Test–retest reliability of performance variables during treadmill rollerski skating

European Journal of Applied Physiology https://doi.org/10.1007/s00421-025-05746-w ORIGINAL ARTICLE Test–retest reliability of performance variables during treadmill rollerski skating Thomas Losnegard1,2 · Paul André Solberg2 · Magne Lund‑Hansen1 · Martin Skaugen2 · Joar Hansen3 · Knut Skovereng4 · Øyvind Sandbakk5 Received: 11 October 2024 / Accepted: 25 February 2025 © The Author(s) 2025 Abstract Purpose We examined the test–retest reliability of rollerski testing across a familiarization trial followed by three separate test trials (T1–T3) conducted within a 14-day period. Methods Ten competitive cross-country skiers performed three sub-maximal tests (5%, speed range 10–16 km h−1) and a maximal speed test until failure (MTF; ~ 5–8 min, 7%, > 10 km h−1) on a rollerski treadmill using the Gear 3 ski skating subtechnique. Reliability was assessed as within-subject typical error, expressed as a coefficient of variation (CV%, [confidence limits]) intraclass correlation (ICC, [confidence limits]), and changes in mean (%). Results The speed at MTF demonstrated a mean CV (T1–T3) of 1.5% [1.1, 2.6] and an ICC of 0.96 [0.87, 0.99], but a systematic familiarization bias from T1 to T2 (1.2% [0.1, 2.3]) and T2 to T3 (2.2% [0.1, 4.3]). Peak oxygen uptake exhibited a mean CV of 2.2% [1.6, 3.8] and an ICC of 0.93 [0.78, 0.98], with no systematic changes from T1 to T2 (− 0.2% [− 2.0, 1.6]) and T2 to T3 (1.8% [− 1.1, 4.7]). VO2 at submaximal load showed a mean CV of 2.1% [1.5, 3.3] and an ICC of 0.94 [0.84,0.99], with no systematic changes from T1 to T2 (− 0.7% [− 2.4, 1.1]) and T2 to T3 (− 0.1% [− 2.4, 2.3]). Conclusion The relatively low CV and high ICC for most measures suggest a high degree of test–retest reliability. However, the systematic mean changes in MTF indicate that familiarization trials are essential to provide valuable information about individual changes. Overall, these reliability measures can be used as a framework by practitioners to discern true changes when testing on a rollerski treadmill. Keywords Oxygen uptake · Performance · Training · Nordic skiing · Cross-country skiing Introduction Communicated by Michael I Lindinger. * Thomas Losnegard 1 Department of Physical Performance, The Norwegian School of Sport Sciences, Ullevål Stadion, Post Box 4014, 0806 Oslo, Norway 2 Norwegian Olympic Federation, Oslo, Norway 3 Section of Health and Exercise Physiology, Inland Norway University of Applied Science, Lillehammer, Norway 4 Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Sciences, Center for Elite Sports Research, Norwegian University of Science and Technology, Trondheim, Norway 5 School of Sport Science, UiT The Artic University of Norway, Tromsø, Norway Physiological variables that determine performance are often evaluated by both researchers and coaches to provide diagnostic information about training-induced changes. In most endurance sports, laboratory tests are conducted in a “sportspecific” manner, aimed at identifying precise traininginduced adaptations. However, reliable results are crucial for the valid interpretation of data from such physiological tests. According to Hopkins (2000) three crucial measures of reliability should be quantified: within-subject variation, retest correlation, and changes in the mean. When a subject undergoes multiple tests, random variation between trials can occur, observed as the standard deviation of individual values. This within-subject variation, also known as typical error, is expressed as the coefficient of variation (CV) of measurement. Vol.:(0123456789) European Journal of Applied Physiology Cross-country skiing, biathlon, and Nordic combined are Olympic sports that utilize the freestyle skiing technique, also known as ski skating. Similar to other endurance sports, higher aerobic metabolic energy turnover (i.e., peak oxygen uptake ( VO2peak) and its fractional utilization) and/or a reduced cost of locomotion (i.e., enhanced work economy/ efficiency) are the primary drivers of performance responses and are, therefore, commonly measured test values. In these sports, large rollerski treadmills are used for “sport-specific” testing, which allows skiers to replicate their skiing technique and simulate on-snow skiing accurately from a biomechanical perspective (Myklebust et al. 2014, 2022). Physiological testing on rollerskis has been extensively utilized for decades (Hoffman et al. 1994; Holmberg et al. 2005; Sandbakk et al. 2010; Losnegard et al. 2013) yet the learning effect and typical error over multiple tests have not been thoroughly examined. Such information is crucial when determining training-induced changes or evaluating experimental interventions. However, to date, only two studies have investigated the test–retest reliability of performance and physiological measurements during treadmill skiing. Losnegard et al. (2013) reported the CV in VO2peak (2.3%), O2-cost (1.2%), and 1000-m time-trial performance (2.7%) while rollerski skating on a treadmill. Bucher et al. (2023) conducted a test–retest reliability study of a comprehensive test battery, including a VO2max test using the diagonal stride technique and a 24-min time-trial test while double poling on a treadmill. The CV was 1.4% for VO2peak and 1.0% for the 24-min time trial. However, since these previous studies only included two trials, less is known about possible learning effects over multiple trials within a short testing period. Cross-country skiing is performed at varying speeds and inclines due to significant variations in terrain. From a testing perspective, this means that the most relevant inclines and speeds must be covered. Nevertheless, there is general agreement that moderate uphill terrain is particularly relevant for testing, considering the importance of uphill performance, the avoidance of excessively high speeds indoors where air drag is absent, and the induction of competitionrelevant speeds during both submaximal and maximal testing (Sandbakk et al. 2010; Losnegard et al. 2013; McGawley and Holmberg 2014; Andersson et al. 2016). On such inclines, the Gear 3 skating sub-technique (i.e., synchronized pole plants for every ski push-off) is the most used subtechnique during races and testing (Andersson et al. 2010; Sandbakk et al. 2011; Sollie et al. 2021). The aim of the present study was to examine the test–retest reliability of performance-determining variables from submaximal and maximal tests using the skating technique during treadmill rollerskiing. We chose a protocol with a constant incline and increasing speed, in which within-subject variation, test–retest correlation, and changes in the mean were investigated. Methods Participants Four female and six male competitive cross-country skiers were recruited (age range 20–30 years). The participants were categorized as Tier 3 according to McKay et al. (2022). All subjects were familiar with testing and training on a rollerski trea (...truncated)