Over the last 15 yr, the use of double poling (DP) has developed substantially (11) and has become a main technique in cross-country ski racing (21). The main reasons for this are considerable advances in ski equipment, better technical preparation of skis and tracks, and improvement in elite skiers' upper body capacity (23). These improvements have further been stimulated by the introduction of several new racing forms, for example, the classical sprint and the mass start events (26), which have increased demands on a skier's ability to reach high DP velocities in certain parts of a classic race, especially in the finish. In previous studies (10,11), we analyzed current technique developments and defined high-performance DP characteristics as having more accentuated flexion-extension patterns in the lower body joints, shorter poling and longer recovery times, and higher forces and impulses of force that provide a more explosive pole push off.
A fundamental question has arisen from these changed characteristics: How do modern elite skiers control DP speed? Earlier studies on different cross-country skiing techniques demonstrated that increases in submaximal velocities were associated with increases in both cycle rate and cycle length (CL), except for DP, which only had increases in poling frequency (Pf) (8,13,16). The lack of increase in CL was explained by this technique's constraints, meaning that propulsive forces can only be applied during the poling phase. At maximal DP velocities, several authors reported decreased (8,13,27) or unchanged (16) CL. Hence, there is a consensus in these studies that skiers controlled speed by adjusting only Pf. On the other hand, it is known that CL can discriminate between skiers with different performance levels (25). An increasing number of studies have shown that DP performance can be improved by an increase in CL due to improved strength or muscle power factors (7,17,19) and/or technique developments in elite skiers (10,11). The combination of these new aspects motivates a reexamination of the mechanisms behind speed control in modern DP.
In comparison to the discussion about Pf and CL adaptations to speed, there are no data that address the effects of speed on joint kinematics, and there are only a few studies that analyze speed adaptations concerning forces and more detailed cycle characteristics during DP. Millet et al. (13) reported an increase in peak pole forces and a decrease in both absolute poling and recovery times. Furthermore, they showed a preservation of relative (% cycle time) poling and recovery times across submaximal velocities followed by an increase at maximal velocity. Nilsson et al. (16) demonstrated similar results as regards absolute poling and recovery times but observed no changes in relative poling and recovery times. Of note is that neither of these studies examined elite skiers performing at a high level.
To increase understanding of how elite skiers control speed in the modern DP technique, the main hypotheses in the present study were that elite skiers adapt to higher speeds by 1) increasing both Pf and CL, 2) generating greater pole force despite shorter poling times, and 3) creating greater and faster a) elbow joint flexion and extension during pole thrust, b) hip and knee flexion around pole plant, and c) hip and knee extensions during the main part of recovery phase.
Twelve elite cross-country skiers and members of the Swedish Junior and Senior National Team (mean ± SD: 23 ± 1.6 yr, range = 20-26 yr; 178.7 ± 5.1 cm, range = 172-186 cm; 71.6 ± 7.0 kg, range = 60-84 kg) volunteered as subjects. Their mean ± SD V˙O2max was 73.1 ± 4.5 mL[middot]kg−1[middot]min−1 (range = 66-80 mL[middot]kg−1[middot]min−1), measured during diagonal roller skiing on a motor-driven treadmill (Rodby, Sodertalje, Sweden) and using an ergospirometry system (AMIS 2001; Innovision A/S, Odense, Denmark) to physiologically characterize the subjects. All subjects were experienced with roller skiing on a treadmill during training as well as testing. Before giving their written informed consent to participate in the current study, all the skiers were fully acquainted with the nature of the whole project. The methods and the experimental protocol of the study were approved by the ethics committee of Umeå University, Umeå, Sweden.
DP cycle definitions.
One double poling (DP) cycle was defined as the period from pole plant to the subsequent pole plant of the right pole determined from pole force data. Each DP cycle (C) was divided into a poling phase (PP), representing the period of pole ground contact and a recovery phase (RP) standing for arm swing. All data were averaged over eight cycles for each analyzed DP speed and each subject.
Pole and plantar forces.
All subjects used carbon racing poles. The right-hand pole, specially constructed for axial force measurements and telescopically adjustable from 140 to 165 cm, enabled the athletes to adjust the pole to their preferred individual length for classic style. The mean pole length during all DP analyses was 152 ± 5 cm (range = 144-157 cm), corresponding to 85 ± 3% of body height. The ground reaction forces directed along the pole axis were recorded at 2000 Hz and were measured by a uniaxial strain gauge load cell (Biovision, Werheim, Germany) with a measurement range of 0.5 to 50 kN, weighing 15 g, which fitted into a short (80 mm) light weight (65 g) aluminum body. Both were installed in the carbon tube directly below the pole grip to minimize moments of inertia during arm swing. The load cell calibration was performed by using a specific calibration apparatus, consisting of a small iron platform, mounted perpendicular to the pole's tube, with 10 different standard weights (5-50 kg) placed on that platform during calibration. Force curve calculations were made with linear regression analysis between the values of the force transducer and the mass of the standard weights in newtons (N). The slope of the linear regression line (R = 0.997) was taken as a factor for converting the pole voltage values into Newton values. Validation of the pole force system was achieved using the procedures proposed by Holmberg et al. (11). The mean absolute errors were about 3.8% when compared with data from an AMTI force plate (AMTI, Watertown, MA). Absolute and relative peak pole force, time-to-peak pole force, and rate of force development were determined. All relative force values were expressed as percent body weight. Vertical plantar ground reaction forces for the right foot were recorded at 100 Hz and were measured by the Pedar Mobile System (Novel GmbH, Munich, Germany). The calibration of the insoles was performed using the Pedar calibration device and has been described elsewhere (11). For detailed analysis, the total foot area was divided into forefoot and rearfoot at 50% of foot length. The relative average forefoot and rearfoot forces were calculated over the poling phase and recovery phase periods.
Elbow, hip, and knee angles (EA, HA, and KA, respectively) were measured by goniometers (potentiometers: Megatron, Munich, Germany; strain gauges: Penny & Giles Controls Ltd, Cwmfelinfach, UK) at 2000 Hz. Calibration measurements were performed five times at 90° as well as at 180° for all three joints, and angle values were calculated from the corresponding mean voltage data that showed linear relationships. A 90° angle corresponded to the forearm perpendicular to the upper arm, the thigh perpendicular to the trunk, and the shank perpendicular to the thigh. One hundred and eighty degrees corresponded with fully extended elbow and knee joints and with the thigh aligned with the trunk. All joint angle variables were collected and calculated from joint angle curves for each cycle. The joint angles recorded at specific events, that is, at pole plant, at peak pole force, at minimum and maximum angles during poling phase, and at recovery phase respectively, are defined in the corresponding figure legends. Mean angular velocity (AV) was calculated as the flexion/extension range of motion (ROM) divided by the flexion/extension duration. Cycle time (CT), absolute (s) and relative (%CT) poling and recovery time, poling frequency (Pf = 1 × CT−1), and cycle length (CL = CT × velocity) were determined and calculated from pole force data.
Data collection and data analysis.
All kinematic and kinetic data except plantar forces (Pedar Mobile System; Novel GmbH) were collected by a complete measurement system (Biovision) consisting of two input boxes with 16 channels connected to A/D converter cards (DAQ 700 A/D card −12 bit; National Instruments, Austin, TX) and two portable pocket PCs (Compaq iPAQ H3800) to store data for further off-line analysis. The processing of all data was managed by IKE-Master Software (IKE-Software Solutions, Salzburg, Austria).
All tests were performed on a motor-driven treadmill (Rodby) specially designed for roller-ski tests. To exclude variations in rolling resistance, all subjects used the same pair of roller skis (Pro-Ski C2; Sterners, Nyhammar, Sweden). DP analyses were performed at a treadmill inclination of 1° like in several previous studies (10,11). The calculation of the individually maximal DP velocity (Vmax) was based on a performed DP incremental test until voluntary exhaustion. The Vmax was calculated using the formula Vmax = Vf + [(t / 240) × 3 km[middot]h−1], where Vf is the speed of the last completed workload (km[middot]h−1), t is the duration of the last workload (s), and 3 km[middot]h−1 is the velocity difference (ΔV) between the last two workloads. Kinetic and kinematic DP analyses were performed at five successive increasing velocities of 9, 15, 21, and 27 km[middot]h−1 and the individual Vmax (29.5 ± 1.3 km[middot]h−1, range = 27.8-31 km[middot]h−1) with separate measurement runs for each speed. All runs included an approximately 15-s acceleration phase, followed by 15 complete DP cycles for data recording. A break of 3 min between each run was taken for fatigue prevention purposes. Before measurements, a standardized warm-up of 10 min was performed by all subjects with DP at 60% and 75% of individual Vmax (alternating 2 × 2 min 60%, 2 × 3 min 75%). All tests of the study were performed during a period of 12 d.
All data were checked for normality, calculated with conventional procedures and presented as mean values (X¯) and SD. Repeated-measures ANOVA were calculated to analyze changes in all defined variables over the five different DP velocities. When global significance over time was determined, Bonferroni post hoc analysis was used to determine changes in the measured variables over velocities. The statistical level of significance was set at P < 0.05 for all analyses. All statistical tests were processed using the SPSS 12.0 Software (SPSS Inc., Chicago, IL) and the Office Excel 2003 (Microsoft Corporation, Redmond, WA).
DP cycle characteristics.
All cycle variables are shown in Table 1 and Figures 1A and 2B. Cycle time (CT) and absolute poling time gradually decreased down to 1.02 ± 0.10 and 0.28 ± 0.03 s, respectively, as speed increased to 27 km[middot]h−1 (all P values <0.05). There were no further decreases at Vmax. Absolute recovery time remained unchanged at 9 and 15 km[middot]h−1. Thereafter, it shortened as speed increased up to 27 km[middot]h−1 (all P values <0.05), with no further change at Vmax. Relative poling and recovery time decreased and increased through 21 km[middot]h−1, respectively (all P values <0.05), with no change thereafter. Pf increased with speed up to Vmax (0.62 ± 0.10 to 1.08 ± 0.11 Hz; all P values <0.05). CL followed the same pattern through 27 km[middot]h−1 (4.14 ± 0.53 to 7.68 ± 0.78 m; all P values <0.05) and was then maintained. Therefore, the Pf/CL index did not change and averaged at 0.14 ± 0.04 Hz with increased velocities.
Pole and plantar force.
All force variables are shown in Table 1 and Figure 3A and B. Peak pole force increased almost twofold from 9 to 27 km[middot]h−1 (all P values <0.05), with no change thereafter. Time-to-peak pole force decreased 62% through 27 km[middot]h−1 (all P values <0.05), with no more change at Vmax while reaching a minimum value of 0.07 ± 0.01 s. At the same time, a nearly fourfold increase in rate of force development up to 3038 ± 956 N[middot]s−1 (all P values <0.05) with no further increase at Vmax was observed. Plantar force showed minor alterations with speed increase. The average relative rearfoot force during the poling phase (ARFFPPrel) increased almost two fold up to 27 km[middot]h−1 (all P values <0.05) and thereafter reached a plateau. In contrast, the average relative forefoot force (AFFFPPrel) was unchanged across all velocities.
All elbow variables are shown in Table 2 and Figure 4 (including events 1-4). EA at pole plant (EAPole_In; event 1) remained unchanged at 105° ± 16° over all velocities and skiers. The first part of the poling phase was characterized by an elbow flexion getting more accentuated with increased velocities. The EA minimum (EAmin_PP; event 3) decreased from 84° ± 25° to 65° ± 19° (P < 0.05), and the elbow flexion ROM (ROME_flex_PP; events 1-3) simultaneously increased through 27 km[middot]h−1 (P < 0.05), both with no further changes up to Vmax. The duration of EA flexion (t) shortened through 27 km[middot]h−1 (all P values <0.05), altogether resulting in a fourfold increase of flexion AV (AVE_flex_PP) up to 406 ± 165°[middot]s−1 (all P values <0.05), which was maintained thereafter. From 21 to 27 km[middot]h−1, the difference between EA at peak pole force (EAPPF; event 2) and EAmin_PP (event 3) increased (1° ± 1° to 5° ± 3°; P < 0.05), and EAPPF started to occur earlier (−0.03 ± 0.01 s; P < 0.05), all with no alterations before and after. The second part of the poling phase was an extension in the elbow joint (events 3-4) characterized by 1) a continuous decrease in extension time (t) and 2) an increase in the elbow extension ROM (ROME_ext_PP) as speed increased to 27 km[middot]h−1, which together caused an almost fourfold increase in extension AV (AVE_ext_PP) up to 548 ± 105°[middot]s−1 (all P values <0.05). There were no more changes at Vmax.
All hip variables are shown in Table 3 and Figure 5A (including events 1-7). HA at pole plant (HAPole_In; event 1) decreased from 161° ± 14° to 133° ± 12° as speed increased to 27 km[middot]h−1 (all P values <0.05) and was then maintained. This was followed by a hip flexion phase where the HA minimum during poling phase (HAmin_PP; event 3) gradually decreased from 126° ± 14° to 96° ± 13° through 27 km[middot]h−1 (all P values <0.05), with no further decrease at Vmax. The hip flexion ROM during poling phase (ROMH_flex_PP; events 1-3) did not change significantly and averaged at 37° ± 9° over all skiers and velocities, whereas the hip flexion time during poling phase (t) shortened as speed increased up to 27 km[middot]h−1 (all P values <0.05). Altogether, this resulted in an almost threefold increase of hip flexion AV during poling phase (AVH_flex_PP) up to 259 ± 55°[middot]s−1 through 27 km[middot]h−1 (all P values <0.05), with no changes thereafter. From 21 to 27 km[middot]h−1, the difference between HAPPF (event 2) and HAmin_PP (event 3) increased from 4° ± 2° to 7° ± 3° (P < 0.05), with no changes before and after, and HAPPF consistently occurred 0.07 ± 0.03 s before HAmin_PP. The following and last part of the poling phase was characterized by a small extension of 3° to 6° (events 3-4) in the hip joint over all velocities, ending with a gradually decreased HA at the end of poling phase (HAPole_Out; event 4) through 27 km[middot]h−1 (all P values <0.05), which was then maintained. The first and major part of the following recovery phase was determined by a further and accentuated hip extension (events 4-6), ending with a maintained HA maximum during recovery phase (HAmax_RP; event 6) across Vsm at 171° ± 8° and a slightly decreased value of 165° ± 7° at Vmax (9 km[middot]h−1 vs Vmax; P < 0.05). Hip extension ROM during recovery phase (ROMH_ext_RP) increased through 21 km[middot]h−1 (all P values <0.05), with no further shift through Vmax, whereas hip extension time during recovery phase (t) shortened from the lowest to the highest DP speed (P < 0.05). Therefore, hip extension AV during recovery phase (AVH_ext_RP) increased twofold up to 107 ± 23°[middot]s−1 through 27 km[middot]h−1 (all P values <0.05) and was maintained thereafter. The recovery phase of the DP cycle was completed by a hip flexion phase (events 6-7 (1′)) that continued to the poling phase of the subsequent cycle down to HAmin_PP (events 1-3 of next cycle). Hip flexion ROM during recovery phase (ROMH_flex_RP) increased more than twofold through 21 km[middot]h−1 (all P values <0.05) and was maintained thereafter, whereas flexion time was unchanged and averaged 0.13 ± 0.08 s across all velocities and skiers. Considering total hip flexion during the cycle (recovery plus poling phase; events 6-7(1)-3), the hip flexion AV (AVH_flex_C) gradually increased up to 300 ± 75°[middot]s−1 through 27 km[middot]h−1 (all P values <0.05), with no further change up to Vmax.
All knee variables are shown in Table 4 and Figure 5B (including events 1-7). The KA at pole plant (KAPole_In; event 1) decreased between 15 and 27 km[middot]h−1 (P < 0.05), with no changes before and thereafter. Pole plant was followed by a considerable knee flexion across all velocities, where KA minimum during poling phase (KAmin_PP; event 3) gradually decreased from 148° ± 12° to 129° ± 13° through 27 km[middot]h−1 (all P values <0.05) and was then maintained. The knee flexion ROM during poling phase (ROMK_flex_PP; events 1-3) was unchanged at 16° ± 6° over all velocities, but the knee flexion time during poling phase (t) shortened through 27 km[middot]h−1 (all P < 0.05), resulting in an almost threefold increase of the knee flexion AV during poling phase (AVK_flex_PP) up to 143 ± 28°[middot]s−1 at Vmax (P < 0.05). The difference between KAPPF (event 2) and KAmin_PP (event 3) increased only between 15 km[middot]h−1 and Vmax from 2° ± 1° to 6° ± 3° (P < 0.05), and KAPPF occurred constantly 0.06 ± 0.03 s before KAmin_PP. The final part of the poling phase was characterized by a knee extension (events 3-4) of 4° to 8° across all velocities and was completed at a decreasing KA at the end of poling phase (KAPole_Out; P < 0.05) across Vsm, with no further changes up to Vmax. The initial recovery phase showed a slight knee flexion (events 4-5) at higher velocities up to 6° followed by an accentuated extension phase in the knee joint (events 5-6). The KA maximum during recovery phase (KAmax_RP; event 6) was maintained at 165° ± 7° across all velocities and skiers. The knee extension ROM during recovery phase (ROMK_ext_RP) increased through 21 km[middot]h−1 (all P values <0.05), with no further change up to Vmax, whereas the knee extension time during recovery phase (t) showed a clear but inconsistent decrease from 9 km[middot]h−1 up to Vmax (P < 0.05). Altogether, this resulted in a fourfold increase of knee extension AV during recovery phase (AVK_ext_RP) up to 63 ± 17°[middot]s−1 through 27 km[middot]h−1 (P < 0.05), with no further changes up to Vmax. The final part of the recovery phase and the DP cycle was characterized by a knee flexion (events 6-7) that continued during the subsequent poling phase (events 1-3 of the next cycle). The knee flexion ROM during recovery phase (ROMK_flex_RP) increased through 27 km[middot]h−1 (P < 0.05), with no further shift up to V, whereas the knee flexion time during recovery phase (t) was maintained at 0.16 ± 4 s across all velocities and skiers. Regarding the total knee flexion phase from KAmax_RP to KAmin_PP (events 6-7/1-3), we found a consistently increasing knee flexion AV during C (AVK_flex_C) up to 124 ± 37°[middot]s−1 through 27 km[middot]h−1 (all P values <0.05), with no further shift up to Vmax.
The main findings of the study are as follows: 1) elite skiers control DP speed by increasing both Pf and CL; 2) increased CL at high speed is achieved by increased pole force, despite reduced poling time and time-to-peak pole force; 3) elbow joint kinematics during the poling phase adapt to higher speeds by decreasing angle minimum and increasing flexion and extension ROM as well as AV; and 4) leg work adaptations are characterized by a) decreasing HA and KA at pole plant and angle minima during the poling phase, b) increasing hip and knee flexion ROM and AV occurring around pole plant, and c) increasing hip and knee extension ROM and AV during the recovery phase.
The constant rise in Pf in the present study is in line with previous research into the DP technique (8,13,16) and has been proposed as the primary method of increasing DP speed (25). The significant increase in CL across Vsm using the DP technique has not been reported in earlier studies. It had previously been shown that CL is maintained at Vsm (8,13,16) and then remained unchanged (16) or decreased (8,13) at Vmax. So what can explain the fact that elite skiers in the present study adapted to increased Vsm through a simultaneous increase of both CL and Pf with a consistent Pf/CL ratio?
First, our subjects performed DP at a higher level, with a 30-50% higher Vmax compared with the earlier studies (Figs. 1A and 2B). The skiers reached a Vmax of 29.5 km[middot]h−1 (8.2 m[middot]s−1) with a CL of 7.5 m at a Pf of 1.08 Hz (64.8 poling cycles per minute). From our point of view, it is unrealistic to attain such high velocities without increasing both Pf and CL. If we assume that skiers who performed at a CL of 4.4-5.7 m at their Vmax in earlier studies (8,13,16) were to ski at a speed of 27 km[middot]h−1, they would have had to increase their Pf by 32-72% (79-102 cycles[middot]min−1) to reach this target speed. This exceeds the reported frequencies that skiers use at racing velocities (11). One suggested explanation as to why skiers seem to limit their increase in Pf is to save energy. Each DP cycle involves a large amount of oxygen-demanding internal work in repositioning the skier between each cycle; this is reported to represent approximately 30% of the total energy cost (V˙O2peak_DP) during DP (9).
A second possible reason for the observed parallel increase in CL and Pf is that a new technical pattern has been developed among world class skiers over the last decade, leading to a more complex movement that increasingly involves the lower body (hip, knee, and ankle joints) and different arm and shoulder patterns (11). This new DP technique was found to be characterized by smaller joint angle minima (elbow, hip, knee, and ankle) and higher flexion AV, more abducted arm position ("wide elbow" pattern) (11), and less trunk flexion. This enables skiers to produce higher pole force and impulse (11) and thereby probably increase their CL. In addition, the longer recovery phase and lower Pf (Figs. 1C and 2A) of this modern form of DP has been shown to be related to a lower heart rate response and blood lactate concentration as well as to longer time to exhaustion (10).
If we look at other sports, we find different solutions to the CL-frequency relationship. Swimmers increase speed by increasing stroke rate (SR) accompanied by a decrease in stroke length (SL) with an increase in the SR/SL ratio (see (24,29)). This strategy can be explained by the high drag forces (F = constant × V) because water is 1000 times denser than air (6). The frequency increase helps to limit the greater reduction in speed during nonpropulsive phases compared with, for example, running (only air resistance) where most of the energy applied to the ground ends up as forward motion against less drag (6). In running, speed increases are controlled by increases in both step frequency and length, accompanied by greater ground reaction forces and rates of force development along with a decrease in ground reaction times (see (12)). This strategy of speed adaptation is in line with our results from examining DP on a treadmill, although in this situation skiers had to overcome rolling resistance without air resistance. The lack of air resistance has to be taken into consideration when interpreting the different results because in the current and previous studies (8,13,16) using roller skis or skis over ground, CL changes with speed. During a DP cycle, air drag forces vary continuously due to the change from a tucked body position during poling to a more upright position during the recovery phase (25). On the basis of data from Svensson (28), air resistance can be estimated as being between approximately 10 N for a tucked position and approximately 20 N for an upright position for the highest DP speed in the current study (8.2 m[middot]s−1). Consequently, the influence of increased air resistance at the highest speeds might have altered our CL versus velocity graph somewhat in comparison to DP over ground. Altogether, today's elite skiers, using a modernized DP technique, follow a speed increase strategy that is similar to various other forms of human terrestrial and aquatic locomotion. These are characterized by a curvilinear relationship between cycle rate versus speed (concave-up shape) and CL versus speed (concave-down shape), with the CL maxima attained at the upper end of the speed range ranging between 60% and 100% of the corresponding maximum speed (5).
The CT and therefore the absolute poling and recovery times decreased across Vsm, with no further change at Vmax in our study. This is in contrast to previous studies that have reported a continuous decrease over velocities (13,16) (Fig. 1A-C). Interestingly, these studies also showed only slightly longer (0.30 s) poling times at each Vmax compared with our study (0.26 s), whereas maximum velocities were approximately 9 km[middot]h−1 lower. As regards absolute recovery time, we found distinctly higher values in all Vsm and Vmax when compared with the aforementioned studies. Of note is that a longer recovery phase has been shown to be one criterion for a modern and fast DP strategy in elite cross-country skiers with lower blood lactate and heart rates (10,11).
The short pole ground contact phases at high velocities (<0.28 s) consisted of short elbow flexion times (<0.11 s), short time-to-peak pole force (<0.08 s), and elbow extension times (<0.17 s). It could be argued these short absolute times represent a critical limit for the full recruitment of both type I and type II fibers for sufficient force generation. Studies on force-velocity characteristics regarding contraction velocity on single fibers in vivo demonstrated that type I fibers need 100-140 ms and type IIA fibers need approximately 55-85 ms to create maximal tension and thereby force (1-3). These time gaps for slow twitch fibers are longer than the elbow flexion time and time-to-peak pole force found in the present study at the highest velocities. Thus, skiers might not be able to fully recruit both fiber types at high velocities. To overcome this, elite skiers use distinct muscle preactivation before pole plant at higher velocities to reinforce the generation of pole force (11). In running, higher preactivity is related to greater stiffness of the tendomuscular system in the subsequent eccentric phase of the foot ground contact, and this relationship becomes even stronger as running speed increases (see (12)). Preactivity is assumed to increase the sensitivity of the muscle spindle through enhanced alpha-gamma coactivation resulting in a potentiation of stretch reflexes, which are able to enhance force production not only in the eccentric but also in the concentric and propulsive phases of running (4,15). These are phenomena that may also play a similar role in cross-country skiing and especially in DP.
As regards to short ground contact times in human locomotion, Minetti (14) has stated that the inverse relationship between contact time and speed means the ability to store and release mechanical energy to and from the elastic structures is a limiting factor. A second problem is the constraint of stopping the foot or the pole with respect to the ground, which would imply that muscle contraction generates less force for propulsion, as it is directly related to progression speed. To circumvent these limitations, Minetti (14) suggests searching for a new pole design that would allow the pushing portion of the pole to slide. This way, muscle contraction speed would be much lower while continuing to slide on the medium.
The ability to produce short ground contact times in combination with high force production seems to be crucial, not only in attaining high maximum velocities but also for movement economy in cross-country skiing. Several studies (e.g., [18,20,22]) have found that endurance performance and peak treadmill performance on level terrain are influenced not only by central factors related to V˙O2 but also by so-called muscle power factors related to neuromuscular characteristics (e.g., voluntary and reflex neural activation, rate and force of myofibrillar cross-bridge cycle activity, muscle force, and running mechanics).
Relative poling and recovery times decreased and increased, respectively, across Vsm and thereafter leveled off. Conversely, Millet et al. (13) demonstrated a preservation of relative poling and recovery time over Vsm with a subsequent increase up to Vmax, accompanied by a shortening of relative recovery time (Table 1 and Fig. 1D). In addition, Nilsson et al. (16) did not report any difference over velocities, suggesting increased ground contact time as a potential strategy for increasing speed. Altogether, this is contrary to our results, which clearly demonstrate skiers' adaptation by a shortening of poling time from approximately 40-30% CT (recovery time: ∼60-70% CT) through 21 km[middot]h−1 and by maintaining both variables as speed increased up to Vmax. In the present study, the constant relative poling time in the last three velocities is approximately 25% shorter and consequently has a longer relative recovery time when compared with the above mentioned studies (Table 1 and Fig. 1D). Such differences between faster and slower athletes have already been reported by Holmberg et al. (10). This new pattern emphasizes that modern competitive skiers use a technical strategy across velocities, in which they first reduce poling time to a "critical" threshold and thereafter try not to fall below a duration that might limit a sufficient production of force.
In accordance with Millet et al. (13), peak pole force in the current investigation increased with Vsm. In addition, the decrease of time-to-peak pole force from 0.21 to 0.07 s over the various velocities caused a fourfold increase in the rate of force development. Our results add to the understanding of the effect of speed increases on force characteristics during DP. Due to the aspect of distinctly reduced poling times at high velocities, as discussed above, it is important to have not only fast but also substantial force production to create a high CL and to not only work with increases in Pf. It has been shown that the better skiers had a higher peak pole force and a shorter time-to-peak pole force at racing speed (11). In the present study, the importance of fast and large pole force production at maximum speed was clearly observed. In comparison to pole force, plantar force showed less alteration with speed. However, a clear load shift from the whole foot to the rear of the foot was observed, whereas forefoot force was unchanged regardless of speed. It can be postulated that the increased rearfoot force during the poling phase was caused by an emphasized pushing-forward movement in the lower leg (slight knee extension) across increased velocities (Fig. 5B). We interpret this as a strategy where the skiers get in a hunched, forward-leaning position with the aims of 1) maintaining forward-backward balance, 2) creating a slight counteraction (action-reaction) in opposition to the backward-directed final pole thrust to generate higher horizontal pole forces, and 3) shifting pressure from the grip wax zone to the gliding zone at the rear part of the ski, thereby reducing gliding friction when DP on snow.
Our data are the first to describe joint kinematics during DP at different velocities and are discussed below using an integrative approach. Elbow position at pole plant remained constant at all velocities compared with a distinct decrease in HA and KA across Vsm (Figs. 4 and 5A and B). Skiers adapted to speed increases with a more flexed body position at pole plant, which perhaps placed the poles in a more advantageous position for generating propulsive force. All the joints investigated showed an emphasized flexion during the poling phase through 27 km[middot]h−1, with decreasing angle minima and flexion times (Tables 2-4) and with increasing flexion AV with the highest values in the elbow joint (344°[middot]s−1). In total, this helped to increase the peak pole force and the rate of force development during the poling phase. Elbow extension during the poling phase, another important propulsive factor, was characterized by a decrease in extension time and an increase in extension ROM, causing a remarkable increase in AV (173-548°[middot]s−1). In comparison, hip and knee extensions were much less distinct, which might be explained by the fact that the lower body only has an assisting function (stabilization and counteraction) regarding the generation of pole force during the final part of the poling phase. It could be suggested that the small knee extension (Fig. 5B; events 3-4) causes a minor pushing-forward movement in the lower leg at the end of the poling phase that may act as a slight counteraction, reverse to the pole push-off direction.
In contrast to the elbow, the hip and the knee joints also have important roles during the recovery phase 1) to perform a body extension, thereby raising the CoG and positioning the body ready for the next pole plant, and 2) to enable the use of the whole body with coactions between arms, trunk, and legs for pole force generation by an extensive hip (trunk) and knee flexion shortly before pole plant (10,11). These important extension-flexion patterns that match pronounced force production increase in range across velocities. It is noteworthy that the maximal HA and KA during the recovery phase did not change across velocities, whereas the extension ROM, starting from the low crunched DP position at the end of the poling phase, increased. This can be explained by the decreasing joint angles at the event "pole-out" (Fig. 5A and B; event 4). The hip and the knee extensions AV of >100°[middot]s−1 during the first two thirds of the recovery phase at racing velocities >25 km[middot]h−1 may be considered a challenge from a coordinative-technical point of view. An accentuated trunk extension requires a high level of timing and back extensor muscle strength; particularly at high velocities, the maximal hip and knee extensions must be reached early enough in the DP cycle to be ready for the subsequent, effective joint flexion toward the next pole plant. From an energetic perspective, this repetitive up-and-down movement is a challenge because the upper body represents approximately 70% of total body mass (30). The hip and the knee flexion with increased ROM and AV continued down to the angle minima during the poling phase and has been shown to be correlated to the production of high pole forces and thus to DP performance (10,11).
The limitations of cross-country skiing research on an indoor treadmill include the estimation of the potential effects of air resistance. One possibility for solving this would be to perform a study in a wind tunnel, which would enable researchers to specifically examine the influence of air resistance on CL, which is likely to increase at higher speeds. Another constraint in the current study is the lack of 3D data that preclude a more detailed analysis of some specific aspects of DP. One of these is the movement of the shoulder joint with the typically more abducted arm position used in modern DP (11) and its possible relation to the generation of pole force. Another aspect is a more complex examination of trunk and pole angles as well as trajectories of CoG in different planes during DP.
In summary, the present study shows that elite skiers are using a strategy to generate higher velocities by increasing both Pf and CL. The increase in CL is achieved by an increased force production during a shortened poling time, approaching physiologically critical values with a limited time for the recruitment of both slow and fast twitch fibers for sufficient force generation. This is managed by increased ROM, shortened flexion-extension times, and increased AV. It is notable that the functional importance of the recovery phase increases with speed due to the required 1) faster and more extensive repositioning of the body before the next poling phase and 2) larger and faster hip and knee flexion toward the next pole plant, thus providing a pronounced and rapid force production across speeds. The strategy described above needs a well-developed ability to produce great pole force over limited time, emphasizing the demands for specific explosive strength and highly developed motor skills.
The authors thank the support of Prof. Dave Bacharach and Prof. John Seifert at St. Cloud State University, St. Cloud, MN. The authors also thank the athletes and the trainers for their participation, enthusiasm, and cooperation in this study. This study was supported by the Swedish Olympic Committee. The results of the present study do not constitute endorsement by ACSM.
1. Buchthal F, Schmalbruch H. Contraction times and fibre types in intact human muscle. Acta Physiol Scand
2. Eberstein A, Goodgold J. Slow and fast twitch fibres in human skeletal muscle. Am J Physiol
3. Garnett R, O'Donovan MJ, Stephens JA, Taylor A. Evidence for the existence of three motor unit types in normal human gastrocnemius. J Physiol
4. Gottlieb GL, Agarwal GC, Jaeger RJ. Response to sudden torques about ankle in man. IV. A functional role of alpha-gamma linkage. J Neurophysiol
5. Hay JG. Cycle rate, length, and speed of progression in human locomotion. J Appl Biomech
6. Hines, EW. Fitness Swimming
. Hong Kong: Human Kinetics; 1998. p. 186.
7. Hoff J, Helgerud J, Wisloff U. Maximal strength training improves work economy in trained female cross-country skiers. Med Sci Sports Exerc
8. Hoffman MD, Clifford PS, Bender F. Effect of velocity on cycle rate and length for three roller skiing techniques. J Appl Biomech
9. Holmberg HC. Physiology of Cross-country Skiing-With Special Emphasis on the Role of the Upper Body
(Dissertation). Stockholm: Department of Physiology and Pharmacology, Karolinska Institutet; 2005. p. 331.
10. Holmberg HC, Lindinger S, Stöggl T, Bjorklund G, Müller E. Contribution of the legs to double-poling performance in elite cross-country skiers. Med Sci Sports Exerc
11. Holmberg HC, Lindinger S, Stöggl T, Eitzlmair E, Müller E. Biomechanical analysis of double poling in elite cross-country skiers. Med Sci Sports Exerc
12. Kyröläinen H, Avela J, Komi PV. Changes in muscle activity with increasing running speed. J Sports Sci
13. Millet GY, Hoffman MD, Candau RB, Clifford PS. Poling forces during roller skiing: effects of technique and speed. Med Sci Sports Exerc
14. Minetti AE. Passive tools for enhancing muscle-driven motion and locomotion. J Exp Biol
15. Nicol C, Komi PV. Significance of passively induced stretch reflexes on achilles tendon force enhancement. Muscle Nerve
16. Nilsson J, Tveit P, Eikrehagen O. Effects of speed on temporal patterns in classical style and freestyle cross-country skiing. Sports Biomech
17. Nilsson JE, Holmberg HC, Tveit P, Hallen J. Effects of 20-s and 180-s double poling interval training in cross-country skiers. Eur J Appl Physiol
18. Nummela AT, Paavolainen LM, Sharwood KA, Lambert MI, Noakes TD, Rusko HK. Neuromuscular factors determining 5 km running performance and running economy in well-trained athletes. Eur J Appl Physiol
19. Österås H, Helgerud J, Hoff J. Maximal strength-training effects on force-velocity and force-power relationships explain increases in aerobic performance in humans. Eur J Appl Physiol
20. Paavolainen L, Hakkinen K, Hamalainen I, Nummela A, Rusko H. Explosive-strength training improves 5-km running time by improving running economy and muscle power. J Appl Physiol
21. Rusko H. Physiology of cross country skiing. In: Rusko H, editor. Cross Country Skiing: Olympic Handbook of Sports Medicine
. Oxford: Blackwell Publishing; 2002. pp. 1-31.
22. Rusko HK. Measurement of maximal and submaximal anaerobic power: an introduction. Int J Sports Med
. 1996;17(suppl 2):89-90.
23. Saltin B. The physiology of competitive c.c. skiing across a four decade perspective; with a note on training induced adaptations and role of training at medium altitude. In: Müller E, Kornexl H, Raschner C, editors. Science and Skiing
. Cambridge: Chapman & Hall; 1997. pp. 435-69.
24. Seifert L, Chollet D, Rouard A. Swimming constraints and arm coordination. Hum Mov Sci
25. Smith GA. Biomechanics of cross country skiing. In: Rusko H, editor. Cross Country Skiing: Olympic Handbook of Sports Medicine
. Oxford: Blackwell Publishing; 2002. pp. 32-61.
26. Stöggl T, Lindinger S, Müller E. Reliability and validity of test concepts for the cross-country skiing sprint. Med Sci Sports Exerc
27. Stöggl T, Lindinger S, Müller E. Evaluation of an upper-body strength test for the cross-country skiing sprint. Med Sci Sports Exerc
28. Svensson E. Ski Skating with Champions: How to Ski with Least Energy
. Seattle (WA): Svensson; 1994. p. 272.
29. Thompson KG, MacLaren DP, Lees A, Atkinson G. The effects of changing pace on metabolism and stroke characteristics during high-speed breaststroke swimming. J Sports Sci
30. Winter DA. Biomechanics and Motor Control of Human Movement
. New York (NY): John Wiley & Sons; 1990. p. 325.