Post-warm-up muscle temperature maintenance: blood flow contribution and external heating optimisation

Ten male cyclists volunteered to participate in this investigation (age 22.0 ± 0.8 years; height 180.0 ± 2.3 cm; body mass 76.5 ± 5.4 kg; mean ± SD). With respect to the training status of the participants, all of them performed at least three cycle-based training sessions per week as well as other regular physical activity, such as running and weight lifting. They completed a general health-screening questionnaire and were all non-smokers, free from injury and from any medication. All subjects were informed verbally of the objectives and procedures of the study before giving written informed consent. In the 24 h prior to each trial, participants were asked to refrain from caffeine, alcohol ingestion and any strenuous exercise. They were also asked to keep a record of their food intake and replicate this prior to the subsequent visit. The protocol and procedures involved were approved by the Loughborough University Ethical Advisory Committee. The study was conducted within the confines of the World Medical Association Declaration of Helsinki for medical research using human participants.

For the comparison of different heating temperatures/methods, a between-group comparison was made to the data presented by Faulkner et al. (2013a, b). Eleven male cyclists (age = 24.7 ± 4.2 years, height = 1.82 ± 0.72 m, body mass = 77.9 ± 9.8 kg; not significantly different from the present experimental group) with the same level of weekly activity were exposed to the exact same warm-up and recovery protocol.

Participants visited the laboratory on three occasions. Prior to the main experimental trials, participants were familiarised with exercise testing and the general measurement procedures used in the present study. On the two remaining visits, participants performed the experimental tests. On each occasion, after 30 min of stabilisation period, participants completed a 15-min cycle sprint warm-up, followed by 30 min of passive recovery. In both experimental trials, during the passive recovery, arterial and venous blood flow of one leg was occluded (NBF), whereas the other leg was used to observe the effect of blood flow on thigh muscle temperature decline (BF). The two visits differed in the surface area over which heat was applied; in one visit, upper and lower leg were warmed up (WHOLE), while in the other visit only the upper leg was kept warm (UPPER). The trials were completed in a balanced order, separated by a minimum of 72 h. In order to make the current results comparable to that from previous studies (Faulkner et al. 2013a, b), all experiments were performed in a temperature and humidity-controlled environment maintained at 17.7 ± 0.3 °C and 54.0 ± 3.0 % relative humidity, which Faulkner et al. (2013a, b) considered a range relevant to athletic events in a Western European climates (originally aiming at the London Olympics).

On the two experimental sessions, participants entered the controlled climatic room environment and were instrumented with skin, rectal and heart rate measurements systems. Participants remained lying on a bench for 30 min, wearing shorts and t-shirt, to allow time for muscle, core and skin temperatures to stabilise. After the stabilisation period, baseline measures for thigh muscle temperature (T _mvl) at three different depths for both legs, quadriceps skin temperature (T _sk-quad), gastrocnemius skin temperature (T _sk-gas), core temperature (T _c) and heart rate (HR) were taken. These parameters were recorded every 1 min during the experimental trials, apart from T _mvl, which was recorded after the stabilisation period (PRE), on completion of the warm-up exercise (POST-WUP) and immediately after the recovery period (POST-REC). After the stabilisation period, participants mounted a cycle ergometer and performed the standardised active warm-up protocol which consisted of 5-min cycling at an external power output of 100 W, followed by five 10-s maximal sprints (10-s Wingate test), each sprint separated by 1 min 55 s of cycling at 75 W (Faulkner et al. 2013a, b). Participants were asked to maintain a cadence of 85 rpm, until the start of each sprint. The maximal sprints were performed at a frictional load equivalent to 10 % body mass. On completion of the warm-up, participants dismounted the ergometer, laid supine on the bench (this took ~30 s), where POST-WUP T _mvl was measured. Following this, a blood pressure cuff was immediately applied proximally around the right thigh to abolish/minimise leg blood flow, and the participant donned the water-perfused trousers (WPT). The mean inflation pressure was 140 mmHg. To ensure that the inflation pressure was supra-systolic, individual blood pressure was measured before the recovery period. During all conditions, the WPT were connected to a controlled water bath (45.0 ± 0.1 °C) and reached a temperature of 43.0 ± 0.5 °C. This temperature was selected after extensive pilot testing, as it was the maximum temperature for which no signs of superficial skin injury or pain were present. Once donned, the participants remained in the environmental chamber in a supine position for 30 min. At the end of the 30-min recovery, the water-perfused trousers were removed and a POST-REC T _mvl measurement was taken. After this measurement, the blood pressure cuff was removed and the participant remained in the lying position until the stabilisation reached baseline parameters.

In addition, to assess the effectiveness of the optimised heating procedure (WPT43) in maintaining post-warm-up T _mvl, POST-REC T _mvl data were compared with T _mvl data from two of our previous studies, where participants donned electrically heated trousers heated at 40 °C (ELEC40) and non-heated tracksuit bottoms (CON) during 30 min of passive recovery subsequent to the same warm-up exercise protocol used in this study. The current group of subjects (WPT43) and those from our previous studies (ELEC40 and CON) were homogenous in terms of gender, age, anthropometric characteristics and fitness level (Faulkner et al. 2013a, b).

A preliminary analysis was conducted to ensure that there were no violations to the assumption of normality of distribution. Data were analysed using 3-way repeated measures ANOVA with a Bonferroni correction. The effect of the independent variables (blood flow, heating procedure and muscle depth) was observed over the time on T _mvl (PRE, POST-WUP and POST-REC), T _c, T _sk-quad, T _sk-gas and HR (PRE, 5WUP, 10WUP, 15WUP, 5REC, 10REC, 15REC, 20REC, 25REC and 30REC). In addition to the p value, the effect size index (partial Eta squared, \(\eta_{p}^{2}\)) was included to assess the proportion of variance in T _mvl in response to the heating procedure (WHOLE versus UPPER). Effect size index is interpreted as 0.10 small effect, 0.30 medium effect and 0.50 large effect (Cohen 1988). One-way between-group ANOVA was conducted to assess the effect of the passive heating (CON; ELEC40; WPT43) on T _mvl. Where significant effects were identified, post hoc multiple comparison tests with Fisher LST correction were conducted. The accepted level of significance was set a priori to p ≤ 0.05. Data are presented as mean ± SD. The software SPSS I20 was used for the statistical analysis.

Resting (PRE) T _mvl at each depth was not significantly different between legs (right versus left), blood flow conditions (BF versus NBF) or heating procedure (WHOLE versus UPPER). However, PRE T _mvl did vary significantly (p < 0.05) across all measured depths (28, 18 and 8 mm): deep-T _mvl (34.7 ± 0.8 °C) was significantly different from mid-T _mvl (33.8 ± 1.4 °C); deep- and mid- T _mvl were significantly different from superficial-T _mvl (32.1 ± 2.4 °C). After the warm-up (POST-WUP), T _mvl increased by 2.7 ± 0.6, 3.2 ± 1.2 and 3.3 ± 1.3 °C above resting values for 28, 18 and 8 mm, respectively (Fig. 1), narrowing the temperature gradient between depths (p < 0.001). POST-WUP T _mvl was unaffected by blood flow conditions (BF versus NBF) and heating procedure (WHOLE versus UPPER).

POST-REC T _mvl was not significantly different (p > 0.05; \(\eta_{p}^{2}\) < 0.10) between heating procedure (deep-T _mvl WHOLE 37.5 ± 0.3 °C versus deep-T _mvl UPPER 37.5 ± 0.2 °C; mid- T _mvl WHOLE 37.5 ± 0.4 °C versus mid- T _mvl UPPER 37.4 ± 0.5 °C; superficial-T _mvl WHOLE 36.6 ± 1.1 °C versus superficial-T _mvl UPPER 37.0 ± 0.8 °C), and there was no interaction between heating procedure and blood flow. Therefore, T _mvl data were collapsed for heating procedure (WHOLE versus UPPER) and are presented in the two conditions BF and NBF. POST-REC T _mvl was lower in BF versus NBF: 0.31 ± 0.3 °C for deep-T _mvl; 0.31 ± 0.4 °C for mid-T _mvl and 0.36 ± 0.4 °C for superficial T _mvl (all p < 0.05, Fig. 2).

Post-recovery period T _mvl values obtained from our previous electrical heating method (ELEC40) and control group (non-heated tracksuit bottoms; CON) were compared to POST-REC BF T_mvl data achieved with the current optimised heating procedure (WPT43) (Fig. 3). In WPT43, the reduction in T _mvl during recovery was significantly attenuated or even eliminated compared to ELEC40 and CON (p < 0.001). Deep-T_mvl decreased by 0.12 ± 0.3, 1.08 ± 0.4 and 1.3 ± 0.3 °C in WPT43, ELEC40 and CON condition, respectively. Mid- T _mvl increased by 0.15 ± 0.3 °C in WPT43, whereas it was reduced by 1.15 ± 0.2 and 1.44 ± 0.2 °C in ELEC40 and CON conditions, respectively. Finally, superficial-T _mvl increased by 1.1 ± 1.1 °C in WPT43 condition and dropped by 0.8 ± 0.5 and 1.4 ± 0.3 °C in ELEC40 and CONT conditions, respectively.

There was no effect of blood flow conditions (BF versus NBF) and heating procedure (WHOLE versus UPPER) on both HR (Table 1) and T _c (Fig. 4) during the trials. In both WHOLE and UPPER conditions T _c increased significantly (p < 0.05) at time points 10WUP, 15WUP, 5REC, 10REC, 15REC, 20REC, 25REC and 30 REC compared to PRE and 5WUP. At time point 10REC, T _c reaches the peak and significantly decreases (p < 0.05) at time points 15REC, 20REC, 25REC and 30 REC compared to 10REC.

Baseline T _sk-quad was 30.9 ± 0.7 °C and not different between legs (right versus left), blood flow conditions (BF versus NBF) and heating procedure (WHOLE versus UPPER). During the 15 min warm-up T _sk-quad did not increase in response to the exercise. During the recovery period (REC), T _sk-quad increased continuously in response to the passive heating, and at 30 min into REC, it was significantly elevated (38.4 ± 1 °C) above PRE (p < 0.05) (Table 1) but not different between conditions (BF versus NBF; WHOLE versus UPPER).

After the stabilisation period (PRE), T _sk-gas was 30.4 ± 0.9 °C with no difference between blood flow conditions (BF versus NBF) and heating procedure (WHOLE versus UPPER). During the warm-up exercise, T _sk-gas decreased to 29.0 ± 1.8 °C (p < 0.05). During the recovery period (REC), T _sk-gas increased continuously in response to the passive heating (p < 0.05). At 10 min into REC (10 REC), T _sk-gas was significantly different (p < 0.001) between heating procedure (WHOLE versus UPPER) but not between blood flow (BF versus NBF). At the end of the recovery period (30 REC), WHOLE T _sk-gas was 37.9 ± 0.7 °C, whereas UPPER T _sk-gas was 34.1 ± 1 °C (Table 1).

Thermal discomfort of the whole body was not affected by blood flow conditions (BF versus NBF) and heating procedure (WHOLE versus UPPER). Over time whole body thermal discomfort varied significantly (p < 0.01). After 30 min of stabilisation period (PRE), thermal comfort was 0 (“comfortable”), during the warm-up exercise increased to 1 (“slightly uncomfortable”) and over the recovery period was 0.7 (between “comfortable” and “slightly uncomfortable”).

A similar trend was shown by legs’ thermal discomfort; it was different (p < 0.01) over time, and after the stabilisation period (PRE) the score was 0 (“comfortable”) and reached the value of 0.6, between “comfortable” and “slightly uncomfortable”, during the warm-up exercise. During the recovery period (REC), heating procedure (WHOLE versus UPPER) did not affect legs thermal discomfort. During REC BF, thermal discomfort increased (p < 0.05) to 1.6, between “slightly uncomfortable” and “uncomfortable”, whereas in NBF, it was 1.8 and not different from thermal discomfort achieved during the warm-up.

Confirming previous results (Saltin et al. 1968; Kenny et al. 2003), baseline T _mvl was different across all measured depths: 34.7, 33.8 and 32.1 °C, for deep-, mid- and superficial-T _mvl, respectively. This is in line with the work of Saltin et al. (1968) who found that at rest deep-T _mvl is around 33–34 °C. In the current study, warm-up exercise increased T _mvl by 2.7, 3.2 and 3.3 °C for deep-, mid- and superficial-depths, respectively, reducing the temperature gradient across the three depths. Additionally, after 30 min of passive recovery, T _mvl of the perfused leg was lower compared to T _mvl of the occluded leg at all measured depths (~0.3 °C for deep- and mid- T _mvl; ~0.4 °C for superficial-T _mvl). In the perfused leg, the optimised external temperature resulted in a small decrease in deep-T _mvl and in an increase in mid- and superficial-T _mvl (0.15 and 1.1 °C, respectively) compared to post-warm-up values.

A central aim of this study was to examine the role of passive heating temperature on the decline in post-warm-up T _mvl and whether a relatively small increase in heating temperature (40–43 °C) would provide a measurable benefit. The heating temperature of the previous trousers was set to a maximum of 40 °C. Since the trousers were electrically heated, it was not possible to control the local heating temperature precisely and further increases in the heating element’s temperature could have caused local overheating and skin burns. During piloting for this study, we observed that an external temperature of 40 °C resulted in a mean leg T _sk of ~37 °C, which was not enough to maintain T _mvl achieved with the active warm-up (Faulkner et al. 2013a, b). The use of circulated liquid heating allowed the safe application of an optimised, higher heating temperature (43 °C) which resulted in a higher quadriceps T _sk (~38 °C; Table 1), thus minimising the temperature gradient between T _sk and T _mvl. The optimisation of the heating procedure lead to a substantial reduction of the deep-T _mvl drop compared to our previous studies (Faulkner et al. 2013a, b), where we reported a ~1 °C decline, and to increases rather than decreases in mid and superficial-T _mvl. The present data indicate that microclimate temperature is a major contributing factor in post-warm-up T _mvl decline and that even relatively small changes in heating temperature have an impact. Furthermore, the greater increase in superficial-T _mvl indicates, as to be expected, that the superficial tissues are most affected by microclimate and ambient temperature, which may be the most important for cycling sprint performance (Faulkner et al. 2013b).

The higher T _mvl in the occluded condition confirms the cooling effect of blood flow. This could be by arterial blood from the core, venous blood from the lower leg, or both. The fact that rectal temperature remains higher than muscle temperature throughout the recovery period would suggest that arterial blood could not be responsible. However, rectal temperature is often higher (~0.3 °C) than arterial blood temperature (Taylor et al. 2014), and oesophageal temperature, better representing arterial temperature, decreases faster than rectal temperature post-exercise (Gagnon et al. 2010). Therefore, it is likely that arterial blood temperature entering the thigh was lower than muscle temperature, during, at least, part of the recovery period; this would support a role of arterial blood flow on post-warm-up muscle temperature reduction, as observed in the present study. However, since in this study arterial blood temperature was not measured, the idea that it may have been lower than muscle temperature remains an assumption.

The temperature difference in T _mvl between perfused and occluded leg was relatively small (~0.3 to ~0.4 °C for different depths) which could be due to the high heating temperature used in this study. In fact, it is possible that the optimised passive heating could also reduce the drop rate in central and peripheral blood temperature post-warm-up, reducing the temperature gradient between blood and muscle and therefore narrowing the temperature difference between the perfused and occluded leg. Nevertheless, the effect of blood flow on T _mvl is still evident, though it would likely have been more pronounced with the electric heating system at 40 °C (Faulkner et al. 2013a, b).

We hypothesised that peripheral blood, together with central blood flow, could contribute to the post-warm-up T _mvl decline. To verify our hypothesis, post-recovery T _mvl was observed in two different conditions: WHOLE and UPPER. However, while we expected a lower thigh T _mvl in the UPPER condition compared to WHOLE condition, thigh muscle temperature was not different between heating interventions. It is possible that the high blood temperature (resulting from the higher heating temperature used) flowing from the thigh to the lower leg, increased the temperature of the lower leg tissue in the UPPER condition. In support of this, skin temperature of the lower leg and foot showed a significant increase of 4 °C during the recovery period even if heat was not directly applied, whereas in the whole leg heated condition (WHOLE), it increased by 7 °C. Therefore, these results indicate that, when using an optimised heating method, heating the calf in addition to the thigh does not further increase quadriceps muscle temperature, compared to heating the thigh only. However, for sports where calf muscles are relevant, lower leg heating may still provide a performance benefit due to increased local T _m.

Blood flow restriction coupled with the use of passive heating resulted in a T _mvl increase at all depths. In this study, blood flow occlusion was used as means to understand the mechanistic effect of circulating blood flow on post-warm-up T _mvl decline, rather than a proposed tool to maintain T _mvl during a period of inactivity. In fact, 30 min of blood flow occlusion is expected to have a detrimental effect on performance, as the restriction of circulation would not allow the wash-out of metabolites accumulated in the leg during exercise, causing muscle discomfort and peripheral fatigue (Bigland-Ritchie et al. 1986; Gandevia et al. 1996; Amann et al. 2013).

As the study of blood flow precluded doing a performance test, it needs to be considered whether the observed improved maintenance of T _m would result a in further sprint–power performance enhancement above those observed in Faulkner et al.’s (2013a, b) muscle heating studies. Optimal T _m is thought to be ~39 °C, with an upper threshold of 43 °C, above which there are impairments to muscle function (Åstrand and Rodahl 1986; McRae and Esrick 1993). Although the relationship between improved sprint performance and increased T _m may not follow complete linearity, Zochowski et al. (2007) and West et al. (2013) showed that shorter recovery periods (down to 10 and 20 min from 45 min) between warm-up and race, with concomitant smaller drops in T _m, did result in a greater improvement in the following sprint performance. The latter strongly indicate that the smaller the post-warm-up T _m drop, the better the subsequent sprint/power performance, thus supporting that the optimised heating method used here has a strong potential of further performance enhancement.