Suppressing Effect of Free Triiodothyronine on the Negative Association between Body Mass Index and Serum Osteocalcin Levels in Euthyroid Population

Previous studies found that thyroid hormones stimulate osteoblast-like cells to secrete osteocalcin. We aimed to investigate the association between serum thyroid hormone and serum osteocalcin in euthyroid population. The study recruited 1152 community-based euthyroid subjects (average age 59 ± 8 years), among whom 677 were women. Serum free triiodothyronine (FT3), free thyroxine (FT4), thyroid-stimulating hormone (TSH), and osteocalcin were measured by electrochemiluminescence immunoassays. After adjusting for age and gender, partial correlation analysis showed that FT3 and FT3/FT4 were both positively correlated with body mass index (BMI) and serum osteocalcin levels (all P < 0.05) and BMI was negatively correlated with serum osteocalcin levels (P < 0.01), while FT4 and TSH were not correlated with serum osteocalcin levels (all P > 0.05). Age, gender, blood pressure, thyroid hormones, and multiple metabolic risk factors were included in the ridge regression model. FT3 and FT3/FT4 were independently and positively associated with serum osteocalcin levels (all P < 0.05), while BMI was independently and negatively associated with serum osteocalcin levels (P < 0.01). The mediating effect model showed that FT3 and FT3/FT4 suppressed the negative association between BMI and serum osteocalcin levels, with suppressing effects of 6.41% and 10.39%, respectively. In euthyroid subjects, both FT3 and FT3/FT4 were positively associated with serum osteocalcin levels, and they further suppressed the negative association between BMI and serum osteocalcin levels.


Introduction
Endocrine functions of the bone have attracted extensive attention in recent years. e bone influences energy metabolism via secreting various cytokines and hormones, among which osteocalcin, specifically secreted by osteoblasts and osteocytes, plays an important role [1]. Osteocalcin is encoded by the human gene BGLAP and is the most abundant noncollagen protein in the bone marrow [2]. As a marker of bone remodeling, osteocalcin regulates the size and shape of bone mineral substances in the vitamin K-dependent c-carboxylated form [3]. Outside the bone, osteocalcin promotes islet β cell proliferation, stimulates insulin secretion, and regulates adipokine expression to influence energy metabolism [4]. e production and secretion of osteocalcin are regulated by local and humoral factors [2]. yroid hormones are important for bone development, maturation, and metabolism [5]. Different types of thyroid hormone receptors are expressed in human osteoblasts [6]. Previous studies found that thyroid hormones regulated the proliferation and differentiation of osteoblasts and the expression of bone turnover factors including osteocalcin [7]. In clinical studies, serum osteocalcin levels were found to increase in hyperthyroidism [8,9], decrease in hypothyroidism [10], and recover to normal levels after treatment [11,12]. yroid hormones and osteocalcin both play important roles in energy metabolism; however, whether circulating thyroid hormones are associated with serum osteocalcin levels remains unclear. Previous studies in Australian and German populations did not reach consensus on the association between serum thyroid hormone levels and serum osteocalcin levels [13,14], in which the bioactive form of thyroid hormone, free triiodothyronine (FT3), was not tested. At present, no study has explored the association between serum thyroid hormones and osteocalcin levels in a Chinese population. In this study, we aimed to explore the associations between serum thyroid hormone levels and serum osteocalcin levels in a community-based euthyroid population and to investigate the potential role of serum thyroid hormone levels in the association between body mass index (BMI) and serum osteocalcin levels.

Subjects.
is study recruited 1152 community-based subjects from October 2015 to July 2016 in Shanghai [15]. Postmenopausal status was defined as at least 1 year of amenorrhea in the absence of other medical conditions [16]. All subjects completed a questionnaire survey and underwent a physical examination and laboratory measurements. e study included Han Chinese adults who volunteered to participate in the study and were able to provide the information required for the survey. According to the 2001 International Physical Activity Questionnaire (IPAQ), the levels of physical activity are classified as light, moderate, or heavy [17]. e exclusion criteria included a valid history of diabetes or cardiovascular diseases, moderate to severe anemia, malignancy, severe hepatic or renal dysfunction, acute infection, antithyroid therapies, abnormal thyroid function, hormone replacement treatment such as thyroid hormones or sex hormones, bone fracture, and use of hypotensive drugs, lipid-regulating drugs, glucocorticoids, lithium, amiodarone, or other medications that affect bone metabolism. All participants signed informed consent forms. e study was approved by the Ethics Committee of the Shanghai Jiao Tong University Affiliated Sixth People's Hospital and was conducted in accordance with the Declaration of Helsinki.

Body Measurements and Laboratory Examinations.
Height and body weight were measured according to standard methods described in our previous study [15], BMI � body weight (kg)/height 2 (m 2 ). Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured as the mean of 3 blood pressure measurements taken at 3-minute intervals.

Statistical Analyses.
All data analyses were performed by using SPSS version 22.0 (SPSS, Inc., Chicago, IL, USA) statistical software. A two-tailed P value <0.05 was considered statistically significant.
e Kolmogorov-Smirnov test was used to evaluate the normal distribution of all continuous variables. Variables with a normal distribution are expressed as the means ± standard deviations. Skewed distributed variables were expressed as medians (interquartile ranges). Categorical variables were expressed as numbers (percentages). Partial correlation analysis was used to explore the correlations among serum thyroid hormone levels, serum osteocalcin levels, and BMI after adjusting for age and gender. Ridge regression analysis was used to explore the associations between serum thyroid hormone levels and serum osteocalcin levels after adjusting for confounding factors. Based on a mediating effect model, the suppressing effect was evaluated to investigate the influence of serum thyroid hormone levels on the associations between BMI and serum osteocalcin levels [18].

Clinical Characteristics of Study Subjects.
e study included 1152 subjects (age range: 27-81 years; average 59 ± 8 years). ere were 475 men and 677 women. Among the women, 96 were eumenorrheic and 581 were postmenopausal. e clinical characteristics of the study subjects are shown in Table 1

Suppressing Effect of FT3 on the Association between
Serum Osteocalcin Levels and BMI. A mediating effect model was further used to investigate the influence of FT3 and FT3/ FT4 on the association between serum osteocalcin levels and BMI ( Figure 2). Age, gender, HbA1c, and HDL-c were adjusted. FT3 and FT3/FT4 suppressed the negative association between BMI and serum osteocalcin levels, with suppressing effects of 6.41% (Sobel test � 1.97, P < 0.05) and 10.39% (Sobel test � 3.05, P < 0.01), respectively.

Discussion
e present study explored the associations between serum thyroid hormone levels and serum osteocalcin levels in a community-based euthyroid population. We found that FT3 and FT3/FT4 were positively associated with serum osteocalcin levels, while FT4 and TSH were not associated with serum osteocalcin levels. FT3 and FT3/FT4 suppressed the negative association between BMI and serum osteocalcin levels.
yroid hormones contribute greatly to bone metabolism. Hyperthyroidism leads to accelerated bone turnover and decreased bone mass and is one of the clinical causes of secondary osteoporosis [19]. Conversely, hypothyroidism leads to dysplasia of bone maturation and epiphyseal development, thus further leading to growth retardation and skeletal deformity [20]. Serum osteocalcin levels in patients with hyperthyroidism were significantly higher than those of healthy controls, while they were significantly lower in patients with hypothyroidism. In patients with hypothyroidism, serum osteocalcin recovered to normal levels after thyroxine replacement treatment [12]. Siru et al. recruited 3338 elderly Australian men, among whom 3117 were euthyroid, 135 were subclinical hypothyroid, and 86 were subclinical hyperthyroid. ey found that there were no differences in serum osteocalcin levels among the three groups, and there were no associations of FT4 or TSH with serum osteocalcin levels in euthyroid subjects [13]. In a German cohort of 2654 subjects, there was a nonlinear U-shaped curve relationship between TSH and serum osteocalcin levels in women, while no association was found between TSH and osteocalcin in men [14]. Similar to the results of Siru et al., we also found no associations of FT4 or TSH with serum osteocalcin levels in euthyroid subjects.
Previous studies found that triiodothyronine (T3) regulated the proliferation and differentiation of osteoblasts and the expression of various bone turnover factors including osteocalcin [7,21]. FT3 is the bioactive form of thyroid hormones; however, whether FT3 influences serum osteocalcin in euthyroid subjects remains unknown. FT3 was not detected in the two previous studies in Australian and German populations. In this study, we investigated the associations of FT3 and FT3/FT4 with serum osteocalcin. A previous study found that osteocalcin levels increase with exercise interventions [22], and physical activity was included as a confounding factor in the regression analysis. Consistent with the results of previous cell experiments, FT3 and FT3/FT4 were significantly and positively associated with serum osteocalcin, suggesting that FT3 and FT3/FT4 may influence serum osteocalcin in euthyroid population. In addition, we used ridge regression analysis instead of  traditional multivariate linear regression analysis to explore the independent influencing factors of serum osteocalcin levels. According to the results of partial correlation analysis, we found that FT3, FT4, and FT3/FT4 were significantly correlated with each other. e principle of a ridge regression analysis allowed us to simultaneously process variables with collinearity in one regression model, by which we could observe the independent influences of FT3, FT4, and FT3/FT4 on serum osteocalcin levels. e potential association mechanism of serum thyroid hormone and serum osteocalcin has been explored in many previous studies. yroid hormone receptors were found to be expressed in both the rat osteogenic-like cell line ROS 17/ 2.8 and the mouse osteogenic-like cell line MC3T3-E1. T3 stimulated the secretion of osteocalcin in vitro [23,24]. In vivo, the mRNA expression of osteocalcin in the femur of rats increased in a dose-dependent manner with T3 injection [25]. e thyroid hormone receptor TRβ1 and sympathetic  Note. K value � 0.4, F � 9.002, P < 0.001. a Men were coded as "0," and women were coded as "1." b Light was coded as "1," moderate was coded as "2," and heavy was coded as "3." adrenal receptor β2 signaling pathways play a key role in mediating T3 to stimulate osteocalcin secretion [26,27]. In the bone, DIO2 and DIO3 synergistically regulate the thyroid hormone signaling [28], which may partly explain our finding that FT3/FT4 is positively associated with osteocalcin. Further studies are needed to elucidate the mechanisms by which adipose tissue, thyroid, and bone interact with each other. Previous studies found that subjects with obesity and nonalcoholic fatty liver disease had higher FT3 and FT3/FT4 levels [29][30][31]. Whether high FT3 and FT3/FT4, which are associated with metabolic disorders, are risk factors or compensatory mechanisms remains controversial. In this study, we found that FT3 suppressed the negative association between BMI and serum osteocalcin levels, suggesting that the high FT3 within the reference range may alleviate the decline in serum osteocalcin levels in obesity. Osteocalcin is a marker of bone remodeling. e obesity-related high FT3 level may induce bone remodeling by regulating osteoblast activity, by which the bone experiences structural changes to adapt to weight gain. In addition, osteocalcin exerts protective effects on energy metabolism; thus, the obesity-related high FT3 may promote the secretion of osteocalcin to compensate for glucose and lipid metabolism.
Our study has some limitations. Firstly, we could not determine causality between serum thyroid hormone levels, serum osteocalcin levels, and the obesity index based on a cross-sectional study. Secondly, the study subjects were Han Chinese adults, and thus our results may not be generalizable to other ethnic populations.
irdly, as potential confounding factors, gonadal hormones are not included in our study, which need to be explored in future studies.

Conclusions
In conclusion, FT3 and FT3/FT4 were both positively associated with serum osteocalcin levels in euthyroid subjects, indicating that FT3 may influence the secretion of osteocalcin in euthyroid subjects. Furthermore, FT3 and FT3/FT4 suppressed the negative association between BMI and serum osteocalcin levels, suggesting that obesity-related high FT3 levels may alleviate the decline in serum osteocalcin levels.

Data Availability
e data used to support the findings of this study are available from the corresponding authors upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.

Authors' Contributions
Xiaomin Nie and Yiting Xu contributed equally to this work. International Journal of Endocrinology 5