Satellite cells (SC) are the muscle stem cell population mainly committed to the skeletal muscle lineage (1–3). When exposed to mechanical loading, SC becomes activated, enters the cell cycle, proliferates, and may eventually fuse and donate its nuclei to existing muscle fibers (3). The donation of myonuclei to existing muscle fibers is required in situations where increased transcription capacity is required, such as repair to damaged muscle cells, or to promote protein accretion over a given physiological cellular myonuclear domain limit. The later concept is based on the theory that each myonucleus controls mRNA transcription and protein synthesis/breakdown over a finite volume of cytoplasm (i.e., myonuclear domain) (4,5). Thus, increases in muscle cell volume (i.e., increase in muscle fiber cross-sectional area [CSA]) expands the myonuclear domain over the physiological range, therefore requiring the addition of new myonuclei to muscle fibers (6–9).
Petrella et al. (9) and others (9–12) have described the existence of a theoretical hypertrophy “threshold,” after which new myonuclei should be added to muscle fibers to increase transcriptional capacity. When muscle fibers transcriptional capacity is not increased, further expansions of muscle fibers CSA may be limited. This threshold seems to be ≥26% of fibers’ CSA, but not less than 15% (9,12). Nonetheless, evidence for such a threshold is equivocal (8,13,14). For instance, Bellamy et al. (8) reported significant increases in type I (~13%) and II (~20%) fiber CSA, with concomitant increments in myonuclei number per muscle fiber, after a 16-wk resistance training (RT) program. In addition, 12 wk of endurance training (ET) increased type I (~12%) and IIA (~16%) fiber CSA with significant addition of myonuclei in type I fibers only (15). Other studies have also failed to detect myonuclei accretion after RT, although muscle fibers achieved the supposed hypertrophy threshold (i.e., 26%) (13,16,17). Taken together, particular original studies do not suggest the existence of a minimum muscle fiber hypertrophy threshold (i.e., upper limit of the myonuclear domain) for myonuclear addition, challenging the myonuclear domain theory. Clustering studies based on the magnitude of muscle fiber hypertrophy may therefore help in determining the required muscle fiber hypertrophy level inducing myonuclei addition.
The variability reported in original studies with regard to a myonuclear threshold may be attributed to several factors such as low statistical power, preintervention myonuclei number and domain size, age, sex, the type and duration of training interventions, and fiber type differences. Meta-analyses summarize data from different studies and allow for testing for the effects of moderator variables (i.e., age, sex, and muscle fiber type), thereby providing a better estimate of the factors affecting the magnitude of the muscle fiber hypertrophy and addition of new myonuclei to muscle fibers. Accordingly, the purpose of the current meta-analysis was to determine a muscle fiber hypertrophy threshold driving the addition of new myonuclei to muscle fibers. To account for the high heterogeneity between studies, we explored the effects of potential moderators (i.e., preintervention myonuclei number and domain size, age, sex, intervention model, and fiber type) on muscle fiber hypertrophy and myonuclear content.
Literature searches were performed on PubMed, Web of Science, and Google Scholar databases and conducted on studies published between January 1980 and March 2016. Combinations of the following keywords were used as search terms related to muscle hypertrophy: “muscle hypertrophy,” “satellite cell,” “myonuclei,” “myonuclear domain,” and “muscle stem cell” and related to training: “resistance training,” “resistance exercise,” “strength training,” “power training,” “endurance training,” and “endurance exercise,” ET studies were included in the present meta-analysis for two reasons. Myonuclei may be added to muscle cells to control both hypertrophic and nonhypertrophic processes. It is also reasonable to assume that hypertrophic and nonhypertrophic processes are enhanced after both ET and RT, although in different magnitudes, because cell’s biological processes should be augmented as individuals impose higher-energy demands and mechanical stress to muscle cells (i.e., training load progression). Therefore, including ET studies data in our meta-analysis increases its external validity, because it may show that increasing myonuclei number may occur even when myonuclear domain is not reaching physiological ceiling and muscle hypertrophy is small. Additional support to this matter is provided in Appendix, Supplemental Digital Content 1, Effect sizes, http://links.lww.com/MSS/B220, which displays data from several studies in which muscle hypertrophy and myonucelar additions were small and similar between RT and ET studies (small cluster studies—described in the Statistical Analysis section).
Furthermore, reference lists of articles were reviewed in which relevant articles to the topic were added to the analyses and new articles published between April 2016 and April 2017 were included when inclusion criteria were reached.
Inclusion criteria were limited to studies involving healthy human subjects, independent of age, sex or intervention (i.e., at least 4 wk of RT, ET, or steroid administration intervention) that evaluated chronic changes in muscle fiber CSA and myonuclear content.
Methodological quality was calculated using the Physiotherapy Evidence Database (PEDro) scale, which is based on the Delphi list (18). PEDro score was estimated by two independent researchers, and only studies that presented a score of ≥4.0 were included for analyses. Disagreements were resolved in a consensus meeting.
Study selection and data extraction
Initially, 1632 studies were retrieved. After removing duplicates and studies unrelated to muscle fiber hypertrophy and myonuclei addition, 274 studies were evaluated for design quality following the procedures previously described. To reduce the selection bias potential, titles and abstracts of all studies were independently evaluated by two investigators (M. S. C. and F. C.V.), and a mutual decision was made as to whether or not they met the inclusion criteria (described earlier). Sixty-three studies were selected, and each study was independently evaluated by the same investigators (M. S. C. and F. C. V.). Twenty studies that met the inclusion criteria were therefore added to the analyses. After reading all articles and reviewing the reference list, an additional two articles were included. Five more articles were published after finishing searching, and finally, 27 studies met the inclusion criteria and outcomes of each study were recorded and used for future analyses (Fig. 1).
Two independent researchers (M. S. C. and F. C. V.) extracted data for the following variables: number of subjects per group, age, sex, intervention model, muscle fiber CSA, myonuclear content, and domain size per muscle fiber. To test for possible coder drift, we randomly selected 30% of the studies for recoding following procedures outlined by Cooper et al. (19). Mean agreement between coders was 95%. Some of the studies did not present mean and SD values of the dependent variables. Despite e-mail contact with the corresponding author of these studies requesting data for muscle fiber CSA and myonuclei number, no information was received in most cases. Thus, when possible, we used a recently validated (20,21) online platform, Webplotdigitizer, (Version 3.9, available at http://arohatgi.info/WebPlotDigitizer/index.html) to obtain mean and SD values from these studies’ figures. Table 1 summarizes studies included in the analyses. Although some studies pooled muscle fiber CSA and myonuclei number data among fiber types, other studies did not pool muscle fiber CSA data among muscle fibers types (e.g., fiber type I and fiber type II) but pooled myonuclei number data across muscle fiber types. However, other studies pooled muscle fiber CSA but did not pool myonuclei number. Whenever muscle fiber CSA and myonuclei data did not present the same level of analysis (i.e., per fiber type or pooled), data were pooled for analyses.
Meta-analytic analyses were conducted using Comprehensive Meta-analysis software version 2.2 (Biostat Inc., Englewwod, NJ). Given that few studies included in the present meta-analysis had a control group, data were analyzed using a prentervention to postintervention model, considering only the intervention groups. In this sense, effect sizes (ES) for muscle fiber hypertrophy and number of myonuclei were calculated for each study, using preintervention and postintervention means, preintervention SD, sample size, and precorrelation to postcorrelation. Because none of the studies included in the meta-analyses reported precorrelation to postcorrelation and that the model used for data analyses requires this moderation, the correlation was estimated as follows:
SD is the SD of the difference score (preintervention to postintervention), defined by SD = square root [(S2pre/n) + (S2post/n)] and S is the standard deviation.
All ES values were corrected for small sample size bias with the following formula:
The between-study heterogeneity was verified with the I2 statistics. Preintervention to postintervention ES values were calculated for each intervention on each study. After hypertrophy, ES values from all the studies were separated using a k-means cluster analysis. Three clusters explained >90% of the data variance and produced significantly different centroids between the three clusters (small, moderate, and large hypertrophy) while maximizing the cubic clustering criteria (i.e., −2.6) being selected for subsequent analyses. The heterogeneity threshold was set as I2 ≤ 25% (low), I2 ≥ 25% to ≤ 50% (moderate), and I2 ≥ 75% (high) (40). Because the heterogeneity of the muscle fiber hypertrophy and myonuclei number per muscle fiber data was high (I2 ≥ 96.58 and I2 ≥ 92.53, respectively), random-effects models were implemented, as recommended (41).
To determine possible sources of the observed heterogeneity, additional analyses were undertaken to test the effects of the following moderators on muscle fiber CSA: age, sex, intervention model, and muscle fiber type. For this purpose, all of the studies were allocated to their respective cluster and then mean ES values of myonuclei number were compared between clusters of muscle fiber CSA. This analysis determined if distinct magnitudes of muscle fiber hypertrophy (i.e., centroids of the three clusters) affected the addition of myonuclei to the muscle fibers. Additional analyses were undertaken to test for the effect of the moderators cited previously on myonuclei number. Whenever ES confidence intervals (CI) of the levels of a given moderator factor did not overlap, levels were considered influential. Because preintervention myonuclear content is a continuous variable, a fixed-effect meta-regression was implemented to test for the effects of this variable on the addition of new myonuclei to muscle fibers.
A sensitivity analysis was performed to identify the presence of highly influential studies which could potentially bias the analyses. Thus, an analysis removing one study at a time was performed, and its effect on between-group comparisons (i.e., overall effect) was examined. Studies were considered influential if the removal resulted in a change of the ES going from significant (P ≤ 0.05) to nonsignificant (P > 0.05), or if their removal caused a large change in the magnitude of the coefficient. This procedure has been adopted elsewhere (42). The significance level adopted was P < 0.05. All data are present as mean ± SE.
Twenty-seven studies investigated the changes in muscle fiber CSA (i.e., hypertrophy) as well as the number of myonuclei after a period of intervention (e.g., RT, ET, with or without nutritional supplementation, or testosterone administration), and therefore were included in the present meta-analysis. This resulted in 62 ES values and a total sample size of 903 subjects. The mean rating of study quality assessed by the PEDro scale was 5.
Muscle fiber CSA
The mean ES for muscle fiber hypertrophy was 0.45 ± 0.03 (95% CI, 0.37–0.52; Fig. 2). However, the funnel plot indicated the occurrence of publication bias because several studies were outside the confidence limits of the plot (τ = 0.13, P = 0.12; Egger’s regression intercepts of 3.17, P = 0.01). Thus, a fill-and-trim procedure was implemented to minimize the magnitude of the publication bias, producing more conservative ES estimates (ES before fill-and-trim: 0.45 (95% CI, 0.37–0.52; ES after fill-and-trim: 0.48 (95% CI, 0.40–0.56)). Because the effect of the fill-and-trim procedure on the ES estimates was trivial, ES estimates before the fill-and-trim procedure were considered appropriate.
Thereafter, a cluster analysis was used to allocate studies according to the magnitude of muscle fiber CSA hypertrophy. The percentage change in muscle fiber hypertrophy in each cluster was classified as follows: small cluster, 10% of hypertrophy, with ES of 0.17 ± 0.02 (95% CI, 0.13–0.22); moderate cluster, 22% of hypertrophy, with ES of 0.59 ± 0.02 (95% CI, 0.55–0.63); and large cluster, 27% of hypertrophy, with ES of 1.03 ± 0.04 (95% CI, 0.95–10; Fig. 3). Clustering muscle fiber hypertrophy data across studies greatly reduced heterogeneity in the model (small cluster: Q = 172.3, df = 26, P = 0.000, I2 = 84.9%; moderate cluster: Q = 56.76, df = 27, P = 0.001, I2 = 52.43%; large cluster: Q = 11.89, df = 6, P = 0.065, I2 = 49.52%).
Effects of age, sex, intervention model, and fiber type on CSA hypertrophy
Muscle fiber CSA increased significantly and by a similar magnitude in younger (i.e., from 17 to 60 yr) and older individuals (i.e., older than 60 yr) (ES values, 0.41 ± 0.04 (95% CI, 0.33–0.49; I2 = 94.65%) and 0.50 ± 0.07 (95% CI, 0.35–0.64; I2 = 98.26%), respectively). Muscle fiber CSA also increased significantly and by a similar extent for both men and women (ESMen, 0.42 ± 0.04 (95% CI, 0.34–0.51; I2 = 96.09%); ESWomen, 0.64 ± 0.12 (95% CI, 0.40–0.88; I2 = 92.66%)). Regarding intervention model, RT (ES, 0.54 ± 0.04; 95% CI, 0.45–0.63; I2 = 97.40%) and testosterone treatment (ES, 0.35 ± 0.06; 95% CI, 0.24–0.47; I2 = 40.50%) produced greater muscle fiber hypertrophy than did ET (ES, 0.14 ± 0.03; 95% CI, 0.06–0.19; I2 = 65.37%). Furthermore, RT presented a trend toward greater muscle fiber hypertrophy than did testosterone treatment because there is a small overall between their respective CI. Although RT produced the highest muscle fiber hypertrophy, it also yielded the highest heterogeneity value (I2 = 97.40%), whereas testosterone treatment and ET showed lower values (I2 = 40.50% and I2 = 65.37, respectively). Interestingly, type I (ES, 0.42 ± 0.08; 95% CI, 0.26–0.58; I2 = 97.17%) and type II muscle fibers (ES, 0.54 ± 0.06; 95% CI, 0.43–0.65; I2 = 96.41%) presented similar magnitudes of increase in muscle fiber size. Pooled muscle fiber analysis resulted in significant increases in CSA (ES, 0.36 ± 0.04; 95% CI, 0.27–0.45; I2 = 92.92%).
The mean ES across all the studies was 0.47 ± 0.05 (95% CI, 0.37–0.57), demonstrating that the number of myonuclei significantly increased after interventions (Fig. 4). Sensitivity analysis (i.e., removing one study at a time and reanalyzing the data) revealed that a number of myonuclei were not highly affected by any single study (data not shown). Although no single study affected the analysis, funnel plot revealed that three studies greatly inflated publication bias (three studies outside the right confidence limit of the funnel plot: τ = 0.21 (P = 0.02) and Egger’s regression intercepts of 3.97 (P = 0.005)). Removing these studies improved the visual profile of the funnel plot, but Egger’s regression intercept remained significant (τ = 0.09 (P = 0.29) and Egger’s regression intercepts of 2.37 (P = 0.005)). Similar to muscle fiber hypertrophy, a fill-and-trim procedure was implemented to minimize publication bias; however, ES estimates did not change based on the random model. Thus, because these studies have unusually high ES, they were not considered for the final analysis.
Myonuclei addition occurred in all three clusters [small cluster: 10% of hypertrophy and ES of 0.25 ± 0.03 (95% CI, 0.18–0.32); moderate cluster: 22% of hypertrophy and ES of 0.46 ± 0.04 (95% CI, 0.36–0.56); large cluster: 27% of hypertrophy and ES of 0.47 ± 0.10 (95% CI, 0.27–0.67)]. However, the moderate cluster showed a significantly greater ES compared with the small cluster (Fig. 5). The heterogeneity data for small, moderate, and large clusters were as follows: Q = 291.42, df = 25, P = 0.000, and I2 = 91.42%; Q = 220.51, df = 25, P = 0.000, and I2 = 88.66%; and Q = 104.80, df = 6, P = 0.000, and I2 = 94.28%, respectively. Although the magnitude of muscle fiber hypertrophy seems to coordinate the addition of new myonuclei to muscle fibers, the high heterogeneity observed within each cluster suggests other sources of variance.
Effects of age, sex, intervention model, and fiber type on myonuclei number
Age did not affect the increase in myonuclei number (ES values, 0.37 ± 0.04 (95% CI, 0.29–0.45; I2 = 93.33%) and 0.35 ± 0.04 (95% CI, 0.27–0.42; I2 = 89.18%), for young and older adults, respectively). In regard to sex, the number of myonuclei increased similarly after intervention for both men (ES, 0.41 ± 0.03; 95% CI, 0.33–0.49; I2 = 92.70%) and women (ES, 0.27 ± 0.08; 95% CI, 0.10–0.45; I2 = 69.00%). Regarding the intervention model, RT (ES, 0.41 ± 0.03; 95% CI, 0.34–0.49; I2 = 93.04%), ET (ES, 0.16 ± 0.05; 95% CI, 0.05–0.27; I2 = 93.27%), and testosterone treatment (ES, 0.34 ± 0.04; 95% CI, 0.25–0.43; I2 = 0.00%) significantly increased myonuclei number. Also, RT showed higher values compared with ET. Fiber type did not affect the addition of myonuclei as it increased similarly for type I (ES, 0.40 ± 0.05; 95% CI, 0.30–0.50; I2 = 88.52%), type II (ES, 0.43 ± 0.06; 95% CI, 0.32–0.55; I2 = 94.75%) and pooled muscle fiber types (ES, 0.27 ± 0.05; 95% CI, 0.20–0.35; I2 = 88.37%).
Effects of preintervention myonuclear domain on myonuclei addition
The meta-regression produced a nonsignificant model (Q = 1.57, P = 0.21) and slope (−0.0001, P = 0.21; i.e., effects not different from zero), but a significant intercept (0.64, P < 0.0001) was observed.
It has been proposed that a theoretical muscle fiber hypertrophy threshold of ~26% is required before the addition of new myonuclei to muscle fibers (9–12). The major findings from our meta-analyses show that myonuclei addition does occur when muscle fiber hypertrophy is ≤10%, although consistent myonuclear addition is observed when muscle fiber hypertrophy is >22%. We also demonstrated that (a) age, sex, and muscle fiber type do not affect myonuclei addition; (b) RT is more effective than ET to induce muscle fiber hypertrophy and myonuclei addition after training; and (c) steroid administration produced the lowest heterogeneity value (I2 = 40.50%) for myonuclei addition. Our findings therefore extend and add new information to the present knowledge regarding the muscle fiber hypertrophy necessary to myonuclei addition.
Results of several studies have indicated that myonuclei addition is necessary for muscle fiber hypertrophy to occur (6,7,9,28,32). In contrast, evidence also suggests that existing myonuclei are able to increase rates of protein synthesis to elicit moderate level of muscle fiber hypertrophy (9–14). Cluster analysis allowed us to determine how distinct levels of muscle fiber hypertrophy (i.e., small, moderate, and large) may affect myonuclei addition. Overall, our data suggest that muscle fiber hypertrophy of ≤10% may induce myonuclei addition, which is lower than the threshold value (≥26%) or even the minimum value (i.e., 15%) reported elsewhere (7,9). The presence of myonuclei addition with such a small amount of muscle fiber hypertrophy indicates that reaching a ceiling of the myonuclear domain is not required. Accordingly, the meta-regression showed that myonuclei addition is not influenced by initial myonuclear domain. The fact that some studies did not find increases in myonuclear content with low-to-moderate magnitude of hypertrophy could be due to the lack of sensitivity in the immunohistochemical procedures to detect small changes in myonuclear content. All studies are based on cross sections, obtained from muscle biopsies, to evaluate muscle fiber size, myonuclear content, and/or domain size. However, it is important to consider that changes in muscle architecture are not occurring in only two dimensions. The theory is that every myonucleus controls a certain cytoplasmic volume (three dimensions) of the muscle fiber, which is not accounted in the two dimensions of the histochemical analyses. Furthermore, it is not clear if myonuclei are homogenously distributed along the muscle fiber length. Thus, it is possible that cross-section slices used in the immunohistochemical analyses may not contain any myonucleus just by chance. Taking these limitations into consideration, our two-dimensional analyses suggested that the moderate and large clusters produced similar ES for myonuclei addition, supporting that 22% of muscle fiber hypertrophy does require robust addition of new myonuclei to muscle fibers. It is important to note that the high heterogeneity values for myonuclei addition among the hypertrophy clusters (large cluster: I2 = 94.28; moderate cluster: I2 = 88.66; small cluster: I2 = 91.42) indicate that other moderate variables may be affecting myonuclei addition, such as the quality of each study, measured here by PEDro scale. When we separated only studies qualified as high quality (PEDro scale higher than 6), seven studies and 28 treatments (data not shown) were identified and the overall heterogeneity test to myonuclei number decreased from high (I2 ≥ 97.62) to moderate (I2 ≥ 65.62), suggesting that “low”-quality study design (e.g., no control group) can definitely influence the myonuclei number results. Accordingly, we moderated the myonuclei number by hypertrophy cluster (small hypertrophy, 10%; moderate hypertrophy, 22%; and large hypertrophy, 27%) only in “high”-quality studies, and thus, the heterogeneity test generally decreases from high to moderate (small: from I2 ≥ 91.42 to I2 ≥ 61.58; moderate: from I2 ≥ 88.66 to I2 ≥ 68.49; and large: from I2 ≥ 94.28 to I2 ≥ 69.88). Taken together, we suggest that ≤10% of muscle fiber hypertrophy can induce the addition of new myonuclei to the muscle fiber; however, significantly higher myonuclei addition is observed with larger muscle fiber hypertrophy (i.e., 22%).
Small-cohort studies have reported that age and sex may affect myonuclear content, although evidence supporting such a notion is equivocal (9,10). SC and myonuclear content, specifically in type II muscle fibers, have been shown to decrease with age (38). In addition, some studies have shown that RT can increase the myonuclear content in younger, but not older, individuals (9,13). Our results contrast these findings because ES CI overlapped when age was considered as a moderator factor, suggesting that the elderly maintain the capacity to increase the number of myonuclei per muscle fiber when data on the topic are pooled. Karlsen et al. (43) did not specifically compared the ability of young and older individuals to add myonuclei to muscle fibers, but showed that regardless of the training status (i.e., sedentary or endurance trained), a large cohort of young and older individuals has similar number of myonuclei per muscle fiber. Furthermore, they showed that myonuclei number has a very high positive relationship with muscle fiber size (r = 0.99). Taken together, our findings and those of Karlsen et al. suggest that myonuclear content is not affected by age but by fiber size, which does not corroborate with the findings of some independent studies (9,13,17,23). Low statistical power may be an issue when assessing the effect of sex in myonuclei addition. For example, Petrella et al. (9) reported that the number of myonuclei increased after RT only in young men (not in women), whereas muscle fiber CSA increased independently of sex. In contrast, our findings point to no differences either in muscle fiber hypertrophy or in myonuclei addition when sex is considered as a moderator factor. The high heterogeneity values obtained at each level of age and sex provide additional support to the lack of appropriate statistical power in small trials. It is important to highlight that, because of the inclusion criteria, only studies investigating muscle fiber CSA and myonuclei number were included in our analyses, suggesting that other studies evaluating muscle fiber CSA or myonuclei number in isolation could exist. Therefore, the effects of age and sex on muscle fiber CSA and myonuclei addition need to be further investigated with larger trials and the inclusion of appropriate control groups to improve the precision of within- and between-subject variability.
Another interesting finding of the present study was the effects of training model and testosterone treatment on muscle fiber CSA and myonuclei addition. Our analyses demonstrated that RT, ET, and testosterone treatment increase muscle fiber CSA and number of myonuclei. However, as would be expected, RT has a greater effect than ET on muscle CSA and myonuclei number. Accordingly, it is a reasonable assumption that myonuclei addition observed with ET may not be directed toward muscle hypertrophy. Considering the importance of muscle protein synthesis to muscle hypertrophy (44–47), such findings are entirely plausible. Although rates of myofibrillar protein synthesis are highly associated with muscle fiber hypertrophy (45,47), most of the protein synthesis is directed toward mitochondrial biogenesis when ET is undertaken (47). Additional myonuclei-driven biological processes can be enhanced after ET to meet the higher-energy demands to muscle cells imposed by the progression of the training load. For instance, McKenzie et al. (29) investigated the effects of carbohydrate and protein supplementation after 10 d of intensified ET in well-trained cyclists. Interestingly, a significant increment in myonuclei number without any significant changes in muscle fiber CSA after carbohydrate supplementation was found. These results suggest that increments in myonuclei number may be necessary to support biological processes other than muscle hypertrophy. Furthermore, our results show that testosterone treatment increases muscle fiber CSA and myonuclei number, confirming previous study findings (35,36,48). Testosterone treatment presented virtually no heterogeneity in the data, whereas RT and ET induced high heterogeneity (I2 > 93%), suggesting that testosterone treatment produces a very homogenous effect on muscle fiber CSA and myonuclei number increases. On the other hand, moderator factors not tested in the present study (i.e., volume, intensity, and frequency of training) may have driven the high heterogeneity after both RT and ET. Taken collectively, we confirm that testosterone treatment can increase muscle fiber CSA and myonuclei addition and that RT is more effective to induce muscle fiber hypertrophy and increase number of myonuclei compared with ET.
Previous reports have suggested that myonuclei addition may present a fiber type–specific response (25). Fry et al. showed that 12 wk of ET significantly increased type I and II muscle fiber CSA, whereas myonuclei number increased only in type I fibers, suggesting that type I fibers are more responsive to ET in regard to myonuclei addition. Bellamy et al. (8) showed that muscle fiber CSA increased after RT independently of muscle fiber type and that myonuclei addition similarly occurred only in type I fibers. The results of these studies demonstrate that muscle fiber hypertrophy can occur independently of fiber type and that myonuclei addition occurs more frequently in type I muscle fibers, independently of training model. However, and in contrast to these findings, our analyses indicate that fiber hypertrophy and myonuclei addition are not fiber type dependent. The reason for such a discrepancy is hard to reconcile, but the ability of meta-analyses to summarize the available data and to minimize possible effects of small trials should be taken into account to support a nonspecific fiber type response for muscle fiber hypertrophy and myonuclei addition.
Although a more consistent myonuclei addition occurs when muscle fiber hypertrophy is >22%, our results challenge the concept of a muscle hypertrophy threshold as evidenced by our analyses reporting significant myonuclei addition with lower muscle hypertrophy increases (i.e., ~10%). Age, sex, and fiber type do not affect myonuclei addition, whereas RT seems to induce higher muscle fiber hypertrophy and increases in myonuclear content compared with ET. Fundamental to future studies, we recommend the inclusion of a reproducibility measure (i.e., coefficient of variation) to myonuclei number and inclusion of a control group to allow for a direct comparison of myonuclei number with any intervention. Furthermore, studies should always include a statistical power analysis to estimate the number of subjects needed to decrease the probability of occurrence of type I or II errors. From a statistical power analyses of our data having β of at least 90% for the interaction effect, type I (α) error rate of 5%, moderate correlation among repeated measures (r = 0.8), nonsphericity correction of 1, and a large partial eta squared of 0.3 (49), a total sample size of 32 subjects (16 per group) was estimated. However, this estimative is also imprecise, because we do not have the random error associated with the myonuclei assessments due to the lack of control groups in most of the trials used herein.
The authors would like to express gratitude to the São Paulo Research Foundation (FAPESP; Grant Nos. 2016/09759-8 and 2015/19756-3) and CNPq (Grant Nos. 448387/2014-0 and 406609/2015-2). The authors declare no conflicts of interest. The authors declare that the results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Almeida CF, Fernandes SA, Ribeiro Junior AF, Keith Okamoto O, Vainzof M. Muscle satellite cells: exploring the basic biology to rule them. Stem Cells Int
2. Dumont NA, Wang YX, Rudnicki MA. Intrinsic and extrinsic mechanisms regulating satellite cell
3. Hawke TJ. Muscle stem cells and exercise training. Exerc Sport Sci Rev
4. Cheek DB. The control of cell mass and replication. The DNA unit—a personal 20-year study. Early Hum Dev
5. Hall ZW, Ralston E. Nuclear domains in muscle cells. Cell
6. Allen DL, Monke SR, Talmadge RJ, Roy RR, Edgerton VR. Plasticity of myonuclear
number in hypertrophied and atrophied mammalian skeletal muscle fibers. J Appl Physiol (1985)
7. Kadi F, Thornell LE. Concomitant increases in myonuclear
and satellite cell
content in female trapezius muscle following strength training. Histochem Cell Biol
8. Bellamy LM, Joanisse S, Grubb A, et al. The acute satellite cell
response and skeletal muscle hypertrophy
following resistance training. PLoS One
9. Petrella JK, Kim JS, Cross JM, Kosek DJ, Bamman MM. Efficacy of myonuclear
addition may explain differential myofiber growth among resistance-trained young and older men and women. Am J Physiol Endocrinol Metab
10. Mackey AL, Esmarck B, Kadi F, et al. Enhanced satellite cell
proliferation with resistance training in elderly men and women. Scand J Med Sci Sports
11. Snijders T, Smeets JS, van Kranenburg J, Kies AK, van Loon LJ, Verdijk LB. Changes in myonuclear domain
size do not precede muscle hypertrophy
during prolonged resistance–type exercise training. Acta Physiol (Oxf)
12. Kadi F, Schjerling P, Andersen LL, et al. The effects of heavy resistance training and detraining on satellite cells in human skeletal muscles. J Physiol
. 2004;558(Pt 3):1005–12.
13. Verdijk LB, Gleeson BG, Jonkers RA, et al. Skeletal muscle hypertrophy
following resistance training is accompanied by a fiber type–specific increase in satellite cell
content in elderly men. J Gerontol A Biol Sci Med Sci
14. Verney J, Kadi F, Charifi N, et al. Effects of combined lower body endurance and upper body resistance training on the satellite cell
pool in elderly subjects. Muscle Nerve
15. Fry CS, Noehren B, Mula J, et al. Fibre type–specific satellite cell
response to aerobic training in sedentary adults. J Physiol
. 2014;592(Pt 12):2625–35.
16. Herman-Montemayor JR, Hikida RS, Staron RS. Early-phase satellite cell
and myonuclear domain
adaptations to slow-speed vs. traditional resistance training programs. J Strength Cond Res
17. Hikida RS, Walsh S, Barylski N, Campos GE, Hagerman FC, Staron RS. Is hypertrophy limited in elderly muscle fibers? A comparison of elderly and young strength-trained men. Basic Appl Myol
18. Verhagen AP, de Vet HC, de Bie RA, et al. The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol
19. Cooper H, Hedges L, Valentine J. The Handbook of Research Synthesis and Meta-Analysis
. New York: Russell Sage Foundation; 2009.
20. Burda BU, O’Connor EA, Webber EM, Redmond N, Perdue LA. Estimating data from figures with a Web-based program: considerations for a systematic review. Res Synthesis Methods
21. Drevon D, Fursa SR, Malcolm AL. Intercoder reliability and validity of WebPlotDigitizer in extracting graphed data. Behav Modif
22. Charifi N, Kadi F, Feasson L, Denis C. Effects of endurance training on satellite cell
frequency in skeletal muscle of old men. Muscle Nerve
23. Dirks ML, Tieland M, Verdijk LB, et al. Protein supplementation augments muscle fiber hypertrophy but does not modulate satellite cell
content during prolonged resistance–type exercise training in frail elderly. J Am Med Dir Assoc
24. Frese S, Ruebner M, Suhr F, et al. Long-term endurance exercise in humans stimulates cell fusion of myoblasts along with fusogenic endogenous retroviral genes in vivo. PLoS One
25. Fry CS, Lee JD, Jackson JR, et al. Regulation of the muscle fiber microenvironment by activated satellite cells during hypertrophy. FASEB J
26. Joanisse S, Gillen JB, Bellamy LM, et al. Evidence for the contribution of muscle stem cells to nonhypertrophic skeletal muscle remodeling in humans. FASEB J
27. Joanisse S, McKay BR, Nederveen JP, et al. Satellite cell
activity, without expansion, after nonhypertrophic stimuli. Am J Physiol Regul Integr Comp Physiol
28. Leenders M, Verdijk LB, van der Hoeven L, van Kranenburg J, Nilwik R, van Loon LJ. Elderly men and women benefit equally from prolonged resistance–type exercise training. J Gerontol A Biol Sci Med Sci
29. McKenzie AI, D’Lugos AC, Saunders MJ, Gworek KD, Luden ND. Fiber type–specific satellite cell
content in cyclists following heavy training with carbohydrate and carbohydrate–protein supplementation. Front Physiol
30. Mackey AL, Holm L, Reitelseder S, et al. Myogenic response of human skeletal muscle to 12 weeks of resistance training at light loading intensity. Scand J Med Sci Sports
31. Nielsen JL, Aagaard P, Bech RD, et al. Proliferation of myogenic stem cells in human skeletal muscle in response to low-load resistance training with blood flow restriction. J Physiol
. 2012;590(Pt 17):4351–61.
32. Olsen S, Aagaard P, Kadi F, et al. Creatine supplementation augments the increase in satellite cell
and myonuclei number in human skeletal muscle induced by strength training. J Physiol
. 2006;573(Pt 2):525–34
33. Reidy PT, Fry CS, Igbinigie S, et al. Protein supplementation does not affect myogenic adaptations to resistance training. Med Sci Sports Exerc
34. Roberts LA, Raastad T, Markworth JF, et al. Post-exercise cold water immersion attenuates acute anabolic signalling and long-term adaptations in muscle to strength training. J Physiol
35. Sinha-Hikim I, Artaza J, Woodhouse L, et al. Testosterone-induced increase in muscle size in healthy young men is associated with muscle fiber hypertrophy. Am J Physiol Endocrinol Metab
36. Sinha-Hikim I, Cornford M, Gaytan H, Lee ML, Bhasin S. Effects of testosterone supplementation on skeletal muscle fiber hypertrophy and satellite cells in community-dwelling older men. J Clin Endocrinol Metab
37. Stec MJ, Kelly NA, Many GM, Windham ST, Tuggle SC, Bamman MM. Ribosome biogenesis may augment resistance training–induced myofiber hypertrophy and is required for myotube growth in vitro. Am J Physiol Endocrinol Metab
38. Verdijk LB, Snijders T, Drost M, Delhaas T, Kadi F, van Loon LJ. Satellite cells in human skeletal muscle: from birth to old age. Age (Dordr)
39. Verdijk LB, Snijders T, Holloway TM, Vank J, Vanl LJ. Resistance training increases skeletal muscle capillarization in healthy older men. Med Sci Sports Exerc
40. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ
41. Huedo-Medina TB, Sanchez-Meca J, Marin-Martinez F, Botella J. Assessing heterogeneity in meta-analysis: Q
statistic or I2
index? Psychol Methods
42. Schoenfeld BJ, Ogborn D, Krieger JW. Effects of resistance training frequency on measures of muscle hypertrophy
: a systematic review and meta-analysis. Sports Med
43. Karlsen A, Couppe C, Andersen JL, et al. Matters of fiber size and myonuclear domain
: does size matter more than age? Muscle Nerve
44. Damas F, Phillips S, Vechin FC, Ugrinowitsch C. A review of resistance training–induced changes in skeletal muscle protein synthesis and their contribution to hypertrophy. Sports Med
45. Damas F, Phillips SM, Libardi CA, et al. Resistance training–induced changes in integrated myofibrillar protein synthesis are related to hypertrophy only after attenuation of muscle damage. J Physiol
46. Phillips SM. Physiologic and molecular bases of muscle hypertrophy
and atrophy: impact of resistance exercise on human skeletal muscle (protein and exercise dose effects). Appl Physiol Nutr Metab
47. Wilkinson SB, Phillips SM, Atherton PJ, et al. Differential effects of resistance and endurance exercise in the fed state on signalling molecule phosphorylation and protein synthesis in human muscle. J Physiol
. 2008;586(Pt 15):3701–17.
48. Sinha-Hikim I, Roth SM, Lee MI, Bhasin S. Testosterone-induced muscle hypertrophy
is associated with an increase in satellite cell
number in healthy, young men. Am J Physiol Endocrinol Metab
49. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods