Genes and Athletic Performance: The 2023 Update

henotypes of athletic performance and exercise capacity are complex traits in uenced by both genetic and environmental factors. This update on the panel of genetic markers (DNA polymorphisms) associated with athlete status summarises recent advances in sports genomics research, including ndings from candidate gene and genome-wide association (GWAS) studies, meta-analyses, and ndings involving larger-scale initiatives such as the UK Biobank. As of the end of May 2023, a total of 251 DNA polymorphisms have been associated with athlete status, of which 128 genetic markers were positively associated with athlete status in at least two studies (41 endurance- related, 45 power-related, and 42 strength-related). The most promising genetic markers include theAMPD1rs17602729 C,CDKN1Ars236448 A,HFErs1799945 G,MYBPC3rs1052373 G,NFIA- AS2rs1572312 C,PPARArs4253778 G, andPPARGC1Ars8192678 G alleles for endurance;ACTN3 rs1815739 C,AMPD1rs17602729 C,CDKN1Ars236448

May 2023, a total of 251 DNA polymorphisms have been associated with athlete status, of which 128 genetic markers were positively associated with athlete status in at least two studies (41 endurance- related, 45 power-related, and 42 strength-related). The most promising genetic markers include theAMPD1rs17602729 C,CDKN1Ars236448 A,HFErs1799945 G,MYBPC3rs1052373 G,NFIA- AS2rs1572312 C,PPARArs4253778 G, andPPARGC1Ars8192678 G alleles for endurance;ACTN3 rs1815739 C,AMPD1rs17602729 C,CDKN1Ars236448 C,CPNE5rs3213537 G,GALNTL6rs558129 T, IGF2rs680 G,IGSF3rs699785 A,NOS3rs2070744 T, andTRHRrs7832552 T alleles for power; and ACTN3rs1815739 C,AR 21 CAG repeats,LRPPRCrs10186876 A,MMS22Lrs9320823 T,PHACTR1 rs6905419 C, andPPARGrs1801282 G alleles for strength. It should be appreciated, however, that elite performance still cannot be predicted well using only genetic testing. Keywords: sports; genetics; genotype; polymorphism; genomics; physical performance; athletes; GWAS; WGS; WES 1. Introduction Athletic success is in uenced by many genetically determined factors, including transcriptomic, biochemical, histological, anthropometric, physiological, and psychological traits, as well as general health status [1–8]. On average, 66% of the variance in athlete status can be explained by genetic factors [9]. The remaining variance is due to environmental factors, such as deliberate practice, nutrition, ergogenic aids, birthplace, the availability of medical and social support, and even luck (e.g., birthdate) [6,10–12]. Starting in the late 1990s, research began to identify DNA polymorphisms associated with predisposition to certain types of sports and exercise-related phenotypes, with initial fo- cus on variants of theACE,ACTN3,AMPD1,PPARA,PPARD, andPPARGC1Agenes[13–23] . Initially, most research was conducted using the candidate gene approach [24–29], which limited progress in the discovery of new genetic markers associated with exercise- and sport-related phenotypes [30]. In addition to the fact that this approach studies only a single genetic variant in isolation, most candidate gene studies in the eld of sports genomics are limited by sample size. This is a potential source of type I error (false positive ndings), underpinning why replication of positive associations in independent cohorts is essential. Genes2023,14, 1235.

Genes2023,14, 1235 2 of 32 Conversely, the genome-wide association approach is considered the most ef cient study design thus far in identifying genetic markers associated with sport-related characteristics. Indeed, the application of this approach has enabled the discovery of hundreds of single nucleotide polymorphisms (SNPs) directly or indirectly associated with exercise and sport, such as height (12,111 SNPs) [31], appendicular lean mass (1059 SNPs) [32], testosterone levels (855 SNPs) [33], handgrip strength (170 SNPs) [34–36], sarcopenia (78 SNPs) [37], and brisk walking (70 SNPs) [38]. It is important to consider that the discovery of 12,111 independent genome-wide (p< 5 10 8 ) signi cant SNPs associated with height (which in combination account for 40% of phenotypic variance) required the study of 5.4 million individuals [31]. Accordingly, such ndings indicate that each DNA locus is likely to explain only a very small proportion of the reported phenotypic variance (~0.0033%). This also suggests that very large samples of athletes and non-athletic controls are needed to detect a substantial number of genetic markers associated with athlete status, with some suggesting those markers be used as part of talent identi cation strategies. To date, ve genome-wide association (GWAS) studies[39–43] , one whole-genome sequencing (WGS) study [44], and one exome-wide association (EWAS) study [45] involving athletic cohorts have been conducted. How- ever, due to their limited sample sizes, these investigations and subsequent replication studies have resulted in the identi cation of only 13 genetic markers. Six markers were associated with endurance athlete status (CDKN1Ars236448 A,GALNTL6rs558129 C, NFIA-AS2rs1572312 C,MYBPC3rs1052373 G,SIRT1rs41299232 G, andTRPM2rs1785440 G alleles)[39,40,42,43,45,46] , ve with sprinter/strength athlete status (AGRNrs4074992 C, CDKN1Ars236448 C,CPNE5rs3213537 G,NUP210rs2280084 A, andGALNTL6rs558129 T alleles) [41,45,47,48], and two with reaction time in wrestlers and athlete status in com- bat/team sports (APCrs518013 A andLRRN3rs80054135 T alleles, respectively) [44]. To overcome the issue of small samples of elite athletes, the use of genome-wide signif- icant (p< 5 10 8 ) markers of exercise-related phenotypes (correlated with sport-related traits) discovered in large cohorts of untrained subjects (for example, UK Biobank, FinnGen, BioBank Japan, etc.) was proposed [49,50]. This approach may signi cantly decrease the risk

andLRRN3rs80054135 T alleles, respectively) [44]. To overcome the issue of small samples of elite athletes, the use of genome-wide signif- icant (p< 5 10 8 ) markers of exercise-related phenotypes (correlated with sport-related traits) discovered in large cohorts of untrained subjects (for example, UK Biobank, FinnGen, BioBank Japan, etc.) was proposed [49,50]. This approach may signi cantly decrease the risk of obtaining false-positive results in the discovery of markers associated with athlete status, which is recognized as a key limitation of sports genomics studies conducted on limited sample sizes. Accordingly, the rst stage of such an investigation is to list can- didate genetic markers previously identi ed in GWAS of exercise-related phenotypes in non-athletic populations. For example, this may include SNPs associated with phenotypes of handgrip strength, brisk walking, physical activity, lean mass, forced vital capacity, haemoglobin, or levels of hormones such as testosterone and IGF1. At the second stage, microarray analysis (with imputation; covering > 10 million SNPs) is carried out in a group of athletes to test the hypothesis that favourable alleles of exercise-related phenotypes are over-represented in those athletes compared to non-athletic controls (Figure). This ap- proach may equally be applied to determine whether favourable alleles could be associated with direct measures of athletic capability, such as weightlifting performance, running times, or scores recorded during competitive events. In this case, the signi cance threshold for the second stage is set atp< 0.05 (to reduce the likelihood of potentially important ndings being overlooked). A recent example of such an approach being implemented is the study by Guil- herme et al. [49], which investigated two correlated phenotypes: brisk walking pace (using UK Biobank participants) and sprint athlete status (using elite Russian sprinters). Brisk walkers perform more physical activity, are taller, have reduced adiposity and demonstrate greater physical performance and strength versus slower walkers, with such traits also recorded more commonly in sprinters than other athletes of other disciplines. Therefore, it was hypothesized that the alleles associated with high-speed walking (discovered in un- trained subjects) would also be over-represented in elite sprinters. Accordingly, 70 genetic markers of brisk walking were

taller, have reduced adiposity and demonstrate greater physical performance and strength versus slower walkers, with such traits also recorded more commonly in sprinters than other athletes of other disciplines. Therefore, it was hypothesized that the alleles associated with high-speed walking (discovered in un- trained subjects) would also be over-represented in elite sprinters. Accordingly, 70 genetic markers of brisk walking were identi ed from the literature [38], of which 15 SNPs had a signi cantly different allele frequency when comparing sprinters with non-athletic con-

Genes2023,14, 1235 3 of 32 trols [49]. The same innovative approach later identi ed 23 SNPs associated with strength athlete status [51] based on genome-wide signi cant markers for handgrip strength in a non-athletic population using the UK Biobank [34,35]. Furthermore, using a panel of 822 testosterone-related SNPs from the UK Biobank study [33], ve DNA-polymorphisms associated with muscle bre size and weightlifting performance were identi ed [50].Genes 2023, 14, x FOR PEER REVIEW 3 of 31 genetic markers of brisk walking were identified from the literature [38], of which 15 SNPs had a significantly different allele frequency when comparing sprinters with non-athletic controls [49]. The same innovative approach later identified 23 SNPs associated with strength athlete status [51] based on genome-wide significant markers for handgrip strength in a non-athletic population using the UK Biobank [34,35]. Furthermore, using a panel of 822 testosterone-related SNPs from the UK Biobank study [33], five DNA-poly- morphisms associated with muscle fibre size and weightlifting performance were identi- fied [50]. Figure 1. Case-control study designs in sports genomics. In this approach, allelic frequencies are compared between athletes and controls (e.g., endurance athletes vs. untrained subjects or endur- ance vs. power athletes). A case-control study may be the first step followed by a genotype–pheno- type study (e.g., identification of VO 2max or weightlifting performance-increasing genotypes among athletes). In some cases, studies begin with a genotype–phenotype approach, and the findings are subsequently validated by a case-control study. Another approach that has proven effective in addressing the possibility of false pos- itive results in sports genomics literature is to perform replication studies in two or more independent athletic cohorts (even with small or moderate sample sizes), followed by a meta-analysis to quantify the overall effect of a polymorphism on athlete status and/or a sport- and exercise-related trait [43,52–65]. However, in some cases, replication is not pos- sible due to the exclusivity of a polymorphism to specific populations based on their geo- graphic ancestry. For example, the rs671 G/A polymorphism of the aldehyde dehydro- genase 2 (ALDH2) gene was associated with strength in athletes and non-athletes from the Japanese

on athlete status and/or a sport- and exercise-related trait [43,52–65]. However, in some cases, replication is not pos- sible due to the exclusivity of a polymorphism to specific populations based on their geo- graphic ancestry. For example, the rs671 G/A polymorphism of the aldehyde dehydro- genase 2 (ALDH2) gene was associated with strength in athletes and non-athletes from the Japanese population [66–68]. Interestingly, the unfavourable (associated with reduced strength) rs671 A allele is not present in Europeans or South Asians (frequency 0%), but common in Chinese, Japanese, and Vietnamese populations (15–25%). This demonstrates a notable challenge seeking to replicate genomic findings in larger samples, as increasing the study sample must also consider the geographic ancestry of participants. This also Figure 1. Case-control study designs in sports genomics. In this approach, allelic frequencies are compared between athletes and controls (e.g., endurance athletes vs. untrained subjects or endurance vs. power athletes). A case-control study may be the rst step followed by a genotype–phenotype study (e.g., identi cation of VO 2maxor weightlifting performance-increasing genotypes among athletes). In some cases, studies begin with a genotype–phenotype approach, and the ndings are subsequently validated by a case-control study. Another approach that has proven effective in addressing the possibility of false positive results in sports genomics literature is to perform replication studies in two or more independent athletic cohorts (even with small or moderate sample sizes), followed by a meta-analysis to quantify the overall effect of a polymorphism on athlete status and/or a sport- and exercise-related trait [43,52–65]. However, in some cases, replication is not possible due to the exclusivity of a polymorphism to speci c populations based on their geographic ancestry. For example, the rs671 G/A polymorphism of the aldehyde dehydrogenase 2 (ALDH2) gene was associated with strength in athletes and non-athletes from the Japanese population [66–68]. Interestingly, the unfavourable (associated with reduced strength) rs671 A allele is not present in Europeans or South Asians (frequency 0%), but common in Chinese, Japanese, and Vietnamese populations (15–25%). This demonstrates a notable challenge seeking to replicate genomic ndings in larger samples, as increasing the study sample must also

with strength in athletes and non-athletes from the Japanese population [66–68]. Interestingly, the unfavourable (associated with reduced strength) rs671 A allele is not present in Europeans or South Asians (frequency 0%), but common in Chinese, Japanese, and Vietnamese populations (15–25%). This demonstrates a notable challenge seeking to replicate genomic ndings in larger samples, as increasing the study sample must also consider the geographic ancestry of participants. This also highlights the possibility that the genetic determinants of some sport- and exercise- related phenotypes are restricted to certain populations, demonstrating that increasing sample size is not as straightforward as simply recruiting participants from multiple countries and/or continents.

Genes2023,14, 1235 4 of 32 As well as the phenotypes of athlete status or competitive performance, several recent studies have investigated a broader range of traits which may relate directly or indirectly to athletic capability. These include exibility, coordination, cardiorespiratory tness, spatial ability, stress resilience, mental toughness, fat loss ef ciency, and cardiovascular and metabolic responses to training, amongst others [69–84]. For example, combat athletes are more likely than untrained subjects to have the warrior (COMTrs4680 GG) genotype [85], whilst chess players demonstrate an increased frequency of an allele linked to improved memory and spatial ability (KIBRArs17070145 T) [86]. Such discoveries demonstrate the broadening nature of sports genomics in recent times, with focus expanding from the traditional domain of investigating what makes elite performers different from the general population into other domains, such as sports nutrigenetics [87–100] and areas of sports medicine, such as genomic variants associated with soft-tissue injuries and sports-related concussion [101–114]. Technological advancement has lowered the cost of conducting genomic studies, increasing accessibility to researchers who wish to investigate the genetic underpinnings of sport and exercise phenotypes. Consequently, sports genomics is a dynamic and continually developing eld, making it important to regularly appraise the contribution of recent advances to the eld. Therefore, the aim of the current review was to summarise recent progress in understanding the genetic determinants of athlete status, and to detail novel DNA polymorphisms that may underpin differences between individuals in their athletic potential. At the time of writing (end of May 2023), the total number of DNA polymorphisms associated with athletic performance since the rst discovery in 1998 is 251 (Figure). Our search for sports genomics publications was based on journals indexed in major databases (i.e., PubMed etc.) using speci c key words (e.g., athletes + polymorphism/genotype etc.). However, not all articles were included in the current review due to language limitations (articles written in languages other than English must contain at least abstracts in English). In addition, papers with very small cohort (less than 25 in athletes/controls), or articles with combined groups of athletes (for example, endurance + power without separation) were

(e.g., athletes + polymorphism/genotype etc.). However, not all articles were included in the current review due to language limitations (articles written in languages other than English must contain at least abstracts in English). In addition, papers with very small cohort (less than 25 in athletes/controls), or articles with combined groups of athletes (for example, endurance + power without separation) were not included. Abstracts of conference proceedings were not considered. In recognition of the fact that many studies in the eld of sports genomics report associations based on the investigation of small sample sizes, we stipulated that only markers where statistically signi cant associations have been reported in at least two studies (two case-control studies and/or one case-control plus one functional study; including those presented in one article) would be included in the present review.Genes 2023, 14, x FOR PEER REVIEW 4 of 31 highlights the possibility that the genetic determinants of some sport- and exercise-related phenotypes are restricted to certain populations, demonstrating that increasing sample size is not as straightforward as simply recruiting participants from multiple countries and/or continents. As well as the phenotypes of athlete status or competitive performance, several re- cent studies have investigated a broader range of traits which may relate directly or indi- rectly to athletic capability. These include flexibility, coordination, cardiorespiratory fit- ness, spatial ability, stress resilience, mental toughness, fat loss efficiency, and cardiovas- cular and metabolic responses to training, amongst others [69–84]. For example, combat athletes are more likely than untrained subjects to have the warrior (COMT rs4680 GG) genotype [85], whilst chess players demonstrate an increased frequency of an allele linked to improved memory and spatial ability (KIBRA rs17070145 T) [86]. Such discoveries demonstrate the broadening nature of sports genomics in recent times, with focus expand- ing from the traditional domain of investigating what makes elite performers different from the general population into other domains, such as sports nutrigenetics [87–100] and areas of sports medicine, such as genomic variants associated with soft-tissue injuries and sports-related concussion [101–114]. Technological advancement has lowered the cost of conducting genomic studies, in- creasing accessibility to researchers who

with focus expand- ing from the traditional domain of investigating what makes elite performers different from the general population into other domains, such as sports nutrigenetics [87–100] and areas of sports medicine, such as genomic variants associated with soft-tissue injuries and sports-related concussion [101–114]. Technological advancement has lowered the cost of conducting genomic studies, in- creasing accessibility to researchers who wish to investigate the genetic underpinnings of sport and exercise phenotypes. Consequently, sports genomics is a dynamic and continu- ally developing field, making it important to regularly appraise the contribution of recent advances to the field. Therefore, the aim of the current review was to summarise recent progress in understanding the genetic determinants of athlete status, and to detail novel DNA polymorphisms that may underpin differences between individuals in their athletic potential. At the time of writing (end of May 2023), the total number of DNA polymorphisms associated with athletic performance since the first discovery in 1998 is 251 (Figure 2). Our search for sports genomics publications was based on journals indexed in major databases (i.e., PubMed etc.) using specific key words (e.g., athletes + polymorphism/genotype etc.). However, not all articles were included in the current review due to language limitations (articles written in languages other than English must contain at least abstracts in English). In addition, papers with very small cohort (less than 25 in athletes/controls), or articles with combined groups of athletes (for example, endurance + power without separation) were not included. Abstracts of conference proceedings were not considered. In recogni- tion of the fact that many studies in the field of sports genomics report associations based on the investigation of small sample sizes, we stipulated that only markers where statisti- cally significant associations have been reported in at least two studies (two case-control studies and/or one case-control plus one functional study; including those presented in one article) would be included in the present review. Figure 2. Sports-related genetic markers discovered between 1998 and 2023. Figure 2.Sports-related genetic markers discovered between 1998 and 2023. According to these criteria, 128 markers could be associated with athlete status (41 endurance-related,

at least two studies (two case-control studies and/or one case-control plus one functional study; including those presented in one article) would be included in the present review. Figure 2. Sports-related genetic markers discovered between 1998 and 2023. Figure 2.Sports-related genetic markers discovered between 1998 and 2023. According to these criteria, 128 markers could be associated with athlete status (41 endurance-related, 45 power-related, and 42 strength-related) from the original 251 iden- ti ed in our literature search. The most promising genetic markers (i.e., most replicated

The article reviews genetic factors influencing athletic performance and summarizes recent findings.