Testosterone deficiency and non-alcoholic fatty liver disease in men with type 2 diabetes mellitus

Testosterone (T) deficiency has a much wider scale than it was believed previously, affecting about less than half of the total male population after the age of 40 years [1].

The role of androgen deficiency in men in the pathogenesis of many diseases associated with metabolic syndrome has been increasingly discussed. Hypogonadism increases the risk of obesity, type 2 diabetes mellitus (DM2), arterial hypertension, and dyslipidemia [1]. In addition, the relationship of androgen deficiency with conditions such as insulin resistance, endothelial dysfunction, cytokine imbalance, and erectile dysfunction has been proven [2, 3, 4]. These relations are bidirectional in nature, and diseases mutually affect one another; therefore, we consider the deficiency of total testosterone (T) as an additional component of the metabolic syndrome [5, 6]. Indeed, the spectrum of disorders of various organs and systems that occur in case of androgen deficiency in men is immensely wide because it affects not only the reproductive system, the brain, the musculoskeletal system,, and the cardiovascular system but also carbohydrate and lipid metabolism indicators. Data on the effect of T on cardiovascular system diseases have remained controversial. However, most studies have proven the importance of T in the formation and development of vascular risks and the positive effect of T replacement therapy in men with hypogonadism [7, 8].

A new risk factor of cardiovascular disease has been identified: nonalcoholic liver disease [9, 10]. This condition is possibly due to the detected nonalcoholic fatty liver disease (NAFLD) 2–3 times more often in patients with obesity and DM2 [11]. Such a close relationship of nonalcoholic liver disease with the components of metabolic syndrome suggests that T deficiency may play an important pathogenetic role in the development of complex diseases under discussion.

However, only few studies have investigated the effect of T deficiency on the liver, and results have shown that this condition contributes to NAFLD development and progression [12, 13, 14]. As such, pathogenetic mechanisms linking NAFLD, hypogonadism, and associated metabolic diseases in men should be revealed.

AIM

This work aimed to assess the effect of hypogonadism on the formation and progression of NAFLD in men with DM2.

METHODS

Study design

An observational single-centre one-stage study was conducted.

Inclusion criteria

The study included 40-to 65-year-old men who had DM2. Hypogonadism was diagnosed in accordance with the diagnostic criteria of the Russian Association of Endocrinologists (2017). Exclusion criteria were primary and secondary forms of hypogonadism. A cohort of individuals with mixed functional or nonclassical hypogonadism was formed.

The patients included in the study consumed alcohol equivalent to not more than three alcohol units per week, so the alcoholic genesis of liver damage was eliminated.

Study conditions

This study was conducted in the Rostov State Medical University.

Study duration

Patients were enrolled in the study from 2018 to 2019.

Description of the medical intervention

All the patients underwent general clinical examination, including anthropometric parameter and blood pressure measurements. A single sample of their venous blood with a volume of 10 ml was obtained after a 12-hour fasting period. Fresh blood serum was used to study biochemical parameters. The blood was centrifuged for enzyme-linked immunosorbent assays, and the serum was frozen at −20 °C. The biochemical blood parameters of the patients were examined to characterise liver function (alanine aminotransferase [ALT] and aspartate aminotransferase [AST]). The de Ritis ratio and gamma-glutamyltranspeptidase (GGTP) contents were determined. The state of carbohydrate and lipid metabolism (glucose and immunoreactive insulin [IRI] on an empty stomach was identified by calculating the insulin resistance index HOMA, glycated haemoglobin (НbА_1с), and lipid profile. The levels of sex hormones, namely, luteinising hormone (LH), total T, sex hormone binding globulin (SHBG), free T, and adipohormones (resistin, adiponectin, and leptin), were measured. Afterward, the participants underwent magnetic resonance imaging (MRI) of the liver.

Primary study outcome

The main endpoints of the study were liver function indicators (ALT, AST, de Ritis ratio, and GGTP), adipohormone (resistin, adiponectin, and leptin) concentrations, and liver fat fraction revealed by MRI.

Subgroup analysis

The patients were divided into two groups based on clinical and laboratory examinations and depending on the presence or absence of hypogonadism.

• Group 1 included eugonadal patients with DM2.
• Group 2 included male patients with DM2 and hypogonadism diagnosed for the first time.

Methods of result registration

Biochemical studies were performed using a Bayer ADVIA 1650 analyzer, and HbA_1c was determined using a Siemens Healthcare Diagnostics DCA 2000+ analyzer. Immunoenzymometric tests were performed using a Zenyth 340 analyzer. Alkor-Bio laboratory kits (Russia) were used to examine the contents of sex hormones (total T, SHBG, and LH). IRI was determined using a laboratory kit (Monobind Inc., USA). The severity of insulin resistance was determined on the basis of glycemia and insulin levels by calculating the HOMA index:

HOMA = fasting glycemia (mmol/L) × insulin level (μU/ml)/22.5.

The de Ritis ratio was calculated as the AST/ALT ratio.

The content of adipohormones was studied using laboratory kits for enzyme-linked immunosorbent assay: leptin (Bcm Diagnostic LLC, Germany), resistin (Biovendor Laboratory, Czech Republic), and adiponectin (eBioscience, Austria).

The liver was subjected to MRI by using a Brilliance 64 Slice multislice helical X-ray computed tomograph (Philips Medical Systems, Netherlands) in accordance with the Dickson method involving the measurement of the proportion of hepatic fat weighted by proton density [15]. This method is considered the only one that quantifies liver fat content [16].

Ethical considerations

All the patients signed an informed consent form approved by the Local Ethics Committee of the Rostov State Medical University (Protocol No. 12/18 of 28.06.2018).

Statistical analysis

Data were statistically analysed using Statistica 11.0. All the variables were checked for normal distribution by using a Kolmogorov–Smirnov test. Most indicators had an abnormal distribution, so data were presented as medians and lower and upper quartiles Me (LQ and UQ). The significance of intergroup differences was assessed using a Mann–Whitney U test for two independent groups. Covariant analysis was performed using an ANCOVA module. Results were considered statistically significant at p < 0.05.

RESULTS

Subjects (participants) of the study

The study included 90 male patients with DM2, and their average age was 54 years [49; 57], and the duration of diabetes was 6 years [2; 10]. Group 1 was represented by 32 eugonadal patients with DM2. Group 2 included 58 male patients with DM2 and hypogonadism diagnosed for the first time. The groups were comparable in terms of age and duration of the course and treatment of DM2.

The comparative characteristics of the examined groups are presented in Table. 1. The laboratory criteria of hypogonadism combined with the clinical symptoms of androgen deficiency constituted the basis for dividing patients with DM2 into groups; as such, the groups differed significantly in the concentrations of total T, free T, and SHBG (p < 0.001), whose content was inversely proportional to that of T. LH levels were moderate to normal and comparable in both groups.

Table 1. Comparative characteristics of the groups of subjects.

Notes: BMI – body mass index; WC – waist circumference; HC – hip circumference; SBP – systolic blood pressure; DBP – diastolic blood pressure; HR – heart rate; SHBG – sex hormone binding globulin; LH – luteinising hormone.

Primary study results

Clinical studies revealed a more pronounced abdominal obesity in patients with hypogonadism, which was manifested as the body mass index (BMI) of 6% higher. The waist circumference (WC) of 6.9% greater than in patients of group 2 compared with the subjects of group 1 (p < 0.01). The hip circumference (HC) of group 2 was larger than that of group 1 (p < 0.001), but the WC/HC ratio in the groups was comparable. Blood pressure indicators should also be considered. Systolic blood pressure (SBP; p = 0.01) and diastolic blood pressure (DBP) levels (p = 0.02) in group 2 significantly increased compared with that in group 1. Their heart rates were comparable.

The comparative analysis of carbohydrate metabolism did not reveal group differences in fasting glycemia (6.9 [6.2; 7.8] mmol/L in group 1 and 7.4 [6.8; 8.3] mmol/L in group 2) and НbА1с (7.6% [6.5; 7.9] and 7.7% [6.6; 7.9], respectively). Conversely, differences in the IRI level and the insulin resistance index of HOMA were significant. The patients of both groups had hyperinsulinemia, which is a characteristic pathogenetic sign of DM2. However, fasting IRI in hypogonadal patients (33.7 [23.3; 49.9] mIU/ml) was more than 1.5 times higher (p = 0.019) than that in eugonadal men (21.5 [16.4; 34.6] mIU/ml). The HOMA index was twice as high (p = 0.009) in patients of group 2 compared with group 1 (6.8 [5.4; 9.6] – in group 1 versus 11.4 [7.2; 16.2] in group 2).

The levels of the total cholesterol (6.3 [5.7; 6.9] mmol/L and 6.3 [5.2; 7.2] mmol/L), low-density lipoproteins (3.3 [2.5; 3.8] mmol/L and 3.3 [2.7; 3.9] mmol/L), and high-density lipoproteins (1.4 [1.2; 1.5] mmol/L and 1.5 [1.3; 1.5] mmol/l) of the patients in groups 1 and 2 did not differ significantly. However, the content of triglycerides (TGs) of the patients with hypogonadism was 25% higher than that of the eugonadal male patients. Thus, the TG levels in groups 1 and 2 were 2.2 [1.9; 2.7] mmol/l and 2.7 [2.3; 3.8] mmol/L (p < 0.001), respectively.

The levels of adipohormones are presented in Fig. 1. Resistin concentrations showed a statistically significant (p < 0.001) 2-fold increase. Likewise, leptin concentrations in patients with T deficiency had a 12-fold increase compared with that in patients with normotestosteronemia. Moreover, differences in adiponectin concentrations in the groups were not statistically significant.

Fig. 1. Concentrations of adipohormones in male patients with type 2 diabetes mellitus, depending on the presence or absence of hypogonadism.

The concentrations of liver enzymes in the subjects are presented in Fig. 2. No significant differences in the levels of total bilirubin, direct bilirubin, and AST contents were observed. However, the ALT concentration and GGTP in male patients with hypogonadism were 24.5% (p = 0.02) and 60.5% higher (p = 0.001) than those in eugonadal patients. GGTP has the maximum sensitivity and specificity in terms of the diagnostics of liver pathology. This result was supported by a statistically significant difference in the ratio of AST/ALT enzymes, as characterised by de Ritis ratio, which was 60.4% significantly higher (p = 0.047) in group 2 (0.74 [0.61; 0.89] units) than in group 1 (0.8 [0.75; 0.94] units). This result confirmed liver damage in the subjects.

The revealed changes in laboratory indicators are supported by instrumental research data presented in Fig. 3. The proton densities of the fat fraction in the liver as revealed by MRI in groups 1 and 2 were 4.12% [2.25; 5.30] and 10.30% [7.78; 14.44], respectively. Thus, the fraction of liver fat was 2.5 times higher in male patients with T deficiency than in eugonadal patients (p < 0.001). These results demonstrated a significantly greater severity of nonalcoholic liver disease in this category of patients.

Fig. 2. Concentrations of liver enzymes in male patients with type 2 diabetes mellitus, depending on the presence or absence of hypogonadism.

Fig. 3. Contents of the hepatic fat fraction in male patients wittype 2 diabetes mellitus2, depending on the presence or absence of hypogonadism.

The difference in the groups in terms of anthropometric indicators was subjected to covariant analysis to assess the effects of intergroup factors and confirm the significance of the effect of T deficiency on liver fat proton density, in addition to the severity of obesity. In this regard, the presence of hypogonadism was considered a factor, BMI and WC were considered covariants, and the quantitative value of the liver fat fraction was a dependent variable. Covariant analysis revealed that the effect of hypogonadism on hepatic fat fraction was statistically significant with correction for BMI (p = 0.019) and WC (p = 0.002).

ADVERSE EVENTS

In the course of the study, no adverse events were registered.

DISCUSSION

Numerous studies [1, 3, 5] have shown that hypogonadism in male patients is accompanied with an increase in body weight and a predominance of the visceral type of obesity. This phenomenon was also demonstrated in our work, which revealed that a group of patients with hypogonadism had higher BMI, WC, and HC than that of eugonadal patients. Moreover, significantly higher levels of SBP and DBP were detected in group 2, indicating higher cardiovascular risks in male patients with DM2 and hypogonadism. This finding also complemented the overall presentation of the metabolic syndrome in this category of patients. Despite the absence of the differences in fasting glycemia and HbA_1c levels in the groups, significantly higher concentrations of IRI and HOMA insulin resistance index were found in male patients with T deficiency. Therefore, hyperinsulinemia could be a result of insulin resistance associated with T deficiency in male patients [17].

Insulin resistance leading to hyperinsulinemia inhibits lipolysis and simultaneously activates lipogenesis in the liver; as a result, fat accumulates in hepatocytes under the influence of adipohormones [11, 18]. More free fatty acids accumulated in liver cells; the more pronounced becomes the insulin resistance of the hepatocytes themselves, thereby stimulating uncontrolled gluconeogenesis and glucose production by the liver. Thus, this explains the relationship of conditions such as DM2, NAFLD, and T deficiency.

The effect of androgens on lipid metabolism remains one of the most controversial issues. We revealed a significantly higher hypertriglyceridemia in patients with DM2 combined with hypogonadism, and this finding agreed with the results of meta-analysis [19]. To date, the effects of androgens on lipid metabolism are still being studied. T increases the activity of hepatic lipase by hydrolysing TG and phospholipids and increases the activity of the SR-B1 scavenger receptor gene, which promotes the selective uptake of cholesterol by hepatocytes and Sertoli cells [20]. Therefore, T deficiency leads not only to dyslipidemia but also to uncontrolled lipid accumulation in liver cells, contributing to the development of NAFLD.

Not so long ago, the attitude changed toward hormones produced by adipose tissues, namely, resistin and leptin, which have been transferred from the category of regulators of fat and carbohydrate metabolism to the category of predictors of cardiovascular diseases. In our study, the concentrations of resistin and leptin in patients with T deficiency were significantly higher than those in men with normotestosteronemia.

After being secreted by adipocytes, resistin not only inhibits insulin-mediated glucose uptake by target tissues but also promotes the formation of foam cells, activates vascular endothelium, and participates in the stimulation of inflammation in various organs, including in the liver, because this substance is also an insulin antagonist; therefore it is considered to be a link between obesity, DM2, hepatosteatosis, and atherosclerosis [21]. Leptin, which has been extensively studied in terms of its metabolic effects, also contributes to cholesterol deposition in macrophages and stimulates atherogenesis and hepatosteatosis. Thus, the high secretory activity of adipose tissues revealed in our study in patients in group 2 disproportionate to minimise intergroup differences in anthropometric parameters, demonstrates the apparent significance of T deficiency as one of the pathogenetic factors in the formation of the metabolic syndrome in male patients with DM2.

The functional state and structure of the liver are important. Thus, the presence of hypogonadism in male patients with DM2 was accompanied with an increase in the concentration of ALT and GGTP and an increase in the fraction of hepatic fat as shown by MRI.

Covariant analysis confirmed the established relationships and showed the statistical significance of the effect of hypogonadism on the liver fat fraction with correction for BMI and WC. Thus, the development of NAFLD was reliably associated with T deficiency in male patients with DM2 and hypogonadism.

Study limitation

The sample size of the study was not calculated before the start of the study, so the extrapolation of the results to the general population was limited. The study was conducted on a group of patients diagnosed with DM2, so the judgement on the development of NAFLD in male patients with hypogonadism and without impaired carbohydrate metabolism was also limited.

CONCLUSION

In contemporary literature, conditions associated with T deficiency have been widely explored, but data on the combination of NAFLD with hypogonadism and DM in male patients are extremely limited. Our results revealed an increase in insulin resistance, compensatory hyperinsulinemia, and hypertriglyceridemia in male patients with hypogonadism compared with those in eugonadal patients. The biochemical signs of impaired liver function, such as an increase in the concentration of liver enzymes ALT, GGTP, and de Ritis ratio, of male patients with DM2 and hypogonadism increased. The hepatic fraction of liver fat also increased, as confirmed by instrumental studies. These results indicated NAFLD progression accompanied with an increase in the production of the adipohormones resistin and leptin in adipose tissues. These factors showed a possible pathogenetic link to the development of conditions such as carbohydrate and lipid metabolic disorders, NAFLD, and T deficiency.

ADDITIONAL INFORMATION

Source of financing. This work was supported by State Assignment No. 14, “Liver Function in Men with DM2.”

Conflict of interest. The authors declare no financial interest and other potential conflicts of interest related to the publication of this article.

Contribution of authors. I. A. Khripun and S. V. Vorobyov created the concept and design of the study, collected the materials, and wrote the text; Ya. S. Allahverdieva collected the materials, analysed the results, and wrote the text. All the authors significantly contributed to this research, prepared the article, and read and approved the final version of the article before publication.