Go to:

Keywords: endocrine regulation, thermoregulation, skin temperature, ambient temperature, sweat production

Go to:

Go to:

Experimental design

This was a single-subject observational case study monitoring changes in body mass, and hematologic, urinary, and biochemical parameters while running a marathon in a climate chamber in order to overcome a total Tambient difference of 100°C. Tambient was set at −45°C at the start of the marathon and steadily increased to +55°C by the end of the marathon. This human experiment was approved by the director of the Brandenburg Technische Universität, Cottbus, Germany.

Subject

The subject was a 30-year old male with a body height of 1.80 m, a body mass of 93.8 kg, and a body mass index of 29.0 kg/m2. He was an experienced ultraendurance athlete (world champion in the Deca Iron ultratriathlon 2008, a world record holder in the 10-day triathlon, a finisher of the Trans Europe Footrace in 2009 and of the Race across Germany in 2010), and intending to run the first marathon in history under these conditions.

Preliminary test

One day before the experiment, a maximal treadmill test was performed to determine maximum oxygen uptake (VO2max). A running speed of 8.0 km/hour was set as the start speed. Every 3 minutes, the treadmill speed was increased by 2 km/hour until volitional fatigue. VO2 was determined using an expired air analysis by Meta Max 3B® (Cortex Biophysik GmbH, Leipzig, Germany) and heart rate was measured using a portable heart rate monitor (RCX5, Polar Electro Oy, Kempele, Finland). Lactate threshold was determined using Ergonizer version 3.2.1 (K. Röcker, www.ergonizer.de) to enable data analysis.

Climate chamber marathon

For the experiment, the athlete ran clockwise laps of 17.6 m in the climate chamber at the CEBra (Centrum für Energietechnologie Brandenburg) of Brandenburg Technische Universität in Cottbus, Germany (www.tu-cottbus.de/cebra). The investigating team was located outside the climate chamber. Each lap was counted using a laser-assisted photoelectric barrier. A computer counted the laps continuously and showed the distance covered at each moment.

During the marathon, Tcore was measured in the rectum along with Tskin at three sites. Before the start, after 10.5 km, 21.1 km, 31.6 km, and at the end of the study, venous blood samples were drawn, urine and sweat were collected, and body mass was measured. Each measurement took on average 10 (8–12) minutes. The athlete changed clothes for comfort during the breaks. He started running with winter boots, winter gloves, a winter jacket, and a winter hat over his running clothes. After 10.5 km (at about −20°C), no changes in clothing were made. After 21.1 km (approximately +5°C), he took off his winter clothes and ran in his usual running clothes. After 31.6 km (around +30°C), he took off his shirt. Ratings of perceived exertion were not recorded. The hydration status of the athlete was not assessed before the marathon and he consumed food and drinks as he was used to doing during his races.

Temperature and humidity in climate chamber

Tambient was set at −45°C at the start and continuously increased to +55°C (Figure 1) at a rate of 1°C per 4.5-minute interval. At the start, humidity was set at 60%. After 210 minutes of running at an Tambient of 0°C, humidity was steadily reduced until it reached 20% by the end of the marathon. Temperature and humidity in the climate chamber were recorded every minute. Percent change was rounded to two decimal digits.

Food and drinks

The day before the climate chamber marathon, the athlete went with his crew to a supermarket to buy conventional food, such as bread, meat, cheese, fruit, and fluids, including mineral water and soft drinks. The athlete consumed food and drinks ad libitum during the marathon. All food and drinks were provided at ambient room temperature. At each point of measurement, his coached prepared the food and drinks as requested by the athlete. Intake of food and drinks was recorded to estimate the fluid and energy intake in a cumulative histogram (Figure 2). The composition of the fluids and solid food was estimated using the information on the products consumed by the athlete and a nutrition chart.23

Estimation of sweat production

While running, sweat was collected using a plastic bag attached to a 17 × 25 cm area on the back of the runner. To estimate total sweat production, the sweat produced on this area of the back was extrapolated to the whole body surface area. Body surface area (BSA) was estimated using the formula BSA = weight (kg)0.425 × height (m)0.725 × 0.007184 following Dubois and Dubois.24

Heart rate

Heart rate was recorded continuously during the marathon using a portable heart rate monitor (RCX5, Polar Electro Oy) to estimate energy expenditure, using the heart rate method.25

Temperature

Tcore in the rectum and Tskin on the forehead, right wrist, and right ankle were measured using an EndoTherm Med® device (EndoTherm GmbH, Arlesheim, Switzerland) every 10 seconds while running. The EndoTherm Med thermoelectric probes sensor has a guaranteed accuracy of 0.0625°C. The sensor for measurement of Tcore in the rectum was encased in a silicone capsule and packed into a fingerstall. The sensors used to measure Tskin were fixed using textile ribbons on defined spots. The sensors were evaluated by EndoTherm Control software (www.endotherm.ch) and showed a temperature variation of ±0.1°C (Figure 3).

Laboratory measurements

Before the start of the marathon and every 10.5 km, body mass was measured and samples of blood, urine, and sweat were collected. Body mass was measured using a commercial scale (Beurer BF 15, Beurer GmbH, Ulm, Germany) to the nearest 0.1 kg, after voiding the urinary bladder. Venous blood samples were drawn in a sitting position. Two serum tubes (Serum Sep Clot Activator, 5 mL, Greiner/Bio-One, Frickenhausen, Germany) and three ethylenediamine tetra-acetic acid (EDTA) tubes (BD Vacutainer®, 3 mL, Becton Dickinson, Heidelberg, Germany) were drawn. One of the EDTA tubes, for the later performed blood count and measurement of lactate dehydrogenase, was stored at room temperature. The two tubes for serum and the remaining two tubes for EDTA plasma were centrifuged immediately at 3000 g for 10 minutes and the plasma was frozen on dry ice.

Urine samples were collected after measuring total urinary volume at the defined times into a Vacuette® system (10 mL, Greiner/Bio-One) and were also stored on dry ice. Sweat samples were collected from the defined skin area on the back. Foil was fixed using Tegaderm film (3M Medica, Neuss, Germany) to the 17 × 25 cm area on the back leading to a small reservoir at the caudal end. After documenting the quantity in mL, the sweat was transferred in empty sterile tubes (7 mL, Greiner/Bio-One) and also stored on dry ice.

After collection, the EDTA blood samples were immediately transferred to the laboratory, and the blood counts were performed within 4 hours of blood sampling. The other serum, plasma, urine, and sweat samples remained stored on dry ice and were transported to the laboratory immediately after the experiment and stored there at −20°C until analysis.

Blood counts were performed using the EDTA blood samples, and included a white blood count with differentiation, red blood count, hemoglobin, hematocrit, and platelets. Hematologic parameters were determined using the Cell-Dyn Sapphire® (Abbott Diagnostics, Wiesbaden, Germany). Osmolality in serum, urine, and sweat was measured using an Osmomat 030 device (Gonotec, Berlin, Germany). Glucose in serum and lactate in plasma were measured using a Biosen 5140 system (Barleben, Magdeburg, Germany). Sodium, potassium, calcium, chloride, creatinine, urea, creatine kinase in serum and urine, C-reactive protein in serum, lactate dehydrogenase in plasma, and glucose in urine were measured using a DxC 800 system (Beckman Coulter, Krefeld Germany). Myoglobin, cardiac troponin I, free triiodothyronine, free thyroxine, thyroid-stimulating hormone (TSH), prolactin, cortisol in serum, and brain natriuretic peptide (BNP) in plasma were measured using an Advia Centaur XP (Siemens, Eschborn, Germany). Insulin in serum was measured using an AxSYM device (Abbott Diagnostics). Insulin-like growth factor 1 and interleukin-6 in serum were measured using the Immulite 2000 (Siemens). Somatotropic hormone in serum, and adrenocorticotropic hormone (ACTH) and renin in plasma were measured using a Liaison® device (diaSorin, Dietzenbach, Germany). Procalcitonin was measured using a Mini-Vidas system (Bio-Merieux, Marcy l′Etoile, France). Aldosterone in plasma was measured using a radioimmunoassay kit (Beckman Coulter). Midregional proatrial natriuretic peptide, midregional proadrenomedullin, c-terminal proendothelin-1, and copeptin in plasma were measured using a Kryptor device (Brahms GmbH, Hennigsdorf, Germany). Sodium, potassium, and calcium in sweat were measured using the AAnalyst 100 (Perkin Elmer, Rodgau, Germany). Albumin in urine was measured using an Immage® 800 device (Beckman Coulter). Metanephrines and normetanephrines in urine were measured using high-performance liquid chromatography (Recipe, Munich, Germany).

Renal parameters

We expressed urine concentration as fractional excretions. This applied to sodium, potassium, urea, and osmolality in urine. We used the general formula: fractional excretion of parameter = [(parameterurine × creatinineserum)/(parameterserum× creatinineurine)] × 100, following Espinel.26 Transtubular potassium gradient was calculated using the equation transtubular potassium gradient = (potassiumurine × osmolalityserum)/(potassiumserum × osmolalityurine) according to West et al.27 Glomerular filtration rate (GFR) was estimated using the four-variable MDRD4 formula (GFR [mL/min/1.73 m2] = 32841 × (creatinine [μmol/L][−1.154]) × age [years][−020]) following Levey et al.28 Creatinine clearance rate was estimated using the Cockcroft-Gault formula.29

Go to:

Pre-event results

In the incremental test, the athlete achieved a maximum speed of 17 km/hour, a VO2max of 58 mL/kg per minute, and a maximum heart rate of 188 beats per minute. Lactate threshold was set at 10 km/hour.

Performance

The athlete invested a total time of 7 hours 10 minutes, including breaks, to complete the marathon, which comprised a total of 2400 laps for 17.6 m in the climate chamber. He was running for 6 hours 30 minutes at a mean speed of approximately 6.5 km/hour, with a mean heart rate of approximately 112 beats per minute, expending total energy of 4434 kcal. Body mass decreased to 93.1 kg (−1.6%). He drank 3.4 L of fluids and consumed 2050 kcal of energy (Figure 2).

Temperature

Tskin at the head showed a total increase of 9°C (minimum 28.5°C, maximum 37.5°C). Tskin at the hand showed a total increase of 9°C (minimum 29.5°C, maximum 38.5°C). Tskin at the ankle showed a total increase of 16°C (minimum 23°C, maximum 39.5°C). Tcore varied by 0.9°C (minimum 37.5°C, maximum 38.4°C). The variations in Tcore/Tskin during the testing phase out of the climate chamber were depicted at every measurement. Tskin at the head was mainly influenced by the transition Tambient out of the climate chamber (Figure 3).

Sweat

Sweat parameters are shown in Table 1. Estimated sweat production increased from 40.3 mL at 10.5 km to 120.8 mL at 21.1 km. At the end of the marathon, total sweat production was estimated to be 685 mL (+888%). The starting sweat osmolality measurement was 215 mOsm/kg and then fell to 89 mOsm/kg, after which the values remained constant, ending at 95 mOsm/kg. Sodium in sweat increased by 56%, after an initial decrease, ending at a level of 57 mg/L. Calcium in sweat decreased initially (−84%) and continued, ending at 35.3 mg/L. Potassium decreased continuously, resulting in a total change of −83%.

Laboratory results

Go to:

Differences in skin temperatures between ankle and wrist

Tcore in mammals is typically in the range of 37°C–39°C. Tcore in our athlete never exceeded this range. The changes in Tcore were similar to the changes reported by Cheuvront and Haymes,4 with deviations due to breaks for measurements and external cooling. Assuming the range of normothermia by Kanouse et al,14 we hypothesized that the increase in Tcore before the first break and after the second break was due to a temporary inability to defend regulated Tcore against a continually increasing Tambient. This increase was only temporary and regressed, demonstrating that the primary mission of thermoregulation is to defend Tcore against a challenging Tambient.14 We found a variation in Tcore and Tskin measurements during the testing phase outside the climate chamber. Tambient out of the climate chamber affected each point of measurement, but mostly Tskin at the head. This may be due to the changing Tambient outside the climate chamber, but it should also be noted that ingestion of cold fluids acts as a cooling element.30 The Tskinvariation developed differently, with the highest difference measured at the ankle with a ∆ of 16°C. Tskin at the ankle was the most distant site from Tcore, but also nearer to a working muscle than, eg, Tskin at the wrist. Therefore, Tskin was the site most able to assimilate Tambient and, therefore, Tskin at the ankle was least defended by regulatory circuits to keep Tcore in range.

Differences in sweat production over time

At a Tambient of −35°C, the Tcore to Tskin gradient was high (head to core 9.7°C; hand to core 8.7°C; ankle to core 14.7°C, respectively) and humidity in the climate chamber was set at 80%. The evaporative capacity seemed to be lower at that time compared with the time in the middle of the marathon. This may explain the sweat production at 10.5 km of 40.3 mL and osmolality in sweat of 215 mOsm/kg. As the Tcoreto Tskin gradient decreased, the sweat volume increased, and serum osmolality was slightly increased at 31.6 km. In temperate or cool conditions, a large skin-to-ambient temperature gradient facilitates radiant and convective heat loss, and reduces blood flow requirements to the skin, which may explain the tolerance for the high Tcore observed during marathons in cool conditions.31 For methodologic reasons, the initial evolution of ion concentration in sweat, as proposed by Shamsuddin et al,32 could not be shown. After the low sweat osmolality at 10.5 km, we suggest that the following stable values we recorded demonstrate good adaption under changing environmental conditions. Furthermore, the index of ion concentration in sweat at the end of the study may indicate a capacity for reabsorption in the eccrine sweat duct.

Continuous increase in BNP and interleukin-6

Pathologic values identified in the laboratory parameters were reassessed against references for marathon runners.15 Thereafter, all hematologic values were within range, except for creatine kinase in serum and creatinine in urine before the start of the marathon. In contrast, serum creatinine was within range, and its value rose to 10% above its initial baseline. Approximately 40% of marathoners experience a transient increase in serum creatinine, and the data suggest that acute kidney injury with a transient and minor change in renal filtration function occurs with the stress of marathon running.33 Following the stages of the Acute Kidney Injury Network,34 our marathoner was in stage 1. At the start, creatine kinase values were outside the adapted level of 6.24 μmol/L suggested by Kratz et al, and rose again to an end value of 10.4 μmol/L.15 However, a massive increase in creatine kinase is common, and is widely accepted as an indirect marker of skeletal muscle damage while running.35–39 There is a suspicion that the extreme increases in serum creatine kinase activity may be related to development of exertional rhabdomyolysis.40Another factor impairing kidney function is the redirection of blood flow, whereby exercising blood is directed more to the working muscles than to the kidneys, which decreases renal perfusion.19 Furthermore, proendothelin-1 was increased by +79%. Maeda et al discussed the possibility of reduced kidney perfusion by an exercise-induced increase in production of endothelin-1 in the kidneys.41 The increase in endothelin-1 causes decreased blood flow to the kidneys via two pathways, ie, a vasoconstrictive action and attenuation of nitric oxide production. BNP increased continuously, and an increase in BNP was considered to be a sign of high cardiovascular performance due to a low blood volume.11,42 Following Hew-Butler et al, BNP may assist arginine vasopressin and atrial natriuretic peptide in the regulation of fluid balance during extreme physical stress.43 Values during the marathon showed a more continual increase of BNP, whereas arginine vasopressin and atrial natriuretic peptide fuctuated more, which supports this hypothesis. The positive effects of atrial natriuretic peptide seen during a marathon could reflect regulation of body temperature by influencing the sweat glands as well as stimulation of lipolysis, compensating for the enormous energy demand.44 Furthermore, we noted that compared with exercise at room temperature, plasma levels of epinephrine, norepinephrine, and cortisol were all higher during exercise in the heat.45

There were increased counts for inflammatory parameters such as leucocytes and interleukin-6. Surprisingly, the plasma C-reactive protein concentration was immeasurably low. Therefore, recovery of plasma C-reactive protein levels possibly occurred during the recovery time after finishing the marathon, as reported by Waśkiewicz et al.46 Cytokines, particularly interleukin-1 and interleukin-6, act as endogenous pyrogens in the brain. Interleukin-1 and interleukin-6 stimulate thermogenesis via synthesis of prostaglandins and corticotropin-releasing factors.47 This induction of thermogenesis by peripheral interleukin-6 seems not to last, but influences the new regulated Tcore.48

Increase in TSH and decrease in ACTH

Several stress hormones, including cortisol, somatotropic hormone, and normetanephrines were released, in urine during prolonged exercise together with increasing Tambient. These findings are compatible with the results reported by Niess et al,49 who suggested that elevated Tambient primarily affected the cardiocirculatory and hormonal systems, resulting in changes in neutrophil and monocyte cell trafficking. In contrast, the acute immune response seems not to be influenced further by additional heat stress during endurance exercise, which is also compatible with our results. TSH increased by 65.3%. It is known that thyroid hormones increase nonshivering thermogenesis in cold conditions by stimulating uncoupling protein-1.50 We found no increase in TSH at 10.5 km during the marathon when Tambient was low. This fact was most probably due to the predominance of endogenous thermogenesis while running. Prolactin increased by 221.9%, together with an increase in TSH of 65.3%. It has been reported that an increase in prolactin is directly related to an increase in TSH during prolonged exhaustive exercise.51 Jimenez et al reported prolactin increases only during passive hyperthermia and not while exercising.52 In the present case, an additive exogenous heat load may have contributed to the increased prolactin levels towards the end of the study, and thus support the use of prolactin as a biomarker for development of central fatigue.53In contrast, ACTH decreased after an initial increase at 10.5 km continuously to the end value. An initial peak of ACTH is most probably explained by the cold environment54 stimulating the release of glucocorticoids, which would then increase the availability of glucose and assist lipolysis. The rise in temperature may play a consequential role, because vasoconstriction and also cortisol must be inhibited for thermolytic reasons in a hot environment.55 The decrease in cortisol after a peak may be explained by this fact. The peak at the end of the study may have been due to the psychologic stress of running in such high temperatures. Alternatively, following the suggestions of Niessner et al,44 the increase in adrenocortical steroids may have compensated for the negative influence of atrial natriuretic peptide on natriuresis and blood pressure, which would explain this peak at the end too. Renin increased steadily to +1146% at 42.195 km, as did aldosterone, which increased by +314%. This metabolic response may support the fact that plasma osmolality, serum sodium concentration, and safe Tcore were maintained. Therefore, we show in this case study that humans were able to exercise in extreme conditions while drinking water ad libitum, as suggested by Nolte et al.56

Limitations and implications for future research

This case study performed under controlled conditions shows the individual response of one athlete. The results might be different in another athlete, and different again in the same athlete on a further attempt. Sweat sampling was performed on the back of the athlete and extrapolated to the total surface of the athlete. However, sweating might be different on different parts of the body, so the total amount of sweat produced in this study might have been higher than reported. The athlete lost 1.6% of his body mass. However, he completed the marathon in 6.5 hours. Whether his performance could have been worse or better with a more regimented drinking/food intake strategy cannot be commented upon without further research. The same study is to be performed in more subjects with the same characteristics to validate the present findings.

Go to:

Go to:

The authors thank Holger Häusler and Lothar Kleinod for their technical assistance with the climate chamber. Furthermore, they thank Andreas Heertsch, business manager of EndoTherm GmbH, Arlesheim, Switzerland, for making valuable equipment available during the study period. We also thank Therese Fässler for her help in translation.

Go to:

 

Disclosure

The authors have no conflict of interest in this work.

 

Go to: