https://orcid.org/0000-0002-4554-1631
Abstract
The exact mechanisms that trigger the onset of puberty are not well known. Adipomyokines are postulated to stimulate the central neural network. In the present study, we investigated irisin levels in girls with central precocious puberty (CPP), slowly progressing precocious puberty (SPPP), or premature thelarche (PT); we also studied prepubertal girls and to determine if this adipomyokine could be used as a marker in this context.
A total of 94 girls including 33 with CPP, 31 with precocious puberty (PP) variants (SPPP or PT), and 30 healthy controls were enrolled to the study. The mean irisin levels were compared between groups. The bivariate correlations of irisin levels with clinical and laboratory parameters were assessed. Multivariate linear regression analysis was performed to determine independent predictive factors of irisin levels.
Irisin levels were higher in the CPP group compared with the other groups (CPP group: 723.25 ± 62.35 ng/mL; PP variants group: 529.60 ± 39.66 ng/mL; and control group: 325.03 ± 27.53 ng/mL) (P < 0.001). Irisin levels were positively correlated with body mass index standard deviation scores (BMI-SDS), height-SDS, weight-SDS, bone age, uterus long axis, ovary size, baseline FSH and LH, and peak LH levels. Multivariate linear regression analysis revealed that irisin levels had the strongest correlation with peak LH. The other independent predictive factor of irisin levels was BMI-SDS.
The mean irisin levels were higher in patients with CPP compared with other groups. The results of this study imply that increased irisin levels may be used as a marker of CPP provided that these findings are confirmed in larger prospective studies.
Puberty begins with the reactivation of the hypothalamus-pituitary-gonad (HPG) axis, which is inhibited in childhood. Onset of secondary sexual features and gaining of reproductive capacity along with physical, psychological, and social growth take place during puberty (1), which is regulated by the interaction among genetic, neuroendocrine, and environmental (adequate nutrition, exercise, and exposure to chemicals) factors (2). The hypothalamic neural network is under suppression by the balance between inhibitor and stimulatory signals such as kisspeptin/neurokinin B/dynorphin, makorin ring finger protein 3, gamma-aminobutyric acid, gonadotropin inhibitory hormone, excitatory amino acids, and neuropeptide Y (3). During the pubertal period, the axis is activated by pulsatile GnRH secretion, which is associated with body weight (fat/muscle), metabolic state, and energy reserves (4). These signals and their mechanisms of action have not been fully elucidated.
In girls, precocious puberty (PP) is defined as the onset of secondary sexual features before 8 years of age (5). Its estimated incidence is between 1/5000 and 1/10 000 (6) and is characterized with accelerated growth rate, bone maturation, and early fusion of epiphysis resulting from increased estrogen levels (7). The treatment of choice for this condition is GnRH analogues (8). Premature thelarche (PT) is defined as the isolated onset of secondary breast development before 8 years of age and has a prevalence of 4.7% (9). It is a benign condition and considered a variant of normal growth, with no treatment needed. In contrast to central precocious puberty (CPP), growth rate and bone age are normal and gonadotropin and estradiol levels are within normal prepubertal range in PT (10). However, 13% of cases with PT may progress to CPP (11). A less severe form of CPP, slowly progressing variant of precocious puberty (SPPP) has been defined previously (12). In contrast to children with PT, in children with SPPP, bone age may be slightly accelerated and ovarian volumes may be slightly enlarged, whereas LH response to GnRH may not be as strong as seen in rapidly progressing CPP. SPPP is also accepted as a benign variant of CPP; these cases should be followed up clinically but treatment is not generally recommended (13). These conditions are common causes of referrals to pediatric endocrinologists. Physical and clinical findings should be carefully evaluated. Clinical findings, bone age, pelvic ultrasound, baseline LH, FSH, and estradiol (E2) levels, and a GnRH stimulation test are used for the differential diagnosis (7). There is no specific test for early diagnosis for these disorders of puberty.
There is a recent trend toward earlier onset of puberty in some regions of the world. Although the exact reasons underlying this trend remain unknown, it is postulated to be associated with the increasing prevalence of obesity in childhood (14). Recent studies reported that PP is closely related with obesity (15). Sexual maturation has been found to be associated with BMI, waist circumference, total body fat, and abdominal subcutaneous tissue (16). In an experimental study, fat-rich nutrition led to obesity, precipitated early puberty, and increased the expression of kisspeptin in hypothalamus in offspring of C57BL/6J mouse (17).
Recent studies suggest that the timing of onset of puberty is related to the metabolic conditions and energy reserve of the body and that the hypothalamic network is activated when the body reaches a certain fat and/or muscle mass (18). However, it is not well known which metabolic signals regulate this relationship. To date, the role of various metabolic signals including leptin, insulin, ghrelin, IGF, nesfatin, and neurokinin-b has been investigated (19-22). These studies reported variable results and the exact mechanisms remain unknown. There is need for further studies on this subject.
Irisin is a recently discovered adipomyokine that plays an important role in energy homeostasis. It is synthetized in the skeletal muscle, subcutaneous, and visceral adipose tissues from fibronectin type III domain-containing protein 5 (FNDC5); this process is promoted by peroxisome proliferator-activated receptor-gamma coactivator 1α (PGC-1α) activity (23). Irisin plays a role in browning of the white adipose tissue and thermogenesis and is found in tissues such as muscle, fat, ovary, uterus, central nervous system (hypothalamus, pituitary gland, and cerebrospinal fluid [CSF]), testis, adrenal glands, and pancreas (24, 25). Circulating and CSF irisin levels have been reported to correlate with BMI, anthropometric parameters, fat mass, and muscle mass (26, 27). Reinehr et al. (28) reported that irisin levels significantly increased during the onset of puberty in obese children in a 1-year follow-up study. Wahab et al. (29) recently reported that FNDC5 and irisin were expressed in the HPG axis in association with the developmental and metabolic status and in a gender-specific manner and that they promoted the expression of GnRH.
These studies investigated the correlation of irisin levels with body fat-muscle mass and BMI in children. However, studies about the relationship between puberty and irisin levels are scarce. In the present study, we aimed to investigate the role of irisin levels in the onset of puberty in nonobese girls with CPP, SPPP, and PT and also in that of healthy girls.
Patients and Methods
Girls with early breast development between the ages of 5 through 8 who presented to our pediatric endocrinology clinic between December 2017 and December 2018 were enrolled in this study. The control group consisted of healthy age-matched prepubertal girls (Tanner stage I) without obesity who presented to our pediatric endocrinology clinic.
The local ethical committee approved the study protocol and the parents of all of the participants provided written informed consent.
Height and weight were measured in all cases. BMI was calculated as weight/height square (kg/m2). Height standard deviation score (SDS) was calculated using Centers for Disease Control and Prevention growth scales (30). BMI-SDS was calculated according to the method described by Rosner et al. (31). Obesity was defined according to the BMI cutoffs reported by the International Obesity Task Force (32).
Puberty stage was determined according to Marshall and Tanner criteria (33). Bone age (BA) and BA-SDS were estimated as reported by Greulich and Pyle (34). An experienced radiologist performed all ovary and uterus ultrasound examinations.
Exclusion criteria were other forms of PP, obesity, chronic diseases, developmental anomalies, medication use, and history of radiotherapy or chemotherapy.
Baseline and stimulated LH and FSH levels were measured using immunochemiluminescence (Abbott Architect i2000sr, Abbott Park, Illinois, USA). Baseline serum E2 levels were measured using double antibody E2 assay (DPC, Los Angeles, CA). The lower limit of sensitivity for the estradiol assay was 10 pg/mL.
GnRH stimulation tests were performed in cases with early breast development. Blood samples were taken for baseline FSH and LH analyses before the injection of GnRH. Afterwards, GnRH (0.1 mg) was given as an IV injection over 1 minute and serial blood samples were taken after 20, 40, 60, and 90 minutes for FSH and LH analyses. A peak LH level > 5 IU/L was regarded as a pubertal response (33). Subjects with a BA to chronological age ratio > 1.2 and a peak LH level > 5 IU/L were considered to have CPP, whereas patients who did not meet these criteria were included in the PP variants group (this group consisted of disorders considered as benign conditions such as PT or possible SPPP) (13, 35). All of the subjects in the CPP group underwent cranial magnetic resonance imaging. Subjects with organic brain disorders on cranial magnetic resonance imaging were also excluded from the study.
Sample collection and storage
In CPP and its variants groups, venous blood samples were obtained after 12 hours of overnight fasting in the morning and collected in dry tubes at 8:00 am during the GnRH stimulation test before the injection of GnRH; fasting blood samples of the control group were recruited the day after their visit for irisin measurement with the ELISA method. Blood samples were centrifuged at 1200g for 10 minutes and serum samples were transferred to Eppendorf tubes using plastic Pasteur pipettes and stored at –80°C until analyzed.
Irisin analysis
Samples were thawed and the Irisin ELISA kit (Aviscera Bioscience Inc., Sunnyvale, CA; catalog no: SK00170-01) was used for the quantitative measurement of irisin in serum samples. The intra-assay and inter-assay precisions of the kit were 4% to 6% and 8% to 10%, respectively. Assays were performed according to the manufacturers’ instructions and standards and solutions were prepared. At the end of the procedures, optical density was measured at 450 nm (Thermo Scientific Microplate Reader, Grand Island, NY). The detection range of this kit was between 1.638 and 1024 ng/mL.
Statistical analysis
The analyses were performed using SPSS 21.0 (IBM Corp., Armonk, NY). The data were expressed as mean ± standard deviation. One-way ANOVA test was used to compare the groups and in case of a significant difference between the groups, and Tukey’s post hoc analysis test was used to determine which groups differed from one another. The bivariate correlations between the variables were analyzed using the Spearman correlation test. Multivariate linear regression analysis was also performed to determine important predictive factors for irisin levels. There were many girls with undetectable estradiol values (n = 52); therefore, we excluded estradiol from the regression models. However, we compared the irisin levels between the girls with undetectable and detectable estradiol levels via Student t test. Additionally, receiver operator characteristic curves were plotted using irisin and presence or absence of CPP. Youden’s index was calculated as sensitivity – (1 – specificity), obtained from coordinates of the curve. The highest value for Youden’s index was used to identify the cutoff value of irisin. A P value of < 0.05 was considered statistically significant.
Results
A total of 33 subjects with CPP (mean age, 7.35 ± 1.15 years), 31 with PP variants (mean age, 7.07 ± 0.67 years), and 30 prepubertal girls (mean age, 6.93 ± 0.70 years) were enrolled. The Tanner pubertal stages II/III ratio in CPP and PP variants groups were 23/10 and 26/5, respectively. Comparison of the study groups with respect to study variables is summarized in Table 1. The mean age of the study groups was not statistically different. The mean height-SDS and weight-SDS of CPP and its variant groups were higher than that of the control group (P = 0.005 and P = 0.001, respectively). The BMI-SDS and peak FSH were similar in all groups. The mean BA was higher in the CPP group (8.63 ± 1.50) compared with the PP variant (7.92 ± 0.98) and control (6.60 ± 1.01) groups (P < 0.001 and P < 0.001, respectively). Ovary size, baseline FSH and LH, and peak LH were higher in the CPP group compared with the other groups and also in the PP variant groups compared with the control group.
Comparisons of the Groups Regarding Clinical and Laboratory Parameters
| Group 1 | Group 2 | Group 3 | P | |
|---|---|---|---|---|
| (CPP) | r (PP Variants) | (Control) | ||
| Age (y) | 7.35 ± 1.15 | 7.07 ± 0.67 | 6.93 ± 0.70 | 0.155 |
| BMI SDS | 0.49 ± 0.60 | 0.59 ± 0.74 | 0.22 ± 0.77 | 0.108 |
| Irisin (ng/mL)a | 723.25 ± 62.35 | 529.60 ± 39.66 | 325.03 ± 27.53 | <0.001 |
| Bone age (y)a | 8.63 ± 1.50 | 7.92 ± 0.98 | 6.60 ± 1.01 | <0.001 |
| Uterine long axis (cm) | 36.68 ± 6.82 | 39.14 ± 35.04 | 29.47 ± 5.67 | 0.285 |
| Right ovary size (mL)a | 2.73 ± 1.82 | 2.19 ± 1.08 | 1.34 ± 0.72 | 0.002 |
| Left ovary size (mL)a | 2.67 ± 1.74 | 1.91 ± 0.77 | 1.08 ± 0.65 | <0.001 |
| Basal FSH (mIU/mL)b | 3.16 ± 1.78 | 1.82 ± 1.77 | 1.43 ± 0.83 | <0.001 |
| Basal LH (mIU/mL)b | 0.68 ± 1.34 | 0.07 ± 0.09 | 0.06 ± 0.06 | 0.003 |
| Peak FSH (mIU/mL) | 13.29 ± 5.95 | 12.13 ± 4.24 | - | 0.376 |
| Peak LH (mIU/mL) | 14.35 ± 11.87 | 2.11 ± 0.80 | - | <0.001 |
| Group 1 | Group 2 | Group 3 | P | |
|---|---|---|---|---|
| (CPP) | r (PP Variants) | (Control) | ||
| Age (y) | 7.35 ± 1.15 | 7.07 ± 0.67 | 6.93 ± 0.70 | 0.155 |
| BMI SDS | 0.49 ± 0.60 | 0.59 ± 0.74 | 0.22 ± 0.77 | 0.108 |
| Irisin (ng/mL)a | 723.25 ± 62.35 | 529.60 ± 39.66 | 325.03 ± 27.53 | <0.001 |
| Bone age (y)a | 8.63 ± 1.50 | 7.92 ± 0.98 | 6.60 ± 1.01 | <0.001 |
| Uterine long axis (cm) | 36.68 ± 6.82 | 39.14 ± 35.04 | 29.47 ± 5.67 | 0.285 |
| Right ovary size (mL)a | 2.73 ± 1.82 | 2.19 ± 1.08 | 1.34 ± 0.72 | 0.002 |
| Left ovary size (mL)a | 2.67 ± 1.74 | 1.91 ± 0.77 | 1.08 ± 0.65 | <0.001 |
| Basal FSH (mIU/mL)b | 3.16 ± 1.78 | 1.82 ± 1.77 | 1.43 ± 0.83 | <0.001 |
| Basal LH (mIU/mL)b | 0.68 ± 1.34 | 0.07 ± 0.09 | 0.06 ± 0.06 | 0.003 |
| Peak FSH (mIU/mL) | 13.29 ± 5.95 | 12.13 ± 4.24 | - | 0.376 |
| Peak LH (mIU/mL) | 14.35 ± 11.87 | 2.11 ± 0.80 | - | <0.001 |
Data are shown as mean ± (SD). P value < 0.05.
Abbreviations: BMI, body mass index; CPP, central precocious puberty; PP variants, premature thelarche or slowly progressing precocious puberty; SDS, standard deviation score.
aPost hoc Tukey’s test revealed that groups 1, 2, and 3 are statistically different from each other.
bPost hoc Tukey’s test revealed that group 1 is statistically different from groups 1 and 2.
Comparisons of the Groups Regarding Clinical and Laboratory Parameters
| Group 1 | Group 2 | Group 3 | P | |
|---|---|---|---|---|
| (CPP) | r (PP Variants) | (Control) | ||
| Age (y) | 7.35 ± 1.15 | 7.07 ± 0.67 | 6.93 ± 0.70 | 0.155 |
| BMI SDS | 0.49 ± 0.60 | 0.59 ± 0.74 | 0.22 ± 0.77 | 0.108 |
| Irisin (ng/mL)a | 723.25 ± 62.35 | 529.60 ± 39.66 | 325.03 ± 27.53 | <0.001 |
| Bone age (y)a | 8.63 ± 1.50 | 7.92 ± 0.98 | 6.60 ± 1.01 | <0.001 |
| Uterine long axis (cm) | 36.68 ± 6.82 | 39.14 ± 35.04 | 29.47 ± 5.67 | 0.285 |
| Right ovary size (mL)a | 2.73 ± 1.82 | 2.19 ± 1.08 | 1.34 ± 0.72 | 0.002 |
| Left ovary size (mL)a | 2.67 ± 1.74 | 1.91 ± 0.77 | 1.08 ± 0.65 | <0.001 |
| Basal FSH (mIU/mL)b | 3.16 ± 1.78 | 1.82 ± 1.77 | 1.43 ± 0.83 | <0.001 |
| Basal LH (mIU/mL)b | 0.68 ± 1.34 | 0.07 ± 0.09 | 0.06 ± 0.06 | 0.003 |
| Peak FSH (mIU/mL) | 13.29 ± 5.95 | 12.13 ± 4.24 | - | 0.376 |
| Peak LH (mIU/mL) | 14.35 ± 11.87 | 2.11 ± 0.80 | - | <0.001 |
| Group 1 | Group 2 | Group 3 | P | |
|---|---|---|---|---|
| (CPP) | r (PP Variants) | (Control) | ||
| Age (y) | 7.35 ± 1.15 | 7.07 ± 0.67 | 6.93 ± 0.70 | 0.155 |
| BMI SDS | 0.49 ± 0.60 | 0.59 ± 0.74 | 0.22 ± 0.77 | 0.108 |
| Irisin (ng/mL)a | 723.25 ± 62.35 | 529.60 ± 39.66 | 325.03 ± 27.53 | <0.001 |
| Bone age (y)a | 8.63 ± 1.50 | 7.92 ± 0.98 | 6.60 ± 1.01 | <0.001 |
| Uterine long axis (cm) | 36.68 ± 6.82 | 39.14 ± 35.04 | 29.47 ± 5.67 | 0.285 |
| Right ovary size (mL)a | 2.73 ± 1.82 | 2.19 ± 1.08 | 1.34 ± 0.72 | 0.002 |
| Left ovary size (mL)a | 2.67 ± 1.74 | 1.91 ± 0.77 | 1.08 ± 0.65 | <0.001 |
| Basal FSH (mIU/mL)b | 3.16 ± 1.78 | 1.82 ± 1.77 | 1.43 ± 0.83 | <0.001 |
| Basal LH (mIU/mL)b | 0.68 ± 1.34 | 0.07 ± 0.09 | 0.06 ± 0.06 | 0.003 |
| Peak FSH (mIU/mL) | 13.29 ± 5.95 | 12.13 ± 4.24 | - | 0.376 |
| Peak LH (mIU/mL) | 14.35 ± 11.87 | 2.11 ± 0.80 | - | <0.001 |
Data are shown as mean ± (SD). P value < 0.05.
Abbreviations: BMI, body mass index; CPP, central precocious puberty; PP variants, premature thelarche or slowly progressing precocious puberty; SDS, standard deviation score.
aPost hoc Tukey’s test revealed that groups 1, 2, and 3 are statistically different from each other.
bPost hoc Tukey’s test revealed that group 1 is statistically different from groups 1 and 2.
Mean irisin level was higher in the CPP group (723.25 ± 62.35) compared with its variant (529.60 ± 39.66) and control groups (325.03 ± 27.53) (P < 0.001 and P < 0.001, respectively) and also in the PP variant groups compared with the control group (P < 0.001). The individual irisin levels for each girl in the 3 groups are shown in Fig. 1 (“scattergram”). Receiver operator characteristic curve analysis revealed that the cutoff value of irisin in girls for CPP was 618.03 ng/mL (sensitivity 93.9% and specificity 98.4%). Additionally, the irisin levels of the girls who had detectable estradiol (n = 42) were higher than those of the girls who had undetectable estradiol (n = 52) (577.08 ± 169.25 vs. 484.23 ± 164.28), but the difference was not statistically significant (P = 0.09).
The comparison of the irisin levels. CPP, central precocious puberty; PP variants group, children with premature thelarche or slowly progressing precocious puberty.
Bivariate correlation analysis revealed that irisin level was positively correlated with BA, height-SDS, weight-SDS, ovary size, baseline FSH, baseline LH, and peak LH (Table 2). Moreover, irisin level had the strongest correlation with peak LH in multivariate linear regression analysis followed by BMI-SDS (Table 3).
Bivariate Correlations Coefficients and P Values Between Irisin and Other Parameters
| Irisin | |||
|---|---|---|---|
| N | r | P | |
| Age (y) | 94 | 0.332 | 0.001 |
| BMI SDS | 94 | 0.386 | <0.001 |
| Bone age (y) | 94 | 0.632 | <0.001 |
| Right ovary size (mL) | 94 | 0.315 | 0.003 |
| Left ovary size (mL) | 94 | 0.419 | <0.001 |
| Basal FSH (mIU/mL) | 94 | 0.456 | <0.001 |
| Basal LH (mIU/mL) | 94 | 0.564 | <0.001 |
| Peak FSH (mIU/mL) | 64 | -0.037 | 0.770 |
| Peak LH (mIU/mL) | 64 | 0.710 | <0.001 |
| Estradiol (pg/mL) | 42 | 0.181 | 0.223 |
| Irisin | |||
|---|---|---|---|
| N | r | P | |
| Age (y) | 94 | 0.332 | 0.001 |
| BMI SDS | 94 | 0.386 | <0.001 |
| Bone age (y) | 94 | 0.632 | <0.001 |
| Right ovary size (mL) | 94 | 0.315 | 0.003 |
| Left ovary size (mL) | 94 | 0.419 | <0.001 |
| Basal FSH (mIU/mL) | 94 | 0.456 | <0.001 |
| Basal LH (mIU/mL) | 94 | 0.564 | <0.001 |
| Peak FSH (mIU/mL) | 64 | -0.037 | 0.770 |
| Peak LH (mIU/mL) | 64 | 0.710 | <0.001 |
| Estradiol (pg/mL) | 42 | 0.181 | 0.223 |
Abbreviations: BMI, body mass index; SDS, standard deviation score.
Bivariate Correlations Coefficients and P Values Between Irisin and Other Parameters
| Irisin | |||
|---|---|---|---|
| N | r | P | |
| Age (y) | 94 | 0.332 | 0.001 |
| BMI SDS | 94 | 0.386 | <0.001 |
| Bone age (y) | 94 | 0.632 | <0.001 |
| Right ovary size (mL) | 94 | 0.315 | 0.003 |
| Left ovary size (mL) | 94 | 0.419 | <0.001 |
| Basal FSH (mIU/mL) | 94 | 0.456 | <0.001 |
| Basal LH (mIU/mL) | 94 | 0.564 | <0.001 |
| Peak FSH (mIU/mL) | 64 | -0.037 | 0.770 |
| Peak LH (mIU/mL) | 64 | 0.710 | <0.001 |
| Estradiol (pg/mL) | 42 | 0.181 | 0.223 |
| Irisin | |||
|---|---|---|---|
| N | r | P | |
| Age (y) | 94 | 0.332 | 0.001 |
| BMI SDS | 94 | 0.386 | <0.001 |
| Bone age (y) | 94 | 0.632 | <0.001 |
| Right ovary size (mL) | 94 | 0.315 | 0.003 |
| Left ovary size (mL) | 94 | 0.419 | <0.001 |
| Basal FSH (mIU/mL) | 94 | 0.456 | <0.001 |
| Basal LH (mIU/mL) | 94 | 0.564 | <0.001 |
| Peak FSH (mIU/mL) | 64 | -0.037 | 0.770 |
| Peak LH (mIU/mL) | 64 | 0.710 | <0.001 |
| Estradiol (pg/mL) | 42 | 0.181 | 0.223 |
Abbreviations: BMI, body mass index; SDS, standard deviation score.
Linear Regression Analysis Coefficientsa
| Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | |
|---|---|---|---|---|---|
| B | SE | Beta | |||
| (Constant) | 372.101 | 111.557 | 3.336 | 0.002 | |
| BMI SDS | 68.489 | 20.105 | 0.405 | 3.407 | 0.002 |
| Bone age | 8.924 | 15.057 | 0.110 | 0.593 | 0.558 |
| Right ovarian volume | -28.330 | 16.522 | -0.453 | -1.715 | 0.097 |
| Left ovarian volume | 23.128 | 18.814 | 0.346 | 1.229 | 0.229 |
| Basal FSH | 4.166 | 9.719 | 0.083 | 0.429 | 0.671 |
| Basal LH | -97.774 | 36.445 | -1.118 | -2.683 | 0.012 |
| Peak LH | 14.774 | 3.685 | 1.322 | 4.009 | <0.001 |
| Peak FSH | 0.417 | 2.893 | 0.022 | 0.144 | 0.886 |
| Age | 20.786 | 25.633 | 0.202 | 0.811 | 0.424 |
| Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | |
|---|---|---|---|---|---|
| B | SE | Beta | |||
| (Constant) | 372.101 | 111.557 | 3.336 | 0.002 | |
| BMI SDS | 68.489 | 20.105 | 0.405 | 3.407 | 0.002 |
| Bone age | 8.924 | 15.057 | 0.110 | 0.593 | 0.558 |
| Right ovarian volume | -28.330 | 16.522 | -0.453 | -1.715 | 0.097 |
| Left ovarian volume | 23.128 | 18.814 | 0.346 | 1.229 | 0.229 |
| Basal FSH | 4.166 | 9.719 | 0.083 | 0.429 | 0.671 |
| Basal LH | -97.774 | 36.445 | -1.118 | -2.683 | 0.012 |
| Peak LH | 14.774 | 3.685 | 1.322 | 4.009 | <0.001 |
| Peak FSH | 0.417 | 2.893 | 0.022 | 0.144 | 0.886 |
| Age | 20.786 | 25.633 | 0.202 | 0.811 | 0.424 |
Abbreviations: BMI, body mass index; SDS, standard deviation score; Sig., significance.
aDependent variable: irisin.
Linear Regression Analysis Coefficientsa
| Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | |
|---|---|---|---|---|---|
| B | SE | Beta | |||
| (Constant) | 372.101 | 111.557 | 3.336 | 0.002 | |
| BMI SDS | 68.489 | 20.105 | 0.405 | 3.407 | 0.002 |
| Bone age | 8.924 | 15.057 | 0.110 | 0.593 | 0.558 |
| Right ovarian volume | -28.330 | 16.522 | -0.453 | -1.715 | 0.097 |
| Left ovarian volume | 23.128 | 18.814 | 0.346 | 1.229 | 0.229 |
| Basal FSH | 4.166 | 9.719 | 0.083 | 0.429 | 0.671 |
| Basal LH | -97.774 | 36.445 | -1.118 | -2.683 | 0.012 |
| Peak LH | 14.774 | 3.685 | 1.322 | 4.009 | <0.001 |
| Peak FSH | 0.417 | 2.893 | 0.022 | 0.144 | 0.886 |
| Age | 20.786 | 25.633 | 0.202 | 0.811 | 0.424 |
| Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | |
|---|---|---|---|---|---|
| B | SE | Beta | |||
| (Constant) | 372.101 | 111.557 | 3.336 | 0.002 | |
| BMI SDS | 68.489 | 20.105 | 0.405 | 3.407 | 0.002 |
| Bone age | 8.924 | 15.057 | 0.110 | 0.593 | 0.558 |
| Right ovarian volume | -28.330 | 16.522 | -0.453 | -1.715 | 0.097 |
| Left ovarian volume | 23.128 | 18.814 | 0.346 | 1.229 | 0.229 |
| Basal FSH | 4.166 | 9.719 | 0.083 | 0.429 | 0.671 |
| Basal LH | -97.774 | 36.445 | -1.118 | -2.683 | 0.012 |
| Peak LH | 14.774 | 3.685 | 1.322 | 4.009 | <0.001 |
| Peak FSH | 0.417 | 2.893 | 0.022 | 0.144 | 0.886 |
| Age | 20.786 | 25.633 | 0.202 | 0.811 | 0.424 |
Abbreviations: BMI, body mass index; SDS, standard deviation score; Sig., significance.
aDependent variable: irisin.
Discussion
The HPG axis is under suppression except during the fetal period, minipuberty, and puberty (36). The exact mechanism of reactivation of the HPG axis after being under suppression during the postnatal period is unclear, but various hypotheses have been postulated (18, 37-39). Among these hypotheses, notable is that a certain body weight (fat and protein mass) is needed for the activation of the HPG axis. Irisin is known to be associated with the metabolic conditions and energy reserves of the body (40); therefore, studies investigating the association between puberty and irisin levels are of importance. To the best of our knowledge, no previous study has investigated irisin levels in children with CPP and PP variants in comparison with healthy prepubertal children.
Irisin is an adipomyokine that contains 112 amino acids and was discovered by Böstrom et al. (41) in 2012. Irisin is a proteolytic product of FNDC5, which is stimulated by PGC-1α activity. It is released from the muscles following exercise and promotes browning of the white adipose tissue, oxygen consumption, and thermogenesis (41, 42).
Studies in adults and children suggest that irisin levels are correlated with BMI and muscle and adipose tissue mass (24, 43, 44). Elizondo-Montemayor et al. (45) compared underweight, normal weight, and obese children between the ages of 2 and 6 with respect to their irisin levels. They found the lowest irisin levels in underweight children and reported that irisin levels were positively correlated with waist circumference and BMI percentiles. Other studies reported that irisin levels were higher in obese children compared with normal weight children and found positive correlations between irisin levels and BMI-SDS, adiposity indices, waist-to-hip ratio, and obesity risk (27, 46-48). The findings of the present study also support this consideration, and serum irisin levels were positively correlated with BMI-SDS in nonobese girls between the ages of 5 and 8.
Another factor that affects irisin levels is puberty. However, to the best of our knowledge, there is only a 1-year cohort study that investigated this association in 40 obese children (28). They found positive correlations among baseline irisin levels, BMI-SDS, and metabolic syndrome parameters. They also reported that the children in whom puberty began during the 1-year follow-up had significantly higher irisin levels. On the other hand, Tas et al. (49) have demonstrated that irisin levels did not differ after the onset of puberty with the progressing pubertal maturation in male adolescents; comparison of the pubertal irisin levels with their prepubertal levels was not performed. We found significantly higher irisin levels with the onset of puberty in nonobese girls. Furthermore, higher irisin levels (compared with the control group) were also demonstrated in girls with PP variants, which has been postulated to result from partial activation of the HPG axis (50).
There is a growing body of evidence regarding puberty physiology. The release of GnRH is regulated by Kisspeptin and its receptor kiss-R1 and is mediated by the excitatory neurokinin B receptor and inhibitory dynorphin receptor. These 3 regulatory neurons play a key role in terms of onset of puberty. An increase in excitatory stimuli and a decrease in inhibitory stimuli lead to an increase in pulsatile GnRH secretion and thus the puberty process begins (3, 51). The metabolic and nutrition status and increase in energy reserve are important factors that precipitate the onset of puberty in girls. Although the exact mechanisms are unknown, they have an excitatory effect on GnRH secretion (4, 52-55). Staiano et al. (56) found an association between sexual development and BMI, waist circumference, total body fat, and abdominal subcutaneous tissue in 382 children between the ages of 7 and 16. Atay et al. (16) reported a significant correlation between PT and BMI-SDS in girls between the ages 4 and 8.
Irisin is an important candidate for serving as a trigger for HPG axis activation and reproductive development by means of transferring information about the metabolic state and energy reserve of the body (18). Irisin/FNDC5 expression has been reported to take place in the components of the reproductive axis such as the hypothalamus, pituitary gland, ovary, and testis, in addition to muscle and adipose tissue (57-61). It has been reported to be colocalized with neuropeptide Y primarily on the ventromedial and arcuate nucleus of the hypothalamus (the area responsible for nutrition, energy homeostasis, and reproduction regulation) (26, 29). Piya et al. (62) reported that serum irisin levels were 20- to 25-fold higher than that in CSF. Ruan et al. (26) found this ratio to be 12.7 in their study. The ratio of CSF/serum irisin level in both of these studies are higher than that reported for leptin (1/100), adiponectin (1/1000), and resistin (1/100) (63, 64).
All of these studies point to the close association of irisin with the HPG axis. They also support the hypothesis that irisin may be a potential messenger that triggers the neurohumoral axis activation needed to activate the onset of puberty.
Wahab et al. (10) very recently observed a putative interaction between irisin-immunoreactive fibers and GnRH-immunoreactive cell bodies; they also observed that irisin stimulated the synthesis and release of GnRH in their experimental study in primates. They reported that FNDC5 and PGC1A transcript levels were significantly higher in the muscle, pituitary gland, and posterior hypothalamus of females compared with males. They also reported that expression of FNDC5 and PGC1A increased with the onset of puberty in both genders. Poretsky et al. (65) reported that irisin (at a concentration of 100 ng/mL) administration led to LH synthesis in human pituitary cells and estradiol synthesis in granulosa cells of the ovary in their in vitro study. We also found positive correlations between irisin and baseline FSH, baseline LH, and peak LH in the present study. However, we did not observe correlations between irisin and estradiol or peak FSH.
The present study has some limitations. First, we did not include obese girls and we did not perform any test to determine body composition of the children; thus, we could not clearly assess the effect of body weight or composition on our study variables. We also did not include boys in whom the rate of CPP is low; thus, our study does not provide information about the effect of gender on the association between irisin levels and puberty. Although we defined a cutoff value for irisin levels, the sample size was small; therefore, we suggest that it would be more appropriate to confirm this finding in larger scale studies. Another limitation of the present study were the potential confounders of increased fat and muscle mass during puberty because higher fat and muscle mass are also associated with increased irisin levels. Therefore, we cannot exclude the contribution of increased fat and muscle mass to the higher irisin levels in the CPP group in this study. The findings of the present study do not provide a cause-and-effect relationship because of the cross-sectional design.
In summary, to the best of our knowledge, the present study is the first to comparatively investigate irisin levels in CPP and its variants and healthy prepubertal girls. We found that higher irisin levels were associated with activation of puberty. Irisin levels were the highest in the CPP group as well as other markers of puberty such as BA, ovary size, baseline FSH, baseline LH, and peak LH. Irisin levels were correlated with markers of pubertal activation such as BA, height-SDS, ovary size, baseline FSH, baseline LH, and peak LH. Multivariate linear regression analysis revealed that irisin had the strongest association with peak LH followed by BMI-SDS. Our findings suggest that irisin is associated with the activation of puberty. However, the findings of the present study do not establish a cause-and-effect relationship and do not provide conclusive information whether irisin stimulates the HPG axis or the increase in irisin levels only accompanies the onset of puberty. There is need for further prospective large-scale studies.
Abbreviations
- BA
bone age
- BMI
body mass index
- CPP
central precocious puberty
- CSF
cerebrospinal fluid
- E2
estradiol
- FNDC5
fibronectin type III domain-containing protein 5
- HPG
hypothalamus-pituitary-gonad
- PGC-1α
proliferator-activated receptor-gamma coactivator 1α
- PP
precocious puberty
- PT
premature thelarche
- SDS
standard deviation score
- SPPP
slowly progressing precocious puberty.
Acknowledgments
Financial Support: This study is supported by the Bezmialem Vakif University Scientific Research Projects Foundation.
Additional Information
Disclosure Summary: The authors have nothing to disclose.
Data Availability
Some or all data generated or analyzed during this study are included in this published article or in the data repositories listed in References.
References
