*Article* **The Influence of Cultural Experiences on the Associations between Socio-Economic Status and Motor Performance as Well as Body Fat Percentage of Grade One Learners in Cape Town, South Africa**

**Eileen Africa <sup>1</sup> , Odelia Van Stryp <sup>1</sup> and Martin Musálek 2,\***


**Citation:** Africa, E.; Stryp, O.V.; Musálek, M. The Influence of Cultural Experiences on the Associations between Socio-Economic Status and Motor Performance as Well as Body Fat Percentage of Grade One Learners in Cape Town, South Africa. *Int. J. Environ. Res. Public Health* **2022**, *19*, 121. https://doi.org/10.3390/ ijerph19010121

Academic Editors: Francesco Campa and Gianpiero Greco

Received: 30 November 2021 Accepted: 17 December 2021 Published: 23 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Abstract:** Fundamental movement skills (FMS), physical fitness (PF) and body fat percentage (BF%) are significantly related to socio-economic status (SES). However, it remains unclear why previous studies have had different findings regarding the direction of the association between SES and FMS, PF and BF%. A suggested explanation is that the direction of the link can be influenced by cultural experiences and traditions. Therefore, the aim of the current study was to investigate links between SES and FMS, PF, BF% of Grade One learners from two different ethno-geographic areas in Cape Town, South Africa. Grade One children (*n* = 191) (*n* = 106 boys and *n* = 85 girls; age (6.7 ± 0.33)) from different socio-economic areas in Cape Town, South Africa, were selected to participate in the study. South African schools are classified into five different quintiles (1 = poorest and 5 = least poor public schools). For this study, two schools were selected, one from quintile 2 and the other from quintile 5. BF% was assessed according to Slaughter's equation. FMS were measured using the Gross Motor Development Test-2 (TGMD-2) and PF via five tests: 1. dynamic strength of lower limb (broad jump); 2. dynamic strength of upper limb and trunk (throwing a tennis ball); 3. speed agility (4 × 10 m shuttle running); 4. cardiorespiratory fitness (20 m shuttle run endurance test (Leger test)) and 5. flexibility (sit and reach test). An analysis of covariance (ANCOVA) found that BF% and WHtR were significantly greater in children with higher SES (Z = 6.04 *p* < 0.001; Hedg = 0.54), (Z = 3.89 *p* < 0.001; Hedg = 0.32). Children with lower SES achieved significantly better TGMD-2 standard scores in the locomotor subtest, compared to their peers with higher SES. In the object control subtest, no significant SES-related difference was found. However, ANCOVA showed that girls performed better in FMS than boys. In PF, the main effect of SES was observed in dynamic strength of trunk and upper limb (throwing) and flexibility, where children with lower SES performed significantly better. No significant difference was found in cardiorespiratory performance (CRP) (Beep test), even though children with lower SES achieved better results. Results from the current study suggest that links between SES, PF, FMS and body fat percentage in children seem to be dependent on cultural and traditional experiences. These experiences should therefore be included as an important factor for the development of programmes and interventions to enhance children's lifelong motor behaviour and health strategies.

**Keywords:** fundamental movement skills; physical fitness; adiposity; children; cultural experiences; socio-economic status

#### **1. Introduction**

The development of fundamental movement skills (FMS) and health-related physical fitness (PF) during childhood presents important health parameters [1] for promoting long-term positive and sustainable health trajectories, especially in obesity prevention [2].

31

In addition, a large body of studies have verified that children with lower PF or FMS, regardless of geographic specification, tend to be overweight or obese [3–7], even at preschool age [8,9]. The aforementioned health trajectories are expounded in the Stodden model [10], which explains that reciprocal relations between motor competence, FMS, health-related physical fitness, self-perceiving of motor competences and physical activity play a key role. However, the Stodden model does not consider the socioeconomic status (SES) of individuals, although SES has been found to be an important indicator in the prevalence of obesity as well as in PF and FMS development [11].

Previous research that focused on the relationship between obesity and SES in schoolage children has not brought clear results. While studies that include children from North America, Australia and Europe suggest that lower SES is significantly associated with a risk of overweightness and obesity [9,12,13], other studies from Brazil or Korea did not find any such relationship [14,15]. Moreover, research done in South America [16], the Arab world [17], and particularly in Africa [11,18,19] showed reversed patterns. It means that children with high SES displayed a higher chance of being overweight and obese. One of the suggested explanations for this research disconformity is that obesity is strongly associated with globalization, i.e., expansion of economic and social interdependence [19,20], which is generally not the same in different world regions.

When we look at the association between motor performance and SES, one of the basic assumptions is that children with lower SES tend to have motor developmental delays [21–23]. However, as noted in previous paragraph, the association between SES and particularly FMS levels seems to be culturally dependent. While in well-developed or Western countries a positive association between SES and FMS was found from pre-school age [24–26], in developing or middle developed countries, including South African (SA) children, the results are not clear [27], since those with lower SES usually performed better in FMS than their peers with higher SES [28]. In addition, Armstrong [29] concluded that in SA children an inverse relation between SES and FMS was observed only for kicking the ball, which is a specific variable performed within the manipulation construct. Moreover, it is important to realize that FMS development is evidently sex-dependent during pre-school and school age. Girls usually have better locomotor skills, while boys perform better in object control [24,30,31], regardless of income status of country [32]. Pienaar et al. [5] found the same pattern in SA first grade children recruited from households with low-to-middle SES. Interestingly, some studies on pre-school [33] and school children [34] found that FMS level is rather associated with SES in girls; however, when considering age, the relation is reversed.

PF as physical readiness showed itself to be also sex-dependent when, regardless of SES, boys outperformed girls in the majority of PF capacity tests (strength, endurance, speed) from preschool age, while girls achieved better results in flexibility [35–37]. However, results from studies investigating the association between PF and SES are inconsistent and copy the trends of results found in the association between SES and FMS. While the majority of previous results suggest that children with higher SES tend to perform better in PF compared to peers with lower SES [29,38,39], some studies did not find any significant association [40]. Moreover, [41] or Freitas et al. [42], for instance, showed that children with higher SES performed better in speed and strength but significantly worse in flexibility and endurance compared to children from lower SES schools.

Previous findings have indicated that the currently accepted relationships between obesity, FMS and PF, for example within the Stodden model, can be strongly influenced by SES, but not always in the same direction and not similarly in both sexes. If we consider obesity prevalence, FMS and PF level from preschool age to be significantly inversely related to the amount and intensity of physical activity [43,44], then sociocultural factors, as suggested by [45], such as the opportunities for leisure time physical activities, personal family support or physical transportation habits [46], might explain why SES is differently associated with motor development or obesity prevalence in school children in culturally different countries. Therefore, the researchers deemed it essential to highlight

the importance of SES in the interpretation of the links between motor development and the prevalence of obesity in children. This information might fundamentally contribute to adequate long-term motor and overall development of children.

The aim of the current study was therefore to investigate the differences in FMS, PF and body fat percentage (BF%) between Grade One learners from different ethno-geographic backgrounds in Cape Town, South Africa.

#### **2. Materials and Methods**

#### *2.1. Participants*

The measurements took place in the first term of the school year. Two Grade One classes from schools in Cape Town, South Africa, were selected to participate in the study. The two schools were based in different socio-economic areas and categorized under different quintiles. All South African schools are classified in five different quintiles based on financial resources (Table 1).

**Table 1.** Background to South African quintile system.


Quintile 1 schools are the poorest, while quintile 5 schools are the least poor [47]. In the current study, school B is categorized under quintile 5 and school W quintile 2. Prior to the data collection, ethical approval was obtained from the Research Ethics committee (#8456) and the study was conducted according to the guidelines of the Declaration of Helsinki. Permission was sought from the Education Department to approach the schools. Written consent from the parents/guardians and assent from the children were obtained for participation in the study. A total of *n* = 191 (*n* = 106 boys and *n* = 85 girls; 6.1 ± 0.4 years old) Grade One children participated in the study.

According to Table 2, all age categories were normally distributed. As reported by the two-way ANOVA, children from school "W" were significantly younger than children from school "B" (F 137 1 = 8.71 *p* = 0.004; η <sup>2</sup>ρ = 0.07). The results also showed no significant difference between girls and boys.



Data are presented as mean ± SD.

#### *2.2. Instruments*

a. Anthropometric measurements:

All anthropometry parameters were measured by one trained examiner using the Eston and Reilly [48] manual. Body height was measured using a portable anthropometer P375 (Co. TRYSTOM, spol. s r.o./1993–2015 www.trystom.cz, accessed on 23 November 2021), with measurements taken to the nearest 0.1 cm. Body weight was measured using a medical calibrated scale TPLZ1T46CLNDBI300, with body weight recorded to the nearest 0.1 kg. Skinfolds were done using subscapular and triceps with the Harpenden skinfold caliper, with an accuracy of 0.2 mm. Waist circumference was measured with metal measuring tape with an accuracy of 0.1 cm.

Body fat percentage was estimated according to [49]'s equations. Previous studies showed that Slaughter's equation is adequate alternative used in children for estimating percentage of body fat where Dual-energy X-ray absorptiometry (DXA) is not available [50–54].

The following equations were used [49]:

For white males with a sum of skinfolds less than 35 mm the following equation was used:

BF% = 1.21 × (tric + subsc) − 0.008 × (tric + subsc)<sup>2</sup> − 1.7

For black males with a sum of skinfolds less than 35 mm the following equation was used:

BF% = 1.21 × (tric + subsc) − 0.008 × (tric + subsc)<sup>2</sup> − 3.2

For all females with a sum of skinfolds less than 35 mm the following equation was used:

BF% = 1.33 × (tric + subsc) − 0.013 × (tric + subsc)<sup>2</sup> − 2.5

For males with a sum of skinfolds higher than 35 mm the following equation was used:

BF% = 0.783 × (tric + subsc) + 1.6

For females with a sum of skinfolds higher than 35 mm the following equation was used:

$$\text{BF\%} = 0.546 \times \text{(tric} + \text{subsc}) + 9.7$$

Note: tric: triceps skinfold; subsc: subscapular skinfold

The waist-to-height ratio index was used as an indirect parameter for estimating abdominal fat. Several studies have indicated that WHtR is useful in clinical and population health as it identifies children with excessive body fat [55] and greater risk of developing weight-related cardiovascular disease at an early age [56]. The waist circumference was measured at the midway between the lowest border of the rib cage and the upper iliac crest to the nearest 0.1 cm [57]. Anthropometric measurements were conducted before lunch time.

b. Fundamental movement skills:

Fundamental movement skills were evaluated with the Test for Gross Motor Development-2 (TGMD-2) [58], which is a valid and reliable measurement of FMS [59–64]. The TGMD-2 assesses proficiency in two motor-area composites (Table 3):



Inter-rater reliability for the TGMD-2 was ensured by two experienced \*Kinderkineticists; both received the same videos of 10 children and had to score them according to the TGMD-2 criteria and manual.

The testing took place in the specific school's hall. A clear demonstration of every skill was given by the assistant at the station. Children had one practice trial and two formal test trials. The formal testing trials were video recorded (consent was given) in order to properly score each participant afterwards according to the criteria of the TGMD-2 manual. The raw scores were converted to standard scores considering the sex and age of participants. Each child received a number for all the measurements and the number was shown on the video.

\*Kinderkinetics is a profession that aims to develop and enhance the total well-being of children between 0–12 years of age, by stimulating, rectifying and promoting age-specific motor and physical development [65].

After the testing, the videos were transferred from the tablets to a memory stick and analysed on a computer by the researcher and assistants.

#### c. Physical fitness:

Physical fitness was measured using five widely accepted tests [66–68], namely broad jump for dynamic strength of lower limbs; throwing a tennis ball for dynamic strength of upper limb and trunk; 4 × 10 m shuttle running for speed agility; 20 m shuttle run endurance test (Leger test) for cardiorespiratory fitness; and a sit and reach test for flexibility. The examiners explained and demonstrated all PF tests to the children before the tests commenced. Detailed descriptions of the PF tests used are available at: (https://ftvs.cuni.cz/FTVS-726-version1-physical\_fitness\_tests\_description.pdf, accessed on 23 November 2021).

#### *2.3. Statistical Analysis*

Normality of data was analysed using the Shapiro–Wilk test as well as coefficients of skewness and kurtosis. Variance–covariance homogeneity was verified using the Box M test, and the regression slopes homogeneity via the significance of between-subjects effects [69]. To accommodate age differences between children with different SES, an analysis of covariance 2 (SES) × 2 (sex) using age as covariate was applied. ANCOVA was used for selected variables, which passed all assumptions for its application (height, weight, skeletal robustness and physical fitness tests).

The effect size was estimated by the partial eta squared (η <sup>2</sup>ρ) with range <0.05 small effect size; 0.06–0.25 moderate effect size; 0.26–0.50 large effect size; >0.50 very large effect size [70,71] and Hedge's g range <0.2 small effect size; 0.21–0.50 moderate effect size; 0.51–0.80 large effect size; >0.80 very large effect size. All data was analysed in NCSS2007 [72].

#### **3. Results**

#### *3.1. Anthropometry*

Since age is significantly correlated with personal height, the current study used analysis of covariance (ANCOVA) (r = 0.47), where age was determined as a covariate. Although children with lower SES were significantly younger, the difference in height in relation to SES between children with lower SES and higher SES remained significant. It means that even though the age of participants was significantly related to height (F 137 1 = 25.63 *p* < 0.001; η <sup>2</sup>ρ = 0.20), children with lower SES were still significantly shorter compared to their peers with higher SES (F 137 1 = 40.23 *p* < 0.001; η <sup>2</sup>ρ =0.30). Furthermore, weight was poorly correlated with age; therefore, a simple two-way ANOVA was performed, which showed that children with lower SES were significantly lighter (F 137 1 = 39.74 *p* < 0.001; η <sup>2</sup>ρ = 0.30). No significant differences were found for height and weight between boys and girls. Body fat percentage and WHtR were not normally distributed (Table 4. BF% was found to be significantly greater in children with higher SES (Z = 6.04 *p* < 0.001; Hedg = 0.54). In addition, girls had a greater BF% compared to boys (Z = 4.41 *p* < 0.001; Hedg = 0.38). The same differences were found in WHtR, where children with higher SES had significantly greater values (Z = 3.89 *p* < 0.001; Hedg = 0.32). In contrast to BF%, no significant differences were found between boys and girls.


**Table 4.** Descriptive height and weight frame indices.

Data are presented as mean ± SD, \*\*\* *p* < 0.001. (a) Significant difference between boys with lower and boys with higher SES. (b) Significant difference between girls with lower and girls with higher SES.

#### *3.2. Fundamental Movement Skills*

In general, children with lower SES achieved significantly better TGMD-2 standard scores compared to their peers with higher SES (F 137 1 = 6.73 *p* = 0.01; η <sup>2</sup>ρ = 0.05). Detailed analysis, however, revealed the effect of SES only in the locomotor subtest, where children with lower SES achieved significantly better scores (F 137 1 = 6.11 *p* = 0.014; η <sup>2</sup>ρ = 0.05). In the object control subtest, no significant difference was found between children with higher and lower SES; however, ANCOVA showed that girls performed better than boys (F 137 1 = 20.78 *p* < 0.001; η <sup>2</sup>ρ = 0.16) (Table 5).


Data are presented as mean ± SD \* *p* < 0.05 \*\* *p* < 0.01 \*\*\* *p* < 0.001. (a) Significant difference between boys with lower and boys with higher SES. (b) Significant difference between girls with lower and girls with higher SES.

#### *3.3. Physical Fitness*

Not all results from the physical fitness tests were significantly related to SES. The effect of SES on muscular strength of trunk and upper limb—throwing (right) (F 137 1 = 24.64 *p* < 0.001; η <sup>2</sup>ρ = 0.18), throwing (left) (F 137 1 = 4.68 *p* = 0.03; η <sup>2</sup>ρ = 0.04) and flexibility (F 137 1 = 12.37 *p* < 0.001; η <sup>2</sup>ρ = 0.09)—was found mainly in children with lower SES. In dynamic strength of lower limbs (broad jump) and agility (shuttle run—4 × 10 m), sex was found to have an effect, but not SES. In both tests, girls achieved lower dynamic strength of lower limbs (F 137 1 = 10.04 *p* = 0.002; η <sup>2</sup>ρ = 0.08) and were significantly slower compared to boys (F 137 1 = 8.16 *p* = 0.005; η <sup>2</sup>ρ = 0.06). No significant difference was found in cardiorespiratory performance (CRP) (Beep test), even though children with lower SES achieved better results (Table 6).


**Table 6.** Physical fitness performance considering SES and sex of children.

Data are presented as mean ± SD \*\* *p* < 0.01 \*\*\* *p* < 0.001. (a): Significant difference between boys with lower and boys with higher SES. (b): Significant difference between girls with lower and girls with higher SES. (c) Significant difference considering sex regardless of SES.

#### **4. Discussion**

The aim of the current study was to investigate the differences in FMS, PF and BF% between Grade One learners from different socio-economic backgrounds in Cape Town, South Africa. After controlling for differences in sex and age, SES was positively associated with height and weight.

Children with higher SES had significantly higher BF% and were heavier. Similar results brought [73] to the conclusion that overweight and obese children in China are from a higher SES. Possible reasons include available amounts of food, less physical activity and a more sedentary lifestyle in children with higher SES. These findings are contrary to the results of [74,75], who found that weight and body mass index in relation to obesity of British children with lower SES were higher compared to their peers with higher SES. In addition, higher weight and body fat are considered as a sign of wealth in certain countries [76,77]. For instance, children with high SES in Sub-Saharan Africa also displayed a higher chance of being overweight and obese [78]. These findings contradict previous studies [9,12,13] which found an inverse association between SES and body fat. A multiethnic study [79] found that obese African black girls had the highest self-esteem compared to Asian or European peers. Specifically, overweight South African black women perceive themselves as more attractive [80,81]. A very recent qualitative study [82] revealed in South African adult participants that fatness is connected with symbols of prosperity and beauty rather than with health problems. A different view of body status is also known from other cultures such as China, where being too thin is the same problem as being too fat [83]. This suggests that socio-cultural environments including ethnicity or race can link SES to weight gain and obesity status differently, as proposed by [84].

#### *4.1. Fundamental Movement Skills*

The results of this study indicate that children with lower SES performed significantly better than their higher-SES peers according to the standard scores of total TGMD-2 and the locomotor subtest of the TGMD-2, but not in the object control subtest. This finding is in contrast with most previous research from the Western world, where children with higher SES outperformed their lower-SES counterparts in FMS [85,86]. For instance, [34], who performed their study in Australia, and [86], who performed theirs in the United States, suggested that the differences could be attributed to lower cardiorespiratory fitness, physical activity levels, absence of weekly physical education, fewer opportunities for perceptual motor experiences and disadvantaged communities that lack facilities. Nevertheless, our study found no significant difference in CRF in relation to SES (see in detail below PF part). Furthermore, our findings are consistent with a recent South African study by [87], which was carried out in a very similar demographic environment and which also stated that rural low-income children had significantly better TGMD-2 standard scores. This negates the notion that children with lower SES naturally perform worse in overall FMS than children with higher SES due to limited access to safe outdoor playing and

equipment [88] or safe places to be active in the community or to sporting equipment at home [89]. A possible reason for this finding is that children with lower SES often engage in unstructured moderate-to-vigorous physical activity with limited teacher facilitation compared to their higher-SES peers, and this might positively influence the development of FMS [87,90,91]. Children with lower SES in South Africa also tend to spend a greater amount of time in active transportation to and from schools [92,93]. Therefore, according to some authors, different findings in terms of FMS levels in South Africa compared to western, educated, industrialized, rich and democratic (WEIRD) countries can be attributed to South Africa's unique socio-cultural environment [94,95].

#### *4.2. Physical Fitness*

In the physical fitness measurement, only the performances in upper limb throwing and flexibility were significantly inversely associated with SES, where children with lower SES achieved significantly better results. These findings support those of [96], who suggested that the relationship between PF and SES has not been consistently clarified.

Some studies, however, did observe a significant positive link between SES and aspects of PF (muscle strength, aerobic fitness, muscular endurance and speed) [97–99]. In contrast, [99] did not find any provable associations between SES and PF, and other studies [39,100] even found that children with lower SES outperformed their peers with higher SES in flexibility and endurance. The results of the current study are more consistent with the conclusions of the latter.

Inverse associations between SES and throwing could be explained by differences in opportunities and content of physical activity [29]. It has been known for more than 70 years that the way children with different SES spend their leisure time depends on their SES environment [101–103]. Children with higher SES usually spend leisure time participating in organized commercial physical activities [104], while children with lower SES tend to play simple group games with cheap equipment in the street [105,106]. In addition, this spontaneous type of PA usually has implicit motor learning characteristics [107], where a high number of repetitions of motor activity without explicit instructions is considered typical. Implicit motor learning for acquiring motor skills such as overhead throwing has been shown to significantly influence automation and accuracy of the movement pattern [108–110]. Therefore, the range of the movement experience and the defined motor pattern, along with how this motor pattern (overhead throwing) was acquired in low SES children, could explain the inverse association between SES and performance in throwing a tennis ball found in the current study. This assumption would also support previous suggestions that children with lower SES seem to have better coordination [97,111–113]. On the other hand, the results of the current study do not support the South African study conducted by [29], who did not find any significant differences in the throwing of a cricket ball in 6–7-year-old children when taking SES into consideration. That study included more than 600 children from five provinces in that age category. Since the participants of the current study were only from one province, sample variability could be a reason for the discrepancy in the results.

In the flexibility measurement, children with lower SES showed better results. These are in line with the findings of [29,39]. This difference could be explained by genetic differences in collagen alleles associated with physical performance/functional tests [114], even though relationships between genotypes and clinical phenotypes are not well defined. Chan et al. [115] found that in African Americans, collagen development COL1A1 COL1A2 responsible for development of bone, cartilage and tendons seemed to be evolutionarily different from European Americans, increasing flexibility in the African American population.

If one looks at differences in performance of each component of PF, it is evident in the research that children with higher SES are stronger and have better muscular explosiveness [97,111] which is in line with the current study's findings.

Results from the cardiorespiratory fitness (CRF) test showed that children with lower SES performed better in the multistage fitness test. However, due to the large variability of

results in each category of children (higher SES, lower SES, boys and girls), the differences were not significant. Nevertheless, studies from Sub-Saharan Africa found better CRF in children with lower SES compared to their higher-SES counterparts [42,105,116]. The better CRF of children with lower SES could be explained by their daily habits and physical activity profile compared to children with higher SES. Prista et al. [117] found that increased physical activity of children with lower SES was mainly due to higher demands of daily physical activities, such as walking, running and playing. VandenDriessche et al. [96] and Micklesfield et al. [103] found that children with lower SES walked to school and engaged in more physical activity on the way to and from school. On the other hand, these children spend less time in moderate-to-vigorous physical activity at school and in clubs [105]. Furthermore, Micklesfield et al. [103] suggested that children with lower SES spend more active time at the household and community level, which implies less sedentary behaviour in this social environment. These findings seem to be dependent on the social and cultural environment because the result of better PF in children with lower SES is contrary to results from studies done in the Western world, where children with lower SES performed repeatedly worse in CRF tests [99,112].

In summary, our findings should be used for the development of further education strategies with the aim of preventing obesity and properly controlling child's motor development and physical fitness, which are influenced by SES differently considering specifics of socio-cultural and ethno-graphic experiences. It means, for instance, the extension of PE classes at least in primary school education, along with changes in content or implementation of active breaks or socialization games. This seems, according to [118], to be positively influenced by the conjunction of school and family environments intervention programs.

#### *4.3. Strength and Limitations*

To the best of our knowledge, this is the first study of its kind to consider traditional and cultural experiences (including ethnographic differences) as an important factor influencing the direction of the links between SES, motor performance and body fat percentage in children. An additional strength of this study is that the sample (two different socioeconomic/ethno-graphic groups) was specifically defined and selected according to the guidelines (www.education.gov.za, accessed on 16 December 2021) stipulated by the Department of Basic Education in South Africa. However, the research sample was selected from a narrow population in the Western Cape; therefore, the results of this study may not be completely representative of all Grade One children in the Western Cape. Furthermore, the absence of biological maturation status of the children in this study might be a limitation because previous studies have suggested that biological maturation influences performance in strength and endurance [119,120]. However, most previous studies did not consider this parameter in similar age samples. In the current study children with higher SES were significantly taller and heavier, which suggests that they might be advanced in their biological maturation. Unfortunately, there is no valid and reliable method in South Africa to assess biological maturation for this age group in multi-ethnic populations. We suggest that future research explore the inclusion of biological maturation when assessing motor performance and BF% in children with different SES.

#### **5. Conclusions**

In contrast to Western countries, children with lower SES in the current study were leaner, had lower BF% and performed significantly better in FMS (specifically in their locomotor skills) compared to their higher SES peers. Furthermore, children with lower SES performed significantly better in dynamic strength of the trunk and upper limb and flexibility compared to children with higher SES. Therefore, we suggest that links between SES, PF, FMS and BF% in children seem to be dependent on country-specific cultural and ethno-graphic experiences. The uniqueness of cultural experiences with regard to SES should be included as an important factor for the development of programmes and interventions to enhance lifelong motor behaviour and health strategies for children.

**Author Contributions:** Conceptualization, O.V.S., E.A. and M.M.; data curation, O.V.S. and M.M.; investigation, O.V.S. and M.M.; methodology, M.M.; project administration, E.A.; writing—original draft, O.V.S., E.A. and M.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was endorsed by the project PROGRES Q19, Social-Sciences Aspects of Human Movement Studies II and financially supported by the institutions involved.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Research Ethics Committee (Humanities) of Stellenbosch University # 8456, 22 February 2019.

**Informed Consent Statement:** Written informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data set is available at https://www.researchgate.net/publications/ create?publicationType=dataset, accessed on 17 December 2021.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Somatotype Profiles of Montenegrin Karatekas: An Observational Study**

**Jelena Slankamenac <sup>1</sup> , Dusko Bjelica <sup>2</sup> , Damjan Jaksic <sup>1</sup> , Tatjana Trivic <sup>1</sup> , Miodrag Drapsin <sup>3</sup> , Sandra Vujkov <sup>4</sup> , Toni Modric <sup>5</sup> , Zoran Milosevic <sup>1</sup> and Patrik Drid 1,\***


**Keywords:** karate; kumite; weight categories; anthropometry; body composition; martial arts; combat sports

#### **1. Introduction**

The origin of karate remains hidden by opaque veil legends, but we still know that karate originates from the Far East, and it was widely practiced by the people who were followers of such different religions as Buddhism, Islam, Hinduism, and Taoism. It was first developed in Okinawa, Japan, in the 17th century when the Japanese took this island and prohibited the usage of all weapons [1]. It gained popularity after the Second World War. Karate is one of the most popular and widely practiced martial arts of today, and only in 2021 (Tokyo, Japan) had its appearance in the Olympic games. It is characterized by two

**Citation:** Slankamenac, J.; Bjelica, D.; Jaksic, D.; Trivic, T.; Drapsin, M.; Vujkov, S.; Modric, T.; Milosevic, Z.; Drid, P. Somatotype Profiles of Montenegrin Karatekas: An Observational Study. *Int. J. Environ. Res. Public Health* **2021**, *18*, 12914. https://doi.org/10.3390/ijerph 182412914

Academic Editors: Francesco Campa and Gianpiero Greco

Received: 21 October 2021 Accepted: 30 November 2021 Published: 7 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

distinguished competitive disciplines: *Kata* and *Kumite* (sports fight). *Kata* means form, and it is a predetermined series of offensive and defensive techniques and movements in standard order, versus one or more nonexistent opponents. Fundamental elements of the *Kata* technique involve rhythm, expressiveness, and *Kime* (a short isometric muscle contraction performed when a technique is finished) [2]. Karatekas that outreach the final are obligated to perform one *Tokui* (free-style *Kata*) and one *Shitei* (fixed *Kata* styles). Athletes have 60–80 s to complete the *Kata* [3]. *Kumite*, on the other hand, represents combat between two karate athletes under certain rules. Strikes are limited to determining areas: face, head, neck, chest, abdomen, side, and back. The duration of the *Kumite* match is 3 min for male and female senior athletes [3]. Judges score kicks and punches—*Ippon* (3 points), *Waza-ari* (2 points), and *Yuko* (1 point). Points are awarded when a technique is executed according to the following principle: good form, vigorous application, sporting attitude, awareness, correct distance, and good timing. *Kumite* competitors are divided into five weight categories for both males and females (<60 kg, <67 kg, <75 kg, <84 kg and >84 kg for males and <50 kg, <55 kg, <61 kg, <68 kg and >68 kg for females). Weight categories in karate and other combat sports can ensure fair competition by complementary opponents of similar body mass and stature [4].

One of the oldest questions in every sport is "what actually makes a successful athlete successful?" Morphological features play an important role in accomplishments in most sports. Body form provides a foundation for the improvement of movement technique and particular physical fitness. When selecting athletes in a particular sport, it is observed whether their physical characteristics fit with a "model" somatic pattern for that sport. That model is based on somatic patterns recorded in athletes who have systematically achieved the best results. Assessment of body composition consists of an assessment of the somatotype, which is based on the relationship between body fat and the lean body content, muscular development, skeleton robustness, and reciprocal ponderal index (height divided by the cube root of body weight) [5–8]. The most commonly used technique in somatotype assessment is the Heath and Carter method [9]. They emphasize that the somatotype is defined as representing the individual's present morphological conformation. Heath-Carter method is used primarily in its anthropometric form in practice, and it is best suited for sports science. Anthropometric measurements are objective and can show body shape, composition, and proportionality. The somatotype consists of three main components in relation to body height: endomorphy, mesomorphy, and ectomorphy [10]. Endomorphy is the first component, and it represents relative fatness or leanness. The second component is mesomorphy and this shows relative musculoskeletal development adjusted for height. Ectomorphy, the third component, is the relative linearity of the build [5]. The knowledge of these characteristics is most informative for coaches and athletes.

Very often, the physical structure is considered as one of the elements for high performance in many sports, as well as in competitive karate [11,12]. In karate, empirical experience states that the athlete's body height and longitudinal dimensions, such as arm and leg length, are some of the main advantages of karate athletes because these measures allow karatekas to raise their legs higher during the kick and they can fight from greater distances [13]. Comparing karate athletes with the general population, they are distinguished by muscular mass with enhanced transverse skeleton dimensionality and reduced adipose tissue. It is known that the body composition of athletes has a great impact on achieving top sports results. Up to date, several studies have dealt with somatotypes in male karate athletes [14–16]. However, there is a lack of evidence regarding female karate somatotype. With this in mind, anthropometric parameters were measured, and the somatotypes of both male and female Montenegrin karatekas were determined.

This study aimed to determine whether there are differences between karate athletes in five male and five female weight categories in different anthropometric measurements and to determine the somatotype profiles of athletes. The results of this study should provide a more specific outline of the morphological biotype best suited to the specific technical requirements for *Kumite* athletes of both genders.

#### **2. Materials and Methods**

#### *2.1. Subjects*

A total of 60 senior karate athletes from Montenegro participated in the National Championships in 2020. For the purpose of this study, we have chosen 51 karate athletes (black belt). According to the calculation, considering that the five weight categories are analyzed, the total sample size should be much larger. However, in this specific case, the total population is 60 competitors, so the classical formula cannot be applied. A cohort of 27 male (21.9 ± 4.7years) and 24 female karate athletes (20.3 ± 3.14 years) of a national level volunteered in this cross-sectional study. The subject sample included healthy, black belt karate senior athletes, with no prior injuries, minimum five year training experience and overall weekly training volume of over 20 h. Measurements were taken in April 2020. All testing procedures were conducted during the karate camp ahead of the National Championship held in Nikši´c (Montenegro). Participants were divided into five official male categories <60 kg (*n* = 5), <67 kg (*n* = 8), <75 kg (*n* = 6), <84 kg (*n* = 4), and >84 kg (*n* = 4) and five female weight categories <50 kg (*n* = 2), <55 kg (*n* = 7), <61 kg (*n* = 7), <68 kg (*n* = 6), and >68 kg (*n* = 2) in accordance with their current body mass, age and gender [3]. All athletes were introduced to all of the testing procedures applied in the current research. All anthropometrical measurements were taken from the participants in the same position, in the morning hours (before breakfast), by the same two experienced graduated students of the Faculty for Sport and Physical Education, University of Montenegro. Informed written consent was acquired from each subject, and all procedures were executed and conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of the Faculty of Sport and Physical Education University of Novi Sad, Serbia (Ref. No. 46-06-02/2020-1).

#### *2.2. Anthropometrical Measurements*

In order to determine somatotypes, ten required measurements were taken as follows: body height and body mass, bi-epicondylar breadths of humerus and femur, four skinfold measurements (triceps, supraspinal, subscapular, and medial calf), and two girths (arm and calf). Body height (cm) was determined using a Martin anthropometer (GPM, Bachenbülach, Switzerland); body mass (kg) was measured with an electronic scale (SECA, Hamburg, Germany) with a sensitivity level of 0.1 kg; skinfolds were taken on the right side of the body using a John Bull caliper (British Indicator Ltd., Weybridge, UK), accurate to 0.2 mm; circumference measurements (cm) were obtained with a steel measuring tape, and wrist girth and bi-epicondylar diameters of the femur and humerus (mm) were measured using a small spreading caliper (SiberHegner, Zurich, Switzerland). Somatotypes were determined using the Carter and Heath method [9].

#### *2.3. Statistical Analysis*

The data obtained are presented as standard deviation (±) and means. One-way analysis of variance (ANOVA) and Tukey's post hoc tests was used to compare group the differences by weight categories. Furthermore, the effect size (h2) was calculated. The level of significance was set at *p*-value < 0.05. SPSS statistics software was used to conduct analyses.

#### **3. Results**

The study involved 27 male and 24 female Montenegrin karate athletes. Anthropometric characteristics and somatotype parameters were measured and presented in tables and charts. Both males and females were divided into five weight categories (male: <60 kg, <67 kg, <75 kg, <84 kg and >84 kg; female: <50 kg, <55 kg, <61 kg, <68 kg and >68 kg). Anthropometric parameters increased within the weight category.

Statistically significant differences in male categories were found between the first category (<60 kg) in body height compared to the last three categories (<75 kg, <84 kg, and >84 kg). The highest athletes were in the <84 kg category. There was no significant difference found between groups in breadths of humerus and femur. In term of arm

girths, there were differences between <60 kg, <67 kg and >84 kg. However, a difference between <60 kg and the last three categories (<75 kg, <84 kg, and >84 kg) in terms of calf circumference was found.

Measuring skinfolds, statistically significant differences were shown only between <84 kg and the first three groups (<60 kg, <67 kg, and <75 kg) in supraspinal skinfold. Other differences in skinfolds were not at a significant level (Table 1).


**Table 1.** Differences between weight categories of male karatekas.

Legend: M—Mean; SD—standard deviation; different from: <sup>a</sup>—<60; <sup>b</sup>—<67; <sup>c</sup>—<75; <sup>d</sup>—<84; <sup>e</sup>—>85; significant differences in bold.

Somatotype analysis of male categories found a difference between the <75 kg and <84 kg in endomorphy. In mesomorphy, there is no difference between the categories. Perceiving ectomorphy, there is a significant difference between the first category and the >84 kg. All male subjects were endomorphic mesomorph, except for category <84 kg, which was endomorphic ectomorphs (Figure 1).

**Figure 1.** Somatochart of male karate athletes by weight categories.

In the female groups, body height increased in relation to the weight category and differed significantly between <50 kg, <55 kg, <61 kg, and the heaviest group (>68 kg), and between the first two (<50 kg, and <55 kg) and <68 kg. The breadth of the humerus shows a difference between >68 kg and all of the other groups (<50 kg, <55 kg, <61 kg, <68 kg). The only difference in the breadth of the femur is between the lightest <50 kg) and the heaviest (>68 kg) category. Measuring arm circumference, there is one difference, between the <50 kg and >68 kg categories. Additionally, there is one difference between the groups in the circumference of the calf, between <50 kg and >68 kg. The categories did not differ significantly in terms of the thickness of the skin folds (Table 3).




**Table 3.** Differences between weight categories of female karatekas.

Legend: M—Mean; SD—standard deviation; different from: <sup>a</sup>—<50; <sup>b</sup>—<55; <sup>c</sup>—<61; <sup>d</sup>—<68; <sup>e</sup>—>68; significant differences in bold.

The somatochart showed that the lightest weight category was predominantly endomorphic ectomorphs. Two weight categories were ectomorphic endomorphs (<61 kg and <68 kg), and the other two weight categories were endomorphic mesomorphs (<55 kg and >68 kg). Somatotype differences in the female karate athletes were observed in the ectomorphy components, between <50 kg and <61 kg, and in mesomorphy between <50 kg and >68 kg. (Figure 2).

**Figure 2.** Somatochart of female karate athletes by weight categories.

#### **4. Discussion**

Accomplishment in most sports depends on the physical, physiological, psychological, and social characteristics of the athlete [17]. This study is focused on physical characteristics of karatekas, and determining whether there is a difference in these characteristics between Montenegrin karatekas in different weight categories.

The somatotype profiles of male and female Montenegrin karate athletes were evaluated in relation to different weight categories. Study results have indicated several differences in somatotype for the female group and some anthropometric characteristics throughout weight categories for both female and male groups of observed karate athletes. Somatotypes of male karatekas were mostly homogeneous. The obtained results showed the predominance of endomorphic mesomorphs, except for athletes in the <84 kg category, who were endomorphic ectomorphs. In contrast to our finding, some other recent studies found dominantly mesomorphic somatotypes in male karate athletes [18,19]. A higher mesomorphy component is significant in that increased muscle mass can be considered as an important benefit for athletes facing severe physical confrontation during training and competition, while increased fat mass reflected in endomorphism may prove useful in affecting absorption and dispersing such forces [20,21].

On the other hand, female karate athletes' results showed different types of somatotypes. Profiling female athletes, three different types of somatotypes in relation to weight category were obtained. Female categories show that the lightest weight category was predominantly endomorphic ectomorphs. Two weight categories were ectomorphic endomorphs (<61 kg and <68 kg), and the other two weight categories were endomorphic mesomorphs (<55 kg and >68 kg). This finding is in accordance with other studies that also examined anthropometric characteristics. Fritzsche and Raschka [14] state that the karatekas who practice *Kata* are exhibit more endomorphs characteristics and *Kumite* athletes take more ectomorph positions in somatocharts.

Karate athletes are characterized by a low percentage of fat tissue and a harmonic body constitution. However, different nationalities have different percentages of fat tissue [22]. A review of data from the literature discovered that elite karatekas are ectomorphic mesomorphs with a small amount of adipose tissue [23–25]. Prominent vertical skeletal development among top-level karatekas is the most influential anthropometric feature [23]. Controlling body composition is obligatory to clarify an athlete's best weight category [26].

In the present study, statistically significant differences in male categories were found between the first two categories in body height compared to the last three categories. In the female groups, body height increased concerning the weight category and differed significantly between <50 kg, <55 kg, <61 kg, and the heaviest group, and in between the first two and <68 kg. Gloc et al. [26] obtained the results which proposed that taller karate athletes with a higher percentage of muscular mass had a better outcome. Morphological characteristics also influence specific motor skills in junior karate athletes [27]. Analysis of bone diameters showed no significant differences in male categories, and in female categories, there are differences in the humerus breadth between the heaviest and all of the other groups; femur breadth was different between the lightest and the heaviest weight category. Azary and Izadi [28] stated that elite karatekas have longer lower limbs compared to non-elite athletes, despite their similarity in body height. They imply that Iranian karatekas have a higher skelic index than Italian athletes. Throughout karate sparring, various techniques are executed, and all of them require explosiveness and high speed to perform. The athletes with longer longitudinal dimensions seem to possess a particular superiority for acquiring points before the opponent, and they are able to use longer limb length to get the upper hand facing the opponent in combat [29]. Skinfolds differed significantly between groups, neither in the male nor the female categories, except between <84 kg and the first three groups in the supraspinal skinfold in male categories.

According to Przybylski et al. [30], the most significant qualificator factor for success in performing karate for each gender appears to be well-built strength based on the morphology of the limbs. In the current study, both male and female karatekas in the heaviest weight category differed significantly from the lighter categories in terms of anthropometric values.

One of the study limitations is presented by the relatively small number of athletes per each weight category. More athletes per each weight group could provide more detailed information regarding somatotypes in karate. Further investigation should be aimed on the dominant techniques typically used within categories and acquiring better insight into whether there is difference in the specific techniques that are used in relation to specific physical characteristics. Furthermore, it could be observed, additionally, whether specific techniques applied occur throughout various weight categories.

National level *Kumite* athletes of both genders in all weight categories were categorized by their physical characteristics in somatotypes in this study. Practicing karate seems to produce general morphological adaptation to the training process. Further studies are needed in order to investigate potential long-term adaptation in terms of the experience of athletes (i.e., national vs. international karatekas), as well as differences in somatotypes between *Kata* and *Kumite* athletes for both genders.

#### **5. Conclusions**

The findings of the study regarding somatotypes and anthropometric characteristics throughout various weight categories in karate should provide important information regarding future training processes, testing, as well as for the identification and selection of karate athletes. There are very few differences between karatekas in different weight categories. Differences were found between the heaviest and lighter categories in terms of body height, breadths, and girths in both male and female categories. There were no differences in the thickness of skin folds. Female categories show heterogeneous somatotypes, but the only significant difference was in ectomorphy between <50 kg and <61 kg. Male groups have similar somatotypes. Most of them were endomorphic mesomorphs. Significant differences between males were found in endomorphy (<67 kg and <84 kg) and in ectomorphy (<60 kg and >85 kg). By studying these characteristics, scientists can give specific details on the functional and morphological somatotype best suited for any sport. The present study could be significant for profiling and selecting karate athletes based on gender, age, and weight categories.

**Author Contributions:** Conceptualization, D.B., Z.M. and P.D.; Data curation, J.S., D.J., T.T., M.D. and S.V.; Formal analysis, D.J. and P.D.; Funding acquisition, D.B., and Z.M.; Investigation, J.S., D.J., T.T. and S.V.; Methodology, D.J. and P.D.; Project administration, T.M. and J.S.; Resources, P.D.; Software, D.J.; Supervision, D.B., Z.M. and P.D.; Validation, T.T. and T.M.; Visualization, D.J.; Writing—original draft, J.S., S.V. and P.D.; Writing—review & editing, J.S., D.B., D.J., T.T., M.D., S.V., T.M., Z.M. and P.D. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work has been supported by the Serbian Ministry of Education, Science, and Technological Development (179011) and Provincial Secretariat for Higher Education and Scientific Research (142-451-2094).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of University of Novi Sad, Serbia (Ref. No. 46–06-02/2020–1).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Conflicts of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

#### **References**


## *Article* **Body Size Measurements and Physical Performance of Youth Female Judo Athletes with Differing Menarcheal Status**

**Marina Saldanha da Silva Athayde 1,\* , Rafael Lima Kons <sup>1</sup> , David Hideyoshi Fukuda <sup>2</sup> and Daniele Detanico <sup>1</sup>**


**\*** Correspondence: marinasaldanha.dsa@gmail.com; Tel.: +55-48-3721-8530

**Abstract:** *Purpose:* To compare body size measurements and physical performance among female youth judo athletes with differing menarcheal status and to identify indicators of physical performance in post-menarcheal girls. *Methods:* Nineteen young female judo athletes (age 13.9 ± 2.3 years) were divided into a pre-menarche (*n* = 7) and a post-menarche (*n* = 12) group. The athletes were evaluated through neuromuscular tests, including standing long jump (SLJ), medicine ball throw (MBT), and handgrip strength (HGS), and judo-specific assessments, including the Special Judo Fitness Test (SJFT) and the Judogi Grip Strength Test (JGSTISO). Furthermore, years of experience in judo and the age at menarche were determined. *Results:* The main results showed higher performance for the post-menarche group for most variables (*p* < 0.05) compared to the pre-menarche group. A multiple linear regression analysis demonstrated that age at menarche, chronological age, and body mass explained close to 70% of JGSTISO, while training experience, chronological age, and age at menarche explained close to 59% of SLJ. Additionally, chronological age and age at menarche explained 40% of MBT, and chronological age and height explained 52% of HGS. *Conclusions:* Age at menarche and somatic growth variables explained moderate proportions of the variance of physical performance, thereby providing evidence that these parameters are the primary indicators of physical performance in young female judo athletes.

**Keywords:** somatic maturity; puberty; combat sports; physical performance; young athletes

#### **1. Introduction**

Adolescence corresponds to the transition period between childhood and adulthood, during which several important biological manifestations occur, such as peak height velocity (PHV), peak weight velocity, sexual maturation, and, specifically for girls, menarche [1]. The range of variability in somatic and biological maturation among individuals of the same chronological age is large and is especially pronounced in adolescents [2]. When considering girls during this period, there is increased production of the estrogen hormone, responsible for stimulating growth and breast development [1], which is usually related to the first menstrual period (menarche) following PHV and considered an indication of biological maturation [3].

The current literature contains several studies on the effects of somatic maturity and growth on physical performance in young male athletes from team sports [4–7]. Most recently, studies have investigated the role of growth and maturity status on physical performance in young male judo athletes [8–12]. Years of formal judo training, growth, and somatic maturity can predict physical performance, when generalized upper and lower limb strength assessments (e.g., medicine ball throw test, handgrip strength, and jump tests) [8,10] and judo-specific tests (e.g., Special Judo Fitness Test and Judogi Grip Strength Test) [8] are considered. Moreover, Giudicelli et al. [11] found a positive relationship

**Citation:** Athayde, M.S.d.S.; Kons, R.L.; Fukuda, D.H.; Detanico, D. Body Size Measurements and Physical Performance of Youth Female Judo Athletes with Differing Menarcheal Status. *Int. J. Environ. Res. Public Health* **2021**, *18*, 12829. https:// doi.org/10.3390/ijerph182312829

Academic Editors: Francesco Campa and Gianpiero Greco

Received: 17 November 2021 Accepted: 1 December 2021 Published: 5 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

between maturity status and handgrip strength test as well as aerobic performance in young male judo athletes, even when chronological age and body mass were controlled. Thus, bearing in mind that judo athletes demand high levels of strength-related performance in the upper and lower limbs and both the aerobic and the anaerobic energy pathways [13], it is essential for coaches to understand the role of maturation during adolescence.

Many young female athletes are now involved in high-level judo competitions. For example, 223 female judo athletes from all continents participated in the 2019 World Cadet Judo Championship (under 18, U18), and 223 in the 2019 World Junior Judo Championship (U21) [14]. However, no studies investigating the effects of maturation on physical performance specifically in young female judo athletes have been conducted. It is known that girls present different somatic and physiological characteristics than boys due to variations in the timing and tempo of the maturation process [3,15] and, consequently, performancerelated characteristics. A recent study with basketball players verified that girls with late maturity status (measured by the onset of menarche) tended to have less experience in the sport [16]. This study also found that body mass and adiposity were the highest predictors for all basketball performance indicators.

In individual sports, such as rhythmic gymnastics, Camargo et al. [17] found that there was an increase in the body fat percentage, fat mass, and fat-free mass 2 years after PHV and the occurrence of menarche. Furthermore, Pinto et al. [18] verified that girls in more advanced maturation stages presented higher values of growth indicators (weight and height) and power output of the upper limbs (through the medicine ball test). However, in individual sports of an intermittent nature requiring high-intensity actions, such as judo, these aspects are not yet fully known, in particular in relation to the physical demands of judo athletes, such as generic and judo-specific assessments.

Understanding the role of biological maturity (e.g., menarche), somatic growth measures (e.g., body size), and training experience in physical performance and their contribution in young female judo athletes can help coaches to design appropriate and individual training programs with consideration of biological development during adolescence. Thus, the purpose of this study was to compare body size measurements and physical performance among youth female judo athletes at different menarcheal statuses (pre- and post-menarche) in addition to identifying indicators of physical performance in postmenarche girls. We hypothesized that post-menarcheal girls are advanced in the growth process and consequently present higher physical performance.

#### **2. Methods**

#### *2.1. Participants*

Nineteen young female judo athletes (age 13.9 ± 2.3 years, range 10.9–17.0 years), purple (*n* = 10) and brown belts (*n* = 9), were divided into two groups: pre-menarche (*n* = 7) and post-menarche (*n* = 12). The athletes who participated in the study were from southern Brazil (Santa Catarina) and primarily of Portuguese ethnicity. The sample size was determined a priori using the GPower 3.1 software, taking as references a probability of 0.05 (minimum error type I), statistical power of 0.8 (minimum error type II), and effect size of 0.5 (mean effect). Thus, the minimum sample size was 21 participants. However, we were able to evaluate 19 athletes representing 48.7% of female athletes who participated in the competition in 2021, considering the age range of this study in the local federation (*n* = 39 athletes).

All athletes trained regularly with technical–tactical training occurring 2–3 times per week during the evaluation period, competed at the state and/or national levels, and had been engaged in formal training for at least 2 years. The girls reported no musculoskeletal disorders or injuries that would influence their maximal physical performance during the assessments. They were in the preparatory phase and therefore not in a period of rapid weight loss. All participants and responsible parties (parents and coaches/trainers) received a detailed verbal explanation of the purpose, methods, and potential risks/benefits of the study, followed by the completion of a written informed consent form. This study

was approved by the Research Ethics Committee of the local university, in accordance with the Declaration of Helsinki.

#### *2.2. Design*

The athletes' assessments were performed during two afternoon testing visits separated by 48 h. During the first testing visit, an interview to determine the years of judo experience and age at menarche, anthropometric measurements (body mass and height), and neuromuscular tests (standing long jump test, medicine ball throw test, and handgrip strength test) were conducted. The physical tests were separated by 20 min intervals. At the second testing visit, the young judo athletes were submitted to sport-specific tests, including the Special Judo Fitness Test (SJFT) and the Judogi Grip Isometric Strength Test (JGSTISO), separated by a 30 min recovery interval.

#### *2.3. Determination of Menarcheal Status, Chronological Age, and Training Experience*

Chronological age was calculated to the nearest 0.1 year by subtracting the date of birth from the date of testing. The number of years of formal training in judo was selfreported by the girls and/or their parents. Age at menarche was obtained through an individual interview with the girls by a female researcher; 12 athletes were determined to be post-menarche and 7 athletes pre-menarche. The body mass was measured using a digital scale Toledo® (0.1 kg accuracy), and height was assessed using a stadiometer scale of 0.1 cm accuracy.

#### **3. Generic Tests**

#### *3.1. Standing Long Jump Test*

For the standing long jump test, we followed the protocol used by Detanico et al. [8]. The girls performed the standing long jump test starting from a standing position by swinging their arms and flexing their knees to provide maximal forward drive. Before the assessment, the participants performed a familiarization/warm-up of 5 min of jogging, followed by 30 s of hopping, and 5 submaximal standing long jumps. A take-off line was drawn on the ground, and the measurement of the jump length was determined using a metric tape measure (Lufkin, L716MAGCME; Apex Group, Sparks, MA, USA) from the take-off line to the nearest point of landing contact (i.e., the back of the heels). Each athlete performed three jump attempts with 2 min intervals, and the longest distance was considered for further analysis.

#### *3.2. Medicine Ball Throw Test*

The procedures adopted for the medicine ball throw test assessment followed the protocol of Vossen et al. [19]. The warm-up and familiarization consisted of two–three submaximal throws prior to the test. The girls remained seated on the floor covered with judo mats and were stabilized with their backs supported against a vertical support, their thighs horizontally supported, their knees flexed at an angle of 90◦ , and their ankles resting on the floor. A 3 Kg medicine ball (Dynamax Inc®., Dallas, TX, USA) was positioned at the sternum of each athlete (point A), who then threw it with both hands without moving her trunk. When an athlete failed to maintain the established body orientation, the throw attempt was disregarded. The distance the medicine ball was thrown from point A up to its first contact with the floor (point B) was measured. Each participant performed three throws, with 2 min intervals, and the greatest distance achieved in the three attempts was considered for further analysis.

#### *3.3. Handgrip Strength Test*

Handgrip strength was measured using a handgrip dynamometer (Carci®, SH 5001 model), following the protocol used by Detanico et al. [8]. The warm-up and familiarization consisted of three rapid contractions performed during a 2 s period. The participants were instructed to sustain a maximal isometric contraction during each measurement (lasting

3 to 6 s). The three contractions were performed with the dominant (self-selected) hand with 2 min intervals, in the standing position with shoulder flexion at 90◦ and the elbow completely extended. The highest value obtained in the three trials was considered for further analysis.

#### *3.4. Judo-Specific Tests*

#### 3.4.1. Special Judo Fitness Test

The SJFT is a judo high-intensity intermittent test developed by Sterkowicz [20]. The girls performed a 5 min warm-up before the test, which consisted of jogging, judo falling techniques (*ukemi*), and repetitive throwing techniques (*uchi-komi*). Subsequently, three athletes of similar body mass (with a maximum variation of 10%) and height performed the SJFT, according to the following protocol: two judokas were positioned at a distance of 6 m from each other, while the athlete being tested was positioned 3 m from the judokas to be thrown. The procedure was divided into three periods: 15 s (A), 30 s (B), and 30 s (C), with a 10 s interval between periods. In each period, the athlete being tested threw the other judoka using the one-arm shoulder throw (*ippon-seoi-nage*) technique as many times as possible. Performance was determined by the total throws completed during each of the three periods (SJFTTT = A + B + C). Heart rate (HR) was measured immediately after the test and then 1 min later by an HR monitor positioned on the chest (Polar® M430— Kempele/Finland). The SJFTINDEX was calculated as the change in heart rate (immediately after the test and 1 min later) divided by SJFTTT. Previous study showed reliability values (Intraclass Correlation Coefficient—ICC) ranging from 0.71 to 0.81 for number of throws, heart rate (0.66–0.86), and SJFT index (0.87) [21].

#### 3.4.2. Judogi Grip Strength Test

The girls were familiarized with the JGST by performing one sustained attempt of 2–3 s grasping a judo uniform (*judogi*) suspended on an elevated horizontal bar. The JGST consisted of sustaining a predefined position of elbow flexion for a maximum time. Athletes performed only the isometric version of the JGST (JGSTISO). The chronometer began with a verbal command and was stopped when the participants could no longer maintain the original position. The reliability of the JGST has been assessed in a previous study, presenting an ICC higher than 0.98 [22].

#### **4. Statistical Analysis**

Data are reported as means and standard deviation (SD). The Shapiro–Wilk test was used to verify the normality of the data. Independent t-tests were used to compare the variables among girls at different menarcheal status. For the t-test, we used the Cohen's d, considering 0.0–0.2 as trivial, 0.21–0.6 as small, 0.61–1.2 as moderate, 1.21–2.0 as large, and 2.1–4.0 as very large [23]. Multiple linear regression analysis (backward stepwise method with criteria for entry of *p* < 0.05 and removal of *p* < 0.10) was used to estimate the relative contributions of age at menarche, chronological age, years of formal training, and growth measurements (height and body mass) to physical performance. All independent variables showed variance inflation factors <2, reflecting no multicollinearity, tolerance >0.1, showing acceptable multicollinearity, and absolute values of correlation coefficients <0.70 [24]. The level of significance was set at 5%, and the analyses were performed using the JASP software (version 0.11.1, University of Amsterdam, Amsterdam, The Netherlands).

#### **5. Results**

Table 1 shows the demographic characteristics, judo experience, body size, and physical performance of female judo athletes of different menarcheal status. It was verified that post-menarche girls were older, more experienced, taller, and heavier and presented higher performance in SJFT (throws and index) and SLJ than pre-menarche girls. The age at menarche ranged from 10 to 13 years.


**Table 1.** Mean ± SD of demographic characteristics, judo experience, body size, and physical performance in female judo athletes of different menarcheal status.

Note: \* *p* < 0.05, SJFT: Special Judo Fitness Test, SJFTTT: total throws of SJFT, SJFTHR: heart rate during SJFT, JGSTISO: Judogi Grip Strength Test, SLJ: Standing Long Jump Test, MBT: Medicine Ball Throw Test, HGS: Handgrip Strength Test.

Table 2 summarizes the indicators of the physical performance tests in post-menarcheal female judo athletes. The age at menarche and body mass (negative predictors) and the chronological age (positive predictor) explained 70% of the variance in JGSTISO performance. Judo training experience and chronological age (positive predictors) and age at menarche (negative predictor) explained 59% of the SLJ performance. Chronological age (positive predictor) and age at menarche (negative predictor) accounted for 40% of the variance in MBT performance, while chronological age and height (both positive predictors) explained 52% of the variance in HGS. For SJFTTT, no predictive analysis was reported because no variable was entered in the model using the stepwise criteria.



Note: JGSTISO: Judogi Grip Strength Test, SLJ: Standing Long Jump Test, MBT: Medicine Ball Throw Test, HGS: Handgrip Strength Test.

#### **6. Discussion**

The results of this study showed that post-menarche youth female judo athletes with advanced somatic growth presented higher performance in a judo-specific test (SJFT) and greater lower limb power output (SLJ) than their pre-menarche counterparts. Chronological age was an indicator in all physical tests, while age at menarche was an indicator for three of the four examined variables in post-menarche girls, thereby demonstrating that age-related maturity has an impact on general neuromuscular and judo-specific physical performance.

Similar to reports in team sports, post-menarcheal female judo athletes self-reported reaching their menarche close to 12 years. For example, Böhme [25] found the mean age at menarche to be close to 12.8 years in athletics, basketball, football, and handball athletes. Menarcheal status is an indicator of sexual maturity and usually occurs from 11 to 15 years of age, following a growth spurt (i.e., PHV) [3,26] due to hormonal alterations [1,27]. From this developmental period, advancement of motor skills and physical performance is expected, especially if there is adequate involvement in physical and sports activities from an early age [28,29]. In this study, it was verified that post-menarche youth female athletes initiated practicing judo earlier than the pre-menarche group, which may be related to their higher physical performance.

It was also found that youth female judo athletes in the post-menarche group were older, taller, and heavier than those in the pre-menarche group. Generally, adolescent girls start their growth spurt quickly, increasing approximately 8–9 cm in height per year [1] and gaining approximately 2.3–2.7 kg of body mass annually [30]. Therefore, the higher chronological age and the advanced somatic growth processes (represented by body size measurements) of post-menarche girls may explain their higher physical performance, particularly, the number of throws in SJFT and SLJ performance. A previous study verified that the number of throws in the SJFT was positively correlated with vertical jump performance in adult male judo athletes [31]. Thus, higher levels of muscle power in the lower limbs of the post-menarche girls may help to explain their higher performance in SJFT. In addition, the greater time of formal training in post-menarche girls most likely contributed to increased muscle power, as Zaggelidis et al. [32] verified that judo training enhances vertical jump performance, mainly due to improvements in the stretch–shortening cycle (SSC).

Adolescent athletes with better aerobic function present higher performance in highintensity intermittent efforts [33]. The ability of children to better maintain performance during repeated high-intensity exercise bouts could be related to a better optimization of oxidative pathways than of glycolytic pathways during exercise and to a lower activation of type II muscle fibers, resulting in greater resistance to fatigue [34]. Although aerobic fitness was not evaluated in this study, it is possible to suggest that post-menarche girls present a higher aerobic condition than pre-menarche girls, especially due to the previously reported correlation between SJFT performance and aerobic capacity [31]. In some neuromuscular tests (HGS, MBT, JGST), there were no significant differences between pre- and postmenarche girls.

When specific indicators of physical tests were investigated in post-menarche youth judo athletes, the age at menarche was found to be negatively associated with JGSTISO, SLJ, and MBT (i.e., the earlier the age at menarche, the higher the performance). The status of menarche represents a great gain in the release of progesterone, estrogen, and, to a lesser extent, testosterone [35,36]. The release of estrogen and testosterone is linked to increased muscle mass and body fat, which have positive and negative influences on physical performance, respectively. In addition, the release of these hormones has been related to increases in lactic anaerobic power [37] and maximum aerobic power due to the growth of body dimensions [38].

Chronological age was a positive indicator of all neuromuscular tests, potentially demonstrating a major influence on physical performance tests in post-menarche youth female judo athletes. Detanico et al. [8] previously verified that chronological age was a positive indicator of judo-specific performance (JGSTISO and SJFT variables) in young male judo athletes. Similarly, a study conducted by Courel-Ibanez et al. [39] detected a higher number of throws in the SJFT in U15 male amateur judo athletes compared to U13 athletes, showing better performance in older boys, especially when utilizing anaerobic pathways. Giudicelli et al. [12] also found that older male judokas (aged 11.0–14.7 years) performed better in most of the physical tests; however, in their study, the maturation attenuated the age effect in most variables and significantly affected upper body strength.

Another interesting finding was that the number of years of formal training was positively associated with SLJ performance, showing improvements in muscle power of the lower limbs in post-menarche girls with judo training experience. A previous investigation found that vertical jump performance discriminated adult judo athletes with different training experience levels (advanced vs. novice) [40], likely due to SSC enhancements [32]. Moreover, muscle power of the lower limbs is an important parameter related to technical– tactical performance during judo matches in senior female athletes [41].

Somatic growth variables (height and body mass) were positive and negative indicators of HGS and JGSTISO, respectively. Taller girls obtained advantages in the HGS test, probably due to the longer forearm and arm [1,42], which may be related to increased force production capacity. This finding exhibits practical relevance, since gripping tasks are an important component of judo performance [43]. Concurrently, body mass showed a negative impact on JGSTISO performance. This test estimates isometric endurance strength in the upper body [22], and to perform it, the athletes must hold onto a bar and suspend themselves (i.e., hold their own body mass). The JGST has been previously shown to have a negative relationship with body mass [44], suggesting that heavier athletes may underperform in absolute terms in this specific test.

Finally, some limitations of this study should be addressed, such as the small sample size, particularly after division into groups according to the menarcheal status. However, the statistical power calculated a posteriori presented values >0.8 for the variables that showed significant differences, thus avoiding type II error. Nonetheless, this is the first investigation examining the effects of menarcheal status and growth on physical performance in young female judo athletes. The current results expose differences in physical performance according to the menarcheal status, showing that post-menarche girls are stronger and perform better than pre-menarche counterparts. This highlights the importance of having pre-menarche girls compete in their own age category and not in higher age groups. Age at menarche, chronological age, growth, and years of formal judo training seemed to explain the performance in post-menarche female athletes. These indicators seem to contribute to competitive success, as it was found that years of formal training, height, and strength tests performance (JGSTISO and SLJ) can discriminate the competitive level in young male judo athletes (national vs. state level) [9]. However, to prevent exposure to early specialization, it is essential to consider the maturation characteristics of female youth judo athletes and individualize training loads during short- and medium-term planning. These actions will contribute to avoid harmful effects of early specialization on physical and mental health during childhood and adolescence. We recommend for future studies to investigate the influence of ethnicity, population size, parents' education, socio-economic and nutritional parameters, as it is known that age at menarche is a sensitive indicator of environmental conditions during childhood.

#### **7. Conclusions**

We concluded that post-menarcheal youth female judo athletes are older, more advanced in the growth process, more experienced in judo, and present higher physical performance when compared to girls who have not yet reached menarche. Chronological age and the age at menarche were shown to be the greatest indicators of neuromuscular and judo-specific performance in post-menarche youth female judo athletes. Furthermore, somatic growth and years of formal training also contributed to neuromuscular performance of the upper and lower limbs.

**Author Contributions:** Conceptualization, M.S.d.S.A., R.L.K. and D.D.; Methodology, M.S.d.S.A., R.L.K. and D.D.; Formal analysis, M.S.d.S.A., R.L.K. and D.D.; Investigation, M.S.d.S.A., R.L.K., D.H.F. and D.D.; Data curation, M.S.d.S.A. and R.L.K.; Writing original draft preparation, M.S.d.S.A., R.L.K., D.H.F. and D.D.; Visualization, M.S.d.S.A., R.L.K., D.H.F. and D.D.; Supervision, D.D. and D.H.F.; Project administration, D.D. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was financed by The Coordination for the Improvement of Higher Education Personnel (CAPES)—PROEX nº: 23038.007266/2021-94.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Research Ethics Committee of the local university (Federal University of Santa Catarina, number: 63053516.4.0000.0121).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data are not publicly available for ethical privacy reasons with the subjects involved in the research.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Analysis of Anthropometric and Body Composition Profile in Male and Female Traditional Rowers**

**Alfonso Penichet-Tomas , Basilio Pueo \* , Sergio Selles-Perez and Jose M. Jimenez-Olmedo**

Physical Education and Sports, Faculty of Education, University of Alicante, 03690 Alicante, Spain; alfonso.penichet@ua.es (A.P.-T.); sergio.selles@ua.es (S.S.-P.); j.olmedo@ua.es (J.M.J.-O.)

**\*** Correspondence: basilio@ua.es

**Abstract:** The anthropometric profile has a fundamental role in rowing performance and young talent detection. The objective of this study was to analyze the anthropometric profile, body composition, and somatotype in traditional rowers, and to analyze which variables can be used as predictors of rowing performance. Twenty-four rowers competing at national level participated in this study, thirteen men and eleven women. Significant differences (*p* < 0.001) were observed in the height of male rowers (large effect size, *d* = 1.8) and in body mass (very large effect size, *d* = 2.4). Also, muscle mass reached a higher percentage in male rowers (*d* = 3.7), whereas the sum of seven skinfolds (*d* = 2.0) and body fat percentage (*d* = 2.0) reached higher values in female rowers, all their difference being significant (*p* < 0.001) with very large effect size. The somatotype of male rowers was ectomesomorph (1.8-4.5-3.0), and the somatotype of female rowers was in the balanced mesomorph (2.8-3.8-2.6). A very strong correlation between height (*r* = 0.75; *p* = 0.002) and rowing performance was found in male rowers. Body mass (*r* = 0.70; *p* = 0.009) and muscle mass (*r* = 0.83; *p* = 0.001) showed also very strong correlation in female rowers. Finally, height was the best predictor of performance for male rowers (R<sup>2</sup> = 0.56, *p* < 0.003) and muscle mass for female rowers (R<sup>2</sup> = 0.68, *p* < 0.002). The anthropometric profile of male and female traditional rowers showed differences to be considered in training programs and talent selection.

**Keywords:** rowing; anthropometry; somatotype; performance; talent identification

#### **1. Introduction**

Rowing is a sport that consists of propelling a boat through the water using one or more oars. The difference with other sports that also use oars is that the oars are fixed to the body of the boat with the rower positioned towards the bow of the boat resulting in the production of different dynamic force components [1,2]. The main classification of rowing modalities differentiates between boats with a mobile seat or a fixed seat [3].

The modality with mobile seat boats is generally called Olympic rowing because only this modality includes Olympic boats. The seat of each rower is placed on rails that allow forward and backward movement. The legs produce almost half the power of the drive (46%) while the trunk around 32% and the arms 22% [4]. The competitions, which can last between 5 and 8 min depending on the type of boat and the category, are generally over the distance of 2000 m in calm waters [5]. On the other hand, in fixed seat boats, the seats do not move in the boat and the technical execution is different since the rowers are supported in the coccyx area. This technical difference that prevents the rower from using the legs in such a wide range of motion implies that the amplitude of the trunk degree is greater than in Olympic rowing [6]. This modality is also called traditional rowing because it is how rowing was originally practiced: Llaüt, with eight rowers and a coxswain [7], and Trainera, with 13 rowers and a coxswain [8]. In addition, traditional rowing courses are not held in parallel lanes, but between two and four lengths with one or three complete tacks, both in calm water and the sea. These technical and competitive differences between modalities, boats, and types of competition result in different functional and physiological demands [9],

**Citation:** Penichet-Tomas, A.; Pueo, B.; Selles-Perez, S.; Jimenez-Olmedo, J.M. Analysis of Anthropometric and Body Composition Profile in Male and Female Traditional Rowers. *Int. J. Environ. Res. Public Health* **2021**, *18*, 7826. https://doi.org/10.3390/ ijerph18157826

Academic Editors: Francesco Campa and Gianpiero Greco

Received: 26 June 2021 Accepted: 21 July 2021 Published: 23 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

where anthropometric characteristics and body composition have a fundamental role in Olympic [10,11] and traditional [7,12] rowing performance.

Most studies about anthropometry, body composition, and somatotype have focused on Olympic rowing [13–17]. Furthermore, some studies have not only compared the different profiles based on weight or age category. The differences between male and female rowers have also been analyzed, finding differences and similarities in anthropometric characteristics that could determine not only training programs but also offering indicators to be able to perform talent detection programs [18–20]. Even De Larochelambert [10] determined which morphologies (tall and thin, tall and robust, small and thin, or small and robust) had a significant effect on speed for both male and female rowers. On the other hand, the research also seem to determine that there are anthropometric characteristics that are related to the level of rowing performance such as height and length measurements [21]. Taller rowers can perform a wider stroke in the water, and a greater stroke range is directly related to increased rowing performance [22]. A similar trend is found with the body mass of the rowers since higher values seem to be correlated with success in competition [14]. Higher body mass can be a disadvantage for performance in other sports where the athlete must shift their own weight. In rowing, the rower is sitting in the boat and his own weight does not seem to have a negative effect on performance. These characteristics are above all in the heavyweight categories because in the lightweight categories the differences and correlations with success in rowing are lower [20]. The studies carried out show that in the heavyweight categories the body mass does not have a negative impact, even a greater weight positively favors power production. However, in the lightweight categories this fact has not been demonstrated as strongly. The profitability of the rower may have a greater impact. Nevertheless, a higher percentage of body fat can be a disadvantage [18]. The body composition of rowers is characterized by a low percentage of fat mass and a mesomorph body type associated with a high development of muscle mass as somatotype [15,16]. It has been widely reported that anthropometric variables and success in rowing are associated, which shows that these characteristics could be used as predictors of performance [23]. Even carrying out a complete body composition study with quantitative and qualitative parameters can be used to plan specific training cycles in different periods of the season [24].

Research in traditional rowing about anthropometry and body composition profile is very limited. Some researchers have studied the relationship of anthropometric characteristics with traditional rowing performance and some of these findings seem to coincide with the Olympic rowing modality, such as a greater body mass and fat-free mass seem to have a positive impact on rowing performance [12]. However, there are some differences such as less muscle mass [25] or lower average height that seem not as important to performance in traditional rowers [8]. Traditional rowing boats require rowers of different heights and weights for hydrodynamic reasons to balance the boat in rough seas [8]. For example, Sebastia-Amat et al. [26] found that only body mass for male rowers and body muscle for female rowers were good predictors of performance in traditional rowing.

Further investigation of these differences between modalities and gender is essential to determine a complete profile of the traditional rower and for following objective criteria in talent recruitment programs. Furthermore, changes in some characteristics of body composition in rowers can be a performance advantage, so control and monitoring of body composition can be crucial for their success in competition [24]. For this reason and because there is also no scientific evidence of comparative studies that carry out a complete study of body composition profile of traditional rowers, the objective of this study is to analyze and compare the anthropometric profile, body composition, and somatotype in male and female traditional rowers. In addition, the present study also aims to analyze the anthropometric variables that influence rowing performance and which of them can be used as predictors of performance. Despite general variations between genders are expected, the differences will allow to create a differentiated profile of rowers competing at the national level and to verify that characteristics such as height and weight, among others, have a relevant role in rowing performance.

#### **2. Materials and Methods**

#### *2.1. Participants*

Twenty-four rowers competing at national level participated in this study, thirteen males (age: 27.3 ± 5.1 years; height: 182.1 ± 6.6 cm, body mass: 75.3 ± 5.3 kg) and eleven females (age: 27.7 ± 4.3 years; height: 169.9 ± 6.7 cm, body mass: 61.9 ± 6.0 kg). The requirement to participate was to have qualified for the national championship, with an experience of at least 3 years, and to regularly train a minimum of six days per week for 2–3 h/day, supervised by one of the authors who perform the physical preparation and monitoring of the athletes who have participated in the study. They were asked to refrain from eating for at least four hours before the measurements, not exercise on the day of the measurement [16] and not high intensity exercise the day before. The hydration guidelines were the same as those carried out for training, no specific hydration guidelines were given. All measurements were made at the same time of the day. Rowers who did not meet the selection criteria were excluded from the study. The Ethics Committee of the University of Alicante gave institutional approval for this study, in accordance with the Declaration of Helsinki (IRB UA-2020-07-21). The subjects were informed about the study and gave their written informed consent.

#### *2.2. Procedure*

The anthropometric assessment followed the guidelines set by the International Society for the Advancement of Kinanthropometry (ISAK) [27]. The measurements were performed by the same researcher with ISAK certification level II under fasting conditions at room temperature (22 ± 1 ◦C) [28,29]. All variables were measured on the right side of the body in duplicate and the mean value was recorded. Intra-observer technical error of the measurement (TEM), 5% for skinfolds and 1% for girths and breadths, was considered for the measurements.

Body mass and height were measured using a scale (model 707, Seca, Hamburg, Germany) to the nearest 0.1 kg and a stadiometer (Harpenden, Burgess Hill, UK) to the nearest 0.1 cm. Rowers were weighted and measured wearing only underwear with bare feet. Height was measured with the rower completely upright and the chin parallel with the ground. Body mass index (BMI) was computed as body mass (kg) divided by height squared (m<sup>2</sup> ). Eight skinfolds (triceps, biceps, subscapular, iliac crest, supraspinal, abdominal, front thigh, and calf) were measured with a Holtain skinfold caliper to the nearest 0.2 mm and six girths (relaxed arm, tensed arm, thigh, medial calf, waist, and hip) were obtained using a Holtain bone breadth caliper to the nearest 0.1 cm (Holtain Ltd., Crymych, UK). The sum of eight skinfolds was examined following validated procedures [30]. Finally, three breadths (humerus, femur, and stylion) were measured with an anthropometric tape (Seca) to the nearest 0.1 cm. Fat, muscle, bone, and residual masses were calculated, as well as somatotype. To calculate the percentage of body fat, the formula of Withers et al. was used [30]. Muscle mass was determined using the methods of Lee et al. [31] and bone mass was calculated following the Rocha model [32]. The anthropometric somatotype was determined using the Carter and Heath equation [33], making a graphic representation of the results in a somatochart.

Once the anthropometric study was completed, the rowers performed an all-out 2000 m test on a rowing ergometer (Model D; Concept 2, Inc., Morrisville, VT, USA) with a coupling adapted for the reproduction of the traditional rowing stroke fixing the seat [25,34] and with a PM5 performance monitor to collect mean power output reached in the test, and its equivalence in time. All the rowers were familiar with the rowing machine and with the drag factor used: 160 for males and 140 for females. The rowers performed a 10-min warm-up before the test at moderate intensity between 70 to 80% of maximum heart rate (above 140 beats per min) at a stroke rate of 18–20 strokes per minute [26]. Power output, stroke rate and time to complete 2000 m rowing ergometer performance test were recorded.

#### *2.3. Statistical Analysis*

Descriptive analysis was presented by the mean, standard deviation (SD), minimum (min), and maximum (max) for all variables. Shapiro–Wilk statistical test was used to verify that the variables followed the normality criterion. Student's *t*-test was used to compare anthropometric data between male and female rowers. Cohen's *d* was used as a measure of the effect size of differences between male and female rowers and interpreted according to modified thresholds [35] for sports sciences [36] as trivial (<0.2), small (0.21–0.6), moderate (0.61–1.2), large (1.2–1.99), and very large (>2.0). Somatotype Attitudinal Mean (SAM) and Somatotype Attitudinal Variance (SAV) were calculated to describe the magnitude of the dispersion of somatotypes in both groups. Somatotype Attitudinal Distance (SAD), the distance in three dimensions between male and female groups, was used to compare somatotype group means. Pearson correlation coefficient (*r*) was used to determine relationships between each anthropometric variable with rowing performance. Effect sizes of relationships were assessed by Pearson's correlations and coefficients of determination: trivial (<0.1), small (0.1–0.29), moderate (0.3–0.49), strong (0.5–0.69), very strong (0.7–0.89), nearly perfect (0.9–0.99), and perfect (1.0) [36]. A stepwise multiple regression analysis (R<sup>2</sup> > 0.5) was used to analyze which anthropometric variables could be used to predict rowing performances. Statistical significance was set at *p* < 0.05. Statistical analyses were performed using Statistical Package for Social Sciences (SPSS v.26 for Windows, SPSS Inc., Chicago, IL, USA).

#### **3. Results**

Body mass, height, and BMI mean values were significantly higher (*p* < 0.05) in male rowers (182.1 ± 6.6 cm, 75.3 ± 5.3 kg, and 22.8 ± 1.3 kg/m<sup>2</sup> ) than female rowers (169.9 ± 6.7 cm, body mass: 61.9 ± 6.0 kg 21.4 ± 1.0 kg/m<sup>2</sup> ) with large to very large effect size, as shown in Table 1. However, the skinfolds of triceps, biceps, iliac crest, front thigh, and calf were significantly higher (*p* < 0.05) in female rowers than in male rowers, with moderate to very large effect size. Therefore, the mean of the sum of skinfolds also showed a larger value in female rowers (88.0 ± 17.6 mm) than in male rowers (58.5 ± 12.4 mm). This difference was statistically significant (*p* < 0.001) with very large effect size (*d* = 2.0). In contrast, most of the girths were significantly higher (*p* < 0.05) in male rowers than in female rowers, with moderate effect size on thigh girth (*d* = 0.9) and very large effect size on relaxed arm (*d* = 2.5), tensed arm (*d* = 3.4) and waist girths (*d* = 2.7). Finally, humerus (*d* = 2.7), femur (*d* = 1.8) and stylion breadths (*d* = 3.0) also reached statistically higher values in male rowers, with large to very large effect size.

Table 2 shows body composition and somatotype profile of male and female rowers which highlights that male rower reached larger values of muscle mass (46.7 ± 2.0%) than female rowers (39.1 ± 2.1%), with significant difference (*p* < 0.001; *d* = 3.7) and very large effect size. However, female rowers achieved higher fat (15.4 ± 3.1%) and residual masses (29.4 ± 1.9%) than male rowers (10.3 ± 2.1% and 26.4 ± 1.9%, respectively). This contrast showed significant differences (*p* < 0.001) and very large (*d* = 2.0) and large (*d* = 1.6) effect size, respectively.

The comparative analysis of the somatotype between male and female rowers indicates that there are significant differences in endomorphy (*p* < 0.001; *d* = 2.0), with very large effect size, and mesomorphy (*p* < 0.001; *d* = 1.8), with large effect size. The mean somatotype of male rowers was mesomorph-ectomorph (1.8-4.5-3.8) and the mean somatotype of female rowers was balanced mesomorph (2.9-3.0-2.9) (Figure 1). Finally, SAM values were 1.1 in male rowers and 0.9 in female rowers where no significant differences between them, and the effect size was small (*d* = 0.2). The difference in SAD between male and female rowers was 1.0.


**Table 1.** Mean values of anthropometric measurements and difference between male and female rowers.

BMI: Body Mass Index; SD: standard deviation; min: minimum; max: maximum; CI: confidence interval; \*: statistically significance between male and female rowers (*p* < 0.05).

**Table 2.** Descriptive data and comparative analysis of body composition and somatotype between male and female rowers.


BMI: Body Mass Index; SD: standard deviation; min: minimum; max: maximum; CI: confidence interval; \*: statistically significance between male and female rowers (*p* < 0.05).

> Figure 2 shows the associations between anthropometric variables and rowing performance expressed in mean power output reached in 2000 m rowing test. The results show a strong correlation with body mass in male rowers (*r* = 0.57; *p* = 0.021) and a very strong correlation in female rowers (*r* = 0.70; *p* = 0.009). However, height was strongly correlated in female rowers (*r* = 0.64; *p* = 0.017) and very strongly correlated in male rowers (*r* = 0.75; *p* = 0.002) with performance. Finally, a very strong correlation was found between rowing performance and muscle mass in female rowers (*r* = 0.83; *p* = 0.001), while in male rowers the correlation was small (*r* = 0.42; *p* = 0.075).

**Figure 1.** Somatochart of male and female rowers and mean somatotypes.

**Figure 2.** Relationships between anthropometric characteristics and rowing performance in male and female rowers.

− −

Table 3 shows the results of the stepwise multiple regression analysis in male and female rowers by which height is the only predictor of rowing performance in male rowers, explaining 56% of variance (R<sup>2</sup> = 0.56, *p* < 0.003). The single predictor of rowing performance in female rowers was muscle mass, explaining explained 68% of variance (R<sup>2</sup> = 0.68, *p* < 0.002). The rest of anthropometric measures did not contribute significatively and were excluded from the prediction equation.

**Table 3.** Stepwise multiple regression model of rowing performance.


SEE: standard error of estimate; W: power.

#### **4. Discussion**

The aim of this study was to analyze and compare the anthropometric profile, body composition, and somatotype in male and female traditional rowers. In addition, the present study aimed to analyze which variables can be used as predictors of rowing performance. As it is the first study that compares the anthropometric profile of traditional rowing between male and female rowers to determine reliable reference values, the selection criteria of the participants were to have classified for the national championship, to have an experience of at least 3 years and to regularly train a minimum of six days per week for 2–3 h/day.

The anthropometric measurements of our study showed that body mass and height mean values were higher in male rowers (182.1 ± 6.6 cm, 75.3 ± 5.3 kg) than female rowers (169.9 ± 6.7 cm, body mass: 61.9 ± 6.0 kg). Results also showed that height and body mass correlate with rowing performance in male and female rowers. Furthermore, height was the best predictor of performance in male rowers (R<sup>2</sup> = 0.56, *p* < 0.003). Although there is no scientific evidence on studies of comparative analysis of a complete body composition profile between male and female rowers in traditional rowing, some of the rowers' characteristics in studies on traditional rowing are consistent with this study. Elite traditional male rowers from the Spanish First League of Traineras (ACT) showed a very similar body mass and height to our male rowers (77.0 ± 7.6 kg and 181.1 ± 3.4 cm) [37]. However, other studies have indicated that elite traditional male rowers were heavier (84.4 ± 6.3 kg) but with similar height (182.5 ± 5.2 kg) [8]. In other studies, traditional male rowers of lower competitive level were shorter (178.4 ± 8.9 cm) but with similar body mass (77.3 ± 7.9 kg) [26]. The winners of the Traineras women's league and the La Concha championship [38] coincide with height (168.2 ± 6.3 cm) and body mass (61.2 ± 4.4 kg) results of our study. However, female rowers in Sebastiá-Amat et al. [26] were slightly shorter (166.3 ± 7.5 kg) and lighter (59.9 ± 8.3 cm). It is generally accepted that height is a very important anthropometric characteristic for rowing performance because a greater height increases the amplitude of the drive in the water [7,39]. The results of the studies on Olympic rowing follow the same trend in both height and body mass. Male Olympic rowers reach heights over 190 cm and weigh more than 90 kg, while female rowers exceed 180 cm in height with a body mass of around 75 kg [14,18,19,40]. These discrepancies may be because the height of rowers can be a differentiating characteristic between higher and lower performance in traditional modalities, while the same does not happen with body mass. However, the rowers of the Trainera boat seem to have a higher average weight than the Llaüt rowers. This may be due to the difference in the number of rowers in each boat and the need for the bow rowers to be lighter, lowering the average weight in the Llaüt for correct navigation. Several studies suggest that traditional rowing boats require rowers with different anthropometric profiles, especially in the bow, due to the hydrodynamics of the boat when competitions are held at sea and the body mass placement of the rowers is important [2,8,38].

In the same way, BMI has reached higher values in male rowers (22.8 ± 1.3 kg/m<sup>2</sup> ) than in female rowers (21.4 ± 1.0 kg/m<sup>2</sup> ). Studies about male traditional rowers have shown values of BMI greater than 23 kg/m<sup>2</sup> [34,37] and 24 kg/m<sup>2</sup> [7,8,26]. However, BMI of our male rowers is similar to lightweight Olympic (22.1 ± 0.3 kg/m<sup>2</sup> ) since the rowers in the present study weighed less than the rowers in both traditional and Olympic rowing studies. Finally, BMI values of our female rowers were similar to other traditional rowing (21.7 ± 2.6 kg/m<sup>2</sup> ) [26] and Olympic rowing studies (21.6 ± 6.1 kg/m<sup>2</sup> ) [19]. In this latest study, Winkert et al. suggested a body composition with high lean body mass and adequate power to body mass ratios instead of a high body mass, because increased body mass and BMI showed a negative effect on career attainment.

The skinfolds and mean of the sum of 8 skinfolds have a larger value in female rowers (88.0 ± 17.6 mm) than in male rowers (58.5 ± 12.4 mm). It is important to know the values obtained from the skinfold measurement, as it is used to predict fat mass. Furthermore, these differences were expected because women have 6 to 11 percent more body fat than men. Studies show that estrogens reduce a woman's ability to burn energy after eating, thus storing more fat in the body [41]. However, female rowers have lower values in girths and breadths, both in the upper body and in the lower body, except for hip girths with very little difference. In contrast to the scientific literature, it seems that the male rowers in our study have lower values in the sum of skinfolds (67.3 ± 15.6 mm) compared to elite traditional rowers [8]. Compared to rowers participating in the 2000 Sydney Olympic Games [18], the sum of skinfolds of the male rowers in the present study is between open-class (65.3 ± 17.3 mm) and lightweight (44.7 ± 8.1 mm). In the case of female traditional rowers, the values are very similar to the values reached by the open-class female rowers (89.0 ± 23.6 mm). The sum of skinfolds of the lightweight female rowers was only 59.7 ± 12.4 mm.

In our study, male traditional rowers have similar values of muscle mass (46.7 ± 2.0%) compared to other traditional rowing studies of competitions of the same distance as the present study: 46.5 ± 2.0% [34], and large values than other studies of competitions over much longer distances where slimmer rowers are needed.: 43.5 ± 2.0% [42] 43.3 ± 2.4% [8]. According to other studies, female rowers achieved a lower percentage of muscle mass (39.1 ± 2.1%). However, muscle mass for female rowers may be a good predictor of performance in traditional rowing in our study (R<sup>2</sup> = 0.68, *p* < 0.002) and in the scientific literature [26]. This may be because women have much less testosterone than men and due to the influence of this hormone on the development of strength and muscles, women are less likely to develop equal strength and muscle size than men [43]. Therefore, the difference in strength between women is greater than between men and this characteristic seems to become a differentiating factor in performance. In female rowers. On the other hand, female rowers achieved higher fat mass (15.4 ± 3.1%) than male rowers (10.3 ± 2.1%), according to the description of elite rowers of González [38], where female rowers reached 16.3 ± 5.5% and male rowers 7.8 ± 1.1%. Studies on elite male rowers showed lower percentages of fat mass (9.9 ± 2.0%) [8] than studies conducted with sub-elite rowers (14.2 ± 4.4%) [25]. The percentage ranges for international Olympic rowers was 6% to 10% and 11% to 15% for male and female, respectively [44].

In the only two studies published to date on anthropometric profile of traditional male rowers, endo-mesomorph somatotypes were found (3.5-4.7-2.4 [8] and 3.3-3.9-2.2 [42]). However, the mean somatotype in the present study is categorized as ecto-mesomorph (1.8- 4.5-3.0) for male rowers, and balanced mesomorph (2.9-3.1-2.9) for female rowers following Carter and Heath [33] where in ecto-mesomorph somatotype, the mesomorphy component is dominant and the ectomorphy component is greater than the endomorphy component; and in balanced mesomorph somatotype, the mesomorphy component is dominant and the endomorphy and ectomorphy components are equal. Our results coincide with the results of Olympic rowers where male rowers had a somatotype defined as ecto-mesomorph (1.9- 5.0-2.5) and female rowers a somatotype categorized as balanced mesomorph (2.8-3.8-2.6). The difference between studies may be due to the competition distances of the rowers

analyzed from each study: 14,816 m [42] and 5556 m [8]. On the other hand, the rowers in the present study had to row over 1400 m, a distance much more like the 2000 m that Olympic rowers must cover.

Results of the present study should be interpreted with caution because the main limitation of this study lies in the sample size. Also, some of the results are the product of predictive equations rather than direct measurements. Therefore, they can be used as references but should be interpreted in the context of individual characteristics and needs. In addition, it is important to bear in mind that the evaluations have been individual and on rowing ergometer, while the athletes compete in collective boats that may require different profiles as mentioned above. Future research should analyze the differences by position in the boat: bow, stern, and rest rowers. The need for more heterogeneous rowers in traditional rowing boats compared to Olympic rowing may yield a more detailed profile by position. Furthermore, it would be interesting to determine the relationships between the anthropometric profile and rowing performance in male and female traditional rowers to define which characteristics might be most relevant to each one.

#### **5. Conclusions**

This study analyzed and compared the anthropometric profile, body composition, and somatotype in male and female traditional rowers, and the role of these variables in the prediction of rowing performance. The results showed that male traditional rowers were significantly taller and heavier, with higher values of girths and breadths, in addition to greater muscle mass. Female traditional rowers reached higher sum of skinfolds and greater fat mass. The mean somatotype for male and female traditional rowers was ectomesomorph and balanced mesomorph, respectively, with significant differences in the mesomorph region of male rowers and the endomorph region of female rowers.

Large values of body mass and height correlated with rowing performance in male and female rowers, highlighting height as the best predictor of rowing performance for male traditional rowers. Furthermore, muscle mass positively correlated in female rowers, being the best predictor for rowing performance.

This study shows a detailed anthropometric description of traditional rowers competing at the national level that can be useful as reference values for coaches and rowers. Furthermore, the study shows different variables that can be used to control training and increase rowing performance, such as body and muscle mass, and to identify potential talents in young athletes thanks to characteristics such as height.

**Author Contributions:** Conceptualization, A.P.-T., B.P. and J.M.J.-O.; Formal analysis, S.S.-P.; Investigation, A.P.-T., B.P. and J.M.J.-O.; Methodology, A.P.-T., B.P., S.S.-P. and J.M.J.-O.; Resources, S.S.-P. and J.M.J.-O.; Supervision, B.P.; Writing—original draft, A.P.-T.; Writing—review & editing, A.P.-T., B.P., S.S.-P. and J.M.J.-O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of University of Alicante (IRB No. UA-2020-07-21, date of approval: 8 August 2020).

**Informed Consent Statement:** Written informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on reasonable request from the corresponding author.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Bioimpedance Vector Patterns according to Age and Handgrip Strength in Adolescent Male and Female Athletes**

**Marcus Vinicius de Oliveira Cattem <sup>1</sup> , Bruna Taranto Sinforoso <sup>1</sup> , Francesco Campa <sup>2</sup> and Josely Correa Koury 1,\***


**Abstract:** Bioelectric Impedance Vector Analysis (BIVA) can be used to qualitatively compare individuals' hydration and cell mass independently of predictive equations. This study aimed to analyze the efficiency of BIVA considering chronological age and handgrip strength in adolescent athletes. A total of 273 adolescents (male; 59%) engaged in different sports were evaluated. Bioelectrical impedance (Z), resistance (R), reactance (Xc), and phase angle (PhA) were obtained using a single-frequency bioelectrical impedance analyzer. Fat-free mass (FFM) and total body water were estimated using bioimpedance-based equations specific for adolescents. Female showed higher values of R (5.5%, *p* = 0.001), R/height (3.8%, *p* = 0.041), Z (5.3%, *p* = 0.001), and fat mass (53.9%, *p* = 0.001) than male adolescents. Male adolescents showed higher values of FFM (5.3%, *p* = 0.021) and PhA (3.1%, *p* = 0.033) than female adolescents. In both stratifications, adolescents (older > 13 years or stronger > median value) shifted to the left on the R-Xc graph, showing patterns of higher hydration and cell mass. The discrimination of subjects older than 13 years and having higher median of handgrip strength values was possibly due to maturity differences. This study showed that BIVA identified age and strength influence in vector displacement, assessing qualitative information and offering patterns of vector distribution in adolescent athletes.

**Keywords:** adolescent athletes; body composition; BIVA; confidence ellipses; fat-free mass; R-Xc graph; tolerance ellipses

#### **1. Introduction**

Strenuous training could be a matter for the competitive adolescent athletes, since high intensity and high training volume impose nutritional and functional risks to body development [1]. Exercise practice has been associated with the development of bone [2] and muscle tissues [3]. Fat-free mass (FFM) is considered a predictor of muscle strength and physical capacities [4–7]. Assessments of body composition contribute to verify the effects of physical activity and sports practice over time.

Muscle strength is another valuable measurement in physically active individuals as it impacts sports performance, daily activities, life quality and is related to low incidence and prevalence of diseases [8]. In order to assess handgrip strength, handgrip dynamometers are easy to use, simple, and not expensive [9]. Muscle strength is also related to gender, chronological and biological age, and body composition, since FFM is important to produce it and fat mass (FM) may limit it in contact sports, for example [10,11]. Handgrip strength has been used in youth soccer and female basketball players for talent identification [12,13].

Bioelectrical Impedance Analysis (BIA) can be used as a non-invasive method to estimate FFM, FM, and total body water (TBW) from electrical body proprieties of resistance (R) and reactance (Xc) while considering individual characteristics, such as sex, age, height, and weight [14,15]. BIA presents good correlation and concordance with dual energy X-ray absorptiometry (DXA) also when analyzing adolescent athletes [16]. However, BIA

**Citation:** Cattem, M.V.d.O.; Sinforoso, B.T.; Campa, F.; Koury, J.C. Bioimpedance Vector Patterns according to Age and Handgrip Strength in Adolescent Male and Female Athletes. *Int. J. Environ. Res. Public Health* **2021**, *18*, 6069. https:// doi.org/10.3390/ijerph18116069

Academic Editor: Paul B. Tchounwou

Received: 8 May 2021 Accepted: 30 May 2021 Published: 4 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

equations are dependent on specific characteristics of the population [15]. For this reason, in recent years, Bioelectric Impedance Vector Analysis (BIVA) has gained relevance for sports [17,18], because its qualitative and semi-quantitative analysis of impedance vectors and impedance components are directly plotted on the R-Xc graph. BIVA graphics are interpreted by impedance vector lengths and their ellipses and phase angle (PhA) [19]. PhA is derived from R and Xc, and it has been interpreted as an index of membrane integrity and water distribution between intra and extracellular compartments [20]. In addition, PhA has been used as a predictor of body cell mass, and for this reason, it has been employed as an indicator of nutritional status [21]. The complementary use of the BIVA and PhA may be helpful in the evaluation of changes of nutrition and hydration status in athletes [22].

Moreover, BIVA provides qualitative information of soft tissue classification and ranking, comparing individual vectors and ellipses to reference populations [23]. In this context, it is important to develop BIVA references for adolescent athletes considering handgrip strength. To the best of our knowledge, there are no studies that relate BIVA and handgrip strength in female and male adolescent athletes.

Considering the importance of body composition and strength to sports practice and for adolescent health, and considering BIVA a useful tool to assess adolescent athletes, the aim of this study was to analyze the efficiency of BIVA, considering chronological age and handgrip strength in female and male adolescent athletes.

#### **2. Materials and Methods**

#### *2.1. Study Design and Subjects*

This was a cross-sectional observational study. Two hundred and seventy-three Brazilian healthy adolescents (*n* = 161, males [59%]), aged mean 12.9 ± 0.9 years participated. All the data were collected at a sports-oriented public school located in the central region of the city of Rio de Janeiro, Brazil (2012–2013). This is an elementary full-time school that, unlike other public schools, offers 120 min of daily sports training and seven sports modalities: swimming, judo, badminton, athletics, soccer, volleyball, and table tennis, in which the students practiced different sports for the same amount of time.

The adolescents were classified as athletes, because they participated in training, skill development, and were engaged in competition, according to the definition described in Sports Dietitians Australia Position Statement: Sports nutrition for the adolescent athletes [24].

The participants were classified according to sex, handgrip strength (high—above median value; low—under median value) and chronological age (≤13 or >13 years). In adolescents, body composition is highly interrelated to biological maturity, due to hormones and growth factors function [1]. In the absence of consistent maturation indicators, adolescents can be divided into ≤13 and >13 years [25]. Mathias-Genovez et al. (2016) [26] showed that in the Brazilian adolescent population, 13 years was the age at which changes in body composition start due to biological maturation.

An a priori power analysis was conducted to determine the sample size using statistical software (G\*Power v. 3.1.9.2, Stuttgart, Germany). The sample size calculation was performed assuming the values of *r* = 40%, α = 5%, and β = 20%, so the number of students estimated by each sex was 126. However, at the end of the study, 161 male and 112 female adolescents participated.

To participate in this study, adolescents and parents agreed to participate after a full explanation of the research objectives. This study was approved by the Ethics Committee of the Pedro Ernesto Hospital (CEP/HUPE 1.020.909).

#### *2.2. Anthropometric and Body Composition Measurements*

Weight was measured with a portable scale to the nearest 0.1 kg (Filizola, Brazil), height was measured with a stadiometer to the nearest 0.5 cm (Sanny, Brazil), and Body Mass Index (BMI = weight[kg]/height<sup>2</sup> [m]) was calculated.

BIA measurements were always performed in the morning, using a tetrapolar analyzer RJL (Quantum 101; Systems, Clinton Township, MI USA), which applies an alternating current of 800 µA at a single frequency of 50 kHz. Participants were in the supine position with a leg opening distant 45◦ from the median line of the body and the upper limbs distant 30◦ from the trunk. Electrodes were applied on the right wrist and ankle after cleansing the skin with alcohol in a thermo-neutral environment of 25 ◦C. To avoid disturbances in fluid distribution, participants were instructed to abstain from foods and liquids for at least 4 h as well as refrain from caffeine intake and intense physical activity 24 h prior to the BIA analysis. Before each testing session, the analyzer was checked with a calibration circuit of known impedance (resistance = 500.0 Ω; reactance = 0.1 Ω; 0.9% error). Resistance (R) and reactance (Xc) were used to calculate phase angle (PhA) [20]. FFM and total body water were assessed using a predictive equation developed by Horlick et al. [27]. The BIA predictive equations used in this study are listed in Table 1. Fat mass (FM) was calculated subtracting FFM from weight, and fat mass percentage was calculated by (FM/weight) × 100.

**Table 1.** Predictive equations used in the present study.


H = height (cm); Wt = weight (kg); R = resistance; Xc = reactance; sex = 0 for females and 1 for males.

#### *2.3. Handgrip Strength*

Handgrip strength was assessed with a hand JAMAR-dynamometer (Asimow Engineering Co., Los Angeles, CA, USA) in both hands alternately, three times, and the mean value was recorded to obtain a single value of HG.

#### *2.4. Bioelectrical Impedance Vector Analysis*

BIVA was developed based on the R and Xc vectors normalized by height (H) [19,28]. The experimental data are plotted in the R-Xc graph and compared with the 95th-percentile confidence ellipses from a reference population. The correlation between R and Xc determines the ellipsoidal form of the bivariate probability distributions [28].

BIVA tolerance consists of plotting the experimental data in a bivariate graph considering the 95th, 75th, and 50th vector percentiles of the Z-score of the reference population. Considering the plotting position of the experimental data, it is possible to suggest an interpretation: abnormal situation, when experimental data are positioned outside of the 95th percentile ellipsis; higher body cell mass, when experimental data are located above the long axis of the ellipsis; hypohydration, when experimental data are positioned to the right of the short axis of the ellipsis. Total body water is inversely related to the length of the impedance vector, and a combination of the vector length and its direction is defined as PhA [28,29] (Figure 1). The reference population for adolescents used in the BIVA analyses was obtained from the dataset of Koury et al. [16].

#### *2.5. Statistical Analysis*

All analyses were performed separately for each sex, and participants were classified according to chronological age (≤ 3 or >13 years) and handgrip strength median. Continuous variables were expressed as mean and standard deviation. An independent *t*-test followed by the Bonferroni post hoc test was used to compare variables between chronological ages. A linear regression model assessed the relation between handgrip strength (outcome) and chronological age, fat-free mass, and PhA (predictors). Univariate linear regression with backward stepwise elimination results were presented as unstandardized B coefficients, 95% confidence intervals (CI), and *p*-value. *p*-value < 0.05 was considered

statistically significant. All statistical analyses were performed using STATISTICA 10 software (Stat Soft. Inc., Tulsa, OK, USA).

For BIVA, the two-sample Hotelling T<sup>2</sup> test was used to compare differences in mean impedance vectors in BIVA confidence analyses, and the Mahalanobis test was used to calculate the distances between ellipses. Confidence and the 50%, 75%, and 95% tolerance ellipses were generated using BIVA software [29].

**Figure 1.** BIVA nomogram pattern, RXc-graph. Resistance (R) and reactance (Xc) were normalized by the height (H, meter) (adapted from Piccoli and Pastore, 2002).

#### **3. Results**

Characteristics of the adolescent athletes according to sex and chronological age are shown in Table 2. Female adolescents showed higher values of R (5.5%, *p* < 0.01), R/H (3.8%, *p* = 0.041), Z (5.3%, *p* < 0.01), and fat mass (53.9%, *p* < 0.01) than male adolescents. Male adolescents showed higher values of FFM (5.3%, *p* = 0.021) and PhA (3.1%, *p* = 0.033) than female adolescents. According to chronological age, older female adolescents showed higher values of weight (19.9%, *p* < 0.01), height (3.2%, *p* < 0.01), BMI (13.5%, *p* < 0.01), PhA (5.1%, *p* = 0.002), FFM (14.9%, *p* < 0.01), FM (37.5%, *p* < 0.01), TBW (15%, *p* < 0.01), and handgrip strength (17.5%, *p* < 0.01). In addition to that, older female adolescents showed lower values of R (6.9%, *p* < 0.01), R/H (10.5%, *p* < 0.01), and Z (6.8%, *p* = 0.002) than younger participants. Older male adolescents showed higher values of weight (17.2%, *p* < 0.01), height (7.3%, *p* < 0.01), FFM (22.2%, *p* < 0.01), TBW (21.5%, *p* < 0.01), and handgrip strength (35.2%, *p* < 0.01); they showed lower values of R (7.5%, *p* < 0.01), R/H (15.3%, *p* < 0.01), Xc (8.9%, *p* < 0.01), Xc/H (16.4%, *p* < 0.01), and Z (7.7%, *p* < 0.01) than younger male adolescents. The different modalities practiced did not present any significant difference in the results of body composition and age.

Handgrip strength values are shown according to sex and chronological age (≤13.0 or >13.0 years) in Figure 2. The median value of handgrip strength was used to stratify female and male participants in groups of low and high handgrip strength. Individuals up to the median of handgrip strength of their sex were classified as low handgrip strength and individuals above the median were classified as high handgrip strength. The median of the female group was 20.6 kgf and that of the male group was 21.1 kgf. Differences were found between older and younger individuals of the same sex (*p* = 0.01) and between male and female participants at older age (*p* = 0.02), but not between younger subjects.


**Table 2.** Descriptive and comparative general characteristics, according to sex and age categories (*n*= 273).

 BMI: body mass index; R/H: resistance/height ratio; Xc/H: reactance/height ratio; PhA: phase angle; FFM: fat-free mass; FM: fat mass; TBW: total body water; HG: handgrip strength. Intra- and intergroup differences were obtained using an independent *t*-test followed by the Bonferroni post-hoc test. Significant differences between sexes and the same age category were marked by \* (*p* < 0.05), \*\* (*p* < 0.01), \*\*\* (*p*< 0.001).

**Figure 2.** Handgrip strength in female and male according to different age classes (≤13 or >13 years).

Table 3 shows that a linear regression model was applied to verify the influence of chronological age, FFM, PhA, and sex on handgrip strength (outcome). For all participants, chronological age (57.2%; *p* = 0.041) and FFM (62.2%, *p* = 0.0001) could explain the handgrip strength. In the female group, only FFM could explain the model in 56.1% (*p* = 0.0001), and in the male group, chronological age (79.2%, *p* = 0.032) and FFM (63.6%, *p* = 0.0001) could explain the handgrip strength.



Linear regression model. \* adjusted by sex. R<sup>2</sup> all = 0.651, R<sup>2</sup> female = 0.386, R<sup>2</sup> male = 0.753.

> Figure 3 shows mean impedance vectors with 95% confidence ellipses for adolescent athletes according to sex and chronological age (Figure 3A) or sex and handgrip strength classification (Figure 3B). Participants showed differences when age and handgrip strength (*p* < 0.05) were compared. Older male and female athletes showed shorter impedance vectors. Similarly, a shorter impedance vector was observed in male and female participants with high handgrip strength. Additionally, when distances between age and handgrip strength ellipses were tested, a significant difference was found only between younger male participants and those with low handgrip strength (*p* = 0.033). In addition, there is a slight overlap in male and female low handgrip strength' ellipses; however, the T<sup>2</sup> test still found a significant difference. Considering age and handgrip strength, 35.6% and 33.7% of the younger female and male adolescents were classified as high handgrip (>median), and 44% and 23.3% of the older individuals were classified as low handgrip strength (<median), respectively.

> The data from female (Figure 4A) and male (Figure 4B) adolescent athletes, considering chronological age and handgrip strength classification, were plotted on the BIVA tolerance ellipses of Brazilian adolescent athlete reference population [16]. Both graphs presented a trend of a higher density of points in the 95% tolerance ellipsis. The frequency of points outside the 95% tolerance ellipsis, above the long axis, was 2% for male adolescents and

0.9% for female adolescent athletes. Only one female older and stronger subject was outside the 95% ellipse.

**Figure 3.** Mean impedance vectors with the 95% confidence ellipses for adolescent athletes sorted by chronological age (**A**) or handgrip strength classification (**B**). Mahalanobis distances (D), Hotelling T 2 -tests, F and *p*-values are included.

**Figure 4.** Mean impedance vectors with the 50, 75, and 95% tolerance ellipses for the female (**A**) and male (**B**) adolescent athletes, according to age and handgrip strength categories.

#### **4. Discussion**

There is a growing interest in BIVA in sports and physical exercise [17]. The present study shows, for the first time, BIVA patterns from female and male adolescent athletes and their associations with handgrip strength. Only FFM was a predictor of handgrip strength for female and male adolescent athletes. So, higher strength in male adolescents could be explained by the higher FFM throughout male development.

Studies in adolescent athletes are centered in male subjects [30–32]. There is only one study about BIVA in female athletes [23]. The present study is the first that shows BIVA responses associated with strength, brings new references for adolescent athletes, and adds knowledge to this field. Studies such as the present one, which assesses general

health, are necessary in order to improve prescription of sports, since it is important to have information on adolescent athletes of both sexes.

Most studies only describe reference values for adult individuals, and thresholds and cutoffs points are needed for all ages and ethnic groups as reviewed by Dodds et al. [33] when analyzing variation in handgrip strength worldwide [33]. In the present study, handgrip strength did not show any statistical difference between female and male adolescents until the age of 13 years. However, it was greater in older male subjects than older female adolescents. In addition, female and male differences accentuated after 13 years of age, which may be attributed to puberty changes [34,35]. FFM/FM proportion may explain the greater strength in older male subjects. FFM is closely related to strength, since FFM is the primary body component that produces it [10]. However, when handgrip strength is standardized by fat-free mass, the difference disappears in this study dataset. Chronological age was important to discriminate male and female individuals by handgrip strength, but it was not a predictor in the linear model in female adolescents.

PhA is often associated with strength and physical fitness in adult athletes [18] and also in male adult and adolescent athletes [31]. PhA was also associated with handgrip strength in healthy adult men [36]. However, this study was conducted in an age range with little PhA variation according to a review of 250,000 subjects in different ages by Mattiello et al. [37]. For this reason, PhA could present a constant behavior in regression models and was not significant in all the analysis. Regarding the role of the somatic maturation on BIVA patterns, Campa et al. [2] identified specific transition periods in which the bioelectrical parameters showed an increase, a decrease, or a plateau. In particular, PhA begins to increase rapidly beginning at two years prior to the maturity offset and continues to do so for the four years following this growth phase [38]. In addition, the vector length shows a sharp decrease up to one year after the maturity offset, which is identifiable with the achievement of the peak height velocity, and then, it reaches a plateau. However, in athletes, the age at peak height velocity can be lower than that measured in the general population [30]. This may represent a common scenario in elite teams, as often there is a tendency to select taller athletes, which is typical in mature adolescents.

BIVA is an effective tool to assess body composition in male and female adult athletes [17,23], although there are no BIVA references to female adolescent athletes and no studies associating BIVA and handgrip strength in adolescent individuals.

In this study with adolescent athletes, BIVA confidence ellipses were sensitive both to age and handgrip strength. Confidence ellipses of older and stronger individuals shifted to the left, indicating increased cell mass and fluid content, which can be attributed to better cell functioning [17], which is consistent with growth development and physical training. It was also noticed that the ellipses of the female group had the same displacement in age and strength categorizations. Ellipses of the male group kept the same general pattern, but there was increased distance in strength categorization.

The hypothesis behind BIVA's greater sensitivity to strength in male adolescents is related to maturity factors, in which the increasing strength is more relevant than chronological age. That means that strength reflects more the increase in body cell mass (especially FFM) and fluid content than age in male individuals. Although there is a slight overlap in both sexes' ellipses in low strength groups, the Hotelling T<sup>2</sup> test was able to identify a significant difference. Since confidence ellipses presented 95% probability, even a slight overlap could not affect the significance of Mahalanobis distance [17]. In this study, from the reference population, tolerance ellipses showed that most individuals were inside the 95% tolerance ellipses. The presence of female adolescents outside the ellipse may be explained by their better training status, which is reflected in higher cell mass; and male adolescents outside the ellipse may be explained by their hypohydration status expressed in long impedance vectors and reinforced by low total body water values (≤50% from weight).

A positive point of this study is a sample size (112 females and 161 males). Additionally, participants were measured in the same physical training conditions. These

characteristics are particularly important to BIVA quality and applicability. Some limitations should be acknowledged. First, the present results refer to adolescent athletes and should not be generalized. Second, the bioelectrical parameters were measured using a foot-to-hand technology at 50 kHz frequency and should not be compared with the different technologies or data obtained at different sampling frequencies. Lastly, unfortunately, in the present study, it was not possible to assess the biological maturity status of the participants. However, our results are in agreement with other studies that used chronological age [26,34,39,40] and maturity status [32,38]. Deuremberg et al. [41] observed that a specific impedance was positively related with age until 13 years for both sexes, after which sex differences became apparent.

The assessment of BIVA patterns may assist in comparing adolescent athletes and identifying changes in body composition and the correlated hydration and cell mass qualitative information. BIVA identified the influence of age and strength in vector displacement. As the results show, handgrip strength may be an easier way to express biological maturity changes because of its correlation to FFM and how easy it is to be obtained. In fact, growth differences in female and male individuals are marked by the higher gain in FFM (and strength) in male than in female adolescents.

Handgrip strength is an acceptable indicator of overall muscle strength and health at any stage of life, from childhood to older age. BIVA is a promising alternative for assessing muscle strength, with potential application in other population groups.

#### **5. Conclusions**

The assessment of BIVA patterns may assist in comparing adolescent athletes and identifying changes in body composition and the correlated hydration and cell mass qualitative information. BIVA identified the influence of age and strength in vector displacement. As the results show, handgrip strength may be an easier way to express biological maturity changes, because of its correlation to FFM and how easy it is to be obtained. In fact, growth differences in female and male individuals are marked by the higher gain in FFM (and strength) in male than in female adolescents. Handgrip strength is an acceptable indicator of overall muscle strength and health at any stage of life, from childhood to older age. BIVA is a promising alternative for assessing muscle strength, with potential application in other population groups.

**Author Contributions:** Conceptualization, J.C.K. and M.V.d.O.C.; data curation, B.T.S.; formal analysis, M.V.d.O.C., F.C. and J.C.K.; investigation, J.C.K., M.V.d.O.C., F.C. and B.T.S.; writing—original draft preparation, J.C.K.; M.V.d.O.C.; B.T.S., F.C. and writing—review and editing, M.V.d.O.C., B.T.S., F.C. and J.C.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Coordenação de Aperfeiçoamento de Nível Superior— Brasil (CAPES)—Finance code 001 and Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro—(FAPERJ) E-26/010.001769/2019.

**Institutional Review Board Statement:** Protocols used in this study were approved by the Ethics Committee of Pedro Ernesto University Hospital (CEP/HUPE 649.202) and the Public Secretariat of Education (07/005.242/14). These protocols align with the Declaration of Helsinki of 1964.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Acknowledgments:** The authors thank all participants involved in this study and appreciate the collaboration of the school in carrying out the experiments, and Antonio Piccoli (Padua University, Italy) for kindly providing BIVA software (*in memorian*).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


International Journal of *Environmental Research and Public Health*

## *Article* **Can Neurocognitive Function Predict Lower Extremity Injuries in Male Collegiate Athletes?**

**Sunghe Ha 1,2,**† **, Hee Seong Jeong 1,2,**† **, Sang-Kyoon Park 3,\* ,**‡ **and Sae Yong Lee 1,2,4,\* ,**‡


Received: 17 September 2020; Accepted: 2 December 2020; Published: 4 December 2020

**Abstract:** The purpose of this study is to demonstrate whether neurocognitive evaluation can confirm the association between neurocognitive level and postural control and to analyze the relationship between neurocognitive level and acute musculoskeletal injury in male non-net sports athletes. Seventy-seven male non-net sports athletes participated in this study. The Standardized Assessment of Concussion (SAC), Landing Error Scoring System (LESS), Balance Error Scoring System (BESS), and Star Excursion Balance Test (SEBT) were used for testing; we collected data related to injury history for six months after testing. Pearson's correlation analysis, logistic regression, and the independent sample *t*-test were used for statistical analysis. The correlation between SAC and SEBT results was weak to moderate (*p* < 0.05). Eleven of the seventy-seven participants experienced acute lower limb injuries. SAC, LESS, BESS, and SEBT results have no effect on the occurrence of acute lower extremity injuries (*p* > 0.05) and were not statistically different between the injured and non-injured groups (*p* > 0.05). Therefore, using the SAC score alone to determine the risk factor of lower extremity injuries, except in the use of assessment after a concussion, should be cautioned against.

**Keywords:** lower limb; men; non-net sports; prevention; screening

#### **1. Introduction**

The occurrence of concussion among male elite athletes participating in contact sports is reported to be higher than that of women, accounting for approximately 66%–76% of the overall incidence [1–4]. In particular, the incidence of concussion was highest in adolescents and young adults [4]. Because of the nature of non-net sports such as contact with other players or objects, shocks to the head, neck, and upper body are frequent, and the accumulation of these shocks causes serious problems, for instance, concussions [5] and the possibility of cognitive decline [6]. According to a recent study analyzing the causal relationship between cognition and musculoskeletal injury, musculoskeletal injury may occur at a high level when participating in sports if the level of cognition is low or lowered owing to concussion [7–9]. Elite athletes who returned to sports after a concussion showed that the likelihood of acute lower musculoskeletal injury was increased compared with non-injured athletes [10,11].

Musculoskeletal injuries cause joint instability, recurrent injury, and other site injuries, as well as premature degenerative osteoarthritis [12] and accelerated retirement [13]. Various field studies have been reported to reduce the incidence of musculoskeletal injury, but several injuries have still been reported [14–16]. Along with current approaches to reduce sports injuries, new and efficient methods are needed in the field of sports. Most studies have suggested a link between cognition and sports injuries using expensive equipment such as computerized neurocognitive testing [17,18], magnetic response imaging [19], and electroencephalography [20]. However, a tool that can efficiently assess risk factors of musculoskeletal injury to athletes through a paper-and-pencil method and that is less expensive and requires less time than the computerized methods currently available is needed.

The Standardized Assessment of Concussion (SAC) was designed to quickly apply, observe, and evaluate the orientation, immediate memory, concentration, and delayed memory of an athlete with a head injury [21]. A reduction in SAC score according to head impact can be said to be a neurocognitive dysfunction [22]. Neurocognitive screening, the third item of the Sport Concussion Assessment Tool-Fifth Edition, is composed of SAC and is the most used method in the field [23,24]. However, its application as an assessment tool for neurocognitive impairment with regard to the accumulation of repetitive shocks during participation in non-net sports is insufficient. Repeated impact on the head results in decreased neurocognitive function, resulting in musculoskeletal injuries when participating in sports.

Due to the notion supported from neuroimage studies suggesting that the motor and neurocognitive process possesses the common neural pathway and resources [25–27], the studies trying to identify an association between neurocognitive function and postural control has been conducted [28,29]. This can explain the possibility of impairment of neuromuscular control if neurocognitive function has been damaged by repetitive impact. The association between neurocognitive function and motor skill may be affected by level of neurocognitive control process (difficulty/complexity of the task). Therefore, different types of field tasks assessing motor skills (e.g., static and dynamic postural control) should be employed to address and evaluate its association with neurocognitive function.

Static and dynamic balances are considered an important aspect of performance and injury risk of the lower extremity in many athletic events. Balance Error Scoring System (BESS), Landing Error Scoring System (LESS), Star Excursion Balance Test (SEBT), etc., used without expensive equipment in a clinical setting are applied to assess static and dynamic postural control capabilities, which are reported to be predictable for ligament sprains of the lower extremity [30–32]. These postural control test methods are not only used as baseline tests to monitor players before the season but also as evaluation criteria for returning to sports.

This study aims to verify whether neurocognitive assessment can identify association between neurocognitive level and postural control and analyze the association between the neurocognitive level and the occurrence of acute musculoskeletal injuries in male non-net sports athletes. This study hypothesized that there would be correlations between the neurocognitive evaluation scores and scores of postural control of the lower extremity, and the neurocognitive evaluation scores could predict acute lower limb injuries.

#### **2. Methods**

#### *2.1. Participants*

Seventy-seven male elite college players of 14 basketball, 22 rugby, 11 baseball, 15 ice hockey, and 15 soccer players participated in this study (height, 180.0 ± 7.4 cm; body weight, 83.9 ± 15.0 kg; age, 19.7 ± 1.3 y). All selected participants had no history of orthopedic acute injury or concussion in the previous six months, were registered as elite athletes, and participated in training and competition.

#### *2.2. Experimental Design*

This study was conducted with the approval of the Bioethics Committee of Yonsei University (7001988-201810-HR-465-03). Informed consent of voluntary participation was received from all study participants. After receiving informed consent, neurocognitive examination to detect injury risk of

the lower extremity was performed using the screening tool, and follow-up investigation on injury occurrence was conducted.

#### 2.2.1. Standardized Assessment of Concussion

As a noninvasive tool for determining brain dysfunction resulting from to sports concussion, the Korean version of SAC, which was verified for reliability and validity, was used for the neurocognitive evaluation test (conformity, 0.88–1.1; external compatibility, approximately 0.55–1.45; separation index, 4.59; separation reliability, 0.95) [33]. SAC is a form of scoring an answer through a tester's question and consists of a mental test, an immediate memory test, a concentration test (speaking in reverse order of numbers), a concentration test (subtracting 7 consecutive numbers from 100), and a delayed memory test. The total score for the Korean version is 37, with higher scores equating to better scores.

#### 2.2.2. Postural Control of the Lower Extremity

Two video cameras (HDR-PJ410; Sony, Tokyo, Japan; EOS 800D; Canon, Tokyo, Japan) were used to evaluate the performance of the lower extremities (sampling rate, 60 Hz). Performing action is an evaluation tool for predicting lower limb injury and returning to rehabilitation. LESS, BESS, and SEBT have been used in the sports field.

In LESS, when the subject is ready for the motion test on a 30 cm box, the subject jumps both feet lightly and lands at 50% of the height and then immediately performs the maximum vertical jump (Figure 1A). A detailed explanation to perform the task successfully was provided to the participants and practice trials were provided three times. No feedback was provided during the task. If the instructions were followed, it was considered to be successful.

**Figure 1.** Experimental set-up to detect the injury risk of lower extremities: (**A**) Landing Error Scoring System; (**B**) Balance Error Scoring System; (**C**), Star Excursion Balance Test.

The BESS uses an action in which the subject closes his eyes and puts both hands on the right and left iliac crests and maintains double-leg standing (feet stand together), single-leg standing (on the nondominant leg), and tandem standing (nondominant foot behind the dominant foot) for 20 s each (Figure 1B). The ground condition was carried out on a flat floor and soft board. Participants were provided with a full description of the movement and warm-up, and practices were conducted thrice for each movement. No feedback was provided during the task.

≤ In the SEBT (Figure 1C), the subjects stood on one leg in the center of the grid, with both hands placed on the left and right iliac crests. They were asked, along eight lines drawn at a 45-degree interval, to extend the legs as far as possible and touch the floor lightly with their toe. The direction of eight lines is as follows: anterior (A), anteromedial (AM), medial (M), posteromedial (PM), posterior (P), posterolateral (PL), lateral (L), and anterolateral (AL). If the subject falls off the fixed hand from the iliac crests and fails to stand on one foot or if the fixed footfalls or the foot fails to return to the starting position, it was considered a failure and was conducted again. Participants were provided with detailed instructions to successfully perform the task. A total of three practice tests were performed,

Differences in reaching distance ൌ | െ |

and no feedback was provided during the task. The task was considered successful if the participants followed the instructions successfully.

For the follow-up of injury occurrence, the injury investigation form used by the International Olympic Committee was used for six months after neurocognitive examination and motion analysis [34]. The injury was defined as occurring acutely in the musculoskeletal system of the lower extremity, except for progressive onset and chronic pain. The history of injury of some participants who have team trainers was directly recorded by the trainer on the online-injury surveillance system, which was developed by YISSEM based on recommendations of International Olympic Committee, and was collected, and soccer participants without team trainers were contacted individually every two months by the author, with their injuries being recorded.

#### *2.3. Data Processing*

The SAC calculated the sub-domains and total score, respectively, for orientation, immediate memory, concentration, and delayed memory.

The LESS scored the error action by observing the initial contact of the ground, the maximum knee flexion, and the overall landing in the sagittal and coronal planes [30]. The lower the error score, the better the landing. The higher the score, the worse the landing (excellent landing, ≤4 points; good landing, 5 points; normal landing, 6 points; wrong landing, >6). There are 17 LESS items for error scoring: (1) the knee flexion angle of <30 degrees at initial contact (IC); (2) thigh in line with the trunk at IC; (3) trunk vertical or in line with the hips at IC; (4) foot landing heel to toe or with a flat foot at IC; (5) center of the patella being medial to the midfoot at IC; (6) trunk lateral flexion at IC; (7) asymmetric initial foot contact; (8) stance width greater than the shoulder width at IC; (9) stance width less than the shoulder width at IC; (10) external rotation of the foot >30 degrees between IC and maximum knee flexion; (11) internal rotation of the foot >30 degrees between IC and maximum knee flexion; (12) knee flexion angle of <45 degrees between IC and maximum knee flexion; (13) thigh not flexing more on the trunk between IC and maximum knee flexion; (14) trunk not flexing more between IC and maximum knee flexion; (15) center of the patella being medial to the great toe during landing; (16) displacement of the trunk, hips, and knees during landing; and (17) overall impression during landing [30]. Items 1–15 receive 1 point each if the above conditions are met. In contrast, item 16 is evaluated as 1 point for average and 2 points for stiff, and item 17 is evaluated as 1 point for average and 2 points for poor [30]. Kinovea (version 0.8.27; Kinovea, https://www.kinovea.org/) software was used to evaluate the knee angle among the items by the same rater (reliability: 0.941).

The BESS was conducted for 20 s. If an error of operation was observed 5 s before, 10 points were scored [35]. If an error was observed after 5 s, 1 point was added for each error operation, but if multiple errors occurred simultaneously, they were treated as 1 point [35]. The lower the error score, the better, and each action score and the total were calculated. There are five BESS items for error scoring: (1) raising the hand from the iliac crest; (2) opening your eyes; (3) any step, stumble, or fall; (4) the hip joint moved to over a 30-degree abduction; and (5) lifting the forefoot or heel [35]. The BESS results between repeat assessments by the same rater were excellent (reliability: 0.980).

The SEBT was calculated by standardizing the distance reached in each direction by the leg length (from the anterior iliac spine to the medial malleolus). The difference in reaching distance between both sides was the difference in distance from the dominant to the nondominant and was presented as an absolute value (Equation (1)) [31]. The length of the lower extremity and the distance to reach it were calculated by the same rater using the ratio method of Kinovea software (reliability: 0.917).

$$\text{Differences in reaching distance } = \left| \text{Domain } \text{leg} - \text{Nondominant } \text{leg} \right| \tag{1}$$

The incidence of injured musculoskeletal injuries in 77 participants collected over a 6-month period was calculated by frequency, type, and cause of injury and recovery period.

#### *2.4. Statistical Analysis*

Spearman's rank correlation analysis was performed for confirming the applicability of SAC to the evaluation tool for postural control of lower extremities. Binary logistic regression was performed to estimate the causal relationship between injury occurrence and SAC, injury occurrence, and postural control evaluation tools, respectively. An independent *t*-test and Mann–Whitney *U*-test were used to verify the difference in SAC score between the groups according to the presence or absence of lower limb damage and the postural control of the lower limb. SPSS 25.0 (IBM Corp., Armonk, NY, USA) was used for all statistical analysis, and the statistical significance level was set to α = 0.05.

#### **3. Results**

#### *3.1. The Correlation between SAC and Evaluation Tool for Postural Control of Lower Limb*

Table 1 shows the correlation between the SAC score and the normalized reach distance of the SEBT's dominant and nondominant legs. A positive correlation was observed between the immediate memory score of SAC and the normalized reach distances of SEBT (P and PL of the dominant leg, and PM and P of the nondominant leg, respectively). A positive correlation was observed between the delayed memory and the normalized reach distance of the SEBT (PM, P, PL, and L of the dominant leg, and M, PM, P, and PL of the nondominant leg, respectively). Among the total scores of SAC and normalized P and PL reach the distance of SEBT of the dominant leg, a positive correlation was observed. The negative correlations were also observed between SAC (delayed memory and total score) and the M-direction difference between the dominant and nondominant legs (Table 2). On the other hand, no correlation was observed between SAC results compared with LESS, BESS, and SEBT results (*p* > 0.05).


**Table 1.** Correlation between neurocognitive testing score and normalized reaching distance in the Star Excursion Balance Test.

A, anterior; AL, anterolateral; AM, anteromedial; L, lateral; M, medial; P, posterior; PL, posterolateral; PM, posteromedial. \* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001.


**Table 2.** The correlation between neurocognitive testing score and differences of reaching distance in the Star Excursion Balance Test.

A, anterior; AL, anterolateral; AM, anteromedial; L, lateral; M, medial; P, posterior; PL, posterolateral; PM, posteromedial. \* *p* < 0.05.

#### *3.2. Injury History for 6 Months after Testing*

A total of 14 cases of acute musculoskeletal injury were reported in 77 participants collected over 6 months. Participants were classified into injured (*n* = 11) and healthy (*n* = 66) groups. One participant reported injuries to the ankle, lower leg, and hip, respectively, and the other participant reported injuries to the ankle and lower leg, respectively. The injured body parts were the ankle (*n* = 5, 35.7%), foot (*n* = 1, 7.1%), lower leg (*n* = 2, 14.3%), knee (*n* = 2, 14.3%), thigh (*n* = 2, 14.3%), and hip (*n* = 2, 14.3%). The types of injury included sprain (*n* = 4, 28.6%), strain (*n* = 3, 21.4%), bruise (*n* = 3, 21.4%), fracture (*n* = 1, 7.1%), ligament rupture (*n* = 1, 7.1%), cartilage injury (*n* = 1, 7.1%), and cramp (*n* = 1, 7.1%). The causes of injury were noncontact injury (*n* = 9, 64.3%), collision with other players (*n* = 4, 28.6%), and collision with moving objects (*n* = 1, 7.1%). The recovery period was 0 days (*n* = 6, 42.9%), 30 days or more (*n* = 5, 35.7%), 1 day (*n* = 1, 7.1%), 2 days (*n* = 1, 7.1%), and 7 days (*n* = 1, 7.1%).

#### *3.3. Predicting Injury Occurrence*

As a result of analyzing the accuracy of the classification of the injury occurrence group by logistic regression, the statistical significance of the individual independent variables for the presence or absence of injury was analyzed. Each result from logistic regression model suggests that the overall model was not found to be statistically significant (*p* > 0.05; Table 3). It was found that the independent variables (SAC, LESS, BESS, and SEBT) did not affect the presence or absence of injury (*p* > 0.05; Table 3).


**Table 3.** Final logistic regression results for the association of the variables with injuries.


**Table 3.** *Cont.*

BESS, Balance Error Scoring System; CI, confidence interval; df, degree of freedom; LESS, Landing Error Scoring System; OR, odds ratio; SAC, Standardized Assessment of Concussion; SEBT, Star Excursion Balance Test.

#### *3.4. Di*ff*erence between Injured and Healthy Groups*

Statistical differences between groups were not observed in the SAC score and lower limb function performance evaluation results (Table 4).


**Table 4.** The results of the independent *t*-test between non-injured and injured groups.

BESS, Balance Error Scoring System; df, degree of freedom; LESS, Landing Error Scoring System; M, mean; SAC, Standardized Assessment of Concussion; SD, standard deviation; SEBT, Star Excursion Balance Test; §, Independent *t*-test value. Values are expressed as mean ± standard deviation.

#### **4. Discussion**

This study aims to assess if neurocognitive assessment can identify risk factors of the lower extremity and to analyze the association between the neurocognitive level and the occurrence of acute musculoskeletal injuries in male collegiate athletes. The major findings of this study are twofold: first, SAC evaluating neurocognitive function of collegiate athletes and dynamic postural control have small to medium correlations, and second, however, lower extremity injuries cannot be predicted

using the SAC. As a result of this study, the correlation between SAC results (immediate memory, delayed memory, and total score) and SEBT result was observed. This result was consistent with our hypothesis. Even though motor skills including gross control such as balance, walking, agility, and flexibility were weakly associated with neurocognitive skills, neurocognitive function has been reported to be associated with motor skills, which may support the result of this study. The strength of link between neurocognitive function and motor skill is influenced by the difficulty of the task [36–38]. Therefore, SEBT, which is dynamic postural assessment tool, can be a difficult motor skill (novelty, complex/difficult task) that can be affected by neurocognitive function [25,39–41]. Since SEBT is more goal directed motor action and need specific (high-order) neurocognitive control process, SEBT is more likely to be affected by neurocognitive level than other balance task. According to the results of a prospective cohort using SEBT, lower limb injuries were reported in high school basketball players with normalized A, PM, and PL reach of 94% or less [31]. In addition, a systematic review has reported that SEBT is associated with an increased risk of injury [42,43]. This study suggests the possibility of predicting lower extremity injuries through correlation between the memory area that is the sub-domain of the SAC and SEBT result.

Based on the results of this study, it was found that lower extremity injuries cannot be predicted using the SAC. This result was inconsistent with our hypothesis. In a study conducted on alpine skiers and snowboarders, there was no difference in neurocognitive scores between the injured and non-injured groups, which is similar to the results of the present study [44]. Although the neurocognitive evaluation was applied to those who did not have a history of concussion within the past six months when recruiting the participants, it is thought that the cumulative shock received by the subjects participating in each of their sports during the six months of data collection may have resulted in impairment that may have affected neurocognitive function. These effects would have decreased physical function and increased the risk of musculoskeletal injury and concussion [45]. Therefore, it is necessary to periodically record and observe SAC scores for athletes who participate in non-net sports.

The criteria for judging sports concussion during training and competition are classified by coaching staff in the field according to the athlete's awareness and loss of consciousness. The understanding of both players and leaders regarding concussion was high, but there was a problem with the classification and management of the actual concussion [46,47]. The recovery period of a simple concussion is reported to occur immediately or within 10 days, depending on the degree [48,49]. However, some athletes do not fully recover from concussion and are more likely to be exposed to other injuries when returning to the field [11,50]. Athletes are reluctant to talk about their symptoms because of the fear of being excluded from the entry list. In the case of the subjects of this study, it is possible that less number of injuries was reported to the team trainer because they are selected to the professional team based on their grades on the university team.

In the SAC results of this study, no statistical difference was observed between the injured and non-injured groups. Ha reported that have no difference in scores of computerized neurocognitive tests between retired contact-sports athletes and control [51]. However, retired contact-sports athletes showed slower gait speeds during dual-task walking because the cognitive task is preceded any movement for walking [51]. A decline in neurocognitive ability may reduce the ability to cope with rapidly changing situations and increase the likelihood of injury. Previous studies have reported that cardiopulmonary training [52] and resistance training [53] improve memory and selective concentration, which are sub-domains of neurocognitive ability. Although athletes who participate in non-net sports repeatedly experience impacts on the head, it is thought that aerobic training and resistance training, in which the athletes regularly participated to improve physical ability, also influenced neurocognitive ability. Therefore, there might be difficulties in predicting injury through changes in neurocognition while participating in sports competitions and training. However, loss of proprioception information due to a past history of musculoskeletal injury and reduction in sports activities after retirements, such as understanding tactics and aerobic exercise, can cause a decrease in neurocognitive ability.

Nevertheless, this study has some limitations. First, the injury follow-up period was as short as six months. A previous study reported that 6–12 months after concussion were greater incidence of the lower extremity injury than 0 to 3 and 3 to 6 months [11]. In future studies, the duration of injury follow-up should be considered. Second, the injured group that was investigated had a relatively small sample size, which is a usual limitation in prospective research. Therefore, further research is needed because the resulting analysis is subject to limitations. Third, baseball and basketball, which have relatively low concussion rates, were included in this study. However, traumatic brain injury has been an important issue in baseball [54]. Some concussed baseball players showed no symptoms when they returned to play, but residual effects on their batting technique were reported [55]. Cognition and perception are the most important factors in playing basketball [56]. If a head injury such as a concussion occurs, these can be affected. Unfortunately, basketball had the highest competition-related rates of concussion for partial-contact sports such as soccer [57]. Fourth, the SAC has so far been used as a cognitive evaluation tool in sports sites [23], but this method was designed about twenty years ago. Because there is a point of contention in the method, which was developed some time ago, further research will be required for the development of a new cognitive evaluation tool that is both valid and reliable. Lastly, follow-up data about the injury incidence were not collected in the same way because the trainer's employment was different, which depended on the participants' team circumstances.

#### **5. Conclusions**

The SAC score of college male non-net sports players alone was unable to predict the occurrence of injury. Therefore, using the SAC score alone to determine the risk factor of lower extremity injuries, except in the use of assessment after a concussion, should be cautioned against.

**Author Contributions:** Study design, S.H., H.S.J., S.-K.P., and S.Y.L.; study conduct, S.H., H.S.J., S.-K.P., and S.Y.L.; data collection, S.H., and H.S.J.; data analysis, S.H., and H.S.J.; data interpretation, S.H., H.S.J., S.-K.P., and S.Y.L.; drafting manuscript, S.H., H.S.J. and S.Y.L.; revising manuscript content, S.H., H.S.J., S.-K.P., and S.Y.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2018S1A5B5A01032989).

**Acknowledgments:** The authors would like to thank all of the participants, as well as Jae Ho Kim and Kun Wang, for their help in collecting the data.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

International Journal of *Environmental Research and Public Health*
