**3. Results**

The studied RG and UC groups did not differ (ES < 0.05; *p* > 0.05) in chronological age, body height, body mass, BMI, and REE (Table 1). Age at menarche was higher in RG compared with UC groups, body fat %, FM, and REE/kg were lower, and LBM, training volume, VO2peak/kg, and Wmax/kg higher in RG in comparison with UC (*p* < 0.05; ES > 0.12). Although mean serum irisin was not significantly (*p* > 0.05) different between the groups, RG had moderately higher (ES = 0.06) irisin values compared with UC (Table 2). The difference in FGF-21 concentrations between groups was only small in magnitude (ES < 0.05; *p* > 0.05). In addition, leptin levels were largely (ES = 0.34; *p* < 0.0001) and resistin concentrations moderately (ES = 0.06; *p* = 0.077) lower in RG in comparison with UC.

Table 3 presents correlations of irisin and FGF-21 concentrations with energy measures. In the RG group, serum irisin concentration was positively correlated with REE (*r* = 0.40; *p* = 0.021) and serum leptin level (*r* = 0.60; *p* = 0.013) (Figure 1). In addition, the relationship between irisin and leptin was independent of age, body FM, and LBM (*r* = 0.57; *p* = 0.001). In the UC group, serum irisin concentration was positively correlated to resistin levels (*r* = 0.31; *p* = 0.036), which remained significant after controlling for age, body FM, and LBM (*r* = 0.57; *p* = 0.016). In the RG group, serum FGF-21 concentration was significantly correlated to body FM (*r* = 0.46; *p* = 0.007) and serum leptin levels (*r* = 0.45; *p* = 0.009) (Figure 1), as opposed to the UC group only to leptin (*r* = 0.54; *p* = 0.014). Finally, irisin was related to FGF-21 (*r* = 0.36; *p* = 0.012) only in RG, and the association between irisin and FGF-21 was independent of age, body FM and LBM (*r* = 0.36; *p* = 0.049).


**Table 1.** Body composition and energy metabolism values (mean ± SD) in rhythmic gymnasts (RG) and untrained controls (UC).

ES, effect size (eta squared); BMI, body mass index; REE, resting energy expenditure; VO2peak/kg, peak oxygen consumption per kg lean body mass; Wmax/kg, maximal power output per kg body mass.

**Table 2.** Energy homeostasis regulating hormone concentrations (mean ± SD) in rhythmic gymnasts (RG) and untrained controls (UC).


ES, effect size (eta squared); FGF-21, fibroblast growth factor-21.

**Table 3.** Relationships of irisin and fibroplast growth factor-21 (FGF-21) with energy measures in rhythmic gymnasts (RG) and untrained controls (UC).


REE, resting energy expenditure; VO2peak/kg, peak oxygen consumption per kg lean body mass. Correlations with *p* < 0.05 are listed in bold.

**Figure 1.** Relationships of irisin levels with resting energy expenditure (REE) (**A**) (*r* = 0.40; *p* = 0.021) and leptin (**B**) (*r* = 0.60; *p* < 0.0001), and relationships of fibroblast growth factor-21 (FGF-21) with body fat mass (FM) (**C**) (*r* = 0.46; *p* = 0.007) and leptin (**D**) (*r* = 0.45; *p* = 0.009) in rhythmic gymnasts.

#### **4. Discussion**

The present study was undertaken to examine the effect of prolonged athletic activity on energy homeostasis regulating hormones, irisin, and FGF-21 in highly trained adolescent RG. We found that serum irisin and FGF-21 concentrations were not significantly different between RG and UC groups. In addition, serum irisin levels were associated with REE and FGF-21 levels with body FM in RG. In contrast, irisin and FGF-21 were not related to energy expenditure and energy availability measures in lean nonathletic UC. These results demonstrate that circulating irisin and FGF-21 levels may play a role in signaling energy status in a setting of a state of long-term high energy expenditure in adolescent athletes.

It has been suggested that the myokine irisin could play an endocrine control of energy metabolism [22,31]. Circulating irisin levels are elevated in obesity as an excess energy state [32] and reduced in anorexia nervosa as a depleted energy state [33], suggesting that irisin levels may reflect energy stores. Indeed, irisin is known to increase energy expenditure by inducing the browning of subcutaneous white adipocytes, which are metabolically favorable for burning energy through thermogenesis [10,17]. Accordingly, irisin is thought to improve glucose and lipid metabolism in response to exercise training [34,35]. Our data linking irisin levels with REE in highly trained adolescent RG are consistent with the literature indicating that irisin may signal energy availability and promote energy expenditure [34]. Similarly, irisin levels were related to REE in young female runners [22]. No such correlation was found in the UC group in our study. One potential explanation for this could be that the UC subjects were in normal body weight and relatively balanced in terms of energy intake and expenditure. Furthermore, irisin concentrations were also not related to different parameters of energy expenditure in patients with anorexia nervosa [10,36]. It could be speculated that the adolescent RG in our study were in a state of subtle energy

deficit, as indicated by their reduced body FM and lower measured REE after correcting for LBM, but not in the state of extreme energy deficit observed in anorexia nervosa. In accordance with young female runners [22], no relationship between irisin with body FM was observed in studied RG, while elevated irisin levels have been reported to be independently associated with obesity risk factors, including body FM in obese adolescents [18]. However, significantly lower (ES = 0.34; *p* < 0.0001) leptin concentrations in RG were related to irisin levels, and this association was independent of age, body fat, and lean masses (*r* = 0.57; *p* = 0.001), demonstrating the muscle-adipose tissue crosstalk in energy homeostasis in adolescent lean females with chronic athletic activity. These seemingly conflicting results demonstrate the specificity of irisin interactions with different markers of energy metabolism in various populations and further studies are needed to clarify the exact role of irisin in energy homeostasis. However, the results of our study and that of Singhal et al. [22] would suggest that irisin concentrations may accentuate the increase in energy expenditure in lean adolescent female athletes, as indicated by the positive associations of circulating irisin levels with measured REE in these individuals.

Studies investigating the effects of chronic athletic activity on circulating irisin levels in adolescent athletes are rare. Earlier studies demonstrated that irisin levels were not significantly different between moderately trained young eumenorrheic runners, amenorrheic runners, and nonathletic controls, although irisin levels were the lowest in amenorrheic runners [19]. Another study found that irisin concentrations were lower in young amenorrheic athletes compared with eumenorrheic athletes and nonathletes [22], while a third study in elite male adolescent tennis players did not observe large variations in irisin concentrations over a competitive tournament season, although it was suggested that irisin may modify overall performance during a long-lasting season [37]. In our study, serum irisin levels were moderately, but not significantly higher (ES = 0.06; *p* > 0.05) in RG compared with the UC group. However, although VO2peak/kg was higher (ES = 0.12; *p* = 0.012) in RG compared with UC, no relationship between irisin concentration and maximal aerobic performance was observed in RG, similar to previous studies demonstrating that maximal aerobic performance does not influence circulating irisin concentrations in blood [21,38]. In accordance with our results, training volume did not modify circulating irisin concentrations in adult highly trained athletes [15,39] and basal irisin may not be a good marker of training volume over a training macrocycle [38]. However, as interval training caused moderate and significant increases in serum irisin levels in previously untrained adults [40,41], circulating irisin may be a more sensitive marker of training intensity rather than training volume in adult athletes [16,34] as well as in exercising adolescents [42]. It is known that training in rhythmic gymnastics is quite intensive involving numerous jumping exercises daily [6,8] and our studied RG athletes were tested during the preparatory period with a relatively high training volume. It has also been suggested that irisin may be a marker of muscle damage [43] and can provide anti-inflammatory protection [34]. Although the biological role of irisin as a moderator of energy metabolism in response to acute training load remains to be fully elucidated [13], circulating irisin levels have been reported to increase as a result of an acute training session in young female athletes [16]. According to the results of our study, it appears that moderately higher irisin levels in highly trained adolescent RG did not reflect training stress in a setting of chronic high energy expenditure state, whereas decreased irisin levels in amenorrheic athletes likely represent an adaptive response to reduce training stress and conserve energy [22].

In accordance with irisin levels, serum FGF-21 concentrations are higher in obesity [44], reduced in anorexia nervosa [45] and related to body FM [22,23], suggesting that FGF-21 levels may reflect energy stores. Accordingly, 12 weeks of aerobic exercise training decreased circulating FGF-21 concentrations, body mass and glucose uptake in overweight and obese men [46]. It has been suggested that short-term energy expenditure results in increases in circulating FGF-21 levels [47], while long-term chronic energy expenditure may lead to decreased FGF-21 levels to preserve energy [22,48]. Accordingly, acute training load with adequate duration increased serum FGF-21 concentrations in young female

athletes [16]. In our study, serum FGF-21 levels were similar between the RG and UC groups in accordance with the previous studies [19,22]. In addition, we found that serum FGF-21 levels were positively correlated with body FM and leptin concentrations in adolescent RG. These relationships suggest crosstalk between muscle and adipose tissue in energy homeostasis and that FGF-21 could be used as a marker of energy stores in adolescent lean females with chronically increased energy expenditure.

Serum FGF-21 concentrations were positively correlated with irisin levels in the RG group. The secretion of FGF-21 and irisin leads to the white adipose tissue browning, uncoupling protein-1-mediated thermogenesis and energy expenditure [19,22]. The secretion of FGF-21 and irisin is increased by the upregulation of peroxisome proliferatoractivated receptor-γ, an exercise-induced transcriptional coactivator that promotes energy metabolism [17,49]. Accordingly, our finding of a positive relationship between irisin and FGF-21 suggests a shared pathway for the regulation of energy metabolism in adolescent athletes with high athletic activity.

This study has some limitations. At first, our cross-sectional design rules out the possibility of identifying causal relationships, particularly from the correlation analysis with some individual outliers. Secondly, a relatively small sample size was used, although the number of individuals in both groups was comparable to previous similar studies with athletes in this area [15,19,31,37]. The main strength of the present study is that, to the best of our knowledge, this is the first study investigating whether specific myokine levels such as irisin and FGF-21 are related to the measures of energy homeostasis in highly trained female adolescent athletes as the participants of this study were international level Estonian rhythmic gymnasts from different sports clubs.

#### **5. Conclusions**

Serum irisin and FGF-21 concentrations were not significantly different between lean adolescent athletes and nonathletic control subjects. Irisin was associated with energy expenditure and FGF-21 with energy availability in lean adolescent athletes with a state of heavily increased energy expenditure, while no relationships of irisin and FGF-21 with energy status measures were observed in lean nonathletic adolescents with normal daily energy expenditure levels.

**Author Contributions:** Conceptualization, J.J., L.R. and V.T.; methodology, J.J., A.-L.T. and L.R.; formal analysis, J.J, L.R. and V.T.; investigation, L.R., P.P. and K.M.; writing-original draft preparation, J.J.; writing-review and editing, L.R., A.-L.T., P.P., K.M. and V.T..; project administration, J.J.; funding acquisition, J.J. and V.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Estonian Ministry of Education and Science Institutional Grant PRG 1428.

**Institutional Review Board Statement:** This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Research Ethics Committee of the University of Tartu (ethical approval code number 274/T-3).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on a request from the corresponding author for researchers who meet the criteria for access to confidential data.

**Acknowledgments:** We thank all the study participants and staff for their assistance.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

