ACTN3 genotype is associated with muscle phenotypes in women across the adult age span

Abstract

The R577X polymorphism in the α-actinin-3 encoding gene (ACTN3) has been associated with elite athletic performance, and recently with differences in isometric and dynamic muscle strength and power in the general population. In this study we sought to determine the association of ACTN3 R577X genotype with muscle strength and mass phenotypes in men and women across the adult age span. Eight hundred forty-eight (n = 848) adult volunteers (454 men and 394 women) aged 22–90 yr were genotyped for ACTN3 R577X. Knee extensor (KE) shortening and lengthening peak torque values were determined using isokinetic dynamometry and fat-free mass (FFM) by dual-energy X-ray absorptiometry. Women deficient in α-actinin-3 (X/X; n = 53) displayed lower KE shortening peak torque (30°/s: 89.5 ± 3.5 vs. 99.3 ± 1.4 N·m, P = 0.011; 180°/s: 60.3 ± 2.6 vs. 67.0 ± 1.0 N·m, P = 0.019) and KE lengthening peak torque (30°/s: 122.8 ± 5.7 vs. 137.0 ± 2.2 N·m, P = 0.022; 180°/s: 121.8 ± 5.8 vs. 138.5 ± 2.2 N·m, P = 0.008) compared with R/X + R/R women (n = 341). Women X/X homozygotes also displayed lower levels of both total body FFM (38.9 ± 0.5 vs. 40.1 ± 0.2 kg, P = 0.040) and lower limb FFM (11.9 ± 0.2 vs. 12.5 ± 0.1 kg, P = 0.044) compared with R/X + R/R women. No genotype-related differences were observed in men. In conclusion, our results indicate that the absence of α-actinin-3 protein (i.e., ACTN3 X/X genotype) influences KE peak torque and FFM in women but not men.

the α-actinins comprise a family of actin-binding proteins, two of which play an important structural role in skeletal muscle: α-actinin-2 and α-actinin-3 (27). α-Actinin-3 is a structural protein of the sarcomeric Z-line that is only expressed in type II muscle fibers (19) and is known to be important for anchoring actin and playing a regulatory function in coordinating muscle fiber contraction (1). The α-actinin-3 protein is absent in ∼18% of healthy individuals of European descent with homozygosity of a common premature stop codon at the R577X polymorphism (rs1815739, C→T transition at position 1747 in exon 16) in the ACTN3 gene (20). The α-actinins have been shown to be highly conserved, and recent evidence has suggested that loss of ACTN3 gene function shows positive selection in humans (13). The ACTN3 R577X genotype has been associated with elite athletic performance (6, 17, 21, 34), and recently our laboratory observed that the ACTN3 R577X nonsense allele is underrepresented in elite-level strength athletes (24), providing further evidence that the absence of the α-actinin-3 protein impairs skeletal muscle force production. Some studies also indicate the possibility of an advantage for X/X homozygotes in endurance-type activities (17, 34), and recent work in rodents supports these initial findings (2, 12, 13).

The role of the ACTN3 R577X polymorphism in normal variation in muscle function in nonathletes has not been clearly established. Clarkson and colleagues (3) demonstrated that women homozygous for the ACTN3 577X allele had lower baseline isometric strength compared with heterozygotes when adjusted for body mass and age but unexpectedly demonstrated greater responses to strength training. Moran et al. (16) observed significantly lower 40-m sprint times in an unselected group of adolescent male X/X homozygotes but found that the R577X polymorphism was not associated with other power phenotypes. Vincent and colleagues (30) demonstrated greater dynamic muscle strength and a higher proportion of type IIX fibers in R/R compared with X/X men. In contrast, our laboratory group (4) has observed unexpected results regarding ACTN3 genotype influences on muscle power in a group of elderly women. We demonstrated that peak power in these older women at baseline was significantly greater in X/X women than both R/X and R/R women, although after a 10-wk strength training program, consistent with our hypothesis, relative peak power change with strength training in the R/R group was higher than the X/X homozygotes (4). In addition, although preliminary, work by San Juan et al. (25) observed that complete deficiency of α-actinin-3 does not effect maximal muscle strength in elderly women. Most recently, Delmonico et al. (5) observed that over a 5-yr period women deficient in α-actinin-3 displayed a significantly greater risk of developing lower extremity limitation with age compared with R/R homozygotes.

Although associations of ACTN3 R577X with performance have been clearly observed in elite-level athletes, studies involving nonathletes have provided more mixed results; therefore we set out to examine the ACTN3 R577X polymorphism in a nonathlete population aged across the lifespan. Given the localization of α-actinin-3 in fast-twitch skeletal muscle fibers, and the importance of such fibers for force-generating capacity, we hypothesized that individuals deficient in α-actinin-3 (X-allele homozygotes) would display lower levels of muscle strength compared with R/X + R/R individuals with the α-actinin-3 protein.

METHODS

Subjects.

The subjects, 454 men and 394 women aged 22–90 yr, included in this study were investigated as part of the Baltimore Longitudinal Study of Aging (BLSA), an ongoing National Institute on Aging (NIA)-funded investigation of normal aging recruited primarily from population-based listings (26). All BLSA subjects underwent a complete medical history and physical examination, and subjects with clinical cardiovascular or musculoskeletal disorders that could be adversely affected by exercise testing were excluded. Before the study, all subjects received a complete explanation of the purpose and procedures of the investigation and gave their written informed consent. The experimental protocols were approved by the Institutional Review Boards (IRB) for Medstar Research Institute, Johns Hopkins Bayview Medical Center, and the University of Maryland. The protocols related to the analysis of genetic data were approved by IRBs at the University of Maryland (College Park, MD).

Body composition.

Body composition variables were obtained using methods previously approved by the BLSA (9). Body mass and height were measured to the nearest 0.1 kg and 0.5 cm, respectively, using a Detecto medical beam scale. Whole body soft tissue composition was measured by dual-energy X-ray absorptiometry (DEXA) with the array mode, as previously described (9). Total body fat and soft tissue fat-free mass (FFM) and total leg fat and FFM (both legs combined) were assessed. Soft tissue FFM was used as a valid indicator of muscle mass based on previous work (8, 33). The scanner was calibrated daily before testing while reliability was assessed by performing two total body scans, 6 wk apart, on 12 older men (>65 yr). The difference between the two scans was ∼0.01% for fat and FFM.

Peak torque.

Peak torque (strength) was measured using the Kinetic Communicator isokinetic dynamometer (Kin-Com model 125E, Chattanooga Group, Chattanooga, TN) and was assessed by using the Kin-Com computer software (version 3.2). Shortening-phase peak torque was measured at angular velocities of 0.52 rad/s (30°/s) and 3.14 rad/s (180°/s) for the dominant knee extensors. The terms “shortening” and “lengthening” are substituted throughout the present study for the more commonly used terms “concentric” and “eccentric,” respectively (8). For each test, subjects performed three maximal efforts, separated by 30-s rest intervals, from which the highest value of the three trials was accepted as the peak torque. Detailed procedures regarding subject positioning and stabilization, gravity correction, Kin-Com calibration, and test-retest reliability are described elsewhere (9, 10).

Genotyping.

Genomic DNA was isolated from EDTA-anticoagulated whole blood using standard methods (Puregene DNA Extraction kit, Gentra Systems, Minneapolis, MN). All subjects were genotyped for the ACTN3 R577X polymorphism using restriction digest methods described previously (15). Sixteen sequence-verified control samples from these same BLSA subjects carrying all three genotypes (R/R, R/X, and X/X) were used for all assays as positive controls, with ambiguous samples regenotyped. Samples with ambiguous results in two assays were excluded from all analyses. All DNA sequencing to confirm positive control samples was completed using standard procedures using an ABI PRISM 3100 Genetic Analyzer.

Statistical analysis.

Two types of analyses were performed, with a three-genotype comparison (X/X vs. R/X vs. R/R) performed initially followed by a dominant-model analysis with subjects grouped as X/X homozygotes compared with R/X + R/R individuals, given that only X/X homozygotes are expected to be α-actinin-3 deficient in type IIX muscle fibers. ANOVA models were used to test for differences in physical characteristics among ACTN3 genotype groups. Physical characteristic data are presented as means ± SE. Dependent variables were analyzed using analysis of covariance (ANCOVA) and were used to compare means between genotype groups. Analyses were performed within each sex group. Significant covariates were included in all models and are listed in each table caption. Linear regression analysis was used to determine the proportion of variance attributable to ACTN3 genotype for muscle phenotypes. Data presented are least squares means ± SE. Statistical significance was defined as P < 0.05; when the omnibus P value was <0.05 for the three-genotype analysis, specific contrasts were examined and those P values are reported.

RESULTS

Subject characteristics are shown in Tables 1 and 2 for both men and women grouped by ACTN3 R577X genotype. Sample sizes were as follows. For men, X/X = 80, R/X = 213, R/R = 161 (n = 454 total). For women, X/X = 53, R/X = 174, R/R = 167 (n = 394 total). ACTN3 R577X genotype distributions (X/X = 16.4%, R/X = 45.4%, R/R = 38.1%) were shown to be in Hardy Weinberg equilibrium for the entire cohort (P = 0.29) and for each sex (women: P = 0.47; men: P = 0.51). Women X/X homozygotes displayed significantly lower levels of body mass, BMI, fat mass, and FFM (P < 0.05; Table 2) compared with R/R women, as well as with women grouped as carrying the R-allele (R/X + R/R).

Table 1.

Subject characteristics by ACTN3 R577X genotype in men

	X/X (a)	R/X (b)	R/R (c)	P Value	R/X + R/R	P Value
n	80	213	161		374
Age, yr	67.5±1.7	66.3±1.0	66.0±1.1	0.77	66.2±0.8	0.47
Height, cm	175.8±0.8	176.6±0.5	175.3±0.6	0.26	176.0±0.4	0.76
Weight, kg	83.6±1.6	85.8±1.0	85.4±1.1	0.49	85.6±0.7	0.24
BMI, kg/m2	27.0±0.4	27.4±0.3	27.7±0.3	0.41	27.5±0.2	0.24
Fat mass, kg	23.7±1.1	25.4±0.7	25.4±0.8	0.38	25.4±0.5	0.12
Total body FFM, kg (n)	55.8±0.6 (77)	56.5±0.4 (193)	57.0±0.5 (146)	0.28	56.7±0.3 (339)	0.15
Total leg FFM, kg (n)	17.7±0.3 (73)	17.9±0.2 (177)	18.4±0.2 (133)	0.12	18.2±0.1 (310)	0.26

Age, height, weight, and body mass index (BMI) data are means ± SE. Fat mass and fat-free mass (FFM) data are least-squares means ± SE. Age and height were included in the model as significant covariates. For the three-genotype (a, b, c) comparison, the omnibus P value is shown for nonsignificant results.

Table 2.

Subject characteristics by ACTN3 R577X genotype in women

	X/X (a)	R/X (b)	R/R (c)	P Value	R/X + R/R	P Value
n	53	174	167		341
Age, yr	63.4±1.9	60.4±1.1	59.9±1.1	0.320	60.2±0.8	0.130
Height, cm	162.5±0.8	163.2±0.5	163.3±0.5	0.506	163.5±0.3	0.290
Weight, kg	65.8±1.7	70.0±1.0	72.3±1.0	a vs. b: 0.041	71.1±0.7	0.006
				a vs. c: 0.002
				b vs. c: 0.119
BMI, kg/m2	24.9±0.6	26.2±0.3	27.0±0.3	a vs. b: 0.069	26.6±0.2	0.013
				a vs. c: 0.005
				b vs. c: 0.146
Fat mass, kg	25.0±1.4	27.7±0.8	29.8±0.8	a vs. b: 0.089	28.8±0.6	0.021
				a vs. c: 0.004
				b vs. c: 0.079
Total body FFM, kg (n)	38.9±0.5	39.8±0.3	40.4±0.3	a vs. b: 0.146	40.1±0.2	0.040
				a vs. c: 0.009
				b vs. c: 0.103
Total leg FFM, kg (n)	11.9±0.2 (50)	12.3±0.1 (160)	12.6±0.2 (146)	a vs. b: 0.149	12.5±0.1 (306)	0.044
				a vs. c: 0.014
				b vs. c: 0.135

Age, height, weight, and BMI data are means ± SE. Fat mass and FFM data are least square means ± SE. Age and height were included in the model as significant covariates. For the three-genotype comparison (a, b, c), the omnibus P value is shown for nonsignificant results, with specific contrast P values shown when the omnibus P value was <0.05.

As shown in Table 3, when covarying for age and height, X/X women displayed significantly lower KE shortening peak torque at 30°/s (P = 0.003) and 180°/s (P = 0.007) and lower KE lengthening peak torque at 30°/s (P = 0.004) and 180°/s (P = 0.004) compared with women R/X heterozygotes. A similar relationship was observed between X/X and R/R women, although not all comparisons were statistically significant (P values of 0.07, 0.09, 0.14, and 0.03, respectively). When examining the dominant model, X/X women displayed significantly lower KE shortening peak torque at 30°/s (P = 0.011) and 180°/s (P = 0.019) and lower KE lengthening peak torque at 30°/s (P = 0.022) and 180°/s (P = 0.008) compared with R/X + R/R women. No significant differences were observed for men in either analysis (Table 3).

Table 3.

Knee extensor peak torque values by ACTN3 R577X genotype in men and women

	X/X (a)	R/X (b)	R/R (c)	P Value	X/X	R/X + R/R	P Value
Men
n	80	213	161		80	374
30°/s Shortening phase, N·m	148.0±5.0	140.6±3.1	142.5±3.5	0.468	148.0±5.0	141.4±2.3	0.244
180°/s Shortening phase, N·m	104.1±3.5	101.7±2.1	103.0±2.4	0.834	104.1±3.5	102.3±1.6	0.654
30°/s Lengthening phase, N·m	194.6±7.4	186.3±4.5	196.7±5.1	0.289	194.6±7.4	190.8±3.4	0.640
180°/s Lengthening phase, N·m	199.3±6.9	189.2±4.2	193.9±4.7	0.434	199.3±6.9	191.2±3.1	0.288
Women
n	53	174	167		53	341
30°/s Shortening phase, N·m	89.5±3.5	101.6±1.9	96.8±2.0	a vs. b: 0.003	89.5±3.5	99.3±1.3	0.011
				a vs. c: 0.073
				b vs. c: 0.092
180°/s Shortening phase, N·m	60.3±2.6	68.5±1.5	65.5±1.5	a vs. b: 0.007	60.3±2.6	67.0±1.0	0.019
				a vs. c: 0.091
				b vs. c: 0.147
30°/s Lengthening phase, N·m	122.8±5.7	141.5±3.1	132.4±3.1	a vs. b: 0.004	122.8±5.7	137.1±2.2	0.022
				a vs. c: 0.144
				b vs. c: 0.040
180°/s Lengthening phase, N·m	121.8±5.8	141.0±3.1	135.8±3.2	a vs. b: 0.004	121.8±5.8	138.5±2.2	0.008
				a vs. c: 0.036
				b vs. c: 0.261

Data are least-squares means ± SE. Age and height were included in the model as significant covariates. For the three-genotype comparison (a, b, c), the omnibus P value is shown for nonsignificant results, with specific contrast P values shown when the omnibus P value was <0.05.

With regard to muscle mass phenotypes, X/X homozygous women displayed significantly lower levels of total body FFM (P = 0.009) and lower limb FFM (P = 0.014) compared with R/R homozygous women, but not compared with R/X women (P = 0.15 for both; Table 2). When examining the dominant model, X/X women displayed lower levels of total body FFM (P = 0.040) and lower limb FFM (P = 0.044) compared with R/X + R/R women (Table 2). Again, no differences were observed in men for either the three-genotype or dominant model analyses. When lower limb FFM differences were covaried in the KE torque analyses in place of height (Table 4), only KE shortening peak torque at 30°/s remained significantly different between the genotype groups, with women X/X homozygotes significantly weaker than R/X + R/R women (P = 0.049). When determining the extent of variability based on the inclusion of genotype in the models, ACTN3 R577X genotype was responsible for approximately 1.0–1.4% of the variation observed in KE shortening peak torque values.

Table 4.

Knee extensor peak torque values by ACTN3 R577X genotype in women including leg FFM as a covariate

	X/X Women	R/X + R/R Women	P Value
n	41	265
30°/s Shortening phase, N·m	85.8±4.1	94.6±1.6	0.049
180°/s Shortening phase, N·m	60.1±2.9	64.8±1.1	0.139
30°/s Lengthening phase, N·m	123.4±6.7	132.2±2.6	0.229
180°/s Lengthening phase, N·m	119.0±6.7	133.0±2.6	0.056

Data are least-squares means ± SE. Age and leg FFM were included in the model as significant covariates.

In a subanalysis examining women only over the age of 50 yr, X/X homozygotes displayed significantly lower KE shortening peak torque at 30°/s (78.4 ± 4.1 vs. 89.1 ± 1.6 N·m, P = 0.015), a tendency toward lower KE shortening peak torque at 180°/s (55.6 ± 2.9 vs. 61.3 ± 1.1 N·m, P = 0.070), and significantly lower KE lengthening peak torque at 30°/s (112.4 ± 6.9 vs. 127.3 ± 2.6 N·m, P = 0.046) and 180°/s (109.0 ± 6.6 vs. 127.0 ± 2.5 N·m, P = 0.012) compared with women grouped as R/X + R/R (age and height included as covariates). Although not significantly different, similar tendencies for lower FFM in the X/X homozygotes compared with R/X + R/R women were also observed (leg FFM: 11.9 ± 0.2 vs. 12.4 ± 0.1 kg, P = 0.062; total body FFM: 38.5 ± 0.6 vs. 39.6 ± 0.2 kg, P = 0.103). Similar to the findings in the entire sample, when FFM replaced height as a covariate, muscle strength values no longer differed significantly between genotype groups (data not shown; P = 0.18–0.55). No significant genotype differences were observed for any muscle phenotype in men over the age of 50 yr.

DISCUSSION

The ACTN3 R577X nonsense polymorphism has been associated with muscle performance (6, 17, 21, 34) and fiber-type proportion (30). The results from the present study add to the growing literature in nonathletes (3–5, 16, 25, 30) by showing that women across the adult age span deficient in α-actinin-3 protein (i.e., X/X) displayed lower knee extensor shortening and lengthening peak torque values compared with women grouped for the R allele (R/X + R/R). Surprisingly, these differences in muscle strength appeared to be driven by lower levels of both total body and lower limb FFM in X/X compared with R/X + R/R women. No such differences were observed in men. The strength and FFM differences were similarly observed in a subcohort of women over 50 yr.

The α-actinin-3 protein is deficient in ∼18% of individuals of European descent due to homozygosity of the nonsense allele at the R577X locus. To date, the X/X genotype has been either underrepresented in elite sprint, power, or strength athletes compared with controls (6, 17, 21, 24, 34), associated with poorer sprint performance compared with R/X + R/R individuals (16), or associated with an increase in functional limitation with aging in women (5). In direct analyses of muscle function, the ACTN3 R577X polymorphism appears to impact muscle power and muscle strength adaptations to resistance training (3, 4). Our laboratory group has observed after a 10-wk strength training program that relative peak power change in older R/R women was higher than X/X women, with similar results observed for older men for absolute peak power change in response to training (4). Clarkson and colleagues (3) observed results contrary to expectations, with young women homozygous for the X allele showing greater absolute and relative gains in one-repetition maximum strength compared with the R/R group following a resistance training program. The authors suggested that since the mutant genotype had the lowest baseline maximal voluntary contraction (consistent with expectations and the present study's findings), the lower initial values might have explained the greater increases in strength for the X/X genotype.

The present study was designed to examine the ACTN3 R577X polymorphism in relation to muscle phenotypes in a nonathlete population aged across the adult lifespan. We hypothesized that individuals deficient in α-actinin-3 protein would display lower levels of muscle strength compared with R/X + R/R individuals with the α-actinin-3 protein. Partially supporting our hypothesis, our results suggest a sex-specific association where women but not men deficient in α-actinin-3 displayed lower KE shortening peak torque and KE lengthening peak torque than women who were not deficient. Although in our study a genotype effect for ACTN3 and muscle strength in men was not observed, Vincent and colleagues (30), using similar strength measurements as our group, did observe that R/R men showed significantly higher relative quadriceps torques at 300°/s compared with X/X men. However, when further examining their data, strength differences were not observed at speeds of 100°/s and 200°/s, which is consistent with the speeds tested in the present study. The present study did not include torque values above 180°/s.

Women X/X homozygotes in our study also displayed lower levels of both total body and lower limb FFM compared with R/X + R/R women, and these differences appeared to explain many of the peak torque differences between the ACTN3 genotype groups. Interestingly, recent data published by MacArthur and colleagues (12) in a mouse model show consistent findings. They observed that Actn3 knockout mice displayed lower total body weight than wild-type mice, and analysis of body composition using DEXA indicated that the lower body weight was predominantly due to lower lean mass in Actn3 knockout mice (12). Similar to the present results, these authors also observed that α-actinin-3 deficient mice exhibited significantly lower grip strength compared with wild-type mice (12). Vincent and colleagues (30) in humans have observed genotype-specific differences for the percentage of type IIX fibers where R/R individuals displayed ∼5% more type IIX fibers than X/X individuals. These authors also observed that the relative surface area covered by type IIX fibers was also significantly greater in the R/R than the X/X genotype group (30). It has been hypothesized by Yang et al. (34) that α-actinin-3 promotes the formation of fast-twitch fibers and the findings of Vincent et al. (30) support this hypothesis. Although we do not have fiber-type data from our subjects, we can speculate that the results of the present study could partially be explained by the ACTN3 R577X polymorphism influencing fiber type distribution, resulting in a greater proportion of type IIX muscle fibers, potentially influencing both FFM and muscle strength in women. There is convincing evidence that the cross-sectional area and the type of myosin expressed in skeletal muscle fiber are related (7), and recent work by MacArthur et al. (12) showed that fast glycolytic 2B muscle fibers in mice deficient in α-actinin-3 display smaller cross-sectional areas. The lower levels of muscle strength observed in the present study for women deficient in α-actinin-3 was partially due to lower FFM, supporting in humans what MacArthur and colleagues (12) observed in the Actn3 mouse model. Furthermore, MacArthur et al. (12) also demonstrated that Actn3 knockout mice display a shift toward oxidative metabolism in fast muscle fibers, resulting in a shift toward a slow phenotype, resulting in differences in fiber diameter and contractile properties. Finally, Chan and colleagues (2) reported an increase in the half-relaxation time in the Actn3 knockout mice, which may hamper the ability to generate forceful contractions, providing yet another possible explanation for the results of the present study. An excellent in-depth review of the issues surrounding ACTN3 R577X genotype has recently been published (18).

As more evidence is collected indicating a role for α-actinin-3 deficiency in muscle function in nonathletes, a key issue to consider is clinical relevance. Recent work by Delmonico and colleagues (5) has demonstrated that over a 5-yr period, 70- to 79-yr-old women X/X homozygotes had an ∼35% greater risk of incident persistent lower extremity limitation (i.e., difficulty walking or climbing stairs) compared with R/R homozygotes, and X/X men of the same age exhibited significantly greater increases in 400-m walk times compared with R/R men. In the present study, a subanalysis in women over the age of 50 yr revealed that X/X women exhibited lower muscle strength than R/X + R/R women, although ACTN3 genotype explained only 1–1.4% of the variability in muscle strength. Skeletal muscle mass gradually declines starting at about age 45 yr (9), and the loss of muscle strength is an independent predictor of mortality in the elderly (14, 22). These recent findings for ACTN3 genotype may indicate that as women lose muscle mass with age, ACTN3 genotype may further reduce muscle mass and strength in X/X women, with potentially important functional implications. Results from the present study suggest that women who are deficient in α-actinin-3 appear to be at a disadvantage for muscular strength, and this may help to contribute to the lower extremity limitations that occur with aging shown by Delmonico et al. (5), although no strength differences (quadriceps torque) were observed among ACTN3 genotype groups in that investigation. A previous report by Delmonico and colleagues (4) found unexpectedly higher muscle strength in X/X compared with R/R older women, although sample sizes were small and the intent of that study was to examine skeletal muscle power adaptations to strength training. The growing evidence that the ACTN3 gene does have an influence on skeletal muscle function, combined with preliminary evidence indicating an influence on muscle adaptation to strength training (3, 4), may point to screening of the ACTN3 gene as a useful tool to identify women who may be at risk for lower muscle strength, lower muscle mass, and future risk of lower extremity limitation. Considerable work remains to fully establish the potential clinical utility for such screening.

Previous candidate gene association studies involving skeletal muscle phenotypes by our group (23, 31, 32) and others (3, 28, 29) have also shown sex-specific differences. Although there is no obvious explanation for the sex differences observed in the present study, Yang and colleagues (34) as well as Clarkson et al. (3) also observed sex-specific genotype differences for ACTN3, with the most striking findings observed in women. As previous groups have suggested, the sex differences may be partially due to sex-specific hormonal differences between men and women. MacArthur and North (11) have suggested that the lower average levels of testosterone in female athletes makes the variation in other parameters, such as ACTN3 genotype, more important in determining athletic ability. Their work involving elite female athletes suggested that the X/X genotype may be more frequent in endurance-trained female athletes compared with controls (34). Although speculative, perhaps the higher levels of testosterone observed in men leads to greater levels of muscle mass that somehow can compensate for loss of α-actinin-3, making the X/X genotype more influential in women, as is suggested by the present results and other recent studies (3, 5, 34).

In conclusion, our results indicate that the absence of α-actinin-3 protein (i.e., ACTN3 X/X genotype) influences knee extensor strength and FFM in women but not men; similar differences were also seen in older women. The strength differences observed between genotype groups in women were partially explained by genotype-related differences in FFM. Future studies are needed to elucidate the mechanisms behind the ACTN3 R577X polymorphism's apparent influence on sex-specific differences in muscle strength and FFM observed in the present study.

GRANTS

The BLSA research was conducted as a component of the Intramural Research Program of the National Institute on Aging. This work was further sponsored by National Institutes of Health Grants AG-021500 and AG-022791.