Field hockey is a team sport with heavy demands on the player's physiology1, 2. As a consequence, elite players show a high degree of leanness3. Furthermore, team mean percentage body fat was found to bear a relation with finishing position in a sample of 12 teams playing the South African Senior Provincial tournament. On the contrary, in Australia, sub elite male field hockey players did not differ from Senior or Youth select sides in mass, height, nor body fat levels4.

Elite level sport tends to self-select individuals with morphological characteristics which respond optimally to its physical demands5. Since field hockey requires players to spend time in a crouched position, having long arms may be an advantage in this activity. Data on 33 male field hockey players from the Montreal Olympic Games showed them to have proportionally longer arm and forearm lengths when compared to a reference Canadian university student sample6. Scott3 did not find any correlation between functional arm length and playing ability in his large sample of players.

The birth-date effect is often found in under-age squads, where age categories are delimited by at least a one-year period, favoring those who are biologically more mature from being born earlier in the year7. This effect may carry on to senior select sides if early selection processes have discouraged late-year birth daters away from high-level competition8. Again, we have not found such information in male field hockey.

The aims of our study are: (a) to establish the anthropometric characteristics of elite junior Argentine male field hockey players who won the International Hockey Federation (IHF) Junior World Cup in 2005; (b) to look for differences in physique, years of playing and birth-date effect between the final players selected to make up the team and those who were not out of the original pre-selected sample; and (c) to establish whether there are any differences in proportional limb lengths between elite junior hockey players and a local reference sample. In this manner we wish to contribute to the anthropometric database on successful elite junior athletes and the characteristics that set them apart for the processes of talent identification.

The study is observational, cross-sectional, and prospective. We measured the players on one occasion, in August 2004, after being selected to form the base from which the final select squad will emerge, to compete in the June 2005 IHF Rabo Hockey Junior World Cup held in Rotterdam, The Netherlands. The final chosen players make up the selection team that is compared to those who were left out.

Thirty five junior male field hockey players from all regions of Argentina (age 19.1 ± 1.0years, weight 70.7±5.4kg, height 176.4±6.4cm), chosen by coaches to make up a pre-selection from which the final 17 players would compete in the IHF Junior World Cup of 2005. In the month preceding their evaluation battery at the national sports centre in Buenos Aires, players were sent an informed consent formulary via electronic mail detailing the scope and intent of the measurements, and were asked to sign it or have their parent or legal tutor do it if underage. Approval for the study was granted by the Ethics Committee at the Medical Department of Club Atlético River Plate. We excluded the four goalkeepers from this analysis, since they have different morphologic characteristics than the other playing positions which could affect the general average, such as higher skinfolds, perhaps because their activity profile is very different from field players; nevertheless their data is shown for descriptive purposes. The other two goalkeepers selected as part of the final team were unavailable for measurement at the time. Selection criteria were based on coaches's subjective observation of playing skills, and the final select squad went on to win the tournament.

Level 2 and Level 3 International Society for the Advancement of Kinanthropometry (ISAK)-certified anthropometrists undertook a full anthropometric profile including height and weight, skinfolds, girths, and structural dimensions (breadths, segment lengths and heights) following procedures outlined by ISAK9, as well as information on playing position and history of training. Lastly, we used the roster of the final select team from the Confederación Argentina de Hockey web page10.

A Level 3 anthropometrist landmarked all subjects prior to being measured. We weighted the players using a mechanical scale (Model p1001, CAM, Buenos Aires, Argentina), and measured height, arm span, and sitting height with wall-mounted stadiometers and a 50cm wooden box. We then used a segmometer (Rosscraft, Buenos Aires, Argentina) for segment lengths and heights; large and small bone calipers (Campbell 20 and Campbell 10, Rosscraft, Buenos Aires, Argentina) for large and small bone breadths; a flexible, non-stretching steel tape (W606PM, Lufkin, USA) for girths; and a Harpenden skinfold caliper (Body Care, Baty International, Harpenden, England) for skinfolds. We took all skinfolds and most girths in duplicate, which are the variables that normally have the largest technical error of measurement. A third measure was taken if a large discrepancy between the first two was found. We measured the players in the afternoon before training, barefooted and in minimal clothing. We included the sum of six skinfolds, a popular indicator of body fatness6, and somatotype11 as derived variables.

We entered data into Microsoft Excel spreadsheets (MSN, Cupertino, USA), and later transferred them to Epi Info 6.0 statistical package (CDC, USA) for analysis, including means, standard deviations, frequencies, Bartlett's test for homogeneity of variance, analysis of variance, Kruskall-Wallis, and Chi square. Statistical significance was set at p<0.05.

It is important to state that measurements were taken almost one year before the Junior World Cup, and that the anthropometric data of the squad that made up the final selection may be different from measurements had they been taken at the tournament. It is not uncommon for athletes this age to undergo morphological changes within a year. Hence care must be used when interpreting the data of the tournament winners.

Technical errors of measurement (TEMs) for girths on the 37 players fall under the 1.0% cut-off value considered to be the acceptable upper limit of error for these measures, and under the 5.0% cut-off for skinfolds12, indicating good reliability in measurement. We established the validity of the anthropometrists’ technique beforehand during four-month long certification courses, when inter-subject TEMs were calculated against Criterion Level 3 and Level 4 anthropometrists.

The descriptive characteristics of the players and goalkeepers are shown in Table 1, Table 2. Data from a local 20 to 30 year-old fitness sample, Argoref13, are also shown to compare heights and segment lengths. This sample, although older, was deemed adequate for comparison of bone structural characteristics, considering that by age nineteen the players have reached or are very close to final bone length growth. In this comparison, there were no differences in height or any of the segment lengths, with the exception of sitting height (p=0,03), where the hockey players had, on average, a larger value for trunk height (Table 1).

Table 1. Basic and skeletal characteristics of players and reference sample

Table 2. Girths, skinfolds and somatotype of players and reference sample

a Sum of triceps, subscapular, supraspinale, abdominal, front thigh, and calf skinfolds.

Within the group of players (Table 3), there were no statistically significant differences between player positions in age, weight, and height, although the Σ 6 skinfolds (p=0.009) and endomorphy (p=0.017) were larger in the goalkeepers. There is the possibility of a Type II error here since the sample sizes for each group are small. We must point out that, although statistically not significant, defenders were, on average, taller (p=0.27) and halves lighter (p=0.17) than all the other players. Defenders also had lower values for Σ 6 skinfolds and endomorphy.

Table 3. Anthropometric differences between playing positions

a (p=0.009).

b (p=0.017).

After goalkeepers were separated from the sample, no statistically significant differences were found between the selected and non-selected players in any of the anthropometric variables nor somatotype, with the exception of humerus breadth which was 0.2cm smaller in the selected sample (p=0.01). This difference in breadth is meaningless in practical terms.

The lack of skeletal differences between the selected players and the normal reference sample refutes the hypothesis that field hockey players may have longer upper limb lengths, as found by Ross et al6, but further confirms the finding of these authors that «this sport has minimal structural limitations, and qualities and factors other than physique enter into the development of the field hockey player.» One such quality is lower back strength, where field hockey players have superior values than a normal poplulation sample17. The difference in sitting height maybe explained by the fact that the hockey players were on average taller than the Argoref sample, but since trochanterion height, which represents leg length, was also greater, we conclude that there might be a slight difference in measurement technique (errors in sitting height measurement are not uncommon, according to Carr et al14) or positioning of the wall-mounted sitting height stadiometer, or accurate assessment of seat height. The older Argoref sample had more weight, endomorphy and mesomorphy than the young field hockey players.

This sample of elite youth field hockey players is similar in height and weight to the Spanish B and Junior sides15, shorter than Dutch 18-19 year old elite players19 and lighter than senior samples like the South African players surveyed by Scott3 and the Argentine Olympic team measured in Montreal 1976 by Carter16, as would be expected when comparing this players with older samples.

The lack of differences in physique between the selected and non-selected players might stem from the fact that the whole group was already an elite different from the rest of the country's players, where again, factors other than physique determine the outcome of the selection process at this level. As before, the possibility of a Type II error is real, with the selected players being slightly taller (p=0.30), heavier (p=0.18), and less mesomorphic (p=0.19), as shown in Table 4.

Table 4. Anthropometric differences between selected and non-selected playersa

a Goalkeepers excluded.

The final selection of players had birth dates that were skewed towards the second half of the year, with eight out of the 17 players being born in the third quarter of the year (Figure 1). Nevertheless, considering the low number of cases in each birth-quarter strata, there were no statistically significant differences between birth-quarter of the selected versus the non-selected players (p=0.11) nor is there an asymmetry in birth dates for the whole sample (p=0.80). This is in contrast to what usually is reported in other sports such as youth soccer across Europe8, where there is an over-representation of those born in the first quarter of the selection year, and in professional basketball in Spain18. The reason why no relative age effect was found in this elite youth field hockey sample is not clear, and merits further research. We can think of a few answers: (a) perhaps the smaller numbers of players of this sport make selection processes not as competitive as in soccer; or (b) that those who did not succeed in soccer turned to other sports like field hockey; or (c) that the nature and competitive schedule of this sport does not place a very high emphasis on the physical and psychological advantages of early maturers. Whatever the reason, this fact is a welcome occurrence for the fairness in competition in youth field hockey.

Figure 1. Birth-quarter in selected and non-selected players.

These players had been playing field hockey an average of 11.2±2.6years, and there were no differences between the selected and non-selected groups (p=0.11). In these players, who on average had been playing for more than ten years, sporting age or experience was not an important factor in being selected to make the final team. This time period is surely ample enough to attenuate any advantages from extra time practicing the sport.

We have described the anthropometric profile of the champion junior field hockey players, who are very similar in structural characteristics to a normal reference adult sample, albeit leaner. Within the pre-selected group, those who were selected to make up the final team did not differ anthropometrically nor in playing history from those who were not selected. It was interesting to find the absence of a relative-age effect in this group, and we do not fully understand the reasons for this, an area which merits further research.

Authors declare that they don’t have any conflict of interests.

Acknowledgements

Federación Argentina de Hockey.

Coach Jorge Lombi.

Dr. Lindsay Carter.

Received 22 November 2010

Accepted 22 February 2011

Corresponding author. fholway@hotmail.com