RETRACTED ARTICLE: Assessing Cardiorespiratory Fitness of Soccer Players: Is Test Specificity the Issue?–A Review

It is important that players and coaches have access to objective information on soccer player’s physical status for team selection and training purposes. Physiological tests can provide this information. Physiological testing in laboratories and field settings are very common, but both methods have been questioned because of their specificity and accuracy respectively. Currently, football players have their direct aerobic fitness assessed in laboratories using treadmills or cycle ergometers, whilst indirect measures (using estimation of aerobic performance) are performed in the field, typically comprising multiple shuttle runs back and forth over a set distance. The purpose of this review is to discuss the applied techniques and technologies used for evaluating soccer players’ health and fitness variables with a specific focus on cardiorespiratory testing. A clear distinction of the functionality and the specificity between the field tests and laboratory tests is well established in the literature. The review findings prioritize field tests over laboratory tests, not only for commodity purpose but also for motivational and specificity reasons. Moreover, the research literature suggests a combination of various tests to provide a comprehensive assessment of the players. Finally, more research needs to be conducted to develop a specific and comprehensive test model through the combination of various exercise modes for soccer players.

1. 1.

   Objective information on soccer player physical status is an important factor for coaches and managers enabling selection and training.

2. 2.

   A comprehensive approach to analyze the physiological profiles of soccer players, with a specific focus on cardiorespiratory fitness provides accurate information in order to assess the complete picture of player abilities.

3. 3.

   A combination of various tests is necessary for a comprehensive assessment of player physiological profiles.

Elite soccer match play is characterized by intermittent high intensity activity, underpinned by high levels of aerobic and anaerobic fitness in players. During a soccer match, the typical distance covered by an elite player is 10–13 km, most of which is walked or at low intensity [1]. This aerobic base is interspersed by high intensity activity including accelerations, sprinting, changes in direction, jumping, side stepping, tackling, and game-specific technical skills [2, 3]. These changes in movement patterns can only be performed providing players have sufficient muscle strength, flexibility, and agility [2, 4]. These high intensity periods are not only the most interesting moments of a soccer game but also the most decisive [5].

However, there are marked differences in physical and physiologic output during a game between individual players, related to position and playing style as well as team tactics [2, 6,7,8]. Therefore, it is important that players and coaches have access to objective data on player physical status during the season for selection and training purposes [9]. This type of data can also be used as feedback and as motivation for players [6, 7]. Tests to assess physical performance can provide this information. In recent years, as professional soccer clubs seek to gain a competitive advantage, increasing their spending for sport science services, a growing number of “soccer-specific” fitness tests have been developed and implemented across the professional game.

Due to the complex nature of soccer match play and the difficulty in isolating specific indicators of performance there is no single test capable of measuring all determinants of physical performance in soccer concurrently. Physical assessments therefore have typically focused on measuring one or two physical components of soccer match play. Physiological testing of soccer players in laboratories and in field settings are common, though both methods have been questioned relating to issues of specificity [10,11,12] and accuracy [13], respectively. Field tests have up until the last decade or so, estimated maximal aerobic capacity with accuracy error of ± 10–15% [14]. The development of portable metabolic assessment systems has enabled direct measures of gas exchange in soccer-specific tests in field environments.

The purpose of this review is to analyze the contemporary research available in physiological testing of soccer players, with the focus on cardiorespiratory testing specifically. The review considers tests carried out in all populations and includes field and laboratory tests.

Performance in soccer relies on a myriad of components, including technical, tactical, mental, and physical skills [15]. In relation to the physical aspect, training, health status, and genetic endowment are fundamental [16]. Sport scientists can, through physiological testing of performers, analyze these components and use this information to create individual profiles of participants to include strengths and weaknesses in relation to other squad players and previous tests [17]. This information is then used to inform and optimize individual training prescription with the aim of rectifying weaknesses. This should be a multifactorial approach that emphasizes sport specific demands, short- and long-term progressions, movement skills, and rest and recovery, acquired from fitness testing which forms the foundation of the physical conditioning program [18].

Balsom [13] described how individual player profiles could also provide objective utility in analyzing the effects of training interventions as well as the readiness of individuals to return to training and match play following injury. Therefore, results from fitness testing can be valuable to coaches and players as a feedback tool that evaluates individual and team responses to the training stimulus. [11]. A comprehensive list of rationale for fitness assessment is provided in Table 1. Several tests have been devised that can either be used as part of overall physiological assessment or to measure key elements of soccer-specific fitness (e.g., agility, speed, power, multiple sprint ability, endurance). Three factors must be considered when choosing a performance protocol: validity, reliability, and sensitivity [19]. Validity relates to the degree to which scores from a test, measures what they are supposed to measure [20]. Three types of validity are associated with performance testing: logical, criterion, and construct. Logical validity is upheld when the test obviously involves the performance being measured [20] but is difficult to truly assess. Criterion validity is an objective measure of validity of which there are two types––concurrent validity and predictive validity [20]. Concurrent validity indicates the protocol is correlated with the criterion measure [20], for example, correlating distance covered in a field test with distance covered during competition. Predictive validity relates to the ability of test performance to predict subsequent performance. An example of such would be to use scores on a maximal oxygen uptake test (VO_{2 max}) to predict performance in a competitive match.

Construct validity is the ability of a test to measure a hypothetical construct, i.e., performance [20]. An example of construct validity would be a test reflecting improvements in flexibility after a flexibility training program. Construct validity can also be established using the known group difference method [20]. For example, as long-distance running requires good aerobic power, a test of aerobic capacity would have construct validity if it could discriminate between groups. Indeed, it has been argued that sports performance is a construct [21]. Reliability considers the ability of the test to produce similar results over different testing times when no intervention is used [22]. Reliability can be assessed by controlled repeated measures that are analyzed using statistical methods [17]. Reliability is expressed as a correlation coefficient between 0.00 and 1.00, the closer the score to 1.00 the less error variance exists and the more the true score is assessed [20].

The technique adopted to determine the reliability correlation coefficient depends on which measures the investigator is attempting to find a correlation between. Interclass correlation (also called Pearson r) computes correlation between two variables, whereas intraclass correlation measures correlation between the same variable in repeated measures such as score in a test-retest scenario [20]. Sensitivity is the ability of the protocol to detect small but meaningful changes in performance [19], specifically the minimum percentage increase required for enhanced performance. A test with low within-subject coefficient of variation (CV) would be able to detect smaller changes between groups or over time [23]. These increases may be very small in elite athletes with CV’s between 0.3–0.4% [24].

Laboratory tests provide a controlled environment, limiting the influence of extraneous variables, so generally yield more accurate and reliable data compared to field tests in relation to isolated elements of fitness [9]. Aerobic endurance performance is dependent on VO_{2 max}, lactate thresholds, and running economy [25] and are most accurately assessed with laboratory protocols utilizing treadmills or cycle ergometers. CV of these tests is typically between 1 and 3% [26].

VO_{2 max} (maximal oxygen uptake) is the highest amount of oxygen that the body can utilize during exhaustive exercise whilst breathing air at sea level [14]. VO_{2 max} is one of the most commonly used indicators of aerobic power and metabolism [27] and is used regularly to measure aerobic performance, VO_{2 max} is considered the gold standard and is the most important measure of aerobic ability.

The primary criterion for attainment of VO_{2 max} is a plateau in VO₂ [28]. Several secondary criteria exist in the case of a plateau in VO₂ not being reached, which include a rise in respiratory exchange ratio (RER) above 1.15, blood lactate concentration above 8 mmol l⁻¹ and increase in heart rate to age-predicted maximum [14].

However, Howley [27] questioned the use of the criteria which originated in studies carried out over 50 years ago [29,30,31]. Because these studies used specific exercise modalities, subjects, and protocols, the application of these criteria to studies with different methodology and participants therefore may not be valid [27, 32]. The use of a plateau in VO₂ as criteria has been subject to criticism. Midgley [33] suggested that unless an absolute plateau in VO₂ is used, a VO₂ plateau only represents a slowing of the rate of change in VO₂––work––rate relationship and not that VO₂ reached a true plateau. In their critique of oxygen uptake criteria, Midgley [33] noted that several criteria (blood lactate, RER, heart rate) necessitate passing a threshold which indicates a maximal effort has been reached. Yet, the large variability between subjects means some participants would attain these criteria sub-maximally [34], whereas others may struggle to achieve specific criteria even with maximal effort [33]. The application of heart rate to age-predicted maximum is also misleading as an indicator of maximal effort [17].

The maximal oxygen uptake of outfield male international soccer players has been reported to range from 50 to 75 ml kg min⁻¹ [15], which supports the view that aerobic energy contributes significantly to soccer performance. Indeed, Bangsbo [6, 7] described how approximately 90% of total energy during a soccer match is derived from aerobic metabolism. Therefore, it is important to ascertain a player’s maximal aerobic capacity. When VO_{2 max} is measured in athletes, it is crucial the test protocol replicates the activity profile of the particular sport [35]. Thus, a treadmill VO_{2 max} protocol would be more valid to evaluate soccer players than a VO_{2 max} test performed on a cycle ergometer.

The average exercise intensity endured by a player during a game is 80–90% of heart rate max (HR max), close to the anaerobic threshold [6, 7, 15]. Ideally, soccer players should be able to maintain high intensity work throughout a 90-min match. Yet, studies have described elite and sub-elite soccer players’ ability to perform high intensity exercise diminishes in the second compared to the first half of games [2, 36,37,38]. Towards the end of games, a reduction in distance covered [6, 7], more low intensity than high intensity work [39], reduced blood glucose and muscle glycogen [36], and lactate concentrations [6, 7] has been reported.

Players with high VO_{2 max} levels also have high glycogen levels required for energy release to perform high intensity actions throughout a soccer match. Smaros [40] described players with higher VO_{2 max} perform the highest number of sprints and are involved more often in decisive plays during a game than those with lower values. These players also have an improved rate of recovery [2] and are better equipped to utilize fat as energy at the same relative workloads, thus are capable of glycogen “sparing” for the most intensive and decisive moments of a game [3]. So, in effect, players with higher VO_{2 max} can run at a higher intensity and greater distances before depletion of glycogen necessitates a reduction in intensity. This would suggest information of players’ VO_{2 max} is beneficial to coaches regarding team selection, individual player roles within the team and tactical decisions.

Studies have also observed how a player’s VO_{2 max} is significantly correlated with the total amount of work performed during a match [41] and distance covered [6, 7, 40]. Helgerud [42] used an intervention design to compare elite under-18 players undergoing 4 × 4––min interval training at 90–95% of maximal heart rate for 8 weeks with a control group performing normal training. Players in the experimental group increased individual VO_{2 max} 6 ml kg⁻¹ min⁻¹ as well as covering greater distance (+ 1700 m) during a game, 100% increase in sprints and 24% more ball contacts. This suggests that VO_{2 max} is sensitive to training regimens in soccer, training at 90–95% of maximal heart rate may be optimal to improve VO_{2 max}, and improvements in VO_{2 max} can directly affect match performance in soccer. Also, a relationship between average VO_{2 max} and team ranking has been shown to exist in several studies. Apor [43] demonstrated a correlation between mean VO_{2 max} and finishing position in the Hungarian First Division Championship. Mean VO_{2 max} values were 66.6, 64.3, 63.3, and 58.1 ml kg min⁻¹ for first, second, third, and fourth placed teams, respectively. Two elite teams in Norway’s top division finishing in significantly different league positions also had significantly different VO_{2 max} values pre-season [5]. Although only two teams in one league were studied which is a limiting factor, these findings suggest VO_{2 max} may differentiate between successful and unsuccessful teams, with higher ranking teams and teams in specific national leagues possessing higher VO_{2 max} [17]. VO_{2 max} values have been reported to differ significantly between playing positions in elite and non-elite players. The study of Wisloff [5] described midfielders as having significantly higher VO_{2 max} values than defenders, when expressed relative to body weight (ml kg min⁻¹). Indeed, studies have consistently reported defenders as having lower VO_{2 max} compared to other outfield players [15, 44,45,46]. Some studies have observed how VO_{2 max} undergoes seasonal variation as VO_{2 max} values improve markedly during pre-season in professional soccer players when there is typically an emphasis on aerobic conditioning [7, 42].

The evidence suggests that VO_{2 max} testing is a useful tool to evaluate soccer players. However, there are several limitations involved in using laboratory VO_{2 max} testing. Firstly, studies have demonstrated an insensitivity of VO_{2 max} tests to indicators of performance and training interventions. Bangsbo and Lindquist [47] compared performance in various exercise tests with a soccer-specific endurance test (Bangsbo test). Results indicated VO_{2 max} score during submaximal running to be insensitive to endurance capacity as measured by the Bangsbo test, in well-trained soccer players. Casajus [48] also reported the inability of VO_{2 max} to reflect training-related improvements in professional players at different times of the competitive season. In agreement with these findings, Raastad [49] also found no difference in VO_{2 max} before and after 10 weeks of training and Omega − 3 supplementation in 50 elite male players.

VO_{2 max} has also been shown to be unable to distinguish competitive level across populations and across leagues in different countries. A study comparing fitness profiles of professional and semiprofessional soccer players in England described no significant difference in VO_{2 max} between competitive level, as well as no significance between senior and junior (~ 16 years) players within the same club [50]. A professional first division Portuguese team was reported to have VO_{2 max} of 59.6 (± 7.7) ml kg min⁻¹ [45], 56.8 (± 4.8) ml kg min⁻¹ was observed in elite Saudi Arabian players [51], 59.1 (± 4.9) ml kg min⁻¹ in elite players in Hong Kong [52], and 57.8 (± 4.0) ml kg min⁻¹ in university players in England [53]. Although these values are broadly comparable and may reflect competitive level, similar aerobic capacity has been shown in professional Italian players (58.9 ± 6.1) ml kg min⁻¹ [54] and English Premier League players (59.4 ± 6.2) ml kg min⁻¹ [55]. Indeed, the study of Faina [54] reported Italian amateur players with VO_{2 max} values of 64.1 (± 7.2) ml kg min⁻¹, significantly higher than in professionals. Casajus [48] showed Spanish First division players to have average values of 66.4 (± 7.6) ml kg min⁻¹ which are among some of the higher values reported in the literature, yet higher values have been reported in German Division 3 players (69.2 ± 7.8) ml kg min⁻¹ [56, 57]. This disparity in players’ aerobic capacity across similar competitive levels in different countries and the higher values reported in players of lower competitive levels––despite time-motion analyses reporting elite players covering more distance at a higher intensity than non-elite [39, 58], suggest VO_{2 max} is a poor discriminant between average and elite soccer players. Of course, the difference in methodologies used to establish VO_{2 max} is likely an influencing factor in results; however, this in itself illustrates the challenges involved in comparison of VO_{2 max} in soccer players.

The inability to detect indicators of soccer performance has been hypothesized to result from fundamental differences between activity patterns in soccer and VO_{2 max} tests and, accordingly, the underlying physiology [17]. Of course, the nature of VO_{2 max} testing (straight treadmill running) does not replicate the activity pattern described in soccer match play, as linear running is not a soccer-specific activity [15]. A test is more reliable and effective when it is specific to the exercise patterns of that sport [59] and validity of the test depends on its ability to elicit similar physiological responses to the actual performance [19]. Indeed, the intermittent activity profile inherent in soccer elicits increased physiological strain [60] and often requires performance at intensities above those observed in VO_{2 max} testing [6, 7]. Studies adopting time-motion analyses underscored how the ability to perform repeated high intensity exercise (intermittently) is fundamental in elite soccer [39]. In support, Aziz [61] observed how measured VO_{2 max} may not be suitable to characterize soccer players’ intermittent endurance capacity, as performance (distance covered/speed) in a soccer-specific intermittent test had no relationship with either VO_{2 max} measured in the same test, as well as a standardized treadmill test nor with VO_{2 max} measured during a multi-stage shuttle run test. VO_{2 max} appears to lack sensitivity, with respect to cardiorespiratory testing of soccer players, therefore may be unable to detect small but worthwhile changes in performance over a season.

Aside from physiological and specificity limitations, more practical and methodological limitations exist. For many soccer teams to implement VO_{2 max} testing, it would require expensive equipment and space, as well as trained personnel that may not be available within their staff. Testing every player would involve multiple time-consuming visits to laboratory facilities at local universities or clinics for many clubs [9, 61, 62]. Also, there are motivational factors that should not be overlooked. High motivation levels and a willingness to push oneself are prerequisites in obtaining true maximal oxygen uptake [31]. When players are not motivated to perform maximally in a fitness test, it is doubtful if the test will provide a valid measure of the performance variable it is designed to measure [63]. Due to the difficulties associated regarding criteria for attainment of VO_{2 max}, care should be taken with terminology and methodology adopted to assess VO_{2 max}. A familiarization session is warranted, robust warm up procedures, and suitable protocols and testing criteria [33].

Laboratory VO_{2 max} testing is not a conclusive measure of physical performance in soccer players and other determinants of performance should be examined. Ziogas [64] argued that velocity at lactate threshold and running economy must also be considered alongside VO_{2 max} when testing the aerobic capacity of elite players. Aerobic capacity is only one part of a complex structure to define overall soccer performance [5]. Indeed, although soccer match play is predominantly aerobic, the most decisive actions, such as the ability to sprint fast to beat an opponent or to jump high, require anaerobic metabolism [6, 7].

Whilst VO_{2 max} measures the maximal ability to consume oxygen during exhaustive exercise, the intensity of exercise that elicits VO_{2 max} cannot be sustained for long [65]. The level at which intense exercise can be prolonged has been referred to as the “anaerobic threshold” which is characterized by the highest exercise intensity, HR, or VO₂ where lactate production and removal is equal [15]. Measurement of the anaerobic threshold using blood lactate has typically focused on the initial rise in lactate above baseline lactate threshold or the 4 mmol L⁻¹ point which is known as OBLA (onset of blood lactate accumulation). Anaerobic threshold in adult male soccer players has been reported to be between 76.6 and 90.3% of HR max [15].

Lactate threshold refers to the VO₂ above which blood lactate exceeds resting values and lactate production exceeds lactate removal during incremental exercise, marking the transition between moderate and heavy exercise [66, 67]. This point represents the onset of lactate accumulation in the blood and may indicate the transition from predominantly aerobic metabolism to anaerobic predominance [68]. Lactate threshold is a more useful indicator of aerobic performance in endurance sports than VO_{2 max} [69, 70]. This suggests that players with a higher lactate threshold would be able to cover more distance at a higher intensity during a game without lactate accumulation than a player who is less aerobically trained [71]. Lactate threshold and OBLA are usually determined during a graded treadmill protocol, and sub-maximal lactate assessment (fixed level between 2 and 4 mmol L⁻¹) can be used to identify changes in aerobic performance in soccer players over time [72]. Lactate threshold has been shown to be sensitive to changes in training regimens in soccer players. Helgerud [42] reported improved lactate threshold from 47.8 (± 5.3) to 55.4 (± 4.1) ml kg⁻¹ min⁻¹ after 8 weeks or interval training in elite under-18 players. Edwards [71] investigated VO_{2 max} and anaerobic thresholds as determinants of training status in elite soccer players, testing once during pre-season and again on completion of the competitive season. Lactate threshold was significantly improved from the first to second test (81 vs 86% VO_{2 max}, respectively) whereas VO_{2 max} showed no significant difference between tests. Lactate threshold has also been found to be sensitive to positional variation––Bangsbo [6, 7] demonstrated the lactate thresholds of elite midfielders and fullbacks to be higher than those of central defenders and goalkeepers.

Lactate threshold can also be used to inform players on the intensity of training and by measuring heart rate during such lactate threshold tests, training intensities can be prescribed in accordance with the aims of training sessions [9]. The protocol used to determine lactate threshold is important because an initial exercise intensity that is too high will elicit an immediate rise in blood lactate which prevents the appearance of a suitable curve and identification of an inflection point [73]. However, several tests may be required in this instance for accurate identification of lactate threshold [17]. Indeed, differences in the activity patterns of soccer-specific exercise and steady-state exercise are a limitation in the use of lactate threshold as a performance measure. The intermittent nature of soccer means players often perform at levels exceeding the lactate threshold intensity. During intermittent exercise, physiological response is dependent on the type of activity and/or exercise protocol being performed [53, 74]. Depending on the protocol used, higher or lower levels of physiological stress can be elicited than in steady-state exercise, resulting in lactate concentrations lower than or above the lactate threshold [17]. Exhaustive intermittent exercise elicits higher blood lactate concentrations than observed in continuous type exercise [75], although recovery periods in intermittent exercise enable lactate removal, dependent on lactate concentration, aerobic capacity, and activity during the recovery phase [15]. Relationships between lactate threshold and other variables (VO₂, heart rate) changes during intermittent activity in comparison to steady-state exercise [44]. This disparity between physiological variables during intermittent exercise will limit the application of lactate threshold tests during intermittent, soccer-specific exercise [76].

The additional physiological demand of performing game skills, above the energetic cost of locomotion, further limit the use of the lactate threshold in soccer [15]. Furthermore, studies have shown lactate threshold to be insensitive to performance measures in soccer. Bangsbo and Lindqvist [47] described no significant relationship between lactate threshold and performance during match play and during a soccer-specific field test in professional players. In another study measuring endurance performance changes over a season in young players, mean running velocity at lactate threshold and OBLA improved significantly, yet no change in lactate threshold (relative to HR_max) was observed between the start of pre-season and early stages of the competitive season [72]. This is typically the period when aerobic conditioning is a focus and when the largest increases in fitness are observed. The lack of change in lactate threshold suggests the improvements in aerobic performance were due to changes in VO_{2 max} or running economy. Lactate threshold appears to change concomitantly with VO_{2 max} [77]. Indeed, a higher level of VO_{2 max} results in lower lactate concentration at a given work intensity. Evidence of the utility of lactate threshold assessment regarding indicators of soccer match performance is lacking. However, it appears to be useful in monitoring change in endurance capacity elicited from training interventions [78]. Although measurement of lactate threshold may be useful in identifying endurance training changes in soccer players, additional assessment of running economy and VO_{2 max} seems to provide more useful information as to the changes in endurance performance related to soccer [72].

Running economy is the ratio between work intensity and oxygen consumption (VO₂) [79]. It is normally expressed as VO₂ at a standardized workload or VO₂ per meter during running [79]. Running velocity at the lactate threshold or at VO_{2 max} is influenced by running economy. Trained runners have been shown to possess greater running economy than recreational runners [80]. Yet, running economy has been shown to vary widely in highly trained subjects with similar VO_{2 max} [79, 81]. Differences in running economy of up to 20% have been reported in elite endurance runners [82], and running economy has shown to be correlated with aerobic ability [81]. Adding support to the inclusion of running economy in fitness testing, Hoff [42] reported that improvements in running economy due to strength training elicited changes in aerobic performance without accompanying changes in VO_{2 max} or lactate threshold. It has been estimated that a 5% improvement in running economy could increase match distance by approximately 1000 m [77]. When improvement of maximal aerobic performance is desired, VO_{2 max}, lactate threshold and running economy must be considered when implementing training regimes [83].

Laboratory VO_2max testing is considered the gold standard to measure maximal aerobic capacity and is a valid tool in the assessment of soccer players. However, the activity elicited during a VO_{2 max} test does not reflect the high intensity intermittent nature of soccer performance. Therefore, VO_{2 max} testing is unable to isolate specific components of fitness and indicators of match performance, but can provide information related to the general physical capacity of a player, as well as differentiate between populations. These factors, as well as the impracticality of adopting VO_{2 max} testing in soccer teams, suggests sport scientists use maximal oxygen uptake assessment during periods of the season when large changes in fitness are expected. Measurement of lactate threshold can be useful during steady-state exercise to show changes in aerobic performance elicited from training interventions. However, the disparity associated with blood lactate accumulation in steady-state and intermittent type activity means its usefulness in soccer is limited. Moreover, as lactate threshold is not sensitive enough to soccer-specific indicators of match performance, it is best used as a determinant of general fitness. Running economy should be included as a measure of aerobic endurance as it influences other determinants of fitness and has been shown to be more sensitive to changes in performance than traditional measures. A comprehensive assessment of aerobic capacity may require that all three measures are considered.

Owing to limitations associated with laboratory testing, several field tests have been devised as practical alternatives to assess the endurance capacity in soccer players. Fitness tests carried out in the field improve the specificity of the test, yet provide less accurate measurements when compared to laboratory tests [11]. Some examples of soccer field tests include the Loughborough Intermittent Shuttle Test [84], 20 m multi-stage shuttle run test [85], and the Yo-Yo tests [6, 7, 86]. Because these tests include soccer-specific activity, they may be more valid than laboratory assessments [11, 13] by better measuring a player’s ability to perform in a soccer situation. Such field tests have included either continuous or intermittent exercise, with and without recovery phases and/or soccer-specific actions. Performance in soccer-specific field tests have been reported to show correlation with VO_{2 max} [87, 88] as well as possessing the ability to differentiate between playing positions [89] and player ability level [39].

The LIST [84] was designed to simulate activity patterns in a soccer match. It involves running between two lines 20 m apart at various speeds related to individual estimated VO_{2 max} dictated by audio signals from a microcomputer. The test comprises two parts––part A and part B. Part A is a set pattern of a 15-min intermittent high intensity running, and part B is an open-ended intermittent shuttle running designed to exhaust the participant within 10 min. Altogether, five 15 min blocks of activity each separated by 3 min recovery constitute part A. Part A involves 3 × 20 m shuttles at walking pace, 1 × 20 m at maximal running speed, 4 s recovery, 3 × 20 m at 55% of individual VO_{2 max}, and 3 × 20 m at 95% individual VO_{2 max}. Test measurements include sprint time, total distance covered, blood lactate, heart rate, and RPE. Examining the test-retest reliability of LIST, Nicholas [84] reported no significant difference in physiological or metabolic variables between tests. Total distance covered (12.4 km) and turns completed (55–60) during LIST was similar to those calculated in competitive soccer matches [1, 15]. Siegler [90] described LIST to elicit similar physiological responses (HR_mean, HR_max, VO_2max) in non-elite players to those observed in elite players as recorded via time-motion analyses. Although LIST simulates the physiology of soccer with regard to distance covered, it does not include many football specific actions such as backward running, jumps, and ball activity, so it is not a valid soccer-specific test. Indeed, a measure of a skill is required in such a protocol for it to be considered a valid test for soccer [19]. Also, because the test is exhaustive, it may be impractical during a competitive season when recovery and preparation are the key.

The multi-stage shuttle run test (MST) was developed by Leger and Lambert [91] and later modified [88]. MST incorporates running back and forth between two lines separated by 20 m, with increases in speed every minute regulated by audio bleeps from a tape recorder. The participant must reach the end line in the shuttle by each bleep, if the participant fails to reach the end line once he/she is warned, and if in two successive shuttles they cannot reach the line the test is terminated, with the total number of shuttles completed typically used as the test score. The test begins at around 8 km/h [88]. VO_{2 max} is estimated based on the shuttle reached at the end of the test using the regression equation VO_{2 max} = (5.857 × speed on the last stage)––19.458 [91]. MST has been shown to significantly correlate with VO_{2 max} directly measured on a treadmill [61, 88] (r = 0.92 and 0.92, respectively). Contrastingly, in non-elite players, Siegler [90] reported significant difference between VO_{2 max} estimated via MST and treadmill VO_{2 max}. Also, the indirect measurement of VO_{2 max} has accuracy of ± 10–15% [26], so it should be viewed with caution. As the shuttle score is discontinuous and cannot be used for statistical analysis, MST score should be expressed as distance covered [15].

MST has been shown to correlate with several soccer-specific match activities in youth players (total distance, distance in speed zones/match activity categories) as measured via GPS technology [92]. However, the study did not measure physiological variables. The MST has been used extensively to assess soccer players in England [46, 50, 55]. Yet, MST has been unable to identify training interventions––as Odetoyinbo and Ramsbottom [93] concluded after finding no significant improvement in MST after 8 weeks of high intensity training in soccer players. Also, studies have reported MST as unable to differentiate between populations and competition level. Edwards [94] reported no difference in VO_{2 max} between academy scholars and recreational players completing the MST. Similarly, Dunbar and Power [50] found no difference between professional and semi––professional players, or between senior and junior (~ 16 years) players within the same club for estimated VO₂ max assessed via MST. However, the accuracy of the results can be questioned as goalkeepers’ results was included in the data, which likely affected the outcome as there exist substantial differences in maximal oxygen uptake between goalkeepers and outfield players [15, 45, 95, 96].

As discussed elsewhere in this review, soccer players require the ability to perform and recover from repeated intense exercise [87]. As a result of a need for a test to measure this physiological attribute, the Yo-Yo tests were developed. The Yo-Yo tests incorporate 2 × 20 m shuttles interspersed with 10 s of active recovery with speed increments regulated by audio signals from a CD player. Participants continue until they can no longer maintain the speed imposed by audio signals, and the distance covered at that moment is the test result [10, 12]. Yo-Yo IR1 begins at lower speed (10 km/h) than Yo-Yo IR2 (13 km/h) with slighter speed increments. The Yo-Yo IR1 assesses an individual’s capacity to repeatedly perform aerobic exercise leading to maximal activation of the aerobic system whilst the Yo-Yo IR2 focuses on the ability to recover from repeated high intensity exercise with a large anaerobic contribution alongside a significant aerobic component [10]. In untrained participants, the Yo-Yo IR1 elicits this physiologic response. In trained participants, the Yo-Yo IR1 typically lasts 15–20 min whilst Yo-Yo IR2 lasts 5–15 min. Accordingly, Yo-Yo IR1 [6, 7] has been reported to be more suitable for recreational players whereas the Yo-Yo IR2 is more applicable to trained subjects [10]. Both Yo-Yo recovery tests allow for quick determination of maximal heart rate in participants, with faster increases in heart rate observed in Yo-Yo IR2 [10]. Increased blood lactate and muscle lactate accumulation is elicited in Yo-Yo IR2, as well as lower creatine phosphate levels on completion of the test and a higher rate of average muscle glycogen utilization during Yo-Yo IR2 [36, 87].

The Yo-Yo IR1 has been reported to have high correlation with several physical match activities in youth [97, 98] and adults [10, 87]. Yo-Yo IR1 has also been shown to be valid and reproducible and differentiate between ability levels in a range of sports. Krustrup [87] investigated the validity and reliability of Yo-Yo IR1 in elite male Danish soccer players. Authors reported a significant relationship between Yo-Yo IR1 test performance and the amount of high intensity exercise performed during soccer match play, and no difference in performance between a test retest of Yo-Yo IR1 within a week of each other (CV = 4.9%). In young soccer players (11–17 years), Deprez [99] showed the Yo-Yo IR1 to be adequately reliable in young players (U13, U15) and highly reliable in U-17-year-olds performing test-retest Yo-Yo IR1 (excellent interclass correlation between 0.82 and 0.94). Another finding was that Yo-Yo IR1 had a good ability to differentiate young elite and non-elite players. However, the inclusion of a small number of non-elite players (N = 20) for comparison with a large cohort of elite players (N = 150) is a limitation of the study, and it is to be expected that elite players with higher training frequency, and ability level, will perform better than sub-elite players on any given soccer-specific field test. In a study with greater equality between participant groups (18 top class vs 24 moderate level professional players), Mohr [39] also showed the Yo-Yo IR1 to differentiate player ability, as top class players performed better in the test than moderate level players (2.26 ± 0.08 vs 2.04 ± 0.06 km). Furthermore, the Yo-Yo IR1 was shown to be sensitive to playing position, as midfielders and fullbacks covered a greater distance than defenders and attackers (2.23 ± 0.10 and 2.21 ± 0.04 vs 1.99 ± 0.11 and 1.91 ± 0.12 km, respectively). Metaxas [100] reported the Yo-Yo IR1 to underestimate VO2 max when compared to treadmill protocols in elite U-20 soccer players. The authors’ suggested that the turns and recovery periods in the Yo-Yo IR1 may have suspended attainment of VO_{2 max}. However, Yo-Yo IR1 appears to be more sensitive to changes in soccer performance than VO_{2 max}. Bangsbo [1] observed Yo-Yo IR1 test performance during different periods of the season in 10 elite soccer players. Authors described how performance on Yo-Yo IR1 improved by 31% during pre-season phase, whereas only a minor change in VO_{2 max} occurred. This is in line with data recorded by Krustrup [87] where seasonal improvements in Yo-Yo IR1 (25%) were observed in elite soccer players with relatively minor changes in VO_{2 max} (7%). Yo-Yo IR1 performance has also been shown to be related to other validated field tests. Castagna demonstrated a strong correlation (r = 0.89) between Yo-Yo IR1 performance and MST in youth male soccer players. This finding suggests that these two tests could be used interchangeably in assessing soccer-specific fitness in young male players.

Yo-Yo IR2 has been studied extensively in a range of sports and has been shown to be sensitive to soccer playing position [89], performance in different seasonal periods, and competitive level [87] and associated to treadmill VO_{2 max}. Krustrup [87] measured 13 normally trained males in repeated Yo-Yo IR2 tests, incremental maximal treadmill test, and various sprint tests. Additionally, 119 Scandinavian professional soccer players performed Yo-Yo IR2 on two to four occasions and found no difference in Yo-Yo IR2 performances carried out within 1 week of each other (CV = 9.6%). The test was correlated to VO_{2 max} and treadmill performance. Authors described how the soccer players improved their Yo-Yo IR2 performance by 27% in the first 4 weeks of pre-season preparation and 42% improvement during the total 8-week pre-season period. This demonstrates the test’s sensitivity in assessing a player’s capacity for repeated high intensity exercise. Additionally, international elite players performed better than moderate elite players (1059 ± 35 vs 771 ± 26 m), whilst central defenders, fullbacks, and midfielders performed better than goalkeepers and attackers (p < 0.05). Carling [101] supports this finding, after analyzing repeated sprint ability and high intensity activity in French professional players reported midfielders complete more high intensity actions with short recovery whilst full-backs performed the most repeated high intensity bouts. The sensitivity of the Yo-Yo IR2 to differentiate between competitive level in soccer players is greater than for the Yo-Yo IR1 [10]. Indeed, international elite level players have been shown to carry out 28% more high intensity running and 58% more sprinting than moderate elite level players during competitive games [39]. It is apparent that high intensity exercise capacity differentiates top level players with those of a lower standard. Yo-Yo IR2 and Yo-Yo IR1 have both shown sensitivity in differentiating players at different levels.

Aziz [61] compared performances in Yo-Yo IR2 with MST and treadmill VO_{2 max} in young soccer players of a national U-18 squad. No significant differences were found in physiological variables (HR max, VO₂, VCO₂, VE, BLa) except RER between the three tests. As treadmill VO_{2 max} is deemed “gold standard,” this demonstrates the validity of Yo-Yo IR2 and MST as field-based tests of aerobic capacity in young soccer players. A significant difference was observed; however, between field tests for distance covered and peak speed and a common variance of just 40% were recorded. This would indicate performance in either test is not interchangeable. This is probably due to protocol differences and the concomitant disparity in physiologic response––the Yo-Yo IR2 begins and finishes at higher speeds and includes 5 s rest periods that provide recovery and a true intermittent activity. Indeed, performance in MST showed strong significant correlation with VO_{2 max} measured during MST and in Yo-Yo and almost correlated with treadmill VO_{2 max}. The largely continuous activity profile of MST with speed increments every minute is similar to treadmill VO_{2 max} protocols adopted (gradient and/or speed increase every minute) which may explain this association. Indeed, other studies comparing continuous and intermittent type field tests have reported marked differences in metabolic and oxygenation response [102] as well as significant differences in VO_{2 max} between intermittent field tests and continuous treadmill protocols [100]. The Yo-Yo tests in this regard provide activity more closely associated to soccer match play which is reflected in similarities in physiological response.

Although these findings suggest the Yo-Yo intermittent recovery tests as effective, and valid measures of soccer-specific fitness, such studies are not exempt from limitations which influence how they are interpreted. A measure of skill is required in a soccer tests to make it a valid test [19, 103]; however, the Yo-Yo tests do not include any measure of skill nor any activity that could be deemed specific to soccer. Also, the turns performed at the end of each shuttle in the Yo-Yo tests may put pressure on the lower limbs’ joint musculature when carried out repeatedly.

The field tests described thus far, although possessing varying degrees of reliability and validity, do not reflect the skills required in soccer match play [103]. They typically incorporate running in a frontal direction, excluding the differing array of movements involved within the sport of football [1, 10, 104]. Furthermore, most research in field testing in soccer has been carried out without a ball. As the literature agrees that testing should incorporate soccer-specific activity, the most soccer-specific activity would include ball actions as part of the test. Chamari [105] found the presence of a ball to be a critical motivational point to assess aerobic performance in soccer players. Perhaps understandably, training solely by running may cause motivational issues in soccer players [15, 41]. Players performing maximal tests in field and lab settings felt more motivated in field testing with a ball [106]. Indeed, it has been advised in professional youth players that a ball be incorporated to increase the specificity and data variation between subjects in agility tests [107].

Another concern in the development of soccer-specific field tests is that soccer is played on a flat surface, and evidence suggests subjects solely running on flat surfaces may not be able to attain exercise intensities close to VO_{2 max} [14]. Treadmills are typically inclined to increase musculature activation enabling true VO_{2 max} to be reached by the participant [14]. Interval training has been performed as uphill running to reach exercise intensities close to VO_{2 max} [42]. However, running with the ball in soccer has been shown to increase energy expenditure by 8% over plain running [108] which could replicate the extra load imposed by an inclined treadmill. Hoff [41] developed the Hoff dribbling test as well as small-sided games in this regard, to substitute the uphill running component with ball dribbling, changes of direction, and backward running, for soccer-specific interval training. The soccer-specific field test incorporated ball dribbling with accelerations, decelerations, jumping, and changes of direction around a specially designed track (Fig. 1). Six male elite players were assessed with Metamax II measuring gas exchange. Players were also assessed during laboratory VO_{2 max} and 5-a-side small group play and compared across conditions.

Soccer-specific dribbling track (Hoff test) to measure VO_{2 max} as implemented by Hoff [41]

Full size image

Results showed the dribbling test and small group play to be as effective as normal running in eliciting work intensities of 90–95% HR max, which has been described as the optimal range to improve VO_{2 max} [41, 83]. However, to achieve these intensities, a high level of organization was required––with constructive instructions and active encouragement. Authors conceded that the small sample of players was used, and because they were volunteers for an endurance experiment, the findings cannot be generalized to wider groups of soccer players. Also, one person replaced cones and hurdles that fell, indicating that predictably, the dribbled ball would be uncontrolled at times by participants undertaking the circuit. Indeed, ball handling skills influence results in fitness tests [107, 109, 110]. Therefore, it is likely the ball dribbling component would interfere with breathing frequency which would in turn affect minute ventilation (VE) and VO_{2 max} attainment. Indeed, all respiratory variables were significantly different in the dribbling test compared to VO_{2 max}.

The study of Castagna [92] reported the Hoff test to have only limited association with match performance and no significant correlation with other validated field tests (MSFT, Yo-Yo IR1) in elite youth soccer players, as well as low relative (ICC = 0.68) and absolute (CV = 18.2) reliability. The labor time required to set up the course, and a limited number of players testable at one time are limitations of the test. However, the Hoff test has been found to correlate well with VO_{2 max}. Kemi [106], using the same dribbling protocol, described by Hoff [41] reported how the test elicited values of VO_{2 max}, max HR, maximal breathing frequency, RER, and oxygen pulse that were not different from those recorded in traditional VO_{2 max} treadmill running in elite male players. Ball dribbling, jumping, accelerations, decelerations, backward running, and turns increased oxygen demand comparably to uphill treadmill running in a laboratory. The only variable that was significantly different between conditions was VE. The cause of the discrepancy in VE may be due to different methods of gas analysis between conditions, as treadmill VO_{2 max} was measured via a stationary metabolic system and the field test used a portable system (Metamax II). Previous studies have also demonstrated differences in VE measurement between portable and stationary gas analysis systems [111]. Authors speculated that working to exhaustion, and thus maximal ventilation, may be more difficult when concomitantly performing technical soccer skills. It was also postulated that differences between mouthpiece (stationary system) and facemask (Metamax II) caused the discrepancy.

However, breathing frequency was similar between tests and Metamax II has previously been validated [112]. Chamari [105] also used a slightly modified version of the Hoff dribbling test [113] to measure aerobic performance and determine correlation with laboratory VO_{2 max} before and after 8 weeks of endurance-centered soccer training in elite youth players. Strong correlations between VO_{2 max} and Hoff test performance (r = 0.68) was reported, as well as the ability of the test to reflect improvements in VO_{2 max}, although authors also encountered difficulty reaching 90–95% HR_max during the dribbling test because of ball loss, resulting in unaccounted periods of rest which allowed activity intensity to decrease. Evidently, the inclusion of soccer-specific technical skills in physiological testing can present obstacles to attainment of maximal values. The relationship between VO_{2 max} and test performance in the study of Chamari [105], although similar, was not strong enough to predict VO_{2 max} from test performance. The ability to reliably predict VO_{2 max} in soccer field tests would be beneficial as during a competitive season, the implementation of maximal tests is discouraged. Also, exhaustive testing induces psychological stress that may vary and influence performance from day to day [10].

These findings suggest that high intensity, soccer-specific testing and training including a ball can induce effective physiological responses in the HR max range to improve VO_{2 max}, deliver a similarly maximal protocol as VO_{2 max}, whilst providing a more motivating exercise for players than straightforward running. However, the issues encountered regarding measurement of respiratory variables and performance of technical skills would suggest simplification or elimination of these from such tests. Removal of a ball dribbling element would consequently reduce load and impact on achievement of VO_{2 max}. Soccer field tests should include changes of direction or accelerations to increase work load and compensate for an inclined treadmill [41]. Many physical fitness tests in soccer are fundamentally continuous exercise, which has limited relevance to ball sports such as soccer [87]. However, Chamari [105] used a soccer-specific intermittent field test without any ball component [47] in male soccer players. Instead the circuit included slalom running, multi directional running, 15 s high intensity periods as well as 10 s low intensity jogs (Fig. 2). Results had no correlation with laboratory VO_{2 max}. An average running intensity of 95% HR_max was elicited in the test which normally correlates with VO_{2 max} when running continuously [26]. Authors suggested the intermittent component of the Bangsbo test resulted in heart rates that were not a true picture of the exercise intensity as typically 1 min of continuous running is required to achieve 95% HR max [42]. The brief high intensity periods may therefore induce anaerobic intensity above that observed in VO_{2 max} testing.

The spatial lay out of the “Foote-Val”

Full size image

Paul and Nassis [98], in their recent review of field tests in youth soccer, reported intermittent-based tests are most appropriate to measure soccer-specific endurance. However, intermittent activity may overestimate VO₂ due to heart rate differences compared to continuous exercise [41]. Indeed Chamari [105] reported how HR_max attained during the high intensity periods of the intermittent Bangsbo test may reflect a response to anaerobic exercise that exceeds VO_{2 max}. Metaxas [100] reported higher VO_{2 max} values in intermittent protocols compared to continuous protocols carried out in either the field or laboratory. These higher values of VO_{2 max} during intermittent protocols have been described elsewhere in the literature [34, 47, 114]. The fact that soccer-specific protocols with intermittent periods elicit higher exercise intensities than VO_{2 max} reflects a weakness in the use of VO_{2 max} in the measurement of a soccer players’ fitness. This issue highlights a requirement for specific testing protocols, chiefly the ability to assess physiological demands within a specific sports condition, i.e., a test incorporating football exercises, is essential to be able to directly assess and structure team/individual training and progression [104]. Therefore, field tests of soccer players must reflect the match activity of soccer––intermittent high intensity activity, with recovery bouts, low intensity periods, and football-specific actions.

Seemingly, field tests used in soccer have typically used comparison and association with treadmill VO_{2 max} as validation criterion. As discussed previously, VO_{2 max} has several limitations regarding sensitivity to soccer. Relationships made between field tests and VO_{2 max} may be of limited use, as VO_{2 max} is a poor marker of physical performance in ball sports such as soccer [1]. Indeed, VO_{2 max} per se, assessed either in soccer-specific field tests or in the laboratory, is not capable of characterizing soccer players’ endurance performance [47, 61]. Field tests have typically attempted to replicate activity patterns during a soccer match and to achieve the highest possible correlation between test performance and endurance performance during a soccer match. Despite their relationship with physiological variables of aerobic fitness, limited evidence exists regarding their relationship with match activity or direct validity [92]. Direct validity is regarded as a precondition of any sport-specific field test [97].

Field tests that have included soccer skills have often used closed skills such as slalom dribbling, which has a limited applicability to soccer match play. Soccer requires performance of multiple skills in a variable environment with a marked perceptual-cognitive component [75]. Indeed, cognitive skill is vital in team sports [115]. In this regard, Bullock [75] developed a novel reactive agility motor skills test (RMST) incorporating physical, technical, and cognitive skills in one protocol and tested amateur male players before and after soccer-specific exercise. Reactive agility enables understanding of decision making, perceptual ability, and physical ability (speed, power, balance) and together with a technical skill component provides an integrative approach to soccer skill testing. Measures of short passing accuracy and trapping the ball comprised the technical element. RMST showed good test-retest reliability with CV between 2.3 and 3.5% for all physical variables, but less reliable for technical and overall performance indices (CV = 9.2–10.7%). After performing 45 min LIST, subjects’ sprint performance and reactive agility time decreased non-significantly, whereas short passing time/accuracy as well as performance index (passing score/test time) improved. This suggests either a learning effect or improved performance elicited from exercise. Other studies have also reported moderate exercise to improve subsequent soccer skill accuracy [116, 117].

These integrative tests of multifactorial components required for successful soccer performance should be the focus of future research, as many soccer field tests include just one outcome measure such as aerobic capacity, which is one of many desired attributes in soccer. However, it is important to remember that no field test will determine performance in a soccer match, as it is difficult to isolate individual physical parameters, because the game demands are so complex [9]. A comprehensive approach comprising physiological, psychological, anthropometric, and skill-related assessment is required [98].

Foote-Val is an incremental, intermittent test based on the spatial organization of Léger’s “20-m shuttle run” test [118] to include direction changes (180°). The purpose of this test is to determine a global index of football players, providing a clear idea of their level, including their physical and technical skills. Notation considers players’ aerobic power and technical capacities in real football conditions (MASS). This test allows VO_{2 max} to be measured in specific conditions and will be influenced by many factors such as running economy, muscular abilities, or technical skills with the ball. It differentiates football players as per their level (Ziogas et al., [119]).

To comply with the distance of 20 m, as in Léger’s 20-m shuttle run test, an optimal trajectory is drawn in the slalom using plastic strips (0.5-m long and 1-cm thick) located 0.5 m from the center of the poles, which does not interfere with the player or the ball. The route of this optimal trajectory aims to help the player to respect the tempo beeps. A marker is placed at 0.4 m from the first pole so that the player would start his slalom at this point and not before. The last step completed by the player provides the Maximal Aerobic Speed Specific (MASS).

However, the test might not be accurate because the ball movement after hitting the plastic board is not controlled. Therefore, the pace, direction, and the distance of running with the ball will vary from player to player. Moreover, the ball control of all the players may not be equal, and the risk of losing the ball during the drill is high between different players with different skill levels. Thus, it may not be the most accurate test to examine players’ aerobic potential.

The maximum progressive laboratory test is carried out on a motorized treadmill, starting at 8 km·h⁻¹ with speed increments of 1 km·h⁻¹ every minute. Immediately after the athlete reached voluntary exhaustion, he should undergo an active recovery lasting 3 min at a speed of 7 km·h⁻¹. Throughout each test, the treadmill was set with a slope of 1%. On the other hand, the field test consisted of a progressive and maximal running test with a total distance of 80 m, in the shape of a square of 20 m. The execution speed of the test was determined by sound beeps like those of the Yo-Yo test of endurance level 2 [86], with initial velocity of 11.5 km/h and load increments of 0.5 km/h every minute, admitting that the Yo-Yo endurance test level 2 aims to estimate the VO_{2 max} in well-trained players to shorten the evaluation time [120]. In each corner of the square, there is a cone, which should be circumvented by the athlete at the time of each beep. The test is always performed counterclockwise and stopped if the athlete did not reach the vertices for two consecutive times in trying to get around the cone at the time of the sound beep. The distance, maximum speed, and total time of each subject are recorded for evaluation.

This test can be used to estimate VO_{2 max}. The field test protocol may be preferred over the treadmill protocol, as the player is exposed to real game conditions since the test is performed in the field, wearing soccer cleats as opposed to the laboratory conditions. Moreover, the field test protocol is easily implemented and, therefore, useful in soccer training planning [121]. Comparing the results with other studies [121, 122], it was observed that data collected in other countries had values of height and weight similar to this study. However, some of the anthropometric characteristics of teams from different countries and leagues showed a wide range of results, especially in body weight [27]. Anthropometric studies of soccer players show that body weight and height are important variables to consider when assessing performance [27]. The values found in this study for distance and speed in the field were superior (1773.33 ± 334.49 m and 15.10 ± 0.64 km·h⁻¹) to those of Castagna [123], with distance values of 1.331 ± 291 m and speed values of 14.15 ± 0.65 km·h⁻¹ for the Yo-Yo endurance level 2. The differences found in our study for distance and speed of the field test and treadmill might have occurred because the initial speed of the field test (11.5 km/h) is greater than the initial speed of the treadmill test (8 km·h⁻¹). Thus, the athlete remained a shorter period in the field test. The high level of blood lactate found in the athletes after the field test (10.0 ± 2.14 mmol/l) is also a criterion for the performance of VO_{2 max} and it shows that the use of anaerobic energy production during the maximal exercise effort [124]. The values for VO_{2 max} (48.55 mL·kg⁻¹·min⁻¹ and 50.19 mL·kg⁻¹·min⁻¹ for field test and treadmill, respectively) were lower than values reported in other studies [125,126,127,128]. However, in the present study, the values for VO_{2 max} on a treadmill and VO_{2 max} in the field test did not differ statistically (p < 0.077). So, it can be concluded that the proposed field test is statistically similar to the test performed on the treadmill. The correlation for VO_{2 max} in the field and treadmill in this study is high (r = 0.748, p < 0.000) and statistically significant. In other studies, [129, 130], the correlations between the field tests and the treadmill tests were inferior to ours. The results of the present study also agree with other researchers that support the idea that a portable telemetric ergo spirometer is a reliable method for determining the aerobic power of a soccer athlete in the field [15, 57, 111, 131]. It seems that the “proposed field test” can effectively contribute in creating the best training plan and, therefore, lead to a higher level of sports performance in modern soccer. The formulas found to indirectly determine the values of aerobic power show that the field test proposed in this study allows the subject to reach values of maximum aerobic power essentially the same as when determined by direct spirometry (Fig. 3).

Proposed Field Test (20 × 20 m)

Full size image

As per the results in this study, it is possible to establish two equations to estimate VO_{2 max} with a field test, one through the maximum speed reached, and another by the distance covered. This finding is an excellent outcome, given the high cost of ergo spirometry equipment, the time that is necessary to train the staff to use it, and the time-consuming ergo spirometry tests in the laboratory. This field test can be adopted by coaches and applied in trained soccer athletes, helping to establish the maximum aerobic power of athletes in Juvenile and Junior categories with lower costs and time saved that can be used in skill training. Another important factor in the field test is the ecological validity of the test since the athlete performs in conditions that are more like those of a real match (i.e., a field test in the grass and wearing soccer cleats). Finally, the test should be considered as an easy and useful tool for coaches and trainers for assessing the athlete’s cardiorespiratory capacity before, during, and after a competitive season [120].

Two progressive maximum tests, Carminatti’s test (T-CAR) and the Vameval test (T-VAM) are used. T-CAR is an intermittent incremental test that is performed as shuttle runs. On the other hand, T-VAM is a continuous incremental test performed on an athletic track. T-CAR includes incremental intermittent shuttle runs performed in a lane set at progressive distances apart [109]. The protocols start at a speed of 9 km/h in a running base of 15 m, which is increased by 1 m at every 90 s stage. Each distance stage (i.e., from 15 m to exhaustion) is composed of five repetitions of 12 s shuttle runs interspersed by a 6 s walk to be performed between two lines set 5 m apart from the start/finish line. The running pace is controlled by a constant timing of 6 s beep, between the parallel lines established on the track and marked by cones. Failure to achieve the shuttle run in time to the prescribed audio cue on two consecutive occasions resulted in termination of the test. T-VAM was performed on a 200-m outdoor running synthetic surface track. Ten cones should be set every 20 m on the track. The test starts at a running speed of 8.5 km/h and increases by 0.5 km/h every minute until exhaustion [132]. Participants adjusted their running speed to the cones placed at 20 m intervals. The test ended when the subject could no longer maintain the required running speed dictated by the audio beep, for three consecutive occasions. The results of the present study showed that the Peak Velocity obtained in T-VAM vs T-CAR were similar and demonstrate a high level of agreement; thus, the maximal variable derived from T-CAR and T-VAM could be exchanged when designing training programs. This will allow coaches to be flexible whilst choosing any of the abovementioned test protocol suitable for their team since both provide similar results.

Carminatti’s test (T-CAR) requires participants to perform repeated bouts of 5612 s shuttle running at progressively faster speeds until volitional exhaustion. The 12 s bouts were separated by 6 s recovery periods, making each stage 90 s in duration. The initial running distance was set at 15 m and was increased by 1 m at each stage (90 s) [109]. The test protocol has an initial speed of 9 km/h over a running distance of 30 m (15 m out and back). The stage length in a single direction is increased progressively by 1 m every set. Each stage consisted of five repetitions; between repetitions, the participants should perform 6 s walk between two lines set 2.5 m from the start line. Running pace is dictated by prerecorded audio cues (beeps) that determined the running speed to be performed between the start and finish lines. The test ends when the participant fails to keep in time with the audio cues for two successive repetitions, or a perceive inability on behalf of the participant to cover more distance at the attained level. Peak running velocity in Carminatti’s test (T-CAR) is calculated from the distance of the last set completed by the athlete divided by the time to complete the stage repetition (Fig. 4).

Carminatti’s test (T-CAR) (30 × 15 m)

Full size image

In the case of an incomplete set, peak velocity is interpolated using the equation: “peak velocity = v + (ns/10) × 0.6,” where v is the velocity of the last fully completed stage and ns is the number of repetitions completed in the partially completed stage.

T-CAR presents physiological indices of aerobic power and aerobic capacity that are associated with the laboratory standard measures [133]. Therefore, the results obtained can be used for a reasonably accurate aerobic-fitness assessment and training prescription in team sport players that possess a physiological background like soccer and futsal players. The test would provide useful information’s to soccer coaches and strength and conditioning professionals to assess changes in the intermittent high intensity endurance of players across the competitive season.

Tests carried out in the field have the advantage of being easily reproduced anywhere whilst requiring minimal equipment. Field-based evaluations should be delivered under standardized conditions to improve reliability. This includes standardization of the surface and environment where possible [9]. Field tests should take minimal time away from the coaching program to avoid conflicts of interest [62]. To reduce the effect of circadian variations, testing sessions should be performed at the same time of day [134]. As soccer is a team sport testing would ideally be conducted on many players simultaneously. Therefore, field tests are most viable for collection of data on a large sample in a relatively short period. Depending on the purpose of testing, it may be practical to ensure the coaching staff are in attendance during fitness tests as presence of the coach (Hawthorne effect) can greatly improve commitment and motivation of the player [62]. It is also important that necessary pre-test measures are taken to ensure players are adequately rested/prepared for testing and not in a detrained or fatigued state. The coach should adopt tapering methods in training in the lead up to testing to ensure players are not in a state of chronic fatigue [16]. It is important participants are familiarized with the test protocol. CV between the first two trials of a reliability analysis is 1.3 times greater than between subsequent trials [135].

It is apparent that the emphasis on fitness testing is more common during pre-season, when physical preparations for the upcoming competitive season are the key [62]. During the season when the focus is more towards regeneration and match preparation, time for physical fitness tests may be limited. However, regular testing between and within seasons has utility in identifying specific areas of physical fitness that may need attention [136].

There are myriad ways to interpret fitness testing data in soccer players. Results can be rated against other team members or squad as well as using established quantitative criteria and Likert scale with simple descriptors [62]. It is useful to include results from previous testing sessions, to show trends in performance over time [9]. Indeed, analysis of individual players’ results and within-subject change scores over the season is fundamental [9, 62]. This kind of comparison over time can both educate and motivate players [6, 7, 9]. Growth and maturation should be considered alongside the “learning effect” when interpreting data. It is vital that feedback of test results is communicated to the players and coach at the earliest opportunity and in a concise manner easily understood by both recipients [9]. The challenge is to use the information acquired from testing to inform training prescription [62].

Routine physical fitness testing in soccer players is warranted and can be used to identify individual strengths and weaknesses, talent selection, fitness profile, training effects, monitor return to training, or competition and for individual training prescription. Testing undertaken in laboratory settings has high accuracy but limited sensitivity to changes in performance and specificity to soccer match play may be adopted to give a general fitness profile of a player and during parts of the season when changes in fitness are expected to be large.

Field tests offer a viable, more practical alternative for soccer teams. Such tests have shown greater sensitivity to performance changes, high correlation with VO_{2 max}, and high validity. Issues regarding accuracy and reliability have been improved as protocols and testing equipment have developed. Presence of a ball in the assessment of physical capacity in soccer players provides greater motivation than without. Intermittent tests with soccer-specific actions increase the validity and the usefulness of data obtained from such tests. Care must be taken regarding test design to minimize the effects of technical skill performance on physiological measures. Questions regarding direct validity of tests still abound and a combination of tests may be required for a comprehensive assessment of players. Close co-operation and communication between coaches, players, and sports science personnel together with well-planned, organized fitness testing will help in delivering the objectives of fitness testing in soccer players.

This critical review discusses the applied techniques and technologies in testing soccer players’ health and fitness variables with a specific focus on cardiorespiratory testing. A clear distinction of the functionality and the specificity between the field tests and laboratory tests is established in the literature. The review findings prioritize field tests over laboratory tests, not only for commodity purpose but also for motivational and specificity. Moreover, the research literature suggests a combination of various tests for a comprehensive assessment of the players. Lastly, more research needs to be conducted to contribute into the setting up of a comprehensive test model through the combination of various specific exercise modes to soccer players.

• 17 October 2019

  The editors have retracted this article [1] because it contains substantial text and figure overlap with the following articles [2–5].

BLa:

Blood lactate

CV:

Coefficient of variation

GPS:

Global Positioning System

HR:

Heart rate

ICC:

Intraclass correlation coefficient

LIST:

Loughborough Intermittent Shuttle run test

MASS:

Maximal Aerobic Speed Specific

MST:

Multi-stage shuttle run test

OBLA:

Onset of blood lactate accumulation

RER:

Respiratory exchange ratio

RMST:

Reactive agility motor skills test

RPE:

Rating of perceived exertion

T-CAR:

Carminatti’s test

T-VAM:

Vameval test

VCO₂ :

Carbon dioxide output

VE:

Ventilation

VO₂ max:

Maximal oxygen uptake

VO₂ :

Oxygen uptake

The authors acknowledge Mr. Daniel Beattie’s effort and help with the review process and additional help with the literature search.

Funding

No financial support was received for the conduct of this review or preparation of this manuscript.

Availability of Data and Materials

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Affiliations

1. Sport Science Program, College of Arts and Sciences, Qatar University, 1396 Al Tarfa, 2713, Doha, 2713, Qatar
   • Monèm Jemni
   •  & Mohammad Shoaib Prince
2. Institute of Clinical Exercise and Health Science, Applied Physiology Research Laboratory, School of Science and Sport, University of the West of Scotland, Hamilton, Lanarkshire, ML3 OJB, Scotland
   • Julien S. Baker

Contributions

All the authors contributed significantly in editing, compiling evidence, synthesizing, proof reading, and revising the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Monèm Jemni.

Authors’ Information

Professor Monèm Jemni is currently working at Qatar University after two decades in the higher academy between Europe and the USA. He is a recognized world figure in football as he is the Chair of the International Science And football Association (ISAFA), a network that promotes science knowledge to back-up football, through interactive sharing of the body of knowledge and experiences between scientists, academics, and professional football practitioners.

Ethics Approval and Consent to Participate

Not applicable

Competing Interests

Monèm Jemni, Mohammad Shoaib Prince, and Julien S Baker declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

The editors have retracted this article because it contains substantial text and figure overlap with other articles (see published retraction note). The first author, Monèm Jemni, has not confirmed whether he agrees or disagrees with this retraction; the second and third authors, Mohammad Shoaib Prince and Julien S. Baker, agree with this retraction.

Reprints and Permissions