Introduction

The popularity of knee-high compression socks in connection with a range of sports, and especially among endurance athletes, has been increasing. Both high-performance and recreational athletes have begun wearing elastic compression socks during a range of activities, in particular, endurance events such as running and triathlons.

It is well recognized that extra-dermal compression reduces the overall cross-sectional area of the lower limb and increases the linear velocity of blood flow in the venous system (Ido et al. 1995; Meyerowitz and Nelson 1964; Stanton et al. 1949). This reduction leads to an improved mean linear blood flow of 0.5 to 2.5 cm s−1 (Litter 1952). Increased velocity of flow reduces venous stasis by reducing venous wall distension, local contact time and increases the clearance of metabolites such as blood lactate (Berry and McMurray 1987; Chatard et al. 2004). Several studies observed an alteration in erythrocyte deformability when lactate concentration is increased while exercising (Brun et al. 1991, 1994a, b; Lipovac et al. 1985; Smith et al. 1997; Yalcin et al. 2003). In this context, it has been shown that low intensity exercise, promoting moderate lactate elevation, increase erythrocyte rigidity (Brun et al. 1994a) as well as high intensity workout (Brun et al. 1991). The ability of the blood to flow freely through the vascular bed is crucial for the maintenance of optimal O2 availability in the exercising muscle. The cellular properties of red blood cells (RBC) control the passage through the microcirculation where blood cells must squeeze through capillaries that are smaller in diameter than the cells. RBC deformability is influenced by internal fluidity, cell surface area-to-volume ratio, and the physical properties of the membrane and cytoskeleton (Smith 1995; Somer and Meiselman 1993). Previous studies have also shown that RBC mechanical properties are sensitive to the alterations in their environment as well as to metabolic disturbances (Chien 1987; Shiga et al. 1990; Yalcin et al. 2003) such as changes in blood lactate and that RBC deformability can be influenced by exercise.

Since extra-dermal compression appears to affect blood lactate concentration and diminish the diameter of blood vessels near to the surface and therefore might hamper the flow of RBC through the capillaries, it is reasonable to expect that applying compression during exercise might lead to an altered RBC deformability. From this point of view, it may be hypothesized that different external pressure of compression socks (0, 10, 20, 40 mmHg) and/or sub-maximal and maximal running exercise, might influence RBC deformability.

Methods

Subjects

Nine healthy, non-smoking, well-trained male runners and triathletes (age 25.8 ± 3.8 years, size 182.2 ± 4.8 cm, body weight 73.4 ± 5.3 kg, peak oxygen uptake 57.7 ± 4.5 mL min−1 kg−1) volunteered and gave their written informed consent to participate in this study, which was approved by the University’s ethics review board. Prior to the testing, the subjects were fully familiarized with the laboratory exercise procedures. On the test days, they were asked to report to the laboratory well-hydrated, having consumed a light breakfast at least 2 h earlier, and not having performed any strenuous exercise during at least the previous 24-h period. The meal was not standardized, but certain foods containing carbohydrates were recommended to the subjects.

Test design

All of the participants carried out five test protocols on a treadmill (Woodway GmbH, Germany). On the first test day, they performed a ramp test in order to determine their individual maximal oxygen uptakes and assess the running speed to be employed in the subsequent sub-maximal tests. The running speed was set initially at 2.8 m s−1, where it was maintained for 5 min, followed by stepwise increments of 0.4 m s−1 every 5 min until exhaustion was reached.

Following this procedure, each participant performed four tests of the same intensity (approximately 70% of peak oxygen uptake) for 30 min, but with different levels of compression, applied in a randomized order. After the 30 min of sub-maximal running, time to exhaustion using a ramp test (increase in incline of 1% every minute) was measured.

Values for statistical data were obtained at baseline, after 15 and 30 min as well as directly after the end of the test. Subjects wore compression socks extending from below the knee to the foot from the same material (94% Polyamide and 6% Lycra). The mean pressure on the calf at its maximum girth was aimed to be 10, 20 and 40 mmHg, respectively. The level of compression applied to each individual was pre-checked before each test five times at B-Level (i.e. at the ankle’s point of minimum girth) and C-Level (i.e. at the calf’s maximum girth) according to international recommendations (Partsch et al. 2006). For this a pneumatic sensor (SIGaT®, Ganzoni, Switzerland) was used to receive the in vivo pressure dimensions at B- and C-Level based on a previous study (Gaied et al. 2006). The mean values are presented in Table 1.

Table 1 Area of pressure measurement as well as the corresponding mean, minimum (Min) and maximum (Max) pressure at B-Level (ankle’s point of minimum girth) and C-Level (i.e. calf’s maximum girth)
Full size table

All participants were asked not to exercise for at least four days between trials, in order to guarantee adequate recovery.

Measurements

RBC deformabiliy was measured based on the ektacytometric principle using laser-assisted optical rotational cell analyser (LORCA; R&R Mechatronics, Hoorn, The Netherlands) and the software Lorca Elongation V2.1. Therefore, 20 μL of blood were withdrawn from the earlobe (pre, 15’, 30’ Max) with a EDTA-coated capillary, directly mixed with 5 mL of a 5% solution of polyvinylpyrrolidone (PVP) [Osmolality 300 mOsmol/L, viscosity 32.0 mPa s (Sigma Chemical, St. Louis, MO)] and directly measured with LORCA. The sampling site ‘earlobe’ was chosen for several reasons. First, it is less invasive as a venipuncture, second it allowed a faster sampling at the various time points (shorter breaks). Furthermore, a recent study from Simmonds et al. (2011) showed that values for RBC deformability were not different between venipuncture and earlobe using a lancet. The samples were then directly subjected to several rotational speeds at 37°C, giving final shear stresses of 0.3, 0.57, 1.08, 2.04, 3.87, 7.34, 13.92, 26.38, 50.00 Pa. Laser diffraction was used to follow the change of the RBC population from biconcave toward an ellipsoid morphology under increased shear stress, and the deformability [elongation index (EI)] of the RBCs was calculated from the major and minor axis of the ellipsoid diffraction pattern.

According to the paper of Baskurt et al. (2009b), Lineweaver-Burke modeling was used to calculate shear stress at half maximal deformation (SS1/2) and the maximum deformability (EImax).

Immediately before, during and after each test, samples of capillary blood were collected in a capillary tube (Eppendorf AG, Hamburg, Germany) from the right ear lobe for subsequent amperometric-enzymatic determination of blood lactate using Ebio Plus (Eppendorf AG) and of arterial partial pressure of oxygen pO2 with AVL OMNI 3 (Roche, Basel, Switzerland). All of these samples were analyzed directly in duplicate and the mean was used for statistical analysis. The earlobe was not pre-warmed for none of the samplings.

Heart rate was measured by using a Polar telemetric system (Polar Wear Link System and Polar S810i HR monitor, Polar Electro Oy, Kempele, Finland).

Oxygen uptake was measured with an open circuit breath-by-breath spirograph (nSpire, Zan 600 USB, Oberthulba, Germany) throughout the testing, using standard algorithms with dynamic account for the time delay between the gas consumption and volume signal. The spirograph was calibrated prior to each test, using calibration gas (15.8% O2, 5% O2 in N; Praxair, Düsseldorf, Germany) targeting the range of anticipated fractional gas concentration administered with a precision 1 L syringe (nSpire). All respiratory data were averaged every 30 s. The highest values for oxygen uptake within the last 30 s of the ramp test were used for statistical analysis. The criteria for VO2peak were (1) plateau in oxygen uptake, i.e., an increase <1.0 mL min−1 kg−1 despite an increase in power output, (2) respiratory exchange ratio >1.10, (3) heart rate ±5% of age predicted maximal heart rate, and (4) maximal capillary blood lactate after exercise greater than 8 mmol L−1. In all cases, at least three of the four criteria were met by all participants.

Statistical analysis

All data were subjected to conventional calculations and are presented as mean values (mean) ± standard deviations (SD). The normal distribution of all sorts of data was also confirmed, so that no further transformation was necessary. ANOVA repeated-measures were employed to compare different time-points and different levels of compression. When an overall difference over time/between levels of compression was indicated, Fisher post-hoc test was used to identify where the changes occurred. An alpha value of p < 0.05 was considered to be statistically significant. The effect size Cohen’s d [defined as (difference between the means)/standard deviation (Cohen 1988)] was calculated for all of the variables and the comparison between 0 and 10, 20, 40 mmHg (Cohen 1988). The thresholds for small, moderate, and large effects were defined as 0.20, 0.50 and 0.80, respectively (Cohen 1988) and the highest effect sizes are documented in Table 2. All statistical tests were carried out with the Statistica (version 7.1, StatSoft Inc., Tulsa, OK, USA) software package for Windows®.

Table 2 Responses of different physiological values when wearing different grades of compression socks during sub-maximal and maximal running (mean ± SD)
Full size table

Results

Mean running velocity at 70% of peak oxygen uptake was 3.6 ± 0.3 m s−1. Lactate values and pO2 values during sub-maximal running and after the ramp test showed no significant differences between the different levels of compression (Table 2). Oxygen uptake, heart rate and time-to-exhaustion determined by a ramp test showed no significant differences between the levels of compression (p < 0.05; best effect size)

Independent of the level of compression, exercise significantly increased RBC deformability at all three time points (15′, 30′, Max) compared to rest. However, it was noticeable that the higher the compression, the lesser was the effect of exercise, showing less significant differences between rest and exercise for low and especially for high shear forces (Fig. 1a–d). Compression itself had no significant influence on RBC deformability at any time point of measurement for any given shear force.

Fig. 1
figure 1

a–d Effects of exercise on RBC deformability for different levels of compression (0, 10, 20, 40 mmHg) and different shear forces (Pa). *Significantly different from “rest” of the same shear force (p < 0.05)

Full size image

As recently suggested by Baskurt et al. (2009b), maximum deformability (EImax) and the shear stress at half maximal deformation (SS1/2) were calculated for an easier comparison of RBC deformability between different compressions (Tables 3, 4).

Table 3 Shear stress required for one-half of the maximal deformation (SS1/2)
Full size table
Table 4 Maximal RBC elongation index (El max) at four time points and different levels of compression
Full size table

Discussion

The present assessment of the influence of different levels of compression on RBC deformability, cardio-respiratory and metabolic parameters during sub-maximal and maximal exercise revealed, in general, no effects at any level. However, exercise itself significantly increased RBC deformability. This effect was blunted by higher levels of compression and the effect of compression on RBC deformability was more marked at high shear stress than at low shear stress.

We can only speculate about the blunting effects of sock compression on RBC deformability during exercise. A possible mechanism could be an alteration of the cytoskeleton due to the compression, which becomes even more noticeable at higher shear forces. As RBC deformability at low shear stress mainly depends on RBC membrane mechanical characteristics whereas at high shear stress, RBC deformability is more related to the cell’s surface area to volume ratio and to intracellular viscosity or mean corpuscular hemoglobin concentration (MCHC) (Baskurt et al. 2009a; Clark 1989), it is more likely that compression affects the latter mentioned properties.

It has been reported that a tourniquet application during sampling, where local vessel compression is greater than sock compression, did not widely affect RBC rheological properties (Connes et al. 2009a), which makes a mechanical effect unlikely. Anyhow, mechanical influences should be further investigated, beside metabolic effects.

Because athletes during endurance events require rapid uptake of oxygen, the ability of RBC to move through capillaries may limit performance (Smith et al. 1999). RBCs are highly deformable, and this physical property plays an important role in facilitating blood flow, particularly in the microcirculation. The erythrocyte’s remarkable mechanical properties originate from the unique architecture of its cell wall. Thereby, RBC mechanical properties are closely related to the structure and physiological status of the cell and are sensitive to the alterations in their environment as well as to metabolic disturbances (Chien 1987; Shiga et al. 1990; Yalcin et al. 2000), like they occur during physical activity (Brun et al. 2010).

Several studies observed an alteration in erythrocyte deformability when lactate concentration was artificially (lactate anion adjunction in blood samples) or naturally increased (during exercise) (Brun et al. 1991, a, b; Lipovac et al. 1985; Smith et al. 1997; Yalcin et al. 2003). A decrease of erythrocyte deformability during exercise has mostly been described in subjects with low physical fitness (Bouix et al. 1998; Yalcin et al. 2003). Blood lactate, which experimentally shrinks the red cells and decreases their deformability, is likely to explain in part this exercise-induced rigidification of erythrocytes, as supported by correlations between lactate concentrations and red cell rigidity at exercise (Brun et al. 1993; Varlet-Marie and Brun 2004). In one study, they even found a threshold value (4 mmol L−1) for this effect (Brun et al. 1991). However, it was shown that also a moderate lactate increase during low intensity exercise results in a transient increase in erythrocyte rigidity (Brun et al. 1994a).

As red cell deformability was generally found to be either decreased or unchanged during exercise in untrained subjects, trained subjects often show an increase of RBC deformability in response to exercise when assessed by ecktacytometry like in the present study (Connes et al. 2009b; Hardeman et al. 1995) but also with other techniques (Caimi et al. 2009; Connes et al. 2004a). This observation has recently been explained by a study on 20 highly trained athletes (Connes et al. 2004a). During a progressive exercise test conducted to VO2max red cell rigidity was found to paradoxically decrease in 11 subjects with normal hemoglobin saturation. The other part (9 subjects) with low hemoglobin saturation during exercise exhibited no change in RBC deformability. Connes et al. (2009b) stated, that changes in RBC rheological properties during exercise are dependent on the fitness level of subjects and also on their “hypoxemic status”. However, in another study, Connes et al. (2004b) observed no change in erythrocyte deformability in trained athletes performing sub-maximal exercise whereas plasma lactate concentration increased. Besides, in vitro experiments show an increase in the deformability of aerobically trained subjects between 2 and 10 mM lactate concentrations and a decrease in the untrained group between 2 and 4 mM (Connes et al. 2004c). Thus, it seems that endurance training influences erythrocyte response to lactate which might explain our results, showing an increase in RBC deformability after sub-maximal and maximal exercise. First of all, all participating subjects were well trained (VO2peak ~ 58 mL min−1 kg−1; 4.4 ± 0.4 m s−1). Second, the sub-maximal exercise at 70% of VO2peak only caused a slight increase in whole blood lactate concentrations, may be having a rather little influence on RBC deformability. However, deformability was still increased after maximal exercise when higher lactate concentrations of ~5 mmol L−1 were reached but which is still in the reported ranges of Connes et al. (2004c). Therefore, our investigation supports the results of previous studies, showing that the deformability of RBC increases in response to exercise in trained subjects.

However, other factors like pO2 and oxidative stress might be involved in the alterations of RBC deformability and might explain the differences between trained and untrained subjects, as it has been shown in other studies (Connes et al. 2004b; Dikmenoglu et al. 2008; Galy et al. 2005; Senturk et al. 2001). In subjects with high hemoglobin saturation (SpO2) during exercise, RBC deformability was increased compared to subjects with low SpO2 (Connes et al. 2004a). This finding of a dependence of SpO2 on the RBC rigidity is supported by the data of Galy et al. (2005). Exercise-induced oxidative stress was also shown to decrease deformability in sedentary subjects; however, this was not the case in trained subjects (Dikmenoglu et al. 2008; Senturk et al. 2001).

It seems that training leads RBC to be more adapted to cope with lactate (and oxidative stress), resulting in a lack of RBC deformability impairment and allowing a higher endurance performance (Connes et al. 2004a).

Limitations of the study and future perspectives

According to the sites of blood sampling no haematological measurements such as MCV and MCHC were performed, but should be investigated in future studies, as changes in these parameters may affect RBC deformability. If there is a direct mechanical effect of sock compression on RBC rheological properties, RBC aggregation measurements should be performed as well, as sock compression acts preferentially on the venous system and RBC aggregation are preferentially formed in veins.

Conclusion

Our assessment of the potential effects of compression socks on RBC deformability and metabolic values during running at a sub-maximal and maximal level revealed no alteration in RBC deformability, oxygen saturation, oxygen uptake, heart rate or lactate concentration at any pressure level from 0 to 40 mmHg.

The experiments carried out comply with the current laws of the country.

The authors declare that they have no conflict of interest.