Total osteocalcin in serum predicts testosterone level in male type 2 diabetes mellitus

Abstract

Objective: To investigate the relationship between total osteocalcin (total OC) and testosterone level in the serum of male patients with type 2 diabetes mellitus (T2DM). Methods: 98 male T2DM were recruited in our cross-section study. Their anthropometric parameters were measured and body mass index (BMI) was calculated. Serum markers including glucose, insulin, HbA1c, testosterone, total OC and other bone metabolic markers were examined. Simple Pearson correlation and multiple regression analysis were performed. Results: Poor level of glucose control was found in the patients (HbA1c 9.19±2.53 and FPG 8.20±2.98). In simple Pearson correlation, total OC was positively related with testosterone (r=0.236, P=0.019), and this relation still existed after considering all the parameters of the patients (Model 1, 2 and 3). Conclusion: In male patients with T2DM, total OC was positively correlated with testosterone, so total OC might predict the testosterone level in the serum.

Introduction

In men, testosterone plays a key role in promoting secondary sexual characteristics and maintaining the function of male reproductive tissues such as the testis and prostate. In addition, testosterone is essential for health and well-being [1] as well as the prevention of osteoporosis [2]. While in type 2 diabetes patients (T2DM), including the disorder of glucose metabolism, there was also a lower testosterone level [3]. However, due to economic cause, testosterone was not a parameter regularly examined in T2DM patients.

Increasing evidence has supported that bone is an endocrine organ in recent years [4]. Osteocalcin (OC) is a very critical factor in the crosstalk between bone and energy metabolism [5], especially the undercarboxylated form. Undercarboxylated OC (ucOC) is able to enhance insulin secretion by β-cells, insulin sensitivity and energy expenditure [6]. Moreover, ucOC also affects the testosterone level. Oury et al did experiments with the co-culture assays and found that ucOC was able to induce testosterone production by the testes [7]. In clinical practice, there was also a positive correlation between testosterone and ucOC [8]. However, total OC is the more frequently examined form of OC in clinical practice rather than ucOC.

If there is a relation between total OC and testosterone level in the serum, it will be very useful to predict the testosterone for the physicians with the total OC level. Moreover most clinical and animal studies show it is ucOC that can regulate the male fertility not the total OC, whether total OC level is related with testosterone level in the serum is still not very clear. So, to investigate the association between total OC and testosterone is both on the benefit of clinical application and further mechanism explore of the bone-derived hormone. Thus a cross section study was performed in T2DM male patients to test the level of total OC and testosterone in the serum and analyze their relationship.

Subjects and methods

Subjects

We screened 40-75 years old male patients with T2DM who visited in our hospital from January 1^st 2011 to December 1^st 2011. Some of them were excluded as follows: 1) those with liver, heart or kidney function failure and osteoporosis; 2) with conditions known to affect testosterone such as hypogonadism; 3) using drugs that would affect osteocalcin such as vitamin D and bisphosphonate. At last, 98 male patients were enrolled in our cross section study. Human investigation was conducted according to the principles expressed in the Declaration of Helsinki. All subjects agreed to participate in the study and gave written informed consent and the protocol was approved by the Ethics Committee of Shanghai Tenth People’s Hospital of Tongji University, School of Medicine.

Anthropometric measurements

Weight measured with a plethysmography scale was scaled when patients wore minimal clothing and height was measured with a stadiometer. Body mass index (BMI) was calculated by dividing the weight by height squared (kg/m²).

Serum biomarkers

After overnight fasting, serum was collected. Biochemical markers, including fasting plasma glucose (FPG) and fasting insulin level (FINS) were measured with standard biochemical methods. Bone metabolic markers including C-terminal telopeptide fragments of type I collagen (CTX) and parathyroid hormone (PTH) were measured with an ELISA (Roche) method. Testosterone, follicle-stimulating hormone (FSH) and luteinizing hormone (LH) were examined with Elecsys and Cobas analyzers (Roche) according to the manufacturer’s instructions.

Serum total OC was measured using an N-MID Osteocalcin ELISA kit (Elecsys, Roche diagnostic Ltd., Switzerland). Intact osteocalcin is unstable due to protease cleavage between amino acids 43 and 44. The N-MID-fragment, resulting from cleavage, is considerably more stable. This assay detects the stable N-MID-fragment.

Testosterone in blood is present as free testosterone (FT) (1-4%) or bound testosterone. The bound testosterone is loosely bound to albumin or to sex hormone binding globulin (SHBG) (60-70%). The term bioavailable testosterone (BAT) refers to the sum of free testosterone plus albumin-bound testosterone, testosterone bound to SHBG is biologically inactive because of the strong affinity between SHBG and testosterone. The Elecsys Testosterone II assay is based on a competitive test principle using a high affinity monoclonal antibody (sheep) specifically directed against testosterone. Endogenous testosterone released from the sample by 2-bromoestradiol competes with the added testosterone derivative labeled with a ruthenium complex for the binding sites on the biotinylated antibody. (http://www.roche-diagnostics.cz).

Statistical analysis

Nonparametric tests (Kolmogorov-Smimo test) were used to examine whether the baseline parameters were followed with normal distribution. If they followed the normal distribution, they were expressed by mean ± SD, otherwise, and they were shown as quantile. Pearson correlation was used to analyze the relation between testosterone or osteocalcin and other parameters in the male T2DM patients. Multiple linear regression analysis was used to set up the models between testosterone and other parameters. Statistical Product and Service Solutions 17.0 (SPSS 17.0) were used for all analyses. Differences were considered to be statistically significant where P < 0.05.

Results

Basic parameters in the enrolled patients

In these 40-75 years old male patients, the duration of T2DM was from months to years (12.00, 123.00). Poor glucose level was found (HbA1c 9.19±2.53 and FPG 8.20±2.98), while the insulin level of most of them was nearly in the normal range (8.24, 22.88). The average figure of this population was overweight or obesity (BMI: 25.05±4.34). Testosterone (13.51±6.89), FSH (8.52±4.88) and total OC (11.31±3.85) were in lower grade while LH (6.99±3.75) was near the upper layer in the subjects. CTX was in normal range (0.21, 0.55) (Table 1).

Table 1.

The baseline of the subjects

Patients	Value	Normal range
Number	98
Age (years old)	60.28±13.61	-
Duration (months)	(12.00, 123.00)	-
BMI (Kg/m2)	25.05±4.34	18.50~24.90
HbA1c (%)	9.19±2.53	4.50-6.30
FPG (mmol/l)	8.20±2.98	3.90-6.10
FINS (mU/L)	(8.24, 22.88)	2.60-24.90
Testosterone (nmol/L)	13.51±6.89	9.90-27.80
FSH (IU/L)	8.52±4.88	1.50-12.40
LH (ng/ml)	6.99±3.75	1.70-8.60
TOC (ng/ml)	11.31±3.85	14.00-46.00
CTX (ng/ml)	(0.21, 0.55)	<0.85

Duration: duration of type 2 diabetes mellitus; BMI: body mass index; FPG: fasting plasma glucose; FINS: fasting insulin level; FSH: follicle-stimulating hormone; LH: luteinizing hormone; TOC: total osteocalcin; CTX: C-terminal telopeptide of type I collagen. Duration, FINS and CTX were not followed the normal distribution, they were expressed as quantile, others were showed as mean ± SD.

Correlation between testosterone level and other parameters in male T2DM

In the relation between testosterone and other parameters, simple Pearson correlation analysis showed testosterone level was only related with total OC in the serum (Table 2A: r=0.236, P=0.019). While, total OC was also positive related with log (CTX) (Table 2B: r=0.325, P=0.002) (Table 2).

Table 2.

Simple Pearson correlation between testosterone (A) or total OC (B) in the serum and other parameters in T2DM

A
	Testosterone
	r	P
Age	-0.144	0.156
Log (Duration)	-0.029	0.790
BMI	-0.186	0.077
HbA1c	0.005	0.963
FPG	-0.022	0.839
Log (FINS)	-0.068	0.523
FSH	0.173	0.111
LH	0.197	0.070
TOC	0.236	0.019
Log (CTX)	-0.037	0.732
B
	TOC
	r	P
Age	-0.116	0.254
Log (Duration)	-0.111	0.302
BMI	-0.051	0.633
HbA1c	-0.209	0.048
FPG	-0.171	0.104
Log (FINS)	0.064	0.551
FSH	0.034	0.755
LH	0.007	0.949
Testosterone	0.236	0.019
Log (CTX)	0.325	0.002

Duration, FINS and CTX were not followed the normal distribution, so we transformed them with logarithm.

The relationship between testosterone and total OC was enhanced by the multi-linear regression analysis. Total OC was positively associated with testosterone level in the serum whatever in the model of patients’ instinct state (age and BMI involved in Model 1: R²=0.089, P=0.011), glucose control condition (age, BMI, Duration and HbA1c involved in Model 2: R²=0.076, P=0.048) or other related hormones (age, BMI, Duration, HbA1c, FSH and LH involved in Model 3: R²=0.144, P=0.012) (Table 3).

Table 3.

Multiple linear regression analysis of testosterone and other parameters

	R2	P
Model 1	0.089	0.011
Model 2	0.076	0.048
Model 3	0.144	0.012

Model 1: age, BMI and TOC were selected. Model 2: age, BMI, Duration, HbA1c and TOC were selected. Model 3: age, BMI, Duration, HbA1c, FSH, LH and TOC were selected.

Discussion

As compared with the normal range of healthy persons, the subjects recruited in our study manifested higher FPG and HbA1c level, which was consistent with most T2DM reported in China [9]. So we concluded that these subjects could represent the most T2DM individuals of our country.

In our results, testosterone was only related with total OC among all the parameters. Theoretically, testosterone will be regulated by FSH and LH, and then a negative relation should be found between testosterone and FSH or LH. Dhindsa et al. found T2DM patients revealed a lower LH and FSH level than those in control group; they concluded a hypothalamic rather than a pituitary defect in the patients [10]. As the previous studies, testosterone might associate with age [11] and BMI [12], FPG [13] and HbA1c [14], while, we did not observe these relations. We inferred in this cross section study, the parameters could not be balanced well, for example, the duration of the T2DM had significant difference. And the sample size was not large enough to get statistical results, since all of the studies that demonstrate the interrelations were much more than 100 recruiters. Except testosterone, total OC was negatively associated with log (CTX). Type I collagen is also secreted by osteoblasts as well as OC and degraded into CTX in the bone matrix, so it reflects the level of bone turnover. Thus, the OC has a parallel declination with CTX. Although ucOC was negative with HbA1c or FPG, the relation between total OC and the glucose metabolism did not get an agreement [15].

Total OC, a small γ-carboxyglutamate protein preferentially expressed by mature osteoblasts, which comprises both the carboxylated and the undercarboxylated forms, has been utilized traditionally as a marker of bone formation or bone turnover. The carboxylated OC binds the calcium ions after release in the bone matrix, which results in the mineralization of the bone [16]. For the ucOC form, recent studies have found it could increase the insulin secretion and sensitivity [6] and at the same time insulin and its receptor could regulate bone acquisition and remodeling [17,18], so ucOC has been considered to play an important role in the crosstalk between bone and energy metabolism. For the phenomenon that testosterone was positively associated with total OC in our study, there would be three explanations. One explanation was that osteocalcin in the serum could affect the testosterone. Among cultured mouse primary Leydig cells in the presence or absence of supernatants of osteoblast cultures or other mesenchymal cell types cultures, they found only osteoblasts could increase the production of testosterone, moreover, osteocalcin was the regulator in the process [7]. Another explanation for the clinical phenomenon was that testosterone affected osteocalcin. Testosterone deficiency could induce bone loss and fragile fracture in senile [19]. Under this condition, the OC secretion of osteoblasts would be affected. At the same time, as the study of ours were a cross section study, we could not exclude that there was other parameter existed that could both affect testosterone and osteocalcin, an example for the parameter was blood glucose level. Low testosterone [3] and low total OC [20] were all reported in T2DM patients. But we found no linear relation between glucose level and testosterone or osteocalcin. So maybe the third consideration could not be supported in the population we studied.

There were some limitations for our study. First, since we focused on the relation between testosterone and total OC, we just did the investigation in patients with diabetes, and we did not examine the population with normal glucose level. Second, there have already been studies about ucOC influencing on testosterone, we did not test the ucOC in the serum as a comparison. Lastly, the number of population is not very large, but after the reference literature review, 98 patients were enough to get the conclusion logically.

Disclosure of conflict of interest

None.