Serum reference value of two potential doping candidates—myostatin and insulin-like growth factor-I in the healthy young male

As a member of the transforming growth factor β superfamily, myostatin is a negative regulator of muscular growth and differentiation. Myostatin knock-out mice have three times more muscle mass than their wild-type counterparts [1]. Better functional performance is also observed without noticeable physiological compromise in these animals. On the other hand, nude mice over-expressing myostatin showed a cachectic phenotype with both skeletal muscle and adipose tissue atrophy [2]. Myostatin is translated into a 310-amino acid peptide, and cleaved as N terminal propeptide and C terminal active peptide. It is then circulated in the blood as an endocrine hormone [2]. The N-terminal propeptide is a potent inhibitor for myostatin. In a human study, a male baby born to a professional athlete mother in Germany was found to have a bulky muscle mass and loss-of-function myostatin mutation [3]. The serum myostatin peptide level was as low as undetectable in this male baby. It is reasonable to hypothesize that inhibiting myostatin can increase muscle mass and strength, and the sports performance of athletes. Another study revealing that the serum myostatin propeptide concentration in a bodybuilder was significantly higher than that in an untrained volunteer supports this hypothesis [4]. Strength training, a mode of exercise that can increase muscle mass, can modulate myostatin. Strength training inhibits intramuscular myostatin mRNA and protein expression in rats [5]. In a human study, similar findings were reported by Roth et al. that muscular myostatin mRNA level decreased 37% in healthy volunteers after a 9-week course of heavy unilateral knee extension strength training [6]. In addition, Walker et al. claimed that the serum myostatin peptide concentration, analyzed with Western blot, decreased 20% in a healthy male volunteer after a 10-week whole body strength training course [7].

Insulin-like growth factor-1 (IGF-1), a downstream molecule of the growth hormone (GH), increases protein synthesis in skeletal muscle. IGF-1 transgenic mice have accelerated skeletal muscle regeneration. The hypertrophic effects of IGF-I infusion on muscle are well documented in animal models and muscle cell culture systems. [8, 9] IGF-1 expression increased in a rat model of cardiomyocyte hypertrophy and in healthy volunteers doing strengthening exercises [10]. Using the C2C12 myotube model, a dose-dependent reduction in myostatin expression was observed on exposure to GH; this action was reversed by the GH antagonist pegvisomant [11]. We previously proposed that myostatin and IGF-1 could serve as a brake and accelerator mechanism for regulating muscle mass or function [12]. Based on this accelerator-brake model, we hypothesized that both IGF-1 and myostatin may correlate with lean or fat-free mass.

Due to their ability to modulate muscle mass and function, myostatin and IGF-1 genes and their protein products might be a target for doping [13]. Myostatin-inhibiting antibody and muscle-specific expression of IGF-I were shown to increase the muscle mass and function of mdx mice, an animal model of Duchenne muscular dystrophy [14, 15]. It is reasonable to expect improved muscle function after injection of exogenous IGF-1, myostatin propeptide, or myostatin neutralizing antibody, which will change their serum concentrations. However, we still do not have their reference value in the general population or in the athlete. Therefore, we conducted a cross-sectional study to investigate the reference values of serum myostatin and IGF-1 concentrations in male student athletes and untrained counterparts, and secondarily to determine whether relationships with body composition exist for each serum marker in the grouped cohort.

Since the athlete group was composed of two sports teams, basketball and weight-lifting, we compared the difference between these two teams. The basketball team had greater height, fat-free mass and muscle mass, and less grip strength than the weight-lifting team. The two serum markers did not differ between these two sports teams, or between the control and these two teams (Table 1).

In this study, we established the reference value of myostatin and IGF-1 and found their relationships with body composition from a cross-sectional cohort. We demonstrated that the reference value of the myostatin serum concentration in healthy young males was 12.3 ± 3.6 ng/ml, which was correlated negatively with age. After adjustment for age, the reference value was negatively correlated with BMI, fat mass percentage, and waist girth. The reference value for IGF-1 was 293 ± 114 ng/ml, which was correlated with age negatively and height positively.

Because myostatin is thought to be a myokine—a muscle secreted cytokine [12], its serum levels have been reported in clinical conditions related to muscle growth rather than in the healthy condition. However, it is difficult to compare the serum myostatin level with previous studies, since the reported levels varied greatly due to the use of different ELISA kits. Using the same ELISA kit as in our study, Ju and Chen revealed that the serum myostatin level was higher in patients with chronic obstructive pulmonary disease (COPD) than in the healthy control (11.85 ± 4.01 vs. 7.46 ± 2.21 ng/ml) [21]. The myostatin level of COPD patients was correlated with BMI negatively and with the serum tumor necrosis factor α level positively. Using the same ELISA kit, Zamora et al. found the mean myostatin serum concentration was 12.3 ng/ml in chronic heart failure patients aged 72.3 years on average [22]. No correlation was found between serum myostatin and parameters of disease severity and prognosis. Wintgens et al. demonstrated that the myostatin level was elevated in decompensated congestive heart failure patients compared with the normal control (49 vs. 32 ng/ml). The level decreased to the normal level after recompensating therapy [20]. Lakshman et al. measured serum myostatin level with a monoclonal antibody-based sandwich ELISA. They reported the myostatin level to be 8.0 ± 2.3 ng/ml in healthy young men with an average age of 26.5 years [23]. The level in older men with age of 66.4 ± 4.7 years dropped significantly to 7.0 ± 2.5 ng/ml. In summary, the serum concentration of myostatin seems to range from 7 to 32 ng/ml in the general population.

Myostatin affects the growth of both muscle and lipid. Our study, recruiting healthy young male subjects, found myostatin was negatively related to BMI, fat mass percentage, and waist girth after adjustment for age. Feldman et al. reported that myostatin transgenic mice had immature adipogenesis resulting in lower body weight, serum triglycerides (TG), and fasting plasma glucose [24]. Two other studies also concluded that administration of recombinant myostatin protein to 3 T3-L1 pre-adipocytes or human mesenchymal stem cells blocked their adipogenesis [25]. In a human study, a 6-month lifestyle intervention reduced the body weight of obese children aged between 6 – 16 years. At the same time, serum myostatin elevation and a decrease in fat mass and BMI were also observed [26]. Furthermore, in a genetic study recruiting 500 non-diabetic Asian Indians in north India, subjects with myostatin polymorphism A55T were reported to have a higher body fat percentage, fat mass, and truncal adiposity [27]. We have observed that a lower serum myostatin concentration is related to metabolic syndrome, central obesity, low high-density lipoprotein cholesterol, and high TG among 246 diabetic patients after adjusting age and gender [28]. Thus, when we determine the normal value of serum myostatin, it is prudent to take fat-related parameters into consideration.

Myostatin exerts its main effect through autocrine/paracrine mode, and excess myostatin peptide will spill-over into the circulation. Like the other members of the TGF-β superfamily, myostatin is regulated by complex interactions with many proteins in extracellular matrix [29]. Therefore, intracellular mRNA and secretion in isolated muscle tissues may not faithfully reflect its circulating levels. Matsakas et al reported that rat myostatin mRNA expression reduced after swimming training for 4 weeks, not after acute bout of exercise; however, no myostatin protein change was detected [30]. Myostatin is also expressed in the cardiac muscle, and this may contribute to the serum myostatin level [31]. Myostatin mRNA expression is increased after infarction and exercise in the cardiac muscle, but remain unchanged in the canine model of dilated cardiomyopathy and chronic heart failure [22, 32]. Myostatin may be a mediator of cardiac development; however, the exact mechanism still need to be elucidated [33].

We previously proposed an “accelerator-brake model” to illustrate the role of myostatin in regulating muscle growth [12]. Myostatin served as a negative feedback molecule during the process of muscular growth or regeneration and limited the skeletal muscle mass [10]. In this model, myostatin is activated only when muscle is growing or regenerating. Our athletes performing strength training had significantly greater muscle mass, BMI, grip strength, and daily energy expenditure than their untrained counterparts. However, the serum myostatin level of the two groups was not significantly different. The main problem might be the timing of blood sampling after exercise training. In another study, the acute response to a single bout of isometric exercise in rats was a rapid increase in myostatin mRNA levels 30 min to 6 h post-exercise, and a return to baseline levels 24 h post-exercise [34]. The athlete’s body composition is at a steady state, so there is no difference in myostatin concentration. Future longitudinal study to observe the dynamic change of muscle mass after strength training may prove the role of myostatin in decelerating muscle growth.

Age is a determining factor of serum myostatin and IGF-1. Lakshman et al reported lower myostatin level in the elderly population aged 66.4 ± 4.7 years [23], which could be explained by “accelerator-brake model”. The negative relation between IGF-1 and age was also well established [35]. Since different age groups have different muscle mass, it is desirable to establish an age-specific norm for the level of these biomarkers. However, the age range of our subjects is too narrow to establish such a norm, so we tried to show their relationship with linear regression model.

IGF-1, a downstream molecule of growth hormone, is a serum protein with insulin-like metabolic activities and growth controlling ability. IGF-1 expression activates muscular satellite cells and counteracts muscle function decline in the mouse model of muscular dystrophy [14]. GH or IGF-1 has been taken by athletes for its anabolic effect to enhance sports performance [36]. Since the concentration of GH is highly variable due to its pulsatile secretion pattern, IGF-1 is suggested as a good indirect surrogate for the GH/IGF-1 axis in the GH-2004 Project. An elevated serum IGF-1 level indicated sustained activation of the GH-IGF-1 axis. The reported normal value of IGF-1 is 297 ± 10 ng/ml, and is related to gender, age, nutrition status, and GH administration [35]. In our study, the serum IGF-1 level was 293 ± 114 ng/ml, which was comparable to previous published reports. Future evaluations of serum IGF-1 levels in other sports are crucial for application in anti-doping.

There are some limitations in this study. First, the sample size was small, so the statistical power was limited. Besides, measurement errors in self-reported physical activity levels might interfere with regression analysis. Although PAR is accurate in estimating energy expenditure compared with other physical activity questionnaires [37], subjectively perceived exercise intensity may vary between individuals. Furthermore, the MET value assigned to each intensity level of physical activity may differ in respondents having different BMI [38]. In addition, since the reference value is only for male athlete in certain sports, further validation in other sports is needed for actual implementation. Finally, because no commercial kit was available, we cannot measure the concentration of the pro-peptide, which inactivates myostatin by binding on it [4].

In conclusion, the reference values of serum myostatin and IGF-1 levels were 12.3 ± 3.6 ng/ml and 293 ± 114 ng/ml, respectively, in the young healthy males. The serum myostatin level was negatively related to body fat percentage; and serum IGF-1 was positively related to height after adjustment for age. It is necessary to detect serum myostatin together with other biomarkers, such as IGF-1, follistatin, and propeptide. Measurement of myostatin alone would not be sufficient. The reference values can provide a basis for future study, especially in the field of doping.