Relative age in Brazilian swimmers and para swimmers

abstract This study aimed to analyze the relationships between swimming competitive events, functional classification and relative age and to detect if the relative age affects performance in Brazilian swimmers and para swimmers. Data were retrieved from public databases and birth dates were classified in four quartiles (Q1 = January-March; Q2 = April-June; Q3 = July-September; Q4 = October-December). Descriptive statistics, Chi-square tests and ANOVA were used. The 50 m freestyle is the competitive event with 2349 swimmers, followed by 100 m (n = 1817) and 200 m freestyle (n = 905), and 200 m butterfly with 42 swimmers. Para swimmers are mainly distributed to S14, S6 and S5 functional classes (n = 140, 87 and 45), and individual medley events were less represented at SM3, SM9 and SM11 (n = 1 swimmer). Most of swimming events (86.36%) and para swimming functional classes (51.43%) are represented by swimmers and para swimmers born in the first and second quartiles. Moderate associations between functional classification and relative age were observed (p < 0.0001, Cramer’s V = 0.277). No effects of relative age on swimmers and para swimmers’ performance were noticed (p > 0.05). The great functional class seems to be related to great para swimmers’ relative age. Quartiles distribution shows the advantage of being born in the first months of the year to be registered among the most talented Brazilian swimmers and para swimmers.


INTRODUCTION
Whether considered from an athlete development or public health perspective, the dynamic factors influencing sport participation and achievement are of key interest to researchers, policy makers, sports organizations and their practioners 1 .In terms of athlete development, relative age has emerged as a consistent influence on both immediate sport participation and longer-term attainment [1][2][3] .The relative age can be defined as the chronological age difference that occurs among athletes of the same age group and the consequences of this birthdate difference are commonly referred to as the relative age effect 2,4 .The relative age effects can reflect outcomes from an interaction between participants' birth dates and the dates used for chronological age grouping 2 and being relatively older within an age grouping is associated with consistent attainment and selection advantages, including an increased likelihood of selection to access further resources with athlete development systems 2,5 .Indeed, relatively older athletes obtain higher levels of sporting success in the short term, while the long-term trend is balanced and even reversed 2,6,7 .Competition performance on short or long-term in individual and team sports seems to be influenced by, among other factors, the relative age 6,7 and the body of scientific literature is smaller in individual than team sports 1,8,9 and males was pointed out as the predominantly studied population 3 .
Competitive swimming, like other sports disciplines 6 , is organized according to standard 1 year-old categories, based on participant's chronological age 9,10 and the prevalence of the relative age effects has been greatly observed 9,10 , with a highperceived precocity in elite competition 2 , detected more often in younger ages, males and some specific events 7 .The different determinants of swimming performance according to the event are related to the different anthropometric characteristics described in elite swimmers, according to the competitive level, specialty, and / or distance event 7,9 .For example, great muscular endurance and low active drag forces would be required for athletes in middle-distance or distance events compared with swimmers' specialists in short distance events 7,9 .Those factors suggest that a great prevalence of the relative age effects could be expected in competitive swimming, but the impact of the relative age on performance could be dependent on the event type (i.e.distance and swimming technique).Consequently, this might have implications on the strength demands and the propulsive forces applied 7 .According to Bezuglov et al. 11 similar effects are reported in Athletics disciplines, which creates discrimination against late-born and late-maturing children.
There is a scarcity of studies considering the relative age on para-athletes, which might be related to the small number of athletes engaged in para sports.Most of the available studies are limited to swimmers' samples when analyzing the relative age, which provides an incomplete picture of the whole phenomenon on athletes and para athletes.Para swimming is one of the most popular Paralympic sports and comprises sport classes based on physical, vision and intellectual impairments 12 .It is possible that para swimmers with impairment progress differently compared to non-disabled swimmers because of differences in training and performance characteristics, and the possibility that they start training at older ages after being impaired 13 .The relative age showed its greatest impact in the years associated with growth and maturational status and after that, its influence faded away 7 .
Whether considered from a public health or athlete development perspective, addressing factors that highlight health behaviours, such as sport participation in children and adolescence, are of interest to policy-makers, sports organisations and practitioners alike 2 .This study aimed to analyze the relationships between swimming competitive events, functional classification, and relative age and to detect if the relative age affects performance in Brazilian swimmers and para swimmers.We hypothesized that: (i) relative age would share variation with swimming competitive events, and the opposite would be noticed in the functional classification in Brazilian swimmers and para swimmers; (ii) quartiles distribution would affect swimmer's performance.Despite the significant previous relative age findings in swimming, this study might contribute to update the relative age analysis in swimmers and para swimmers performance considering a large amount of swimming events and functional classes, and guide coaches and competitive institution bodies to review the training, competition and talent selection process.

METHODS Participants
The present study is based on a short cohort database collected from the Brazilian Aquatic Sports Confederation from 2018 to 2020 and the Brazilian Paralympic Committee in 2021.Data collection was quite straightforward, and no specific hindrances were encountered.Swimmers registered in these public databases were added in the analysis.For the purposes of this study, 11.601 valid samples were considered.Swimmers with foreign backgrounds registered in the birth region were excluded from the analysis.

Relative Age Effect (RAE)
Databases included gender, birth month and place, competitive events and respective times, club affiliation and functional class.To analyze RAE, swimmers and para swimmers' birth month was assigned in birth quartiles.Thus, the first quartile (Q1) comprised sample born in January, February and March; the second (Q2), sample born in April, May and June; the third (Q3), sample born in July, August and September; the fourth (Q4) encompassed those who were born in October, November and December.

Classification of para swimmers
Para swimmers were selected if they met the classification conditions employed by the International Paralympic Committee or the sport's governing body.In swimming, the functional classification occurs as follows: physical-motor -S1 to S10, SB1 to SB9 and SM1 to SM10; visual -S11 to S13 or B11 to B13, SB11 to SB13 or B11 to B13 and SM11 to SM11 or BM11 to BM13; intellectual; S14, SB14 and SM1416.Regarding to data collection, the para swimmers' samples were included in all functional classes.

Statistical procedures
Median, interquartile range (IQR), minimum, maximum and percentiles were presented as Brazilian swimmers and para swimmers' descriptive statistics samples.The non-parametric chi squarechi-square of independence (expected values equal or more than five in at least 0.8 of the cells) or maximum likelihood ratio chi-square tests (expected values less than five in more than 0.2 of the cells) were used to determine associations between the relative age, swimming event and functional classification in swimmers and para swimmers.X 2 for each cell was calculated to show the direction of difference.The association between nominal variables relied on Cramer's V coefficient and the strength of association was characterized characterized as follows: 0 to 0.1 -little if any association; 0.1 to 0.3 -low association; 0.3 to 0.5 -moderately strong association and > 0.5 -very strong association 14 .
Two-way independent measures analysis of variance (ANOVA) was used to verify the interaction between quartiles and competitive event or functional classification effects on performance.Q-Q plots were checked for approximately normal distribution.We also performed a Robust ANOVA (RA) with 20% trimmed mean for dealing with outliers, unequal variance and non-normal distribution.To respect the data independence, athletes were included only once for each event participated.Partial eta squared (η 2 p) quantified the effect size and was interpreted as 0.01-0.06(small), 0.06-0.14(moderate), and > 0.14 (large 15 ).Simple main effects were used to verify the effect of one independent variable for each level of another independent variable.Only main events and functional classification with sufficient samples for analyzes were included in ANOVA.We set the alpha level at 0.05 and all statistical procedures were conducted in the software JASP 0.16.3.0 and Jamovi 2.3.

RESULTS
The freestyle is a competitive event with a high number of registered swimmers considering the short and middle-distance events.There is an overrepresentation of swimmers born in the second quartiles considering the official competitive events.Para swimmers are mainly distributed in the S14, S6 and S5 functional classification groups.There is an underrepresentation of swimmers born in the third and fourth quartiles.
Table 1 shows the association between the official competitive events and the relative age of swimmers, with no association identified.Expected values represent the distribution if quartiles are not associated with swimming stroke.X 2 for each cell figure in the parentheses.
Table 2 shows the association between the functional classification in paralympic swimming and the relative age, with an association identified with small magnitude.
Table 3 presents show the simple main effects of quartile for each main event.But there was no effect of quartile and swimming interaction in event time (Two-way ANOVA: df = 57; F = 0.692; p = 0.963; η 2 p = 0.004; RA: Q: 75.693; p = 0.1340).
Table 4 shows the simple main effects of quartile for each functional classification.However, there was no effect of quartile and functional classification interaction in swimming events time (df = 27; F = 0.847; p = 0.688; η 2 p = 0.055; RA: Q: 59.116; p = 0.1210).

DISCUSSION
The key findings of the present study showed that most of swimmers and para swimmers born in the 1st and 2nd quartiles and most swimmers were registered in the short and middle-distance freestyle events and the 200m butterfly was the less representative.Moderate associations were just noticed on para swimmers, which were partially in agreement with our hypotheses.No effect of quartile was noticed in swimmers and para swimmers' performance to refute our hypothesis.Despite the prevalence of the relative age effects in swimming has been previously reported, this is the first study that used national swimmers and para swimmers' public databases to identify shared variance and effects of relative age in performance in both groups.
The meaningful swimmer's representation, who were born in the 2nd quartile and the less amount of para swimmers who were born in the 3nd and 4nd quartiles of the year seem to evidence a discrimination against relatively younger children solely because they were born in certain months of the year 11,16,17 and corroborate with previous studies 6,9,18 .In fact, swimming is a sport in which the greatest prevalence of the relative age effects has been observed, with increased participation in championships for those born in the first quartiles 9,10 and a high perceived precocity in elite competition 1 .The increased proportion of swimmers in the 2nd quartile was already noticed in previous studies and this trend can be explained using different aspects (e.g., cultural importance of sports 19 .It would be logical to think that athletes born in the first months of selection year would be highly representative 1,4 , with a greater chance of reaching the highest professional sport level 5,6,20 .To reduce and eliminate relative age effects related inequalities, direct policy, organisational and practioner interventions are required 1 , since the relative age can only emerge from the simultaneous interaction of an age grouping policy and an individual's birth date 3 .In a sense our results add weight to that assumption for both, swimmers, and para swimmers, although further studies are needed to investigate whether early swimming dropout could be a circumstance somehow related to the inevitable differences in maturational development of young athletes, which can expose them to failure and frustration resulting in lower personal expectations of success and higher likelihood of quitting 1,7,20 . Swimming event distance and type revealed no associations with relative age, which has also not affected swimmers' performance.Previous studies mentioned that the impact of the relative age on each swimming event was inconsistent and was only visible in 100 m women butterfly after 12-years old10.Lorenzo-Calvo et al. 7 noticed differences in the influence of the relative age effects depending on the event technique and distance, as medley and butterfly swimming events and shorter event performances seemed to be most affected, contrarily to Brustio & Boccia 21 that found greater relative age effects in events with greater emphasis on metabolic requirements.Abbot et al. 10 revealed that the relative age affected performance times and the use of its raw values criteria to identify talented swimmers may be problematic.According to Brustio et al. 6 , when the level of competitiveness increased the effect size of the relative age effects also increased, suggesting that different trends in quartile distribution can be observed when considering corrected performances, minimizing or at least reducing the possible bias provided to those born near the selection data.Despite the absence of clear association between competitive swimming events and relative age, the influence of task constraints (i.e., strokes or distance) on the relative age effects should also be considered when designing an intervention with young developing athletes 7 .Because most swimmers rapidly specialize in a particular swimming technique, analysing the relative age in each competitive event is much more accurate than the analysis of the relative age effects on mean swimming performance 9 .The non-effects of quartiles in swimming events performance registered in the present study might be related to data grouping and respective variations in each swimming event.
Moderate association was noticed between functional classification and relative age, which has also not affected para swimmers' performance.Previous studies revealed that the most successful para swimmers can continue training and improving performance in their mid -to late twenties 12 .These ages of peak performance are older than those previously reported in non-disabled swimmers and tend to decrease as the event duration increases 22 .In the long-term, studies suggest that the relative age effects-related and maturation inequalities may be temporary and transient 1,2,6 , but definitive explanations for why they reduce and even reverse at the elite adult stage remain somewhat speculative and deserving of further attention 1 .Since previous studies have shown that classification broadly describes the type and severity of impairment and influences the relationship between age and competitive swimming performance 12 , it seems necessary to analyze the relative age effects of each class with a larger sample size.
This study is pioneer in presenting data of registered Brazilian swimmers and para swimmers in recent years and has reinforced some of the previous evidence noticed when analysing the relative age distribution in swimmers and has also clarified the absence of data on this topic in para swimmers.However, some limitations should be observed such as the number of swimmers per swimming event and para swimmers per classification, which hampers deep statistical analysis.Further longitudinal studies are recommended to analyze the influence of relative age combined to training-related and socio-economiccultural factors available in public databases on performance using linear and non-linear modelling approaches.

CONCLUSIONS
The present study findings indicated that most of Brazilian swimmers and para swimmers were born in the 1st and 2nd quartiles.A moderate association between relative age and functional classification was noticed.There was no quartile effect noticed in swimmers and para swimmers' performance.Further studies should predict swimmers and para swimmers' performance through linear and non-linear modelling using beyond the relative age, the environment and task constraints data accessible in large public databases.

COMPLIANCE WITH ETHICAL STANDARDS Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.This study was funded by the Amazonas Research Support Foundation (FAPEAM), Infrastructure Program for Young Researchers -First Project Program (PPP) Nº 004/2017 under grant number 062.01554/2018.).The present work also had the support of the National Council for Scientific and Technological Development (CNPq) for funding the Scientific Initiation Scholarship (PIB-S/0055/2022).

Table 1 .
Chi-square test between quartiles and main events in swimmers.

Table 2 .
Likelihood ratio between quartiles and functional classification in para swimmers.
Df: degrees of freedom

Table 3 .
Simple main effects of quartiles for each main event.
Df: degrees of freedom

Table 4 .
Simple main effects of quartiles for each functional classification.
Df: degrees of freedom