Physical performance and loading for six playing positions in elite female football: full-game, end-game, and peak periods

1 INTRODUCTION

The vast majority of scientific studies on football have been performed with male players, but in recent years, the interest in women's football has increased considerably.¹ Most studies of female football players have focused on injuries and injury prevention, as well as the reporting of anthropometrics and test performance, and, to a lesser extent, the demands of the game.^1-5

Women's football has undergone a professionalization during the last decade, and a larger number of female players can now train and compete at professional level.⁵ As match performance has been shown to be related to training status⁶ and match demands are higher in matches at a higher level,^{7, 8} it would be expected that the increased professionalism in women's football has increased match demands. A recent large-scale study, on the other hand, found only minor differences in match demands between female football players at different standards,⁹ and the authors suggested that this discrepancy with previous studies could be due to the lower number of participants in the older studies leading to underpowered analysis. Another explanation may be the difference in the tracking systems used, as the older studies used video-based time-motion analysis, in contrast to the study by Scott and colleagues⁹ using GPS-tracking, and different tracking systems have been shown to produce different values.^{10, 11}

Across tracking systems, total distance covered is typically in the range 9.2-11.3 km, whereas distance covered with high-speed running is in the range 1.2-2.7 km and the sprinting range is 160-460 m.^{6-9, 12-14} A number of studies have investigated match demands and found differences between playing positions. But the majority of previous studies have used a less specific categorization of positions, using defenders, midfielders and attackers,^{6-8, 14, 15} and more specific categorization of the outfield positions is warranted,⁴ as has been applied in a number of the recent studies.^{9, 13, 16} Although presenting data for each position, the main aims of the studies by Bradley and colleagues¹³ and Scott and colleagues⁹ were to compare genders¹³ and female players at domestic or international level.⁹ Thus, large-scale studies on match demands relative to position in women's elite football are warranted.

In addition to data on external load, a number of studies have analyzed internal load using heart rate (HR) monitors and blood lactate measurements. HR_mean was in the range 152-186 bpm and HR_peak in the range 175-212 bpm, corresponding to 87 and 97% of HR_max,⁶ whereas blood lactate values were in the range 1-11 mmol/L after the first and second half.¹⁷ None of the large-scale studies, however, have evaluated internal load, for example, by measuring heart rate, in conjunction with external load during matches and related this to positional differences.

The aim of the present study was, therefore, to investigate the position-specific match demands and heart rate response of elite female football players from the entire Danish Women's League.

2 METHODS

3 RESULTS

3.1 Movement pattern in full-game

Total distance covered ranged from 7812 to 12 823 m for outfield players and was 14%, 13% and 8% higher in CM, EM and FB, respectively, than in CD (P < .001), with no other differences between outfield positions (Table 1). Limited differences were observed between outfield positions in speed zones below 18 km/h; greater distances in EM than in CD in SZ2 and SZ3, and greater distances in CM compared to CD and FW in SZ1 and SZ2, but only compared to CD in SZ3. In SZ4, FB, EM, and FW covered more distance than CD, whereas for SZ5 FB, CM, EM, and FW covered a greater distance than in CD. HSR was higher in FB, CM, EM, and FW than in CD, whereas VHSR was higher in FB, CM, EM, and FW compared to CD and in EM compared to CM, and sprinting was higher in EM compared to CD and CM (Table 1).

Table 1. Full-game velocity and distances covered, heart rate responses, accelerations, and decelerations

	GK	CD	FB	CM	EM	FW
Total distance [m]	5214 ± 949	9274 ± 762	10 053 ± 639†	10 572 ± 880†	10 519 ± 963†	9745 ± 988
	(3510-6806)	(7812-10537)	(9117-10871)	(8317-12823)	(8474-12484)	(8426-11784)
HSR [m]	99 ± 70	1088 ± 261	1529 ± 369†	1518 ± 499†	1786 ± 527†	1561 ± 372†
	(25-273)	(631-1567)	(1114-2258)	(854-3370)	(1120-2917)	(1083-2047)
VHSR [m]	31 ± 31	442 ± 135	717 ± 242†	623 ± 252†	863 ± 299†,§	737 ± 223†
	(5-106)	(191-695)	(399-1175)	(315-1381)	(354-1320)	(399-1117)
Sprinting [m]	1 ± 3	19 ± 17	46 ± 48	33 ± 31	91 ± 81†,§	56 ± 45
	(0-9)	(0-63)	(0-199)	(0-108)	(0-245)	(4-148)
Peak speed [km/h]	21.5 ± 1.2	27.5 ± 2.3	28.2 ± 3.2	27.8 ± 2.0	27.6 ± 2.1	29.2 ± 3.2
	(19.4-22.9)	(23.7-30.6)	(23.7-32.0)	(23.7-30.8)	(23.1-30.3)	(25.1-31.6)
HR mean [bpm]	148 ± 10	169 ± 9	171 ± 11	170 ± 10	173 ± 8	170 ± 8
	(130-159)	(154-185)	(144-185)	(146-187)	(157-188)	(159-186)
HR mean % of HR max	79 ± 5	87 ± 4	89 ± 3	89 ± 3	89 ± 4	87 ± 3
	(70-85)	(78-93)	(79-93)	(82-93)	(78-93)	(80-93)
HR peak [bpm]	181 ± 11	192 ± 8	190 ± 9	190 ± 11	193 ± 13	194 ± 6
	(165-193)	(172-216)	(171-206)	(169-208)	(176-235)	(181-204)
HR peak % of HR max	96 ± 4	98 ± 2	99 ± 1	99 ± 2	99 ± 1	98 ± 3
	(90-100)	(92-100)	(96-100)	(91-100)	(96-100)	(91-100)
% of total time with HR 160-180 bpm	23 ± 16	50 ± 14	47 ± 20	45 ± 18	53 ± 15	50 ± 14
	(0-42)	(17-71)	(18-80)	(13-83)	(25-82)	(20-65)
% of total time with HR > 180 bpm	2 ± 3	26 ± 25	32 ± 27	35 ± 25	30 ± 21	29 ± 23
	(0-8)	(0-79)	(0-82)	(0-82)	(0-71)	(0-80)
High accelerations [n]	2.8 ± 1.5	6.7 ± 3.7	8.0 ± 4.9	10.0 ± 6.8	7.1 ± 5.4	12.1 ± 7.0
	(1-6)	(2-17)	(3-23)	(0-26)	(0-18)	(0-25)
Sum of accelerations [n]	695 ± 164	864 ± 114	878 ± 136	945 ± 140	871 ± 116	884 ± 126
	(441-857)	(701-1143)	(730-1248)	(705-1209)	(678-1092)	(709-1140)
High decelerations [n]	3.3 ± 1.3	12.6 ± 4.3	17.4 ± 4.6†	16.4 ± 5.5†	22.5 ± 6.7†,‡,§	18.7 ± 3.9†
	(2-5)	(4-22)	(8-27)	(8-30)	(9-37)	(9-23)
Sum of decelerations [n]	738 ± 139	887 ± 101	895 ± 137	946 ± 135	893 ± 123	921 ± 148
	(517-887)	(698-1133)	(710-1226)	(735-1232)	(670-1116)	(712-1220)

Note

• Mean ± SD and range (min-max).
• Abbreviation: GK, goalkeepers; CD, central defenders; FB, full-backs; CM, central midfielders; EM, external midfielders; FW, forwards.
• ^† Different from CD.
• ^‡ Different from FB.
• ^§ Different from CM.

The number of VHSR was higher in EM and FW than in CD (54 ± 10 and 49 ± 18 vs 28 ± 12; P < .05), with no differences between other outfield positions. In addition, no difference was observed between positions in peak speed (range: 27.5-29.2 km/h; Table 1). Limited differences between positions were observed for number of accelerations. A higher number of moderate accelerations were observed in CM and EM than in CD (200 ± 42 and 198 ± 48 vs 162 ± 24; P < .05), whereas no positional differences were observed in high or low accelerations. For decelerations, no differences were observed in low decelerations, whereas a higher number of moderate decelerations was observed in CM and EM than in CD. For high decelerations, a higher number was observed in EM than in CD, FB, and CM, whereas FB, CM, and FW were higher than CD (Table 1; Figure 1).

HR_mean was 169-173 bpm, corresponding to 87%-89% of HR_max, and HR_peak was 190-194 bpm, corresponding to 98%-99% of HR_max, with no difference between positions. Percentage of total time spent in HR zones did not differ between positions in any of the HR zones (Table 1).

In GK, total distance covered was ~50% of that for outfield players, with the vast majority (66%) in SZ1. Distances were different from all outfield positions in SZ1, SZ2, SZ3, SZ4, and SZ5, but only EM was different from GK in Sprinting (Table 1). GK had fewer moderate accelerations and moderate and high decelerations than outfield players. Only CM and FW had more high accelerations than GK. HR_mean was 148 ± 10 bpm, corresponding to 79 ± 5% of HR max, which was lower than for outfield players, whereas HR_peak was 181 ± 11 bpm, corresponding to 96 ± 4%, which was lower than in CM, EM, and FB.

4 DISCUSSION

The major findings of the present study were that competitive elite female matches impose very high physical demands on all outfield playing positions, including central defenders, full-backs, central midfielders, external midfielders and forwards, resulting in high aerobic loading throughout matches and marked declines in physical performance toward the end of games. For all outfield playing positions, pronounced decrements were observed in the last 15-minute period of the game in peak speed (~15%), intense accelerations and decelerations (~20%), and intense running bouts (~30%), whereas heart rates were similarly high. Moreover, the present study revealed that the overall physical match demands are much higher for central midfielders, external midfielders, full-backs and forwards, compared to central defenders, with the largest differences observed in the full-game values for high-speed running and sprinting as well as the number of moderate and intense accelerations and decelerations. The physical demands of the most intense 5-minute period of the game were also higher for central midfielders, external midfielders, full-backs and forwards, compared to central defenders, albeit the differences were smaller.

Total distance covered for the various outfield positions was 9274-10 572 m, which is slightly higher,¹⁵ in the same range^{6-9, 16} or slightly lower¹³ than previously reported in elite women's football. The greater total distance covered in FB, CM, and EM than in CD found in the present study is similar to previous findings,^{8, 15, 16} whereas no difference was observed between FW and CM, as observed by Datson and colleagues.¹⁶ The relevance of total distance covered for evaluating physical match performance has been questioned in several studies, as similar total distance covered has been found across playing levels,^{7, 8} whereas HSR distance varies more between performance level and has been shown to be closely related to fitness and training status.⁶

High-speed running was similar between FB, CM, EM, and FW, all of which were greater than CD (1518 ± 499 to 1786 ± 527 m vs 1088 ± 261 m). For VHSR (>18 km/h), the greatest distance was observed for EM (863 ± 299 m), which was higher than for CD and CM. In addition, FB, CM, and FW covered a greater distance with VHSR than CD. These observations are comparable to previous findings showing the greatest distances covered in the highest speed zones for EM or FW.^{9, 13, 16} Due to differences in speed thresholds, it is difficult to compare values between all studies. However, Bradley and colleagues ¹³ used similar thresholds and presented slightly higher values for VHSR for all positions, possibly as a result of the higher standard of the matches.

Sprint distance (>25 km/h) did not differ between positions, except for a greater sprint distance in EM than in CD and CM. In the study by Datson and colleagues,¹⁶ shorter sprint distance was found in CD than in all other positions, with no other differences. The lowest sprint distances covered were also observed in CD in the present study, whereas EM and FW displayed the greatest sprint distances, as in the study by Datson and colleagues,¹⁶ though in both studies, the values did not reach statistical significance. Nevertheless, this indicates similar trends in the positional demands of the game. Sprint values were lower in the present study for all positions than those presented in the studies by Bradley and colleagues ¹³ and Datson and colleagues,¹⁶ which used similar sprint thresholds. This may be due to the difference in the level of the players, as the present study included players from the entire league, whereas these two studies investigated players in international matches and higher match demands have been observed in international compared to domestic matches, especially for highest running speeds.⁷ Large variation was observed within each position, and various factors, such as training status, the standard of players, fatigue development, tactical decisions, individual playing style, may influence the demands of the game.^{6-8, 17}

Preparation for the most intense periods is critical for a high performance level,²⁰ as these periods are linked to crucial moments such as goalscoring.²¹ Data on these intense periods during women's football matches are scarce in the scientific literature. 13%-17% of total distance covered with VHSR and 32%-63% of total distance covered with sprinting were observed during the most intense 5-minute period of the game, though a 5-minute period equates to approximately 5.6% of total playing time. This highlights the importance of preparing for the most intense periods and not using averages of whole-match values. When analyzing differences between positions, only CD stands out in intense running speed, with lower distance in sprinting than EM, lower VHSR than FB, EM, and FW, and lower HSR distance than FB, CM, and EM. Thus, only minor differences between positions were observed during the most intense periods. Values for peak 5-minute periods may, however, have been underestimated by up to 25%, since data were analyzed in predefined 5-minute periods and not rolling averages.^{22, 23} Overall, the present values for HSR are similar to those for players participating in the Champions League.¹³

HR_mean was 87%-89% of HR_max and HR_peak was 98%-99% of HR_max, which for both variables is similar to or slightly higher than previously reported in female players.^{6, 12, 17} In the present study, individual HR_max was obtained as the highest HR during matches or Yo-Yo IR1 testing. Only 62 of the 95 players included completed the Yo-Yo IR1 test, and of these 62 players a total of 57 had their highest heart rate measured during the Yo-Yo IR1 test. Thus, higher HR_max may have been obtained if more players had completed the Yo-Yo testing, leading to a slightly lower relative heart rate. Time spent with high heart rate was high, with 26%-35% of the total time spent with HR above 180 bpm. This demonstrated that the external load highly taxes the players’ aerobic system. No differences were found between positions in HR_mean, HR_peak or percentage of time in HR zones, as also described by Krustrup and colleagues,⁶ indicating that the players’ training status matches their match demands, as minor differences were found between positions in external load.

To evaluate fatigue development and changes during the game, we compared the first and second halves as well as the first and last 15-minute periods. A decrease in total distance was only observed in CM, whereas 6%-11% and 10%-12% drops were observed in HSR and VHSR for both midfield positions, underpinning that these positions have the overall highest match demands (figure 3). Fluctuations in external load throughout the match were evident for all positions, with all having lower HSR and VHSR in the last 15-minute period compared to the first, whereas all except FW experienced a decrease in total distance. This is in agreement with previous studies showing lower distance covered at the end compared to the beginning of the match.^{7, 8, 13, 15, 16} A drop in the number of low, moderate, and high accelerations and decelerations was also observed for some of the positions. A drop in acceleration and deceleration performance toward the end of a match has also been observed in men.²⁴ These observations support the observation that football players’ capacity to perform is decreased likely as a result of fatigue toward the end of a match.

Despite covering less distance in total and in HSR and VHSR, performing fewer intense runs and lower peak speed, all indicating lower intensity, heart rate response did not differ between the first and last 15-minute periods. This may be due to a deteriorated fluid balance and lowering of plasma volume leading to a higher HR to maintain cardiac output.²⁵ HR_mean and HR_peak were, though, slightly lower in the second than in the first half, and less time was spent in the higher HR zones for more positions.

Even though this study is the first large-scale study to present data on external and internal load in women's football for an entire league, the study has a few limitations that should be acknowledged. Only two thirds of the participants completed the Yo-Yo test, and since 92% of those completing the Yo-Yo test obtained their highest HR during the test, this may have influenced the relative HR measurements. Heart rate should be related to individual HR_max, and individualized HR zones would have been preferable to absolute values. Due to limitations in the software, this was unfortunately not possible. Individualized thresholds for external load have also been debated by a number of researchers,^26-28 but were not applied in the current study. Moreover, analyses were performed using predefined 5-minute periods and not rolling averages, which may have limited the values in peak 5-minute periods.^{22, 23} Positional data were collected using Polar Team Pro system, which has only undergone few independent scientific validity and reliability tests.^{18, 19} In an internal validity test of Polar Team Pro system against other more thoroughly validated systems (Tracab and ZXY, Chyron Hego), Polar Team Pro was considered valid and reliable for measuring distances covered during longer periods, whereas decreasing validity and reliability was observed for high speeds and short periods as observed for other tracking systems (data not shown—Elstrup, personal communication). A major strength in the present study is that all 8 teams from the Danish Women's League were included; hence, a large sample size could be used for analysis. However, due to methodological decisions on inclusion and exclusion criteria, some positional subgroups had more than a two-fold sample size compared to other subgroups. This may potentially have influenced statistical analysis, where the lack of a significant finding may have been due to low statistical power.

5 CONCLUSION

Taken together, the present study revealed that competitive elite female matches are highly physically demanding for all outfield playing positions, with high aerobic loading throughout matches, multi-sprint peak-intensity periods and marked declines in high-speed running, peak speeds and intense accelerations and decelerations toward the end of games. The overall physical match demands are much lower for central defenders than for the four other outfield playing positions, though this difference is minimized in peak-intensity periods.

6 PERSPECTIVES

The results of the present study clearly emphasize the relevance of monitoring physical match demands for the entire game, but also with a specific focus on the peak-intensity periods and the final periods of the game. The high and multi-faceted physical demands placed on all outfield positions during competitive elite female games highlight the importance of proper physical preparation and recovery strategies to be able to cope with the demands. Coaches should pay certain attention to individualized position-specific physical demands in relation to the team's tactical plan and secure a proper training stimulus related to the individual players movement pattern. Moreover, special emphasis on preparation for peak periods (worst-case scenarios) is important.

CONFLICTS OF INTEREST

Peter Krustrup is employed by the Danish Football Association as Fitness Coach for the Danish Women's national team.