*ACTN3 (The Sprint Gene)*

In this review, fourteen studies were included which examined the association between the *ACTN3* rs1815739 (*R577X*) polymorphism and skeletal muscle phenotypes. Carrying the X allele was often associated with lower baseline muscle strength and function (Table 2). For example, in a study conducted by Kikuchi et al. [48], homozygosity for the X allele was associated with significantly poorer performance in the chair stand test compared to RR carriers (*p* = 0.024). Ma et al. [58] also found XX homozygotes to perform significantly worse in HG strength (*p* = 0.012), 5 m walk (*p* = 0.011) and TUG (*p* = 0.039) tests and to also have a significantly higher frailty index (*p* = 0.004). Similar results were observed by Judson et al. [46] in a group of 4163 females where RX and XX genotypes were significantly associated with fall incidence (*p* = 0.049, *p* = 0.02 respectively). In contrast, Delmonico et al. [37] found female XX homozygotes to have significantly higher absolute and relative KE peak power and peak velocity than carriers of the R allele (*p* < 0.05).

Individuals carrying the XX genotype were also shown to have significantly lower improvements in one repetition maximum (1RM) bench press and leg extension, vertical jump and sit-to-stand performance in response to speed and power training when compared to RR carriers (all *p* < 0.05) [63]. Pereira et al. [64] also demonstrated XX carriers to have significantly poorer improvements in 10 m sprint times in response to high speed and power training compared to RR homozygotes (*p* = 0.044). Similarly, female XX carriers were observed to have significantly lower improvements in relative KE peak power following RT compared to RR homozygotes (*p* = 0.02) [37]. In the male population, change in absolute KE peak power post RT approached significance when comparing RR and XX genotypes (*p* = 0.07) [37]. In contrast to the above studies, Delmonico et al. [38] found male XX homozygotes had a significantly greater increase in 400 m walk time when compared to RX/RR carriers (*p* = 0.03).

In a study conducted by Zempo et al. [77] XX homozygotes were observed to have significantly lower thigh muscle CSA compared to RR carriers (*p* = 0.04). Interestingly, in a secondary analysis comparing a middle age group with an old age group, XX homozygosity was only associated with low thigh muscle CSA in the old age group (*p* < 0.05), suggesting that the influence of ACTN3 deficiency is heightened with age [78]. Similar results were noted in 2017 by Cho et al. [32], where sarcopenia prevalence was significantly associated with the XX genotype (*p* = 0.038). In contrast, Lima et al. [54] found X allele carriers to have significantly more relative total FFM than RR homozygotes (*p* = 0.04).

Three studies found no significant differences in muscle phenotypes between *ACTN3* rs1815739 genotypes [30,39,59].

### *ACE*

The relationship between the *ACE* rs1799752 (insertion/deletion) polymorphism and skeletal muscle traits has been extensively investigated since the original study of Montgomery et al. in 1998 [79]. Thirteen articles are included in this review. Firstly, Charbonneau et al. [31] found that carriers of the DD genotype had significantly greater total FFM (*p* < 0.05) and lower limb muscle volume (*p* = 0.01) than II homozygotes. Similarly, in a study of 246 Brazilian females, Lima et al. [54] noted DD homozygotes to have a significantly greater SMI than I allele carriers (*p* = 0.044). These findings were further strengthened by Da Silva et al. [34], who demonstrated sarcopenia prevalence to be significantly higher in II genotype carriers compared to D allele carriers (*p* = 0.015) (Table 2). Interestingly, Lima et al. [54] showed that in response to RT, only *ACE* II homozygotes significantly increased AFFM (*p* < 0.001).

The II genotype was also associated with lower muscle strength and functional performance. For example, within a group of 431 Japanese individuals, Yoshihara et al. [76] found II homozygosity to be associated with significantly lower HG strength compared to D allele carriers (*p* = 0.004). Homozygosity for the I allele was also shown to associate with significantly poorer performance in the 6-min walking test and 8 ft TUG test (*p* = 0.008, *p* < 0.001 respectively) when compared to ID/DD genotypes. Furthermore, in response to RT, DD carriers achieved significantly greater improvements in 1RM bench press and sit-to-stand performance (*p* = 0.019, *p* = 0.013 respectively) [63]. Giaccaglia et al. [40] also found that DD genotype carriers achieved significantly greater improvements in concentric KE strength in response to RT compared to II homozygotes (*p* < 0.05). Similarly, Pereira et al. [64] observed that DD homozygotes became significantly quicker performing 10 m sprints (*p* = 0.012) compared to II carriers. Buford et al. [29] also reported that a 12-month exercise intervention evoked significant improvements in 400 m walking speed (*p* = 0.018) and short physical performance battery test (SPPB) scores (*p* = 0.015), but only in D allele carriers. Interestingly, II homozygosity was also significantly associated with developing mobility limitation at a 45% faster rate when compared to ID/DD carriers (*p* = 0.01) [52].

As with the *ACTN3* rs1815739 genotypes, three studies found rs1799752 genotypes to have no significant influence on skeletal muscle traits [30,39,59].

### *APOE*

Three studies demonstrated significant associations between the Apolipoprotein E (*APOE*) gene and muscle phenotypes (Table 3). A 6-year follow-up study conducted by Melzer et al. [60] found that e4 carriers displayed significantly slower gait speed and chair stand performance (*p* = 0.006, *p* = 0.015 respectively) at baseline and significantly slower chair stand performance (*p* = 0.034) at the end of the 6-year follow-up, compared to e3 carriers. The *APOE* e4 allele was also shown to be associated with a significantly larger decline in HG strength between the ages of 75 and 79 over a 4-year period, compared to non-carriers (*p* = 0.015) [68]. Furthermore, carriers of the e4 allele had significantly lower HG strength at age 79 compared to non-carriers (*p* = 0.006). Interestingly, the effect of the e4 allele on HG strength was significantly larger at age 79 than age 75 (*p* = 0.033), suggesting that the e4 allele becomes increasingly influential with age. In a 3-year follow-up study conducted by Verghese et al. [71], males carrying the e4 allele showed a significantly more rapid decline in gait speed and greater risk of disability than male non-carriers (*p* = 0.04, *p* = 0.007 respectively).

#### *UCP2* and *UCP3*

Three studies reported significant interactions between Uncoupling Proteins 2/3 (*UCP2*/*3*) polymorphisms and skeletal muscle traits. Firstly, in a group of 432 Caucasians, Crocco et al. [33] found carriers of the CC genotype of the *UCP3* rs1800849 polymorphism to exhibit significantly lower HG strength than carriers of the T allele (*p* = 0.010). Dato et al. [35], then showed that individuals carrying the AA genotype of *UCP3* rs11235972 polymorphism have significantly lower HG strength than GG homozygotes (*p* < 0.001). In 2015, Keogh et al. [47] demonstrated that GG carriers of *UCP2* rs659366 polymorphism perform significantly worse in the 8 ft TUG test compared with AA/GA genotypes (*p* = 0.045). However, post RT intervention, GG homozygotes of *UCP2* rs659366 had the greatest improvements in 8 ft TUG performance (*p* = 0.023).

#### Genome-wide Studies

Other genes that demonstrated significant associations with muscle phenotypes included the PR domain containing 16 (*PRDM16*) gene, Zinc finger protein 295 (*ZNF295*) gene and C2 calcium dependent domain containing 2 (*C2CD2*) gene (Tables 2 and 3) [44,70].

Moreover, a recent GWAS by Hernandez-Cordero et al. [80] evaluated genetic contribution to ALM in the UK Biobank dataset, comparing middle-aged (38–49 years) and elderly (60–74 years) individuals. A total of 182 genome-wide significant regions, many with multiple variants within them, were associated with ALM in middle-aged individuals. Of these, 78% were also associated with ALM in elderly individuals. Variants at three genes, *VCAM*, *ADAMTSL3* and *FTO*, had previously been associated with lean body mass in the UK Biobank [81]. Hernandez Cortez et al. also confirmed, in vitro, a functional role for *CPNE1* and *STC2* in myogenesis. In addition, the study highlighted five genomic regions, containing multiple genes, that are associated with muscle mass in both mice and humans.

#### **4. Discussion**

To the best of the authors' knowledge, this is the first systematic review to collate literature on genetic associations with muscle phenotypes relevant to sarcopenia. To date, most research targeting genetic associations with muscle phenotypes has not focused on elderly subjects, and thus, the genetic mechanisms underpinning the age-related changes in skeletal muscle traits are largely uncharted.

Given that the deterioration of skeletal muscle with advancing age can have profound consequences for patients and public health systems, improving our understanding of how genes influence this process is of paramount importance. This review has enhanced our knowledge surrounding the key genes and gene variants that may prove crucial in further developing our understanding of the pathogenesis of sarcopenia and improving prognosis and treatment interventions alike.

#### *4.1. Summary of Findings*

The systematic literature search identified 24 genes and 46 DNA polymorphisms whose expression was significantly associated with muscle phenotypes in older adults. Ten of these DNA polymorphisms (rs154410, rs2228570, rs1800169, rs3093059, rs1800629, rs1815739, rs1799752, rs7412, rs429358 and 192 bp allele) were significantly associated with muscle phenotypes in two or more studies. The complex and multifactorial mechanisms underpinning muscle regulation suggest that the accrual and loss of muscle mass and muscle strength is not reducible to one single gene or gene variant. The dynamic interactions between inhibitory and promotory pathways within the human body further highlight the importance of a holistic approach when considering genetic associations with skeletal muscle traits.

Nevertheless, the findings of this systematic review demonstrate that the most compelling current evidence in the field exists for the *ACTN3*, *ACE* and *VDR* genotypes.

#### 4.1.1. ACTN3 (The Sprint Gene)

The *ACTN3* gene is among the most extensively researched genes in relation to muscle phenotypes, and appeared most frequently within this review. The ACTN3 protein encoded by the *ACTN3* gene forms an integral part of the sarcomere Z-line in fast twitch muscle fibres and further aids in coordinating myofiber contractions [82,83]. Up to 20% of humans are deficient in this protein, due to homozygosity for the premature stop codon at the rs1815739 polymorphism [84]. This significant proportion of ACTN3 deficiency among the population suggests that X allele status is a key factor in variability in muscle phenotypes. In this regard, much of the research surrounding the *ACTN3* genotype has focused on athletic performance [85]. Association studies have repeatedly found reduced X allele frequency among elite sprint/power athletes [85–87]. This suggests that the presence of ACTN3 is crucial for the optimal generation of force. Considering that fast twitch muscle fibres are particularly susceptible to age-related atrophy [88], it is plausible that regulation of this protein may also be an important factor in understanding age-related changes in muscle phenotypes. To date, however, limited research has been conducted within elderly populations, with the result that the true impact of the *ACTN3* gene on age-related changes in muscle phenotypes remains inconclusive. Despite this, fourteen of the studies included in this review examining the *ACTN3* genotype reported promising findings. Carriers of the X allele were often found to display lower skeletal muscle mass, strength and functional abilities. This was particularly evident among the Asian population. All five cross-sectional studies that examined Asian participants found significant associations between X allele status and muscle phenotypes [32,48,58,77]. No such association was found in the other three cross-sectional

studies that targeted Caucasian individuals [30,39,59], therefore suggesting ethnicity may determine the degree to which *ACTN3* genotypes effect aging muscle. This coincides with existing research whereby X allele frequency and fast twitch fibre composition have been shown to vary across different ethnic groups [89–92]. The Asian population have the highest frequency of the X allele [89], while having the lowest percentage of fast twitch muscle fibres [90–92], two likely contributing factors in the ethnic group having the highest sarcopenia prevalence globally [93]. Unlike above, X allele status was significantly associated with training adaptation within Caucasian, North-American and South-American individuals. Thus, the inconsistencies within this review highlight the need for future research to provide clarification on how ethnicity, *ACTN3* genotypes and muscle phenotypes are associated within the elderly.

### 4.1.2. ACE

Like the *ACTN3* gene, the *ACE* gene has been widely researched within athletic populations, and knowledge within older populations is limited. There are, however, compelling molecular pathways controlled by the *ACE* gene that suggest its importance in age-related changes in muscle phenotypes. The ACE is expressed by skeletal muscle endothelial cells, and catalyses the production of angiotensin II, known to enhance skeletal muscle hypertrophy [94,95]. To date, research in relation to muscle phenotypes has centred around the *ACE* rs1799752 polymorphism. The D and I alleles have been associated with higher and lower ACE activity respectively [96–98]. The D allele is suggested, therefore, to associate with greater muscle performance. To support this hypothesis, recent studies have focused on the rs1799752 polymorphism in elite athletes, with interesting findings. The I allele has been repeatedly associated with endurance performance, while the D allele associates with strength/power capabilities [99,100]. Findings from this systematic review further strengthen these observations. The D allele was consistently associated with higher baseline muscle strength and functional performance, as well as greater improvements in muscle strength and function in response to RT. Evidence of the association between the *ACE* rs1799752 polymorphism and muscle mass is less definitive. While the D allele was often associated with greater amounts of FFM, contradictory findings were also in evidence, and thus, further research is needed in this area to reach a consensus. Like with *ACTN3* genotypes, frequency of the I and D allele of the *ACE* gene are highly determined by ethnic background. Asians have been shown to have the highest frequency for the undesirable I allele [101], while African-American have the lowest [101], aligning with global sarcopenia prevalence estimates where Asians and African-Americans have the highest and lowest risk respectively [93]. While evidence in this review is insufficient in highlighting a true ethnic impact on the association between *ACE* genotypes and aging muscle phenotypes, the disparity in allele frequency among different ethnicities is promising.

#### 4.1.3. VDR

The true significance of the association between the *VDR* gene and muscle phenotypes is currently unknown. While the *VDR* gene has been extensively researched, findings are often contradictory. Furthermore, due to its crucial role in regulating calcium absorption, much of the existing research has focused on the association between *VDR* genotypes and bone health [102]. However, the *VDR* gene is also known to stimulate changes in muscle protein synthesis through its key regulatory role in the transcription of messenger RNA [103], and thus, the potential of the *VDR* gene as a candidate gene for muscle phenotype associations has been suggested. More specifically, the rs2228570 polymorphism is the only known VDR polymorphism where variation results in structural changes within the VDR protein due to differences in translational initiation sites [104]. The *VDR* f allele results in a full length VDR protein of 427 amino acids [105], while a *VDR* F allele results in a truncated VDR protein with three amino acids less [106]. Interestingly, three of four studies that examined the rs2228570 polymorphism in this review found F allele carriers to perform significantly worse across a range of muscle phenotypes [45,67,74], suggesting the potential importance of the rs2228570 polymorphism.

While compelling evidence exists supporting the importance of the *VDR* gene for muscle phenotypes, many studies have failed to replicate earlier results, and thus, the strength of this association remains to be established [107,108]. Unlike for *ACTN3* and *ACE* polymorphisms, evidence of an ethnic influence on *VDR* polymorphism frequency is conflicting [109,110]. As with most genetic association studies, much of the research surrounding *VDR* polymorphisms and muscle phenotypes has been conducted using Caucasian subjects. Only nine articles examining *VDR* genotypes were included in this review, seven of which focused on Caucasian individuals [26–28,42,45,62,67]. Furthermore, as with the *ACTN3* and *ACE* genes, limited research has been conducted within an elderly population, further limiting the transferability of findings for older adults.

#### 4.1.4. Other Genes of Interest

Other genes with convincing molecular pathways and findings, that warrant future investigation include the *IGF1*/*IGFBP3*, *TNF*α, *APOE*, *CNTF*/*R* and *UCP2*/*3* genes.

### 4.1.5. IGF1 and IGFBP3

The *IGF* family of genes encode peptides that are crucial in regulating cell proliferation, apoptosis and differentiation [111]. The mitogenic effect of IGF1 is integral to the facilitation of growth in multiple tissues, including skeletal muscle [112]. Considering that advancing age is associated with a decline in circulating IGF1 levels, the *IGF1* gene is a likely candidate to effect muscle phenotypes among the elderly [113]. The current review found significant associations between *IGF1* variants and skeletal muscle mass and strength. Associations were particularly convincing in longitudinal studies, suggesting that the *IGF1* 192 polymorphism may be particularly influential in the strength-training response of skeletal muscle phenotypes as opposed to baseline measurements.

The function of IGF1 is mediated through interactions with binding proteins, mainly, IGFBP3. Research has demonstrated that IGFBP3 is the most prolific potentiator of IGF1, therefore suggesting its importance in explaining inter-individual variation in muscle phenotypes [114]. While only Yang et al. [75] have investigated the impact of the *IGFBP3* gene in an elderly population, the significant findings of that study combined with the relevant gene mechanisms warrants future research.
