press and pulldown, and also at the actual center of the weight loads. Each strength test started using 10 reps at a weight load assumed to be approximately 50% of 1RM. After 3 min of rest: 5 reps at approximately 60% 1RM, then 3 reps at approximately 70% 1RM, 2 reps at approximately 80% and at least 1 rep at estimated 1RM with 3 min rest in between. From there on: 1 rep at a weight load increased by 2.5 – 10 kg from the subsequent lift, followed by 3 min of resting, until reaching 1RM. The time spent in each lift, as well as the work distance was measured. As the external force of each lift is represented by the weight of the lifted bars, the power output can be calculated and expressed as N m s^{–1} or watt (W).

Normality was tested by use of -plots and found to represent normal distributions for the main variables (TT performance, maximal strength and VO2_{peak} DP). Values were thus expressed descriptively as mean ± SD. Inter-individual variability was expressed as coefficient of variance (CV). Correlations were expressed as the correlation factor _{xy}, whereas the correlation factor in partial correlations corrected for sex has been denoted _{xy–z}. The practical (clinical) implication of the relations displayed by the

Test results in the different variables are presented both as total and per sex in

Test results (

TT_{DP} (s) |
899.4 ± 152.7 | 17.0 | 1032.8 ± 134.6 | 13.0 | 799.4 ± 61.5^{∗∗} |
7.7 |

_{2max} running |
||||||

L⋅min^{–1} |
4.48 ± 1.0 | 22.3 | 3.54 ± 0.39 | 11.0 | 5.22 ± 0.63^{∗∗} |
12.1 |

Ml⋅kg^{–1}⋅min^{–1} |
64.0 ± 10.4 | 16.3 | 53.8 ± 4.4 | 8.2 | 72.2 ± 5.1^{∗∗} |
7.1 |

Ml⋅kg^{–0.67}⋅min^{–0.67} |
259.6 ± 46.4 | 17.9 | 213.9 ± 17.9 | 8.4 | 296.2 ± 22.8^{∗∗} |
7.7 |

_{2peak} DP |
||||||

L⋅min^{–1} |
3.93 ± 0.89 | 22.6 | 3.10 ± 0.36 | 11.6 | 4.54 ± 0.62^{∗∗} |
13.7 |

Ml⋅kg^{–1}⋅min^{–1} |
55.7 ± 9.1 | 16.3 | 47.3 ± 5.2 | 11.0 | 62.0 ± 5.3^{∗∗} |
8.5 |

Ml⋅kg^{–0.67}⋅min^{–0.67} |
226.6 ± 40.5 | 18.2 | 188.0 ± 19.8 | 10.5 | 255.5 ± 24.1^{∗∗} |
9.4 |

Fract util DP (%VO_{2max}) |
86.9 ± 7.3 | 7.9 | 87.9 ± 5.9 | 6.7 | 86.0 ± 8.4 | 9.8 |

_{DP} |
||||||

Ml⋅kg^{–1}⋅meter^{–1} |
0.183 ± 0.023 | 12.6 | 0.192 ± 0.020 | 10.4 | 0.177 ± 0.023 | 13.0 |

Ml⋅kg^{–0.67}⋅meter^{–1} |
0.742 ± 0.087 | 11.7 | 0.763 ± 0.084 | 11.0 | 0.727 ± 0.089 | 12.2 |

MAS (km h^{–1}) |
18.6 ± 4.1 | 22.0 | 14.9 ± 1.7 | 11.4 | 21.4 ± 3.0^{∗∗} |
14.0 |

%VO_{2peak} |
79.0 ± 9.0 | 11.3 | 81.1 ± 4.3 | 5.3 | 78.3 ± 11.4 | 14.6 |

Km⋅h^{–1} |
14.6 ± 2.7 | 18.5 | 12.1 ± 1.2 | 9.9 | 16.5 ± 1.7^{∗∗} |
10.3 |

Km⋅h^{–1} calc. (MAS⋅%VO_{2peak}) |
14.6 ± 2.7 | 18.5 | 12.1 ± 1.3 | 10.0 | 16.5 ± 1.8^{∗∗} |
10.4 |

_{DP} |
||||||

Peak (N) | 381.8 ± 124.0 | 32.5 | 277.6 ± 80.7 | 28.8 | 459.9 ± 87.7^{∗∗} |
19.1 |

Average during CT (N) | 169.3 ± 47.8 | 28.2 | 145.6 ± 37.7 | 25.9 | 187.1 ± 47.7^{∗∗} |
25.5 |

RFD (N⋅s^{–1}) |
2620 ± 1233 | 47.0 | 2063 ± 1230 | 59.7 | 3038 ± 1091^{∗} |
35.9 |

Freq. at LT (St⋅meter^{–1}) |
0.239 ± 0.055 | 23.0 | 0.286 ± 0.047 | 6.5 | 0.204 ± 0.027^{∗∗} |
13.2 |

Freq. at LT (St⋅s^{–1}) |
0.938 ± 0.100 | 10.7 | 0.956 ± 0.137 | 14.3 | 0.925 ± 0.060 | 6.5 |

CT (s) | 0.353 ± 0.073 | 20.7 | 0.388 ± 0.092 | 23.7 | 0.327 ± 0.040^{∗} |
12.2 |

1RM pull-down (kg) | 86.2 ± 19.3 | 22.4 | 66.0 ± 8.8 | 13.2 | 99.7 ± 10.3^{∗∗} |
10.3 |

1RM leg-press (kg) | 278.1 ± 54.9 | 19.7 | 235.6 ± 37.5 | 15.9 | 303.7 ± 47.8^{∗∗} |
17.0 |

Power pull-down (W) | 439.2 ± 122.3 | 27.8 | 323.1 ± 54.1 | 16.7 | 516.5 ± 87.9^{∗∗} |
17.0 |

Power leg-press (W) | 609.7 ± 157.2 | 25.8 | 466.6 ± 70.2 | 15.0 | 694.0 ± 130.3^{∗∗} |
18.7 |

_{DP}, double poling time trial on roller skies; S, seconds; BW, body weight; Kg, kilograms; VO

_{2max}, maximal oxygen uptake; L⋅min

^{–1}, liters per minute; Ml⋅kg

^{–1}⋅min

^{–1}, milliliters per kg BW per minute; Ml⋅kg

^{–0.67}⋅min

^{–0.67}, milliliters per kg BW raised to the power of 0.67 per minute; DP, double poling; VO

_{2peak}, peak oxygen uptake during DP; Fract Util, fractional utilization of VO

_{2peak}vs. VO

_{2max}; C

_{DP}, oxygen cost of DP at LT; Ml⋅kg

^{–1}⋅meter

^{–1}, milliliters per kg BW per meter; Ml⋅kg

^{–0.67}⋅meter

^{–1}, milliliters per kg BW raised to the power of 0.67 per meter; MAS, maximal aerobic speed calculated as peak oxygen uptake during DP divided on C

_{DP}; km⋅h

^{–1}, kilometers per hour; LT, lactate threshold; N, Newton; CT, contact time; RFD, rate of force development; N⋅s

^{–1}, Newton per second; Freq, frequency; St⋅meter

^{–1}, strokes per meter; S, seconds; 1RM, one repetition maximum; W, watt.

^{∗}

^{∗∗}

Independent of sex, strong correlations were found between 1RM pulldown and TT performance (_{xy} = 0.83, _{xy} = 0.81, _{xy–z} = 0.62, _{xy–z} = −0.56, _{xy} = −0.65, _{xy} = 0.85,

Correlations with time trial performance (

_{xy} |
_{xy–z} |
||||

VO_{2max} running |
|||||

L⋅min^{–1} |
–0.77 | 10.9 | <0.01 | –0.37 | 0.07 |

Ml⋅kg^{–1}⋅min^{–1} |
–0.77 | 11.0 | <0.01 | –0.31 | 0.12 |

Ml⋅kg^{–0.67}⋅min^{–0.67} |
–0.79 | 10.4 | <0.01 | –0.38 | 0.06 |

_{2peak} DP |
|||||

L⋅min^{–1} |
–0.78 | 10.7 | <0.01 | –0.42 | 0.05 |

Ml⋅kg^{–1}⋅min^{–1} |
–0.77 | 11.0 | <0.01 | –0.38 | 0.03 |

Ml⋅kg^{–0.67}⋅min^{–0.67} |
–0.80 | 10.4 | <0.01 | –0.44 | 0.02 |

_{DP} |
|||||

Ml⋅kg^{–1}⋅meter^{–1} |
0.40 | 15.8 | 0.04 | 0.24 | 0.23 |

Ml⋅kg^{–0.67}⋅meter^{–1} |
0.28 | 16.5 | 0.14 | 0.19 | 0.34 |

MAS (km h^{–1}) |
–0.80 | 10.3 | <0.01 | –0.48 | 0.01 |

%VO_{2peak} |
0.22 | 16.8 | 0.26 | 0.16 | 0.44 |

Km⋅h^{–1} |
–0.78 | 10.7 | <0.01 | –0.40 | 0.04 |

Km⋅h^{–1} calc. (MAS⋅%VO_{2peak}) |
–0.77 | 10.7 | <0.01 | –0.39 | 0.04 |

_{DP} |
|||||

Peak (N) | –0.75 | 11.5 | <0.01 | –0.41 | 0.04 |

Average during CT (N) | –0.45 | 15.5 | 0.02 | –0.19 | 0.33 |

RFD (N⋅s^{–1}) |
–0.42 | 15.4 | 0.03 | –0.19 | 0.35 |

%of 1RM pull-down | –0.65 | 13.4 | <0.01 | –0.56 | <0.01 |

Freq. at LT (St⋅meter^{–1}) |
0.55 | 14.3 | 0.01 | –0.07 | 0.73 |

Freq. at LT (St⋅s^{–1}) |
–0.18 | 16.9 | 0.36 | –0.48 | 0.01 |

CT (s) | 0.69 | 12.5 | <0.01 | 0.62 | <0.01 |

1RM pull-down (kg) | –0.83 | 10.5 | <0.01 | –0.50 | 0.02 |

1RM leg-press (kg) | –0.53 | 15.3 | 0.01 | –0.09 | 0.68 |

Power pull-down (W) | –0.81 | 10.7 | <0.01 | –0.49 | 0.02 |

Power leg-press (W) | –0.68 | 13.2 | <0.01 | –0.27 | 0.21 |

_{2max}, maximal oxygen uptake; L⋅min

^{–1}, liters per minute; Ml⋅kg

^{–1}⋅min

^{–1}, milliliters per kg BW per minute; Ml⋅kg

^{–0.67}⋅min

^{–0.67}, milliliters per kg BW raised to the power of 0.67 per minute; DP, double poling; VO

_{2peak}, peak oxygen uptake during DP; C

_{DP}, oxygen cost of DP at LT; Ml⋅kg

^{–1}⋅meter

^{–1}, milliliters per kg BW per meter; Ml⋅kg

^{–0.67}⋅meter

^{–1}, milliliters per kg BW raised to the power of 0.67 per meter; MAS, maximal aerobic speed calculated as peak oxygen uptake during DP divided on C

_{DP}; km⋅h

^{–1}, kilometers per hour; LT, lactate threshold; N, Newton; CT, contact time; RFD, rate of force development; N⋅s

^{–1}, Newton per second; Freq. % of 1RM pull-down, percentage of one repetition maximum in pull-down during one full DP cycle; Frequency. St⋅meter

^{–1}, strokes per meter; 1RM, one repetition maximum; W, watt.

Within sex correlations with time trial performance.

_{xy} |
_{xy} |
|||

_{2max} running |
||||

L⋅min^{–1} |
–0.39 | 0.16 | –0.49 | 0.11 |

Ml⋅kg^{–1}⋅min^{–1} |
0.16 | 0.57 | –0.70 | 0.01 |

Ml⋅kg^{–0.67}⋅min^{–0.67} |
–0.10 | 0.71 | –0.66 | 0.02 |

_{2peak} DP |
||||

L⋅min^{–1} |
–0.57 | 0.02 | –0.46 | 0.13 |

Ml⋅kg^{–1}⋅min^{–1} |
–0.26 | 0.33 | –0.51 | 0.09 |

Ml⋅kg^{–0.67}⋅min^{–0.67} |
–0.43 | 0.09 | –0.53 | 0.08 |

_{DP} |
||||

Ml⋅kg^{–1}⋅meter^{–1} |
0.53 | 0.03 | 0.07 | 0.82 |

Ml⋅kg^{–0.67}⋅meter^{–1} |
0.40 | 0.12 | 0.08 | 0.81 |

MAS (km h^{–1}) |
–0.62 | 0.01 | –0.56 | 0.06 |

%VO_{2peak} |
0.51 | 0.04 | –0.27 | 0.40 |

Km⋅h^{–1} |
–0.18 | 0.51 | –0.71 | 0.01 |

Km⋅h^{–1} calc. (MAS⋅%VO_{2peak}) |
–0.23 | 0.42 | –0.74 | 0.01 |

_{DP} |
||||

Peak (N) | –0.65 | 0.01 | –0.31 | 0.33 |

Average during CT (N) | –0.17 | 0.54 | –0.26 | 0.41 |

RFD (N⋅s^{–1}) |
0.13 | 0.63 | –0.38 | 0.22 |

% of 1RM pull-down | –0.44 | 0.10 | –0.18 | 0.62 |

Freq. at LT (St⋅meter^{–1}) |
0.38 | 0.14 | –0.23 | 0.47 |

Freq. at LT (St⋅s^{–1}) |
0.46 | 0.07 | –0.75 | 0.01 |

CT (s) | –0.21 | 0.44 | 0.85 | <0.01 |

1RM pull-down (kg) | –0.52 | 0.05 | –0.52 | 0.12 |

1RM leg-press (kg) | –0.24 | 0.39 | 0.02 | 0.95 |

Power pull-down (W) | –0.47 | 0.08 | –0.76 | 0.01 |

Power leg-press (W) | –0.27 | 0.34 | –0.47 | 0.21 |

_{2max}, maximal oxygen uptake; L⋅min

^{–1}, liters per minute; Ml⋅kg

^{–1}⋅min

^{–1}, milliliters per kg BW per minute; Ml⋅kg

^{–0.67}⋅min

^{–0.67}, milliliters per kg BW raised to the power of 0.67 per minute; DP, double poling; VO

_{2peak}, peak oxygen uptake during DP; C

_{DP}, oxygen cost of DP at LT; Ml⋅kg

^{–1}⋅meter

^{–1}, milliliters per kg BW per meter; Ml⋅kg

^{–0.67}⋅meter

^{–1}, milliliters per kg BW raised to the power of 0.67 per meter; MAS, maximal aerobic speed calculated as peak oxygen uptake during DP divided on C

_{DP}; km⋅h

^{–1}, kilometers per hour; LT, lactate threshold; N, Newton; CT, contact time; RFD, rate of force development; N⋅s

^{–1}, Newton per second; Freq. % of 1RM pull-down, percentage of one repetition maximum in pull-down during one full DP cycle; Frequency. St⋅meter

^{–1}, strokes per meter; 1RM, one repetition maximum; W, watt.

The skiers with the highest 1RM pulldown also had the highest PF (_{xy} = 0.78, _{xy} = 0.48, _{xy} = −0.71,

Correlations with maximal strength in pull-down (

_{xy} |
_{xy–z} |
||||

_{DP} |
|||||

Ml⋅kg^{–1}⋅meter^{–1} |
−0.36 | 21.5 | 0.08 | −0.18 | 0.40 |

Ml⋅kg^{–0.67}⋅meter^{–1} |
−0.20 | 22.5 | 0.36 | 0.02 | 0.94 |

MAS (km h^{–1}) |
−0.74 | 15.2 | <0.01 | −0.21 | 0.32 |

_{DP} |
|||||

Peak (N) | 0.78 | 14.2 | <0.01 | 0.50 | 0.01 |

Average during CT (N) | 0.39 | 21.1 | 0.05 | 0.09 | 0.64 |

RFD (N⋅s^{–1}) |
0.31 | 21.8 | 0.13 | 0.02 | 0.92 |

% of 1RM pull-down | −0.54 | 19.3 | 0.01 | −0.40 | 0.05 |

Freq. at LT (St⋅meter^{–1}) |
−0.71 | 16.1 | <0.01 | −0.20 | 0.35 |

Freq. at LT (St⋅s^{–1}) |
−0.09 | 22.9 | 0.66 | 0.08 | 0.72 |

CT (s) | −0.48 | 20.1 | 0.02 | −0.29 | 0.17 |

Power pull-down (W) | 0.92 | 8.9 | <0.01 | 0.77 | 0.02 |

_{DP}, oxygen cost of DP at LT; Ml⋅kg

^{–1}⋅meter

^{–1}, milliliters per kg BW per meter; Ml⋅kg

^{–0.67}⋅meter

^{–1}, milliliters per kg BW raised to the power of 0.67 per meter; MAS, maximal aerobic speed calculated as peak oxygen uptake during DP divided on C

_{DP}; km⋅h

^{–1}, kilometers per hour; N, Newton; CT, contact time; RFD, rate of force development; N⋅s

^{–1}, Newton per second. % of 1RM pull-down, percentage of one repetition maximum in pull-down during one full DP cycle; Freq, frequency; St⋅meter

^{–1}, strokes per meter; 1RM, one repetition maximum; W, watt.

The main findings in the present study are the correlations between roller skiing DPTT performance and maximal strength in pull-down, both independent and dependent of sex. Higher maximal upper body strength was related to higher PF in DP, lower DP frequency, and shorter CT.

The novelty of the present study was the finding of a strong correlation between maximal strength (1RM) in pulldown

For the strength variables, strong correlations were found between 1RM pulldown and TT performance (_{xy} = 0.83) and maximal power output in pulldown and performance (_{xy} = 0.81). SEE was 10.5 and 10.7, respectively. The ^{2} values indicate that both variables predicts TT performance by 69%, and the SEE shows this to be outside a margin of approximately 10.5% of either 1RM or power output results. The 10.5% corresponds to 9 kg in pulldown. This implies that if one skier was at least 9 kg’s stronger than another in pulldown, he or she would perform better in TT. Regarding DP characteristics, PF (_{xy} = −0.75), PF during DP as a percentage of 1RM (_{xy} = −0.65) and CT during DP (_{xy} = 0.69) correlated best with TT performance. The relationship between PF and TT is in accordance with previous studies demonstrating that faster skiers had higher PF, or that higher PF related to peak skiing speeds (_{2}_{max} in running and VO_{2}_{peak} DP expressed as ml⋅kg^{–0.67}⋅min^{–1} (_{xy} = 0.79 and _{xy} = 0.80, respectively). The SEE value of 10.4%, implies that if one skier had at least 23 ml⋅kg^{–0.67}⋅min^{–1} higher VO_{2}_{peak} DP than another, he or she would perform better in TT. There was also a strong correlation between velocity at LT and performance (_{xy} = 0.78). All of these correlations were found in the heterogeneous cohort including both sexes.

When corrected for sex, the aerobic endurance variables decreased substantially in predicting TT performance. The two variables CT (_{xy–z} = 0.62) and PF as a percentage of 1 RM pull down during DP (_{xy–z} = −0.56) expressed the best correlation with TT performance when corrected for sex. When correcting for sex, the cohorts are more homogeneous since males and females results are clustered in to two groups. This was apparent when comparing coefficient of variance (CV) (_{2}_{max} (running) and VO_{2}_{peak} DP when including both sexes, were both 18%. When separated into males and females, the CV values were cut in half. This phenomenon was not so obvious regarding strength and DP characteristics.

Both normal and partial correlations were performed in the present study. When partial correlations were still significant, this would strengthen the normal correlations by showing that it was not confounded by sex. However, the partial correlations only showed to what extent the normal correlations were confounded by sex, and so correlations with TT performance were also analyzed within sexes. These correlations should be handled with caution, due to the low number of skiers within each sex. The division into two separate groups also caused a greater degree of homogeneity in almost all variables. As a result of this and the low number in each group, correlations across sexes were weakened or disappeared. Those analyzes are merely included for informative reasons, but not addressed further in this discussion. Maximal strength in pulldown had approximately the same correlation with TT performance in both males and females. However, the relationships between the utilization of this maximal strength and TT performance seemed to differ, as CT correlated well in females but not males, and PF in DP correlated well in males but not females.

Both males and females had approximately the same correlation between MAS and TT performance. However, TT performance seemed to depend mostly on VO_{2}_{peak} but not C_{DP} in females. In males, TT performance seemed to depend mostly on C_{DP}, but not VO_{2}_{peak}.

The importance of a high VO_{2}_{max} in running for TT performance in the present study, is in accordance with several previous studies (_{2}_{max}; 80–90 ml⋅kg^{–1}⋅min^{–1} and 70–80 ml⋅kg^{–1}⋅min^{–1} for men and women world-class cross-country skiers, respectively (_{2}_{max} values in the present study were 53.8 ± 4.4 ml⋅kg^{–1}⋅min^{–1} and 72.2 ± 5.1 ml⋅kg^{–1}⋅min^{–1} for women and men, respectively. Regarding the skiers cost of skiing, the present study found that C_{DP} did not correlate well with TT performance (_{DP} in the present study is further highlighted when including C_{DP} in the MAS equation (VO_{2}_{peak} DP/C_{DP}). MAS did not correlate better with TT performance than VO_{2}_{peak} DP alone (_{xy} = −0.80), independent of sex.

Maximal oxygen consumption at LT in% of VO_{2}_{peak} DP did not correlate with TT performance in the present study. However, velocity at LT correlated strongly (_{xy} = −0.78) with TT performance. This is in accordance with _{2}_{peak}, while LT in% of VO_{2}_{peak} alone did not explain LT velocity. The same results were echoed in the present study. When applying the same equation for velocity at LT (MAS LT%), this correlated nearly perfect (_{xy} = 0.99) with the actually measured LT velocity. This implies that it is not LT _{2}_{peak} that predicts TT performance.

Although C_{DP} correlated weakly with TT performance in the present study, variables previously shown to affect work economy in other sports (_{DP}. That maximal strength _{xy} = −0.65).

The skiers with the highest 1RM pulldown also had the highest PF (_{xy} = 0.78) during DP (_{xy} = 0.48), and the lowest DP frequency measured as strokes per meter (_{xy} = −0.71). Also CT correlated with LT velocity (_{xy} = −0.53), indicating that the fastest skiers had the shortest CT, although CT did not correlate with TT directly. Therefore, this shortened CT in faster skiers might be basically explained by the higher skiing speeds. A shorter CT and a lower frequency allows for a shorter contraction time and a longer transit time during each DP cycle. This could theoretically lead to better circulation and thus O_{2} and substrate deliverance as well as better clearance of lactic acid (

Even though DP may be considered a whole body exercise involving muscle mass from feet to neck, the leg press results in the present study seemed to have much less impact on TT performance and MAS than pulldown. This does not necessarily imply that lower body muscles do not have an impact on DP. Based on EMG activity in lower body muscles,

Power output was calculated as the product of force and work distance divided by time. The power output results in leg press in the present study may seem low. This is due to the measurements of work distance being performed vertically when the lifting direction is diagonal in the leg press apparatus. In studies were the lifting direction is vertical like squat in e.g.,

Since the male and female participants in the present study represented a higher and a lower TT performance level, a comparison of the results from the two sexes may be used to discuss the importance of factors predicting DP performance. The male and female skiers were at the same age, and being recruited from the same teams and high schools, their training background was relatively similar. TT performance was 23% better in males than in females. The sex difference in the present study was therefore in accordance with results from _{2}_{peak} can be expressed as ml.kg^{–1}.min^{–1}, and DP as ml.kg^{–1}.m^{–1}, the product of denominations equals m min^{–1}, which may also be expressed as km.h^{–1}. The gender difference in MAS, could therefore be explained by 18% difference in VO_{2}_{p}_{eak} DP, and a none significant 8% difference in C_{DP}. These differences is somewhat lower than presented in

Males where 34% stronger in pulldown than females, which is a somewhat lower difference than the 50% reported in

The practical implications of the present study is to acknowledge maximal upper body strength as a possible performance determining factor in DP. We suggest including maximal strength training in the cross-country skiers training programs, but the effect of this needs further evaluation in future studies. We recommend few repetitions (2–5) in 3–5 series with maximal mobilization in the concentric phase, with relatively long (2–3 min) pauses in between. These principles have in previous studies been shown to improve work economy as well as maximal strength (

In conclusion maximal upper body strength was shown to have a significant impact on DP roller skiing performance, both dependent and independent of sex, and both dependent and independent of C_{DP}. Higher maximal upper body strength was related to higher DP peak forces, lower DP frequency and shorter CT.

The datasets generated for this study are available on request to the corresponding author.

This study was carried out in accordance with the recommendations of the Institutional Review Board (IRB) at the University of South-Eastern Norway with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the Institutional Review Board (IRB) at the University of South-Eastern Norway.

AS, J-MJ, ØS, JH, GP, and MB participated significantly in the planning and designing of the study. AS, J-MJ, ØS, and JH participated in the data analysis and writing of the manuscript. AS, ØS, J-MJ, and MG participated in the data collection and analysis. All authors read and approved the manuscript.

JH was employed by the company Myworkout. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

We wish to thank all the participants in this study for their great effort.

_{DP}

oxygen cost in double poling

contact time

double poling

heart rate

_{max}

maximal heart rate

lactate threshold

maximal aerobic speed

peak force

respiratory exchange ratio

_{peak}

peak respiratory exchange ratio

double poling time trial

_{2}

oxygen uptake

_{2max}

maximal oxygen uptake in running

_{2peak}DP

peak oxygen uptake in double poling.