INTRODUCTION

Soccer is considered one of the most popular sports worldwide and surely the most famous in Europe. Female soccer has been growing its popularity, especially on the last two decades where professionalism has increased markedly, with elite players now employed on either a professional or semi-professional basis (^{Datson, et al., 2014}). There is an abundance of research on the physical aspects of male soccer (e.g., match analysis, fitness characteristics, and training patterns) but yet research on the female game is largely confined to match analysis of elite senior competition and fitness characteristics (^{Alexiou & Coutts, 2008}; ^{Andersson et al 2010}; ^{Mohr et al, 2008}; ^{Krustrup et al, 2010}; Datson et al, 2018; ^{Emmonds et al, 2018}). Currently, elite women's soccer players are exposed to greater training volume and competition demands than ever before. Consequently, these increased physical requirements may have implications for both physical performance levels, injury patterns and on athlete's wellness. Training load (TL) quantification allow coaches and staff to better manage load during training sessions to avoid cumulative fatigue, higher risk of injury and to best prepare the athlete's physical fitness to the next match (^{Oliveira et al., 2020}; 2019a, 2019b; ^{Halperin & Emanuel, 2020}; ^{Ehrmann et al., 2016}; ^{Peart et al., 2018}). Therefore, nowadays it is critical to quantify internal training and match load on women's soccer to provide data to the practitioners and technical staff, to allow for better planning, periodization, assessment and workload control (^{McLaren, Smith, Spears, & Weston, 2016}). The ability to recover between bouts of high intensity activity and subsequently, repeating these efforts has been reported as a critical physical ability of modern-day soccer players (^{Gabbett & Mulvey, 2008}). Consequently, the physical preparation of elite women soccer players has become an important component of professional training to deal with the increasing energy demands of match play (Krustrup et al., 2005; Carling et al., 2008; Iaia et al., 2009; Peart et al., 2018). To optimize athletic performance, physical training should be prescribed to suit each athlete's individual characteristics. However, in team sports like soccer, training sessions are often conducted in a group, which reduces the likelihood that players are receiving specific training based on their individual characteristics (^{Impellizzeri et al., 2005}). For example, some authors (^{Arcos et al., 2016}; ^{Benjamin et al., 2020}; ^{Clemente et al., 2017}; Emmonds et al., 2018) demonstrated that soccer players with superior fitness levels did not receive sufficient training stimulus to further increase their fitness when training in a team environment using small-sided games alone. It has also been suggested that players with inferior fitness levels may be overstressed during team-based training sessions leading to increased fatigue, injury, and a reduction in performance (Krustrup et al., 2010; ^{Giménez et al., 2019}, Ehrmann et al., 2018). These findings may suggest that individuals within the same soccer team may not receive an appropriate level of training stimulus when a team-based training approach is undertaken. Therefore, to overcome the limitations associated with team-based training, it has been suggested that a simple system that quantifies an individual's response to training (ie, internal TL) is developed, so that coaches can monitor and modify training according to individual players' needs (^{Chen et al., 2010}; Impellizzeri et al., 2019; ^{Foster et al., 2001}). The assessment of internal TL requires quantification of the intensity and duration of the physiological stress imposed on the athlete (Foster et al., 2001; Foster, 1998). Over the last decade, rated of Perceived Exertion (RPE) has shown to be a good method of assessing the internal load in soccer players (^{Gaudino, et al., 2015}; Arcos, Méndez-Villanueva, Yanci, & Martínez-Santos, 2016), however, about women's soccer, information has still been scarce although scientific literature on women soccer athletes is growing (Alexiou & Coutts, 2008). Understanding the internal response to external loads placed upon players is particularly relevant to practitioners (Impellizzeri et al., 2019; ^{Mclaren et al., 2017}). Internal response to training can be measured practically using session-RPE (s-RPE), which provide a valid quantification of relative exercise intensity and internal load across a range of exercise modalities (Chen et al., 2010; ^{Scantlebury et al., 2017}). This method is also cost- and time-effective, allowing to monitor internal TL in clubs where resources can be limited (^{Wright et al, 2020}).

The attention given to the post-match and training sessions recovery status of soccer players has been increasing recently (^{Thorpe et al., 2015}, 2016a, 2016b; ^{Rabbani et al., 2019}). Tracking recovery of soccer players requires monitoring tools that are sensitive to match and training sessions fatigue to help practitioners making decisions on a daily basis (Thorpe et al., 2016b). Hooper index (HI) is another easy-to-use, non-cost, and non-invasive monitoring tool that allow to monitor the players' recovery and wellness (Hooper & Mackinnon, 1995). HI seems promising for tracking fatigue during in the season in professional soccer (Rabbani et al, 2019). HI has been reported in recent literature as a reliable method for the monitoring of athlete wellness providing further information concerning the detail of player fatigue, stress, muscle soreness and sleep perception and it has been used in a large number of soccer studies (^{Clemente, et al., 2017}, 2019; ^{Moalla et al., 2016}; ^{Oliveira et al., 2020}, 2019a, 2019b; Rabbani et al., 2019; Thorpe et al., 2016a, 2016b).

However, compared to internal measures, such as heart rate (HR), RPE and to the best of author's knowledge, HI has been little used in elite women's soccer as a measure of athlete recovery and wellness. Thus, the aim of this study was to analyse and compare s-RPE and HI between training sessions and match days from a women's Portuguese BPI League team. It is intended to get information of TL organization that coaches and researchers may use to help design and implement optimal training programs for appropriate physical fitness development in female professional soccer players, without causing negative disturbances.

MATERIAL AND METHODS

RESULTS

A total of sixteen elite women soccer players from the same team participated in the Portuguese BPI League in the 2019/20 season. Table 1 summarizes the characteristics and comparisons of the study sample in training and match days.

Table 1.  TL during the MD minus for squad average, mean ± SD. 

Regarding RPE, results showed differences between MD-5 vs MD-2 (p<0.001; g= 1.85 [1.07, 2.71]) and MD-5 vs MD (p=0.027; g= -0.87 [-1.59, -0.18]). Also, MD-4 showed differences between MD-2 (p=0.001; g= 1.45 [0.71, 2.24]) and MD-4 vs MD (p=0.003; g= -1.30 [-2.01, -0.58]). MD-2 showed difference between MD (p<0.001; g= -2.10 [-3.00, -1.29]).

In training duration, there were differences between MD-5 vs MD-4 (p=0.020; g= -1.54 [-2.34, -0.79]), MD-5 vs MD-2 (p<0.001; g= 2.47 [1.61, 3.44]) and MD-5 vs MD (p=0.011; g= 1.23 [0.52, 2.00]). Also, there were differences between MD-4 vs MD-2 (p<0.001; g= 3.49 [2.46, 4.66]) and MD-4 vs MD (p=0.006; g= 1.34 [0.62, 2.12]).

In addition, there were differences between MD-5 vs MD-2 (p<0.001; g= 2.09 [1.27, 2.98]) and MD-4 vs MD-2 (p<0.001; g= 1.97 [1.18, 2.84]) for s-RPE.

Table 2.  HI during the MD minus for squad average, mean ± SD. 

Regarding HI categories, there were significant differences between MD-4 vs MD (p=0.049; g= 1.03 [0.33, 1.77]) for fatigue, between MD-4 vs MD-2 (p=0.040; g= 0.52 [-0.15, 1.21]) for DOMS and between MD-4 vs MD (p=0.042; g= 0.86 [0.16, 1.58]) for total HI.

DISCUSSION

In this study, the aim was to analyse and compare the internal TL between training sessions and match-days from the same women's Portuguese League team. It is noteworthy to report that soccer team used for this study was consistently ranked among the top five seeds in the national Women's Soccer League (BPI League).

Our results showed that there are substantial differences for RPE and s-RPE when comparing training sessions, MD-5> MD-4> MD-2. Regarding RPE, there was a decrease over the training microcycle, with differences between MD-5 and MD-4 compare with MD-2 and between MD-5, MD-4 and MD-2 compare with MD. In terms of training methodology and periodization, there would be a pattern in the microcycle. The pattern on the three training sessions was usually composed of a first training session (MD-5) in which the exercises of small-sided games, exercises of ball possession games and the circuits of specific strength and specific resistance were predominated. In MD-4, it was used exercises that requested higher intensities of aerobic resistance training, with the tactical games with larger spaces and a higher number of players. Finally, in MD-2, standardized exercises and the achievement of the goal were sought, fundamentally with the use of a few athletes and, consequently, requesting the use of speed, acceleration, and speed on the skills execution.

Regarding the s-RPE, there was also a decrease in TL over the microcycle, with differences between MD-5 and MD-4 compare with MD. From the authors' knowledge there was a paucity in studies that address the weekly training load distribution pattern in female's soccer, however the results found are in line with those reported in male's soccer (^{Malone et al., 2015}; ^{Stevens et al., 2017}; ^{Oliveira et al., 2020}, 2019a, 2019b).

The RPE values obtained are higher in matches, revealing that the matches' internal loads are the stimulus with the greatest impact on the perceived fatigue of the players, although HI did not present higher values in the following day of the match. According to these results, matches were rated hard to very hard, which corroborates those presented by ^{Wright et al. (2020}) who also stated that compared to matches, soccer training sessions were rated substantially easier for all exertion types. The fact that the training sessions' RPE was lower than matches are in accordance with results of both female's and male's soccer studies, which reported that the training sessions intensity is lower than matches intensity (Wright et al., 2020; ^{Malone et al., 2015}; ^{Stevens et al., 2017}; ^{Oliveira et al., 2020}, 2019a, 2019b).

A relevant observation was the fact that the matches' RPE values present a higher standard deviation which may be explained by less player effort due to their field 'position, status (starter, non-starter) or, due to the fact that there were differences in the perception that each one has in relation to their general physical condition response to the effort made.

The same pattern as RPE was not found in s-RPE, in which the match presents lower values, although not statistically significant. Interestingly, regarding s-RPE values, there were no differences between MD-5, MD-4 and MD. These results were not usual in the literature on male teams (^{Thorpe et al., 2016}; ^{Stevens et al., 2017}; ^{Oliveira et al., 2019}a), but this may be due to the number of weekly training sessions for female teams is inferior to that of the male teams. However, the literature available with female soccer suggested that, in relative terms, physiological demands (percentage of heart rate and percentage of VO_2max) during match-play were similar to those obtained for male soccer (^{Stolen et al., 2005}). However, ^{Krustrup et al. (2010}) observed an average reduction of more than 4% for most starters in elite female soccer matches. The observed decrement appears to be somewhat higher than the 2-3% reduction determined for male players (Krustrup et al., 2006; ^{Mohr et al., 2004}; ^{Rebelo et al., 1998}), which may be because the study investigated sprint performance after competitive matches and not friendly matches as in the other studies (Krustrup et al., 2010). Moreover, it could also be due to differences in training status or gender-specific fatigue development (^{Kent-Braun et al., 2002}).

Regarding training duration, differences along the microcycle MD-4 > MD-5 > MD-2 > MD were observed. These findings were probably due to the training planning and periodization strategy performed by the coach who privileges the MD-5 with high-intensity activities, like small-sided games that require more training duration. From the information provided to us by the team coach, in MD-4, the training session privileged low volume exercises to stimulate aerobic capacity and focus on tactical team issues to develop the squad' match model. The MD-2, last training of the week, was characterized by incorporating exercises that privilege sprint and short high-speed runs in low volume short game situations with the intention of preparing the squad for the match.

To the authors' knowledge, this is the first study investigating the variance of HI across microcycles in elite female professional soccer players. Although some authors (^{Scott et al., 2020}) claimed that the ordinal wellness ratings do not meaningfully report to within-player previous or subsequent external loads in training or matches, numerous authors (^{Alexiou & Coutts, 2008}; ^{Rabbani et al., 2019}; Los ^{Arcos et al., 2016}; Polito et al., 2017; ^{Clemente et al., 2017}; ^{Oliveira et al., 2020}, 2019a, 2019b; ^{Impellizzeri et al., 2019}) claimed that the HI showed to be a useful measure for monitoring match- and training load- induced fatigue in soccer players. In the present study, the results of the HI indicated that fatigue, DOMS and total HI were highest on MD-4. ^{Ispirlidis et al. (2008}) reported significant decreases in performance and biological alterations up to 72 hours after soccer matches, which corroborates with the highest fatigue, stress and sleep quality levels on MD-5 and MD-4.

In general, Hooper's categories indicated good mean levels, with low DOMS. As far as we know, there are no studies with reference to HI in female soccer players. When compared to female basketball players (^{Clemente et al., 2019}), the present values are higher, which indicates that soccer players had less recovery capacity, although they had fewer weekly training sessions than the basketball players in the study by Clemente et al. (2019). The same study analyzed the internal load during training sessions that occurred two or three days before a match. In our study results indicated higher internal loads during training sessions that occurred on MD-4, which was the session training with more significant differences, either with MD-2 (DOMS) or with the match (fatigue and total HI). These data could show that regarding fatigue and DOMS, the MD-5 and MD-4 training were very close to each other regarding to the time variable, as there was only one day between the two, it could show an error in the planning and periodization of training although, in most cases, it is not the responsibility of the technical team as they feel obliged to train these days due to the club conditions and limited facilities that clubs offer to their teams. However, from the analysis of weekly TL distribution, it appeared that there was a load decrease (tapering phase) in the last training session before a match (MD-2), which was accompanied by an increase in wellness status on a match day.

In general, Hooper's categories indicated good mean levels, with very low DOMS, stress, and fatigue and very good sleep quality. The results of this study corroborate with ^{Manzi et al. (2010}), who reported the use of a tapering strategy before matches in professional basketball, with lower s-RPEs on the day before a match. It was noteworthy that Manzi et al. (2010) and ^{Clemente et al. (2019}) showed higher loads in some training sessions a s-RPE of approximately, 600-800 AU, while the average s-RPE of the present study is lower. Although wellness status can be influenced by extra-training factors (Haddad et al., 2013; Hooper and Mackinnon, 1995; ^{Scott et al., 2020}), the tapering phases before matches may have contributed to improve players' overall wellness, since the load decrease on the day before a match was accompanied by improvements in the players' total HI, DOMS, fatigue, and sleep quality on a MD. Due to differences between sports (soccer vs. basketball), future studies should investigate the impact of the training session and matches per week on soccer players' performance, as well as the relationship between wellness status and sport performance.

The results from the present study should not be generalized because it has been reported that a higher workload is performed by players at higher competition level compared with players at moderate levels in female players (^{Mohr et al., 2008}; Mohr et al., 2004; ^{Gabbett & Mulvey, 2008}) and the analyzed games are of national level. Also, the small sample size does not allow to generalize the results.

Moreover, there are other contextual factors that could influence the interpretation of our results such as player positions, player status, match results, match location, and quality of opponents (^{Oliveira et al., 2020}).

In addition, we were also unable to quantify external load within this study which may limit our understanding of the specific types of activity that result in a given internal response (for example, we do not know if a higher s-RPE in a match or training session is due to greater total distance covered or the requirement to repeat more bouts of high-speed running). Therefore, we suggest that future studies would consider such contextual factors. Also, it would be relevant to include some global/local positioning system measures to better understand in internal and external load would relate to each other.

Other suggestion for future studies is to analyze intra- and inter-week variations of wellness such as in the study of ^{Nobari, Fani, et al. (2021}).

Meanwhile, the present study represented the actual competitive national level from 2019-20 in-season.

CONCLUSIONS AND PRACTICAL APPLICATIONS

The results showed that there is a higher TL (RPE and s-RPE) on the first and second training sessions of the week (MD-5 and MD-4) than in the last training session (MD-2). Furthermore, matches presented the highest internal workload. Results from HI showed that sleep, fatigue, stress and DOMS was well fair controlled by coaches and staff from the beginning of the microcycle until the MD. This means that TL was adjusted to improve players' wellness until the day of the match.

Through easy, valid and low-cost tools such as the RPE, s-RPE and HI, this study provided relevant information to the staff teams of Portuguese elite women's league clubs to better prepare their athletes for the match, especially for those that trained three times a week. For coaches and practitioners, the study generates reference values for professional players which can be considered when planning training sessions.