The present study was conducted as a double-blind, counterbalanced, placebo-controlled crossover study. The experimental protocol was completed in 4 visits, separated by at least 48 h and not more than 72 h (Figure 1). All test trials were conducted on an indoor climbing wall and the evaluation of each participant took place at approximately the same time of day to minimize the effects of diurnal variations on the measured variables. The interference of uncontrolled variables was reduced by instructing all participants not to exercise within 48 h before the test trials and to avoid alcohol and caffeine ingestion during the experimental period. Participants were asked to maintain their normal dietary habits and habitual lifestyle before and during the experimental period.

Before beginning each assessment, each participant completed a standardized 10 min warm-up that included a general activation of the cardiovascular system, coordination, and light dynamic stretching. The first visit served as a familiarization trial for the test procedures and for determination of the onsight climbing ability, and each participant was advised of the purpose, benefits, and risks associated with the investigation. During the familiarization trial, the body weight and body fat percentage were determined with a digital scale (Tanita BC 601; Tanita Europe, Amsterdam, Netherlands) and height was measured (to the nearest 0.1 cm) using a stadiometer. The onsight climbing ability was determined by first asking the subjects about their most difficult ascent by top rope or red point style, whichever was higher, rated on the UIAA (Union Internationale des Associations d’Alpinisme) scale. The participants then had to climb a predefined route at the self-reported level of difficulty in top rope style. If this route was accomplished without falling, the route difficulty was increased for the next route until no better attempt was possible. If the predefined route was not accomplished, the degree was decreased accordingly for the next attempt. A break of 20 min was given between the attempts. The best attempt was rated as the participant’s onsight climbing performance within this study. The UIAA scale for free climbing difficulty currently extends from I to XII (Draper et al., 2016). For statistical analyses and a better international comparability, the UIAA climbing scale was converted to the metric IRCRA (International Rock Climbing Research Association) scale, according to recommendations for the statistical analysis of sports-climbing grades (Watts et al., 1993; Draper et al., 2016). On the following three visits, the participants completed the test trials in a crossover design, either with compression forearm sleeves (COMP), with non-compressive placebo forearm sleeves (PLAC), or without forearm sleeves (CON). The three test trials (COMP, PLAC, and CON) included three different measurements (hand grip endurance, finger hang, and lap climbing) after completion of the standardized 10 min warm-up. Between the measurements, each subject was required to take a 10 min rest break.

An a priori power calculation indicated that 21 participants were needed to detect significant differences between the conditions COMP, PLAC, and CON, based on an estimated α level of 0.05 and a power of 90% (based on blood lactate concentration results after maximum-intensity climbing bouts (forearm compression sleeves vs placebo sleeves) from an earlier study with an effect size of η_p ² = 0.10 (Engel et al., 2018)). A total of 26 (13 male, 13 female) recreationally and moderately trained climbers volunteered to participate in this study. Of those, two dropped out during the study due to busy schedules. The results presented were obtained from the remaining 24 participants (12 male, 12 female). The participants had a mean (±SD) age of 29.1 ± 6.6 years and showed climbing abilities ranging from IRCRA 13 to 18 (14.8 ± 1.4). Characteristics of male and female participants are shown in Table 1.

All participants underwent a medical screening before entering the study. Participants had to be in good health, with no pre-existing cardiac or pulmonary conditions. Exclusion criteria included acute infections, alcohol consumption at any time during the test period, and chronic medication intake, as well as acute muscular injuries or restrictions. Inclusion criteria were a top rope onsight climbing ability (CA) of at least UIAA grade VII- (IRCRA 13), at least 3 years of climbing experience, an average training load of at least 2 days per week with 3 h per session within the last 3 months, and a self-report of sports climbing as a predominant discipline. All participants gave their written informed consent prior to participation, and all procedures were approved by the ethical committee of the German Sports University Cologne in accordance with the Declaration of Helsinki.

During warm-up and the test trials, the participants wore either compression forearm sleeves (VERTICS.Sleeves, Vertics, Wiesbaden, Germany), non-compressive placebo forearm sleeves (Kiprun Forearm Sleeves, Kalenji/Decathlon S.A., Villeneuve d'Ascq, France), or no forearm sleeves (as control condition). The compression forearm sleeves were made from a 75% polyamide and 25% spandex fabric and exerted a graduated compression from distal (22.4 mmHg) to proximal (12.4 mmHg) (Engel et al., 2018). The placebo forearm sleeves consisted of 83% polyester, 13% polyamide, and 4% spandex and were non-compressive.

The placebo and graduated compression sleeves were visually similar, and the participants were informed that the effectiveness of the two types of forearm sleeves was being tested; they remained unaware of differences in capacity for compression. The investigators were informed about the capacity of compression of both forearm sleeves. The forearm circumference was used to select the size of the forearm sleeves in accordance with the manufacturer’s instructions.

An adjustable Jamar^® Plus+ (Patterson Medical/Sammons Preston, Illinois, United States) digital hand dynamometer was used for handgrip measurements. The participants were seated, without leaning, with an elbow flexion at 90°, a slight shoulder abduction of about 15°, and with their forearms in a neutral supination/pronation position (Trampisch et al., 2012). For the hand grip endurance measurement, an audible signal was given every 3 s to indicate the change between maximal effort and relaxation. Each participant was required to complete 10 trials of maximum voluntary contraction (MVC) (Ozimek et al., 2016). MVC had to be built up from zero to maximum effort within 3 s and was followed by a 3 s rest for each trial. The participants were asked to grip the handle of the dynamometer with maximum effort and were verbally encouraged throughout the test. We determined the following variables over the 10 MVC trials: peak force (F_max), lowest force (F_min), and fatigue index (FI). The F_max and F_min were defined as the highest and lowest forces, respectively, achieved within the 10 MVC trials, whereas FI was calculated as FI (%) = [(F_max–F_min)/F_max] × 100. The variables F_max and FI were used for further statistical analyses. We eliminated a potential hand dominance by averaging the right and left hand values for F_max and FI, and both variables were defined as non-specific muscle endurance parameter dominance (Watts, 2003; Baláš et al., 2012; Kim and Kim, 2016).

The participants’ muscle endurance (resistance to fatigue) of the finger flexors was assessed by having the climbers hold onto a 4 cm ledge with their arms straight (Ozimek et al., 2016). The hold on the ledge was carried out with four fingers in an open grip; the thumb was not included in the grip and was positioned at the bottom of the ledge (Baláš et al., 2012). The elbow, shoulder, and hip joints had to remain fully extended while the participant was hanging on the ledge. The time to failure was determined as the point when the climber could no longer maintain this position on the ledge and was defined as a specific muscle endurance parameter (hang time in seconds).

The third measurement was lap climbing (LC) to assess each participant’s sports climbing endurance performance. In the LC test, the participants were required to climb a predefined route as many times as possible on top rope belay, with no rest between the ascents. The route was set at one UIAA grade below the assessed maximum onsight climbing level (climbing ability VIII > LC route VII). The climbing routes were chosen on a 15 m high vertical to slightly overhanging wall. Only homogeneous, mainly regular, and topographically similar routes were included to induce high levels of perceived exertion and muscle fatigue (Button et al., 2018). The climbing routes consisted of climbing hand holds, with only a few differences in the types of hand holds. The route difficulty was established mainly with buckets, ledges, and crimps. Only routes without pronounced “cruxes” (demanding regions of the route where more advanced climbing actions could be required for further progress) were included in the LC test (Button et al., 2018). The potential influence of recovery strategies on climbing performance was standardized by forbidding the participants from chalking or shaking their hands during climbing or lowering (Baláš et al., 2016). The LC was to be accomplished until a fall occurred due to fatigue. The total covered climbing distance (LCD) in m and the climbing time (LCT) in min were assessed by video analysis. Once the subject had to cease due to fatigue, capillary blood samples (20 μL) were collected from the earlobe directly and after 2, 4, 6, and 8 min after the LC trials to assess blood lactate values. Capillary blood samples were taken while the participants were sitting, without any further physical activity, next to the climbing wall. Blood lactate measurements were conducted directly after the LC trials (Biosen S-Line, EKF-diagnostic GmbH, Magdeburg, Germany). The maximum post-exercise lactate concentration (La_max) was used for statistical analyses. The rate of perceived exertion (RPE) and forearm muscle pain were also assessed using Borgs’ RPE scale (Borg, 1982) and the visual analog (VAS) graphic rating scale (Lee et al., 1991), respectively.

Near-infrared spectroscopy was assessed during the hand grip strength and endurance measurements, as well as during the finger hang. The oxygenation and blood volume changes were assessed using a continuous-wave near-infrared (NIR) spectrophotometer (NIRS; OxyMon MKIII, Artinis Medical System, BV, Elst, Netherlands) and Oxysoft software (Artinis Medical System, BV). The OxyMon MK III instrument generates NIR light with wavelengths at 765 and 855 nm; a sampling rate of 10 Hz and a fixed differential path-length factor of 4.0 were used, as suggested by van Beekvelt et al. (2002). The Lambert-Beer law, in which a path-length factor is incorporated to correct for scattering of light in the tissue, was used to convert the changes in absorption at the discrete wavelengths into concentration changes in oxygenated and deoxygenated hemoglobin/myoglobin/cytochrome oxidase (Barstow, 2019). The spectrum of these hemes compounds overlaps, so differentiation between changes was not possible using NIRS measurements. However, since exercise performance studies are mainly interested in the total amount of O₂ consumed during exercise, the results are not affected by this aspect (van Beekvelt et al., 2002). We therefore reported the sum of the main absorbing chromophores in skeletal muscle as [heme] in the present study, following the suggestion of Barstow (2019).

The flexor digitorum profundus (FDP) muscle was chosen for NIRS measurements because it has been proposed as the most important forearm muscle for sports climbing performance (Michailov, 2014; Fryer et al., 2015) and has been used in previous research on hemodynamics in sports climbers (Philippe et al., 2012; Fryer et al., 2015; Baláš et al., 2018; Feldmann et al., 2020). NIRS optodes were placed between the medial epicondyle of the humerus and the styloid process of the ulna at 1/3 of the proximal distance (Schweizer and Hudek, 2011; Baláš et al., 2018). Following recent recommendations for the location of the FDP (Fryer et al., 2016), the participants were asked to squeeze their thumbs and the first fingers together, and the investigator then palpated the FDP muscle to locate the middle of the muscle belly. The specific point was marked with a skin-friendly permanent pen to avoid variations in the placement of the optodes over all test trials.

The optodes were placed in a template holder on the skin with an interoptode distance of 40 mm, and the template holder was fixed to the skin using dark opaque tape to prevent possible ambient light interference. The effectiveness of the optodes is affected by the presence of excessive adipose tissue in the body (van Beekvelt et al., 2001). However, all participants had a generally low percentage body fat (Table 1), and the forearms are not a major site of subcutaneous body fat. Therefore, we assumed that subcutaneous adipose tissue thickness did not interfere with data collection (Fryer et al., 2015).

The NIRS outcome measures were oxygenated [heme] (oxy[heme]) and deoxygenated [heme] (deoxy[heme]) concentrations. The sum of the oxy[heme] and deoxy[heme] values were the total [heme] concentrations (total[heme]), and the tissue oxygen saturation (StO₂) was calculated as oxy[heme]/(deoxy[heme] + oxy[heme]). In accordance with the approach reported in earlier studies (MacLeod et al., 2007; Philippe et al., 2012; Fryer et al., 2015; Baláš et al., 2018), during the hand grip strength and endurance measurements, the changes in the tissue oxygenation and blood volume (ΔStO₂, Δoxy [heme], Δtotal[heme]) were calculated from the maximum concentrations of StO₂, oxy[heme], and total[heme], while the minimum concentrations were used to represent the mean deoxygenation in the 3 s contraction periods and the mean reoxygenation in the 3 s relief periods (Figure 2A). The mean deoxygenation (ΔStO_2Deoxy, Δoxy[heme]_Deoxy, Δtotal[heme]_Deoxy) and reoxygenation (ΔStO_2Reoxy, Δoxy [heme]_Reoxy, Δtotal[heme]_Reoxy) concentration changes were calculated over all 10 MVC trials. Because of missing NIRS data, the results of the variables ΔStO_2Deoxy, Δoxy[heme]_Deoxy, Δtotal[heme]_Deoxy, ΔStO_2Reoxy, Δoxy[heme]_Reoxy, and Δtotal[heme]_Reoxy in hand grip strength and endurance measurements are presented for the remaining 16 full data sets only.

The statistical analyses of the continuous finger hang incorporated the changes in tissue oxygenation and blood volume during the continuous finger hang exercise (ΔStO₂, Δoxy[heme], Δtotal[heme]) for all 24 full data sets. The changes were calculated as the decrease from the maximum concentrations of StO₂, oxy[heme], and total[heme] at the beginning of the finger hang to the minimum concentration (MacLeod et al., 2007) (Figure 2B).

Data are presented as means ± standard deviations. The Shapiro-Wilk test was used to identify all departures from normal distribution. The effects of treatments (COMP, PLAC, and CON) on the parameters F_max, FI, hang time, LCD, LCT, La_max, and all NIRS output parameters were tested by one-way repeated measures ANOVA, with sex (male or female) as a between subject factor. Violations of the assumption of sphericity were corrected using Greenhouse–Geisser adjustments. Two-tailed paired t-tests were utilized as post hoc tests to indicate significant differences. A Bonferroni procedure was used (p*) to retain an α = 0.05, and the significance level was set at p ≤ 0.05 for all comparisons. Effect sizes were calculated using partial η squared (η_p ²), and were interpreted as small (0.01), medium (0.06), and large (0.14). For the ordinal parameters RPE and VAS, the Friedman test was used to identify differences between treatments (COMP, PLAC, CON). Dunn-Bonferroni tests were used for post-hoc comparisons of the means with corrections for multiple tests to retain an alpha level of 0.05. Kendall’s W (coefficient of concordance) was used to interpret the effect sizes. Student’s t-test was calculated for differences in IRCRA climbing ability between female and male participants and Cohen’s d (d) was used to calculate effect size. The alpha level was set at p ≤ 0.05, and all analyses were conducted using SPSS 28 (IBM Corp., Armonk, NY, United States).

We found no significant main treatment effects for F_max (p = 0.442, η_p ² = 0.036) and FI (p = 0.695, η_p ² = 0.016) (Table 2). The condition sex also had no influence on the intervention-induced changes in F_max (p = 0.384, η_p ² = 0.043) and FI (p = 0.349, η_p ² = 0.047), but the F_max values were significantly lower for female than for male participants (p < 0.001, η_p ² = 0.542) (Table 2). Conversely, the IRCRA climbing abilities did not differ between female and male participants (female 14.3 ± 1.0, male 15.3 ± 1.6; p = 0.070, d = –0.776). We also found a significant main treatment effect for the NIRS-related parameters ΔStO_2Deoxy (COMP –4.0 ± 2.2, PLAC –3.0 ± 1.4, CON –2.8 ± 1.8 [%]; p = 0.049, η_p ² = 0.194), Δoxy[heme]_Deoxy (COMP –10.7 ± 5.4, PLAC –6.7 ± 4.3, CON –6.9 ± 5.0 [μmol]; p = 0.014, η_p ² = 0.263), ΔStO_2Reoxy (COMP 3.5 ± 1.9, PLAC 2.4 ± 1.2, CON 2.3 ± 1.9 [%]; p = 0.028, η_p ² = 0.225), and Δoxy[heme]_Reoxy (COMP 10.2 ± 5.3, PLAC 6.0 ± 4.1, CON 6.3 ± 4.9 [μmol]; p = 0.011, η_p ² = 0.274) (Figures 3A–D). The parameters Δtotal [heme]_Deoxy (COMP –5.4 ± 7.4, PLAC –1.9 ± 5.4, CON –2.6 ± 5.1 [μmol]; p = 0.302, η_p ² = 0.082) and Δtotal[heme]_Reoxy (COMP 5.8 ± 7.4, PLAC 2.7 ± 5.6, CON 3.3 ± 5.2 [μmol]; p = 0.424, η_p ² = 0.059) had no significant differences (Figures 3E,F). The condition sex showed no influence on NIRS-related parameters in hand grip strength and endurance measurements (ΔStO_2Deoxy: p = 0.123, η_p ² = 0.139; Δoxy[heme]_Deoxy: p = 0.352, η_p ² = 0.072; Δtotal[heme]_Deoxy: p = 0.774, η_p ² = 0.018; ΔStO_2Reoxy: p = 0.090, η_p ² = 0.158; Δoxy[heme]_Reoxy: p = 0.364, η_p ² = 0.070; Δtotal[heme]_Reoxy: p = 0.573, η_p ² = 0.039) (for data and pairwise comparisons see Table 3).

No significant main treatment effect was detected for the time to failure in the finger hang (p = 0.327, η_p ² = 0.049) and the condition sex had no influence on the changes in hang time (p = 0.955, η_p ² = 0.002) (Table 4). We also found no significant main treatment effect for the NIRS-related parameters ΔStO₂, Δoxy[heme], and Δtotal[heme] (Table 4). The condition sex showed no influence on NIRS-related parameters in the finger hang for ΔStO₂ (p = 0.790, η_p ² = 0.011), Δoxy[heme] (p = 0.887, η_p ² = 0.005), and Δtotal[heme] (p = 0.589, η_p ² = 0.024) (Table 4).

We found no significant main treatment effects for LCT (p = 0.169, η_p ² = 0.078), LCD (p = 0.188, η_p ² = 0.073), and La_max (p = 0.245, η_p ² = 0.062) (Table 5). The condition sex had no influence on changes in LCT (p = 0.210, η_p ² = 0.068), or LCD (p = 0.147, η_p ² = 0.084), or La_max (p = 0.858, η_p ² = 0.007). A further comparison of treatments revealed no influence of the wearing of compression forearm sleeves on RPE (p = 0.950, W = 0.002) or VAS (p = 0.431, W = 0.035) (Table 5).