Relationships between strength and endurance parameters and air depletion rates in professional firefighters

The aim of this study was to quantify the physical demands of a simulated firefighting circuit and to establish the relationship between job performance and endurance and strength fitness measurements. On four separate days 41 professional firefighters (39 ± 9 yr, 179.6 ± 2.3 cm, 84.4 ± 9.2 kg, BMI 26.1 ± 2.8 kg/m2) performed treadmill testing, fitness testing (strength, balance and flexibility) and a simulated firefighting exercise. The firefighting exercise included ladder climbing (20 m), treadmill walking (200 m), pulling a wire rope hoist (15 times) and crawling an orientation section (50 m). Firefighting performance during the simulated exercise was evaluated by a simple time-strain-air depletion model (TSA) taking the sum of z-transformed parameters of time to finish the exercise, strain in terms of mean heart rate, and air depletion from the breathing apparatus. Multiple regression analysis based on the TSA-model served for the identification of the physiological determinants most relevant for professional firefighting. Three main factors with great influence on firefighting performance were identified (70.1% of total explained variance): VO2peak, the time firefighter exercised below their individual ventilatory threshold and mean breathing frequency. Based on the identified main factors influencing firefighting performance we recommend a periodic preventive health screening for incumbents to monitor peak VO2 and individual ventilatory threshold.

depletion from the SCBA has hardly been researched yet 19,20 . Moreover, the relationship between fitness variables and air depletion has not been established yet.
Together with well-researched parameters of completion time and physical strain, we propose that air depletion from the SCBA can provide additional, extremely valuable information. To our knowledge, there is no study researching firefighting performance as a combination of these three parameters. For this study, we therefore defined a simple formula to quantify the demands of the simulated firefighting exercise adding time needed for the exercise, heart rate and air depletion from SCBA. The aim of our study was 2-fold: (1) Quantification of the physical demands of a simulated firefighting exercise by the simple formula taking into account completion time of the exercise, heart rate and air depletion rate. (2) Establishment of the relationship between firefighting performance and highly standardized fitness measurements in order to identify the most relevant physical and physiological attributes to fulfill the job demands of a professional firefighter. From this approach we expected to characterize firefighting more in detail. We hypothesized that firefighters with lower air depletion from the SCBA, fast completion time and lower physical strain during the simulated firefighting exercise possess a higher fitness level.

Materials and Methods
Subjects. Forty-one male career firefighters (39 ± 9 yr, 179.6 ± 2.3 cm, 84.4 ± 9.2 kg, BMI 26.1 ± 2.8 kg/m 2 ) of the Munich Airport volunteered to participate in the research. Full written and verbal details about the study were provided. Informed written consent was obtained from all participants prior to testing. The ethic statement for this study was approved by the Dean of the Faculty of Sports and Health Sciences of the Technical University of Munich. All tests were conducted according to the Declaration of Helsinki. All participants were in possess of a valid G26.3 medical examination for operational fitness, a mandatory periodically medical health check for firefighters in Germany.

Material and methods.
The tests were conducted on four days each separated by at least 4 days in the following order: Day 1 -Test of VO 2peak during maximal treadmill running and anthropometric evaluation. Day 2 -Flexibility, balance, muscular strength and muscular endurance testing. Day 3 -Respiratory protection exercise (REPE standard ) with SCBA. Day 4: Respiratory protection exercise (REPE spirometry ) with a spirometry mask.
All subjects wore functional sportswear and -shoes during the VO 2 peak testing, muscular strength and endurance testing.
Anthropometric evaluation. Body mass (kg) was recorded with the nearest 0.1 kg on a scale with shoes removed. Body height was measured by a tapeline with the nearest 0.1 cm of the maximum distance from the floor to the vertex of the head with shoes removed. Body Mass Index (BMI) was calculated by the following formula: Bodyweight in kilograms divided by height in meters squared (kg/m 2 ). VO 2 peak testing. Minute ventilation (V E ) and gas exchange (oxygen consumption -VO 2 , carbon dioxide output -VCO 2 , respiratory exchange ratio -RER) were measured breath-by-breath with the Cortex Metamax 3B (Cortex Biophysics GmbH, Germany). An incremental exercise test based on the Ellestad Protocol 21 was conducted on a motorized treadmill (Life Fitness, Integrity Series, Germany) to determine peak oxygen uptake (VO 2peak ), minute ventilation (V E) and heart rate maximum (HR max ). The test was terminated when subjects reached volitional fatigue and were not able to continue running. VO 2peak and HR max were taken as the highest 30s-average during the final minute of the test. In addition, two thresholds were determined based on the test: ventilatory threshold 1 (VT1) and respiratory compensation point (RCP). The VT1 was determined from the V-slope method 22 in combination with the break point of the ventilatory equivalent for O 2 against VO 2 23 . The RCP was identified by the break points of the ventilatory equivalent for CO 2 and the end tidal CO 2 concentration against VO 2 23 . These two thresholds were then used to establish three physiological intensity zones that correspond to the heart rates at the following exercise intensities: Zone 1, 2 and 3 were represented by the percentage of time subjects experienced HR below VT1 (Zone 1), HR between VT1 and RCP (Zone 2) and HR above RCP (Zone 3), respectively.
Flexibility, balance, muscular strength and endurance testing. A description of standardized fitness tests can be found in Table 1. All tests were completed sequentially with a break of at least 4 minutes between each test. A standardized warm up of 20 minutes on a cross-trainer (Life Fitness, Integrity Series, Germany) preceded the tests.
Simulated firefighting exercise test protocol. The simulated firefighting exercise was completed twice by each subject. One trial was with wearing full gear and the second trial was with full gear, but without the facemask, and wearing a portable metabolic measurement system. Respiratory Protection Exercise Standard (REPE standard ): This exercise is a standardized, mandatory and periodically performed ability test for professional German firefighters. The test was conducted as prescribed by German firefighting test regulations 24 . Subjects were tested in a purpose-built practice area, wearing full personal protection gear (clothing, helmet, gloves, belt, facial mask, boots) and SCBA. The SCBA cylinders were filled with 300 bar according to the standard protocol for the fire services. The weight of the protection gear and SCBA was approximately 22 kg. The tasks included ladder climb (20 m), a 200 m treadmill walk, pulling a wire rope hoist (15 times) and crawling a 50 m orientation section in the dark with bottlenecks and a narrow tunnel. Subjects were instructed to perform the REPE safely and as fast as possible but in a pace similar to the work at a real fire emergency scene. The tasks were performed in succession without interruption but with individually chosen pace and possible breaks in case of exhaustion. During the REPE standard , heart rate was measured continuously (Polar, Scientific RepoRts | 7:44590 | DOI: 10.1038/srep44590 Finland) and ratings of perceived exertion 25 as well as air depletion from the SCBA were taken at the end of the exercise. Individual task time and total performance time were recorded.
Respiratory Protection Exercise with Spirometry (REPE spirometry ): The exercise protocol for the REPE spirometry was identical with the testing of the REPE standard but included spirometric measurements. This measurement provided additional information in terms of respiratory variables during firefighting. The standard facial mask of the SCBA was replaced by the mobile spirometry mask of the Cortex Metamax device to measure VO 2 , VCO 2 , VE, RER and ventilatory equivalents (VE/VO 2 ; VE/VCO 2 ). These variables were measured breath-by-breath and were then used to define the metabolic demands of the REPE. Subjects still wore the SCBA (without facial mask) to simulate the weight of their equipment.
Firefighting performance formula. We defined a simple formula to quantify the demands of the simulated firefighting exercise adding time needed for the exercise, mean heart rate during exercise expressed as percentage of the treadmill determined HR max and air depletion from SCBA. We included the three variables in our formula because we defined optimal firefighting performance due to three important key aspects: (1) How much time do firefighters need to complete a given simulated firefighting exercise ? (2) What are their physiological responses to the chosen operation speed? (3) How much air do they consume from their SCBA due to operation speed and work intensity? We defined this as the time-strain-air depletion (TSA) formula resulting in a TSA score: As the impact of every single factor on overall TSA-score is not clear at the moment, we used z-transformations to prevent different weighting of the parameters due to their different absolute values and normal distributions. The resulting z-scores allow us to compare and sum up the three parameters resulting in the TSA-score. As this score is based on the function of a z-score, the TSA-score indicates the resultant firefighting performance in relation to the sample mean, with the distance measured in standard deviations. A TSA-score of 0 represents the average. We ranked performers according to their TSA-scores into 5 categories based on standard deviations: "Outstanding" (TSA < − 2), "Above Average" (TSA − 1 to − 2), "Average" (TSA − 0.99 to + 0.99), "Below Average" (TSA 1 to 2), and "Poor" (TSA > 2). Individual performance scores for the TSA should be kept at a minimum achieved through fast completion time, low heart rate as well as low air depletion during the exercise.

Data analyses.
Statistical calculations were carried out with SPSS version 23.0 (IBM Corporation, USA).
Descriptive statistics (means, standard deviations (SD)) were calculated to define subjects with respect to physical characteristics and performance in the tests. For legpress, handgrip and one-leg standing, data were taken as the average of left and right. Data were assumed to be normally distributed if the Shapiro-Wilk's test was > 0.05.

One-leg standing with eyes closed
Subjects had to show their one-leg standing balance with eyes closed for 15 seconds. Subjects stand on the supporting leg with the free leg raised and bent 90°. Assessment scheme: 2 points -Subject remained unmoving for 15 s. 1 point -Subject performed the exercise with compensation movements. 0 points -The supporting leg leaves its position (jumping as compensation movement) or the free leg was moved back to the ground or eyes were opened.

Sit and Reach Testing
This test, performed on a traditional 32.4-cm-high and 53.3-cm-long box, was used to obtain flexibility assessments for lower back and hamstring muscles. The subject sat on the floor with its legs fully extended with the bottom of the bare feet against the box. The subject placed one hand on top of the other, slowly bended forward and reached along the top of the ruler as far as possible holding the stretch for two seconds. The distance reached by the subject's finger tips (cm) was recorded. The test was performed three times. The average of the three distances were calculated

Standing Long Jump
Subjects placed their toes behind the takeoff line and were instructed to jump as far as possible forward with arm swing being allowed. The best jumping distance (cm) out of three trials was registered 1-RM testing legpress (one-legged) Maximum strength was obtained using a predictive one-repetition (1-RM) formula as described by Brzycki 42 . Values (in kg) were taken from the left and the right leg and the maximum values of both legs averaged

Hand grip strength
The grip size of the force dynamometer (Jamar: Lafayette Instrument, Lafayette USA) was individually adjusted to fit the proximal interphalangeal joint of the third finger. In a standing position, with the elbow bent 90 degrees alongside the body, the subjects squeezed the dynamometer as hard as they could. The best of three trials on each hand was registered for the maximum (in kg) and the maximum values of both sides averaged

Push-ups (reps to fatigue)
The test was taken as the number of times the firefighter could perform push-ups with shoulder-width space between his hands at a rate of 30 lifts per minute

Partial Curl-Ups
The partial curl-ups test was used to measure muscular endurance of the abdominal muscle 43 . The total number of properly performed curl-ups (reps to fatigue) was recorded.

Shoulder Press (20 kg)
In a sitting upright position, subjects grasped two 10 kg dumbbells with a pronated grip, vertically pushed the attachment from chin level up to straight arms overhead, and then pulled back to the starting position. The test was taken as the number of times the firefighter could raise and lower the dumbbells in a seating position at a rate of 25 lifts per minute (metronome set at 50)

Rowing
In a seating position, the subjects grasped two 7.5 kg dumbbells with a pronated grip. The test was taken as the number of times the firefighter could row in a seating position by abducting arms (90°) in sagittal plane with two 10 kg dumbbells. The weight was lifted between the spina iliaca anterior superior and the chin at a rate of 30 full lifts per minute (metronome set at 60). The number of completed lifts was registered Table 1. Description of balance, flexibility, strength and muscular endurance testing.
As all data was normally distributed, parametric tests were consequently carried out. The alpha level was set to 0.05. A paired t-test was calculated to show up differences between the two firefighting exercises REPE standard and REPE spirometry . Reference values according to Cohen 26 were used to interpret the correlations. Values from 0.10-0.29 were considered 'small' , 0.30-0.49 'moderate' and ≥ 0.50 'strong' . The combination of physical characteristics that best predict TSA was determined by multiple regression (Enter Method). The combination of variables that resulted in the highest explained variance that predicted the largest portion of the variance was then selected.

Results
Aerobic fitness, muscular strength, flexibility and balance testing. . Mean air depletion from the air cylinder averaged 161.7 ± 28.7 bar. In the first part of the REPE standard (ladder climb, treadmill walk and hoist), mean air depletion was 85.6 ± 16.8 bar which corresponded to 28.6 ± 5.6% of the capacity of a nominal 30 min-cylinder. In the second part, the orientation section crawling, air depletion ended up in 76.3 ± 19.1 bar (25.5 ± 6.3%). In total subjects consumed 54.1 ± 9.9% of the capacity of a nominal 30 min-cylinder.
Respiratory demands of the REPE spirometry . Mean heart rate during the REPE spirometry was 144.3 ± 12.7 bpm corresponding to 79.8 ± 7.3% of HR max . Exercise total time of the REPE spirometry averaged 797 ± 122 s (13.3 ± 2.0 min). Mean heart rates (p = 0.433) and exercise total time (p = 0.858) showed no significant differences between REPE standard and REPE spirometry . The mean oxygen consumption for the whole REPE spirometry exercise was 2.13 ± 0.32 l/min. Among the different exercise elements, ladder climb required the highest absolute oxygen uptake (2.51 ± 0.39 l/min). Corrected for body mass, mean VO 2 was 25 ± 3 ml/min/kg across the whole exercise, 30 ± 4 ml/min/kg during ladder climb, 27 ± 6 ml/min/kg during hoist and 26 ± 6 ml/min/kg both during treadmill walk and the orientation section crawling. The two most demanding tasks required 38 ± 6 ml/min/kg over 20 seconds during orientation section crawling and 38 ± 5 ml/min/kg at the ladder climb (Fig. 1).
Mean minute ventilation during the whole exercise was 67.5 ± 13.1 l/min. Hoist showed the highest mean minute ventilation rate (74.5 ± 17.4 l/min), followed by the orientation section (70.9 ± 14.5 l/min) and the ladder climb (60.9 ± 16.7 l/min). Mean breathing volume values during the exercise were 2.08 ± 0.33 l. Ladder climb and hoist required breathing volumes of 2.4 l, the treadmill walk 2.28 ± 0.45 and the orientation section 1.9 ± 0.33 l. The mean breathing frequency was registered with 34.1 ± 4.8 breaths per minute. The orientation section required the highest number of breaths per minute (38.9 ± 5.8), followed by hoist (32.0 ± 5.5), ladder climb 29.1 ± 4.9) and treadmill walk (27.4 ± 5.1) (see Fig. 1). Furthermore, respiratory exchange ratio (RER) averaged 1.08 ± 0.08 across the total exercise. 36.0 ± 21.7% of total exercise time subjects had a RER < 1.0 and 64.0 ± 21.7% of time a RER ≥ 1.0.

Variable
Mean ±SD  Relationship between TSA-score and fitness characteristics. Thirteen firefighters obtained a TSA-score of − 0.99 to + 0.99 (average), 9 firefighters a TSA-score between − 1 and − 2 (above average) and 6 firefighters a score smaller than − 2 (outstanding). Furthermore, 6 firefighters obtained a score between 1 and 2 (below average) and 7 subjects a TSA-score of more than 2 (poor) (Fig. 2). As there was no significant difference between mean HR and mean completion time of REPE standard and REPE spirometry , we assumed that strain and duration of both exercises were comparable. Therefore, we used all REPE standard and REPE spirometry variables in addition to variables from treadmill and muscular strength, flexibility and balance testing to find the most predictive parameters for firefighting by multiple regression. Based on our performance model (Table 3), multiple regression identified three main factors that show a great influence on optimal firefighting performance in terms of the TSA-score (70.1% of total explained variance): relative VO 2peak from maximum treadmill testing, mean breathing frequency and the percentage of time spent in Zone 1 during REPE standard . Figure 3 shows the relationship of all three parameters to TSA-score. To better understand the characteristics of firefighters with respect to different TSA-scores, Table 4 shows TSA-parameters, the main identified variables by regression and additional variables for all categories of performers. It can be noted that outstanding performers had significantly higher VO 2peak (p = 0.001) and significantly lower mean heart rates during REPE (p = 0.001) while completing the exercise faster (p = 0.001) compared to average, below average and poor performers. The differences of VO 2peak -levels and time spent in zone 1 of different TSA-performers are highlighted in Table 4. The poorest performers also showed an increased perceived exertion when rating the BORG scale after the exercise. Furthermore, the outstanding performers were the only subjects performing the REPE parcours without spending any time in Zone 3 and showing the highest fraction of time spent in Zone 1.

Discussion
The results of this research describe, for the first time, firefighting performance as a combination of operating speed (time to complete the circuit), physical strain and air depletion during a simulated firefighting exercise.
Firefighting is a physically demanding occupation. Several authors offered evidence for high physiological responses above 80% of peak relative oxygen uptake (VO 2 ) during the completion of simulated task circuits 1,5,7 .
Others reported values between 47% and 80% of relative peak oxygen uptake 4,8,15,20 . In our study, relative VO 2 averaged at 56% of VO 2peak across the exercise, which was towards the lower end of the range of average values reported from other studies. The values reported for single firefighting tasks within circuits varied from 23 ml/ min/kg for boundary cooling 1 to 44 ml/min/kg for tower stair climbing 27 . Literature indicates stair climbing 6,27 and victim rescue 28 to be the most arduous tasks, requiring a VO 2 of 38-43 ml/min/kg over 20 seconds. These findings are comparable to our values determined for ladder climb and orientation section crawling, although

Figure 1. Physiological responses during the REPE: Heart rate (HR), peak oxygen uptake absolute (VO 2peak absolute) and relative (VO 2peak relative), minute ventilation (VE), breathing volume (BV) and breathing frequency (BF) during ladder climb, treadmill walk, hoist, orientation section and the overall exercise.
Data are shown as means ± standard deviations (SD). *Significant difference between tasks (P < 0.05). measured VO 2 rather represent the lower end of the reported ranges. During the REPE circuit, HR averaged at 79.2 ± 6.6% of HR max determined on the treadmill. These findings were consistent with values reported from other studies ranging from 61% 29 to 95% 4 of HR max . The most common physiological responses in terms of mean HR during the firefighting exercises averaged between 80% and 90% of HR max 3,7,9 . However, in our study, we analyzed not only mean heart rate but also the time spent in the three defined physiological intensity zones. These zones indicate the contribution of different energy sources to total exercise performance and provide more detailed information on cardiovascular load during firefighting. Subjects spent most time of their exercise time in Zone 2, the aerobic-anaerobic metabolic transition zone (69% of time), whereas the aerobic fraction (Zone 1) represented approximately 22% of exercise time. The smallest fraction (9% of time) was Zone 3 representing an anaerobic metabolism and indicating the onset of hyperventilation. A high fraction of time in Zone 3 will lead to subject's rapid fatigue, whereas a high fraction of time in Zone 1 confirms a subject's good aerobic metabolism 30 . To our knowledge, only one other study analyzed the three physiological intensity zones in the same way and found a distribution of energy metabolism of approximately 84% Z1, 12% Z2 and 2% Z3 31 . These results show important differences to our findings, however, it should be noted that the mentioned study investigated prolonged (> 120 min) wildland firefighting and exercise time was almost 10 times longer than the simulated firefighting tests we investigated.
Subjects showed a RER ≥ 1.0 during 64% of total exercise time indicating a major contribution of anaerobic energy due to more CO 2 being produced than O 2 consumed. An increase of RER above 1.0 would only be expected, if VO 2 exceeded 89% VO 2peak 32 . During the maximum treadmill testing in our study, we established an average RER of 1.08 not before subjects ran at an intensity of 90% to 97% of VO 2peak . However, during the firefighting exercise, subjects reached an average RER 1.08 already at an intensity of 56% of VO 2peak . These observed RERs were out of line with RERs found during other moderate activities between 50 and 60% VO 2peak . For example, Davis et al. 33 reported average RER values of 0.85 during treadmill running at exercise intensities of 58% VO 2peak . Although the mean values for VO 2 were relatively low in our study, RER averaged 1.08 across the total exercise, representing an unexpected high level close to maximal exertion of subjects which would correspond to a RER of 1.15. One primary reason for the differences in RER in relation to VO 2 can be the different muscular strain that occurs during a running treadmill test compared to a simulated firefighting test. Firefighting includes many start-and-stop motions similar to game sports. Previous studies showed that the muscular strain during game sports affected metabolic parameters such as RER differently compared to respiratory parameters like VO 2 34,35 . This unusual VO 2 /RER relationship has also been found by Harvey et al. 15 Table 3. Multiple regression model (using Enter-Method) to predict optimal firefighting performance (TSA-score). Multiple r = 0.850, r 2 = 0.723, corrected r 2 = 0.701, standard error = 1.250. during firefighting exercises. Another possible explanation for that can be the influence of a firefighter's turnout gear and equipment (e.g. SCBA) on different physiological variables. According to Perroni et al. 36 , wearing additional weight in terms of full protective clothing and the SCBA reduces a subject's VO 2peak by ~27% from 55 to 43 ml/min/kg. Therefore, and based on our findings, we suggest that assessing only mean VO 2 values from a simulated firefighting test will fail to represent the demands of firefighting. The pattern of RER provides considerable inside into the true metabolic demands which we found were hidden when assessing only mean VO 2 -values. Mean minute ventilation (67.5 ± 13.2 l/min) across the whole REPE spirometry exercise was lower compared to the values reported from Holmer and Gavhed 27 . Mean breathing frequency was highest averaging 38.9 ± 5.8 breaths per minute during the last task, the orientation section crawling. In contrast, subjects mean breathing frequency in the final minute of maximum treadmill testing averaged at 41 ± 5 breaths per minute. Accordingly, subjects were close to their maximum breathing frequency during the orientation section crawling, which indicates the strenuous nature of the exercise. These findings are underlined by McArdle, Katch, & Katch 37 showing that breathing frequency increased to 35 to 45 breaths per minute during strenuous exercise. Subjects consumed 54% of the capacity of a nominal 30 min-cylinder during the REPE standard circuit which lasted 801 s (13.2 minutes). However, these rates of compressed air consumption from the SCBA would have depleted the air supply after 1464 s (24.4 min) and that is before the nominal time of 30 min described for the cylinders. These findings are in line with the data reported from Williams-Bell et al. 20 who determined similar air consumption rates (~51%) and mean VO 2 values (24 ml/min/kg) for a firefighting exercise of 12.1 minutes duration. For example, the best firefighter regarding his TSA-Score (− 4.81) completed the exercise in 10 minutes, consumed 90 bar of compressed air while showing a mean heart rate of 74.2% HR max . He had a VO 2peak -level of 49 ml/min/kg, spent 72% of time in zone 1 and breathing frequency averaged at 19 breaths per minute. With the shown values he would not have depleted the air supply before the nominal time.
The reasons for the selection of time for completion, heart rate and air depletion rates for our firefighting performance formula were primarily based on the rationale nature of firefighting. When arriving at an emergency scene, firefighters have to work as fast as possible in order to e.g. save lives or prevent the spread of fires. Furthermore, previous data showed that less fit firefighters experienced higher physiological strain near HR maximum not being able to sustain operating speed and therefore not being able to complete firefighting tasks successfully 12 . Finally, firefighters can run out of air supply due to the limited amount of air compressed in the SCBA. Based on these considerations each of the three factors is thought to be important so that all three parameters were included into our TSA firefighting performance formula. Since at present we do not know which factor is most important or whether one factor is more important than another, we used z-transformations in order to avoid unintended weighting of one of the three parameters.    The results demonstrate that performers of different TSA-levels showed significant differences in maximal endurance parameters, the capacity to work below their ventilatory threshold 1 and breathing variables. This is a strong argument that a high firefighting performance comes along with a good aerobic metabolism and confirms our initial hypothesis: firefighters with lower air depletion from the self-containing breathing apparatus, faster completion time and lower physical strain during the simulated firefighting exercise possess a higher aerobic fitness level in terms of VO 2peak . Indeed, this can clearly be proven by a low TSA-score which can be seen as evidence for the usefulness of the new developed firefighting performance formula. For practical application, the TSA-formula also can be used without z-transforming all three parameters. The product of time for exercise completion, mean heart rate and air depletion from the SCBA (Time*Strain*Air Depletion) is highly related to our z-transformed TSA-formula (Time + Strain + Air Depletion) (r s = 0.974, p = 0.000). However, further research is needed to validate the model, as a variety of weighting options may even improve predictability of firefighting performance.
Out of all parameters we measured we identified the most important firefighting determinants by means of multiple regression. We found a combination of laboratory (VO 2peak ) and occupation-specific parameters (breathing frequency and time spent in intensity zone 1 during the simulated firefighting exercise) that predicted TSA-score best, accounting for 70% of the observed variance. These results are in line with the results of Sothmann et al. 12 , who were one of few research groups recommending a subset of screening tests for firefighters incorporating simulated work tasks as well as fitness measurements. In their study the combination of test items such as hose drag, high rise pack, arm cranking and lifting accounted for 50% of the variance associated with the completion of a work task circuit. However, like other previous studies, Sothmann et al. 12 only focused on the completion time of the simulated firefighting tasks for the classification of performance. Other authors found combinations of aerobic fitness and strength parameters accounting for variances around 60% 2,9,14 and 70% 13 . As one of the main differences between our study and previous research, we could not identify any strength parameters as important performance predictors in the regression model. This is surprising given that the characteristics of typical firefighting tasks such as chopping, carrying heavy equipment require the intense use of upper body muscular strength. Except push-ups (p = 0.039, r s = − 0.40), we found no significant correlations between TSA-score and muscular strength, flexibility and balance variables. Moreover, integrating push-ups into the regression model could not increase the predictive power of 70% of explained total variance. However, with regard to the strength fitness profile our subjects showed a comparable strength level to other studies 11,14,16,38 .
Although muscular strength and flexibility in this study did not show significant relevance for the predictive power of job demands, both should be essential components of firefighting training in order to decrease the risk of job injuries.
The results of multiple regression support the idea that aerobic fitness, in terms of VO 2peak and the time spent in zone 1, considerably contributes to how fast (time) and effective (low air depletion from SCBA and minimal physical strain) a firefighter can perform his tasks. Treadmill determined VO 2peak has been established previously to be important for firefighters 1,5,9,10,28 and thus a high level of VO 2peak is postulated. Recommendations for a minimum relative VO 2peak -threshold varied between the suggestions of O' Connell 6 with 39 ml*kg −1 *min −1 and the recommended values by Gledhill and Jamnik 10 with 45 ml*kg −1 *min −1 . Our results also emphasized the importance of VO 2peak , as a high VO 2peak is related to a faster operating speed, lower strain and lower air depletion from SCBA. Based on our results, we now recommend a slightly higher minimum VO 2peak of 46 ml*kg −1 *min −1 as this value was identified for subjects showing at least average performance in terms of TSA-scores.
In addition to VO 2peak as one of the primary determinants of aerobic endurance performance 30 , the time spent in zone 1 was identified to be the second most important fitness factor. Therefore, heart rate kinetics and the contribution of aerobic energy sources need to be considered to play a major role in preparation and shaping of fit and healthy firemen. Furthermore, we found a strong correlation (r = 0.69, p = 0.001) between VO 2 at VT1 on the treadmill and the time spent in Zone 1 across the REPE exercise. This means that subjects with a high percentage of time spent in zone 1 possessed a high VO 2 at VT1. Those subjects can work at a higher exercise intensity while still covering the energy demand aerobically. Lemon and Hermiston 28 pointed out that firefighters with a higher VO 2peak and a high VT1 (as %VO 2peak ) are able to supply a greater percentage of the total energy demand aerobically which results in more work efficiency in terms of total physiological demands on the organism. These findings can help to design endurance exercise programs for firefighters more detailed by focusing not only on VO 2peak -training but also improving VT1. According to Jones and Carter 30 , an improvement in VT1 with training is a clear marker of an enhanced endurance capacity.
In our study, VT1 averaged at 49% of VO 2peak across all subjects. This is a strong indication to extend basic aerobic endurance training, as values between 50 to 60%VO 2peak are related to a low basic endurance level 39 . As suggested by Farrell et al. 40 , aerobically better trained subjects can exercise at 75-85% of VO 2peak while still covering their energy demands aerobically. We therefore recommend a VT1 at 60-80% VO 2peak for professional firefighters as it would allow better metabolic adaptation to physical work at this level 41 . It would also enable to increase Zone 1-fraction and reduce the physical strain during firefighting, respectively. Furthermore, breathing rate can be sustained at lower intensity levels and the blow off of the extra CO 2 produced by the buffering of lactic acid metabolites is reduced. Oxygen needs can then be primarily met by an increase in tidal volume instead of increased breathing frequency. Moreover, increased breathing frequency was identified to have a negative effect on TSA-score based on the results of multiple regression.

Conclusions
Firefighting is a physically demanding activity challenging both the aerobic and anaerobic system. While other studies researching firefighting activity focused on VO 2 and HR, we strongly emphasize to also take RER values and VT1 into account when assessing the fitness level of firemen. Based on the results of our study, we recommend a 3-fold fitness analyses for firefighters that allows for designing optimized, detailed and individualized exercise programs for firefighters: 1. Conducting a maximum treadmill test to determine VO 2peak , VT1 and RCP 2. Conducting a simulated firefighting exercise to determine physical strain with the help of three physiological intensity zones 3. Using our new developed model TSA: HR+ + time AD to characterize performance during specific firefighting simulation.