Introduction: The study aimed to determine if hepatic steatosis assessed by fatty liver index (FLI) was an independent risk factor for male low testosterone level and whether the FLI was the strongest risk factor for low testosterone level in two different age groups. Methods: Two cross-sectional studies were performed. A total of 3,443 male participants (aged 46–75) were recruited into study A (part of lONgitudinal study (REACTION)). Then a total of 267 male participants (aged 25–45) were recruited into study B. Serum total testosterone (TT) and sex hormone-binding globulin (SHBG) levels, indicators for assessing hepatic steatosis were measured. The Pearson correlation and regression analysis were performed to investigate the risk factors for low testosterone level. Results: The FLI had the strongest negative correlation with serum testosterone in the study A (r = −0.436) and B (r = −0.542). Compared with patients with a FLI lower than 30, the risk for low testosterone level increased by 3.48-fold in subjects with a FLI higher than 60 adjusted for potential risk factors in study A. In study B, the odds ratio of low testosterone level in patients with potential hepatic steatosis was 4.26 (1.57–11.60) after adjusted for age and homeostasis model assessment of insulin resistance (HOMA-IR) and 0.59 (0.14–2.60) after adjusted for age, HOMA-IR, waist circumference, body mass index, and SHBG. Conclusions: FLI was the strongest risk factor for male low testosterone level independent of insulin resistance in male populations of different ages; however, the association can be modulated by SHBG levels in the young. Significance Statement: In the study, FLI was the strongest negative risk factor for low testosterone level in the Chinese adult male population. The results suggested that hepatic steatosis assessed by the FLI was the main risk factor for male low testosterone level, independent of age, insulin resistance, smoking, and drinking status; however, the association of FLI and TT levels can be modulated by SHBG levels. Taken together these findings indicate that clinical physicians should pay more attention to the FLI index and hepatic steatosis, so that they can take advantage of them for assessing the risk of developing of low testosterone level in the male population.

Testosterone, as a vital steroid hormone in humans, plays an important role in the physiological processes of growth and development, sexual desire maintenance, reproductive function, material metabolism, and senescence. Low testosterone level has become prevalent in the recent years, especially within the middle-aged and elderly male population. A study estimated that there were over 2.4 million 40- to 69-year-old males with testosterone deficiency in the USA [1]. Moreover, a study in China demonstrated that about a quarter of the population had testosterone deficiency [2]. Apart from sexual dysfunction, patients with low testosterone levels have also been reported to have an increased risk of developing cardiometabolic disorders, Alzheimer’s disease, and other systemic diseases, which can seriously affect their quality of life [3, 4]. Therefore, early prevention and intervention of low testosterone level is of paramount importance.

Hepatic steatosis refers to the ectopic deposition of fat in the liver without the excessive consumption of alcohol and is considered as the early stage of non-alcoholic fatty liver disease (NAFLD) [5]. It is well known that hepatic steatosis is very closely linked to diabetes mellitus, hypertension, cardiovascular disease, and chronic renal disease [6‒8]. Numerous studies have demonstrated that low serum testosterone levels are associated with obesity, visceral obesity, and metabolic syndrome [9, 10]. There are three forms of testosterone that circulate in the body including those tightly bound to sex hormone-binding globulins (SHBG) (60%), those loosely bound to albumin (38%) and to a much lesser degree the free form (2%) [11]. It has been reported that up to 90% of hepatic cirrhotic men had low levels of serum testosterone, which was inversely correlated with the severity of liver disease [12]. Therefore, it is reasonable to speculate that serum testosterone levels are likely to be directly influenced or modulated by liver metabolism and functional states.

Recently, some cross-sectional studies have shown that serum testosterone levels were associated with fatty liver [13‒16], and this association remained unaltered even after controlling for visceral fat and insulin resistance [15]. However, other cross-sectional studies in patients with type 2 diabetes mellitus have reported that total testosterone (TT) levels were not related to NAFLD, but SHBG levels were independently associated with NAFLD [17‒19]. Therefore, whether hepatic steatosis is associated with male low testosterone level independent of obesity, other metabolic factors, and sex hormones still remain controversial presently. However, recent studies have paid a lot of attention to the effect of testosterone deficiency on hepatic steatosis; the influence of excessive liver fat accumulation on serum testosterone levels has not obtained enough attention.

The fatty liver index (FLI), was first proposed as a clinical indicator of hepatic fat accumulation based on body mass index (BMI), waist circumference (WC), gamma-glutamyl transferase (GGT), and triglycerides (TG) in 2006 [20]. Recent studies have revealed that the FLI was associated with hypertension, metabolic syndrome, diabetes mellitus, cardiovascular disease, chronic kidney disease, and colorectal adenoma [21‒26]. Therefore, the FLI is a simple and precise index for measuring hepatic steatosis and is simultaneously an indicator for metabolic diseases. As a result, clinicians should pay more attention to the FLI as clinical indicator because of its simplicity, precision, safety, and availability. However, it must be borne in mind that to the best of our knowledge, no research studies have as yet demonstrated the relationship of the FLI and low testosterone level in the general male population. Moreover, it is known that aging is one of the risk factors associated with hepatic steatosis, and some studies report that aging also correlates with serum TT levels. Whether the association of TT levels and the FLI modulated by aging and whether the association remains unchanged in different age groups still remains to be fully elucidated.

For this study, we aimed to examine the relationship between serum sex hormones levels and the FLI in two different age groups, while investigating whether hepatic steatosis assessed by the FLI was the main risk factor for low testosterone level, independent of age, insulin resistance, visceral obesity, or other sex hormones.

Subjects

This paper reports the findings of two cross-sectional studies. Both studies gathered information on the TT and metabolic characteristics of male participants. Study A with a larger sample size was performed to find the association between FLI and serum TT levels. Study B which was able to provide more information about sex hormones, and insulin resistance was performed to further evaluate whether FLI is a risk factor of low testosterone level independent of SHBG and insulin resistance in a lower age range group.

Study A obtained data from a population-based cross-sectional study in Ningyang County (Taian, Shandong Province, China) from June to November 2011. We recruited nearly 11,000 persons participated in the cross-sectional study and all participants were local-registered residents aged 40 years and older who have lived there for at least 5 years. The patients aged from 46 to 75 years old were recruited to the study A. Study B obtained data from results of physical examinations of subjects from one community (Feicuijun in Ji’nan, China) performed in Shandong Provincial Hospital in China from August 2015 to November 2015. The individuals aged from 25 to 45 years old were initially recruited. A total of 1,531 people participated in the investigation. All participants in above studies were asked to complete a self-reported questionnaire and provided an overnight fasting blood sample. In our studies, the exclusion criteria consisted of female, no information on vital statistics (such as age, height, weight, or WC) or missing data on serum TT, blood pressure, liver enzymes, fasting serum glucose, or lipid levels, taking medications that might affect TT level (such as androgens, steroid hormones), and severe hepatic or renal disorders, or tumors that might affect TT level (such as brain cancer or prostate cancer), and as well individuals with excessive alcohol consumption (≥140 g/week for men). At last, a total of 3,443 participants were recruited in this study A and 267 men were finally included in study B (Fig. 1). The comparison of baseline characteristics of included patients and excluded patients are shown in online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000533962).

Fig. 1.

Flowchart of the analysis.

Fig. 1.

Flowchart of the analysis.

Close modal

Data and Specimen Collection

Height and WC were measured adjusting by 0.1 cm. Weight was measured adjusting by 0.1 kg. BMI was obtained by dividing weight in kilograms by the square of height in meters. In study A, the blood pressure was measured in sitting position after a 5 min rest for three times and then took an average. Past medical history, smoking status, and alcohol consumption were obtained by a questionnaire. The diseases (hypertension, diabetes mellitus, and coronary heart disease) were based on a previous diagnosis by a physician. Smoking status and alcohol consumption were defined as never, ever, and current.

Venous blood samples were collected from all patients after at least 10 h overnight fast. And samples were separated and preserved in −80°C. The concentrations of total cholesterol (TC), TG, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), fasting plasma glucose (FPG), 2 h plasma glucose, and glycosylated hemoglobin (HbA1c) were measured directly with an ARCHITECT ci16200 Integrated System (Abbott). Serum TT, luteinizing hormone (LH), and follicle-stimulating hormone (FSH) concentrations were measured by electrochemiluminescent procedures (Cobas E601; Roche) at the clinical laboratory of Shandong Provincial Hospital in study A. The concentrations of insulin (INS), TT, estradiol (E2), sex hormone-binding globulin (SHBG), LH, and FSH were measured by an electrochemiluminescence immunoassay (Cobas8000; Roche, Basel, Switzerland) in study B. The homeostasis model assessment of insulin resistance (HOMA-IR) was calculated with the following formula: fasting glucose (mmol/L) × FINS (μIU/mL)/22.5. The free testosterone index (FTI) was calculated with the following equation: testosterone (ng/mL) × 347/SHBG (nmol/L).

Study Outcome Definition

Male patients with TT level ≤346 ng/dL are suggestive of low testosterone level [27]. However, given that testosterone levels decline with age, we defined low testosterone level in young patients based on age-dependent cutoffs (age 25–29, 413 ng/dL, age 30–34, 359 ng/dL, age 35–39, 352 ng/dL, age 40–44, 350 ng/dL) [28]. The calculation formula of fat accumulation indicators were listed as follows: FLI = [e0.953 × loge (TG)(mg/dL)+ 0.139 ×BMI(kg/m2) + 0.718 × loge (GGT) + 0.053 ×WC− 15.745/(1 + e0.953 × loge (TG) (mmol/L)+ 0.139 ×BMI(kg/m2) + 0.718 × loge (GGT)+ 0.053 × WC− 15.745] × 100 [20]; HSI = 8 × (ALT/AST ratio) + BMI (+2, if female; +2, if diabetes mellitus) [29]; LAP = (WC[cm]-65) × TG(mmol/L) [10].

Statistical Analysis

All data were recorded in a computer database and analyzed using the SPSS 22.0 program. Continuous variables were presented as mean ± standard error of the mean, and the categorical variables were presented as percentage. The differences between groups were assessed by single-factor analysis of variance (ANOVA) for continuous variables and χ2 test for categorical variables. Correlations between the analyzed continuous variables were calculated using a Pearson correlation coefficient and correlation matrix was drawn via “Corrplot” packages in R (4.1.1). All p values were 2-sided, and less than 0.05 were considered statistically significant.

Characteristics of the Studied Population

In study A, we recruited a total of 3,443 males ranging in ages from 46 to 75 years old in the final analysis. Four hundred and fifty-two participants were diagnosed with low testosterone level in the total sample. When comparing these participants to those without low testosterone level, the men with low testosterone level had comparable AST and alcohol consumption status, but they had significantly higher TC, TG, LDL-C, ALT, GGT, INS, FPG, 2h plasma glucose, HbA1C, HOMA-IR, SBP, DBP, LH, and FSH levels, as well as having a higher prevalence of diabetes mellitus. These men were also significantly younger and had lower levels of HDL-C, as well as having a lower proportion of current smokers. In relation to the population with low testosterone level, all fat accumulation indicators including BMI, WC, HSI, LAP, and the FLI were significantly higher than participants without low testosterone level (Table 1).

Table 1.

Baseline characteristics of studied population A

CharacteristicsTT ≤346 ng/dL (n = 452)TT >346 ng/dL (n = 2,991)p value
Age, years 56.74±0.34 58.40±0.13 <0.001 
BMI, kg/m2 26.91±0.18 24.58±0.06 <0.001 
WC, cm 94.12±0.48 88.35±0.18 <0.001 
HSI 34.84±0.24 31.43±0.09 <0.001 
LAP 65.31±2.85 35.15±0.69 <0.001 
FLI 56.12±1.25 34.80±0.46 <0.001 
TC, mmol/L 5.25±0.05 5.00±0.02 <0.001 
TG, mmol/L 2.16±0.08 1.39±0.02 <0.001 
LDL-C, mmol/L 3.08±0.04 2.98±0.02 0.023 
HDL-C, mmol/L 1.32±0.02 1.42±0.01 <0.001 
ALT, U/L 22.18±0.51 19.34±0.19 <0.001 
AST, U/L 23.66±0.47 23.17±0.16 0.329 
GGT, U/L 45.81±2.42 35.23±0.73 <0.001 
INS, μIU/mL 10.00±0.60 6.69±0.13 <0.001 
FPG, mmol/L 7.25±0.12 6.38±0.03 <0.001 
2hPG, mmol/L 11.77±0.33 9.39±0.10 <0.001 
HbA1C, % 6.65±0.07 6.08±0.02 <0.001 
HOMA-IR 3.29±0.21 1.99±0.05 <0.001 
TT, ng/dL 244.46±4.84 605.57±3.30 <0.001 
LH, mIU/mL 8.89±0.44 7.49±0.08 0.002 
FSH, mIU/mL 16.58±0.98 10.69±0.15 <0.001 
SBP, mm Hg 144.77±0.95 142.38±0.38 0.022 
DBP, mm Hg 85.86±0.54 83.68±0.21 <0.001 
DM, % 15.27 9.43 <0.001 
Hypertension, % 21.62 20.22 0.534 
CHD, % 6.93 8.07 0.453 
Smoking status, %   0.001 
 Never 55.27 45.70  
 Ever 10.26 8.61  
 Current 34.47 45.70  
Alcohol consumption, % 0.865 
 Never 38.86 37.04  
 Ever 16.00 15.70  
 Current 45.14 47.26  
CharacteristicsTT ≤346 ng/dL (n = 452)TT >346 ng/dL (n = 2,991)p value
Age, years 56.74±0.34 58.40±0.13 <0.001 
BMI, kg/m2 26.91±0.18 24.58±0.06 <0.001 
WC, cm 94.12±0.48 88.35±0.18 <0.001 
HSI 34.84±0.24 31.43±0.09 <0.001 
LAP 65.31±2.85 35.15±0.69 <0.001 
FLI 56.12±1.25 34.80±0.46 <0.001 
TC, mmol/L 5.25±0.05 5.00±0.02 <0.001 
TG, mmol/L 2.16±0.08 1.39±0.02 <0.001 
LDL-C, mmol/L 3.08±0.04 2.98±0.02 0.023 
HDL-C, mmol/L 1.32±0.02 1.42±0.01 <0.001 
ALT, U/L 22.18±0.51 19.34±0.19 <0.001 
AST, U/L 23.66±0.47 23.17±0.16 0.329 
GGT, U/L 45.81±2.42 35.23±0.73 <0.001 
INS, μIU/mL 10.00±0.60 6.69±0.13 <0.001 
FPG, mmol/L 7.25±0.12 6.38±0.03 <0.001 
2hPG, mmol/L 11.77±0.33 9.39±0.10 <0.001 
HbA1C, % 6.65±0.07 6.08±0.02 <0.001 
HOMA-IR 3.29±0.21 1.99±0.05 <0.001 
TT, ng/dL 244.46±4.84 605.57±3.30 <0.001 
LH, mIU/mL 8.89±0.44 7.49±0.08 0.002 
FSH, mIU/mL 16.58±0.98 10.69±0.15 <0.001 
SBP, mm Hg 144.77±0.95 142.38±0.38 0.022 
DBP, mm Hg 85.86±0.54 83.68±0.21 <0.001 
DM, % 15.27 9.43 <0.001 
Hypertension, % 21.62 20.22 0.534 
CHD, % 6.93 8.07 0.453 
Smoking status, %   0.001 
 Never 55.27 45.70  
 Ever 10.26 8.61  
 Current 34.47 45.70  
Alcohol consumption, % 0.865 
 Never 38.86 37.04  
 Ever 16.00 15.70  
 Current 45.14 47.26  

Continuous variables are presented as mean ± standard error of the mean and categorical variables are presented as percentage.

TT, total testosterone; BMI, body mass index; WC, waist circumference; HSI, hepatic steatosis index; LAP, lipid accumulation product; FLI, fatty liver index; TC, total cholesterol; TG, triglyceride; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; INS, insulin; FPG, fasting plasma glucose; 2hPG, 2 h plasma glucose; HbA1c, glycosylated hemoglobin; HOMA-IR, homeostasis model assessment of insulin resistance; LH, luteinizing hormone; FSH, follicle-stimulating hormone; SBP, systolic blood pressure; DBP, diastolic blood pressure; DM, diabetes mellitus; CHD, coronary heart disease.

In study B, a total of 267 males ranging in ages from 25 to 45 years old were recruited in the final analysis. Of these, 34 participants were diagnosed with low testosterone level measured by TT levels. When comparing these to participants without low testosterone level, the men with low testosterone level had comparable age, ALT, AST, TC, LDL-C, HDL-C, URIC, FPG, HbA1C, FSH, LH, and FTI but had a significantly higher average BMI, WC, FLI, HSI, LAP, GGT, ALP, TG, INS, HOMA-IR, as well as significantly lower E2, TT, and SHBG levels (Table 2).

Table 2.

Baseline characteristics of studied population B

CharacteristicsLow TT level (n = 53)Normal TT (n = 214)p value
Age, years 34.30±0.92 33.19±0.43 0.253 
BMI, kg/m2 27.74±0.50 25.29±0.23 <0.001 
WC, cm 97.13±1.27 89.46±0.67 <0.001 
HSI 37.89±0.68 34.68±0.36 <0.001 
LAP 82.96±8.81 43.89±2.82 <0.001 
FLI 64.94±3.32 42.17±1.87 <0.001 
TC, mmol/L 5.16±0.10 4.99±0.06 0.167 
TG, mmol/L 2.56±0.25 1.70±0.09 0.002 
LDL-C, mmol/L 3.22±0.08 3.13±0.04 0.337 
HDL-C, mmol/L 1.11±0.03 1.17±0.02 0.073 
ALT, U/L 36.64±3.48 32.20±1.48 0.201 
AST, U/L 28.68±2.68 26.16±0.74 0.208 
GGT, U/L 45.40±5.08 36.34±1.78 0.040 
ALP, U/L 99.15±3.18 92.12±1.40 0.031 
URIC, μmol/L 386.87±11.30 385.32±5.24 0.896 
INS, μIU/mL 15.67±0.96 10.97±0.38 <0.001 
FPG, mmol/L 5.45±0.09 5.48±0.09 0.909 
HbA1C, % 5.37±0.06 5.43±0.09 0.756 
HOMA-IR 3.89±0.30 2.69±0.10 <0.001 
FSH, mIU/mL 5.19±0.36 4.84±0.15 0.325 
LH, mIU/mL 4.54±0.29 5.14±0.13 0.050 
E2, pg/mL 28.61±1.13 31.75±0.69 0.037 
TT, ng/dL 323.75±6.83 570.36±11.72 <0.001 
SHBG, nmol/L 22.47±0.85 37.26±1.13 <0.001 
FTI, % 53.02±1.97 57.13±1.14 0.101 
CharacteristicsLow TT level (n = 53)Normal TT (n = 214)p value
Age, years 34.30±0.92 33.19±0.43 0.253 
BMI, kg/m2 27.74±0.50 25.29±0.23 <0.001 
WC, cm 97.13±1.27 89.46±0.67 <0.001 
HSI 37.89±0.68 34.68±0.36 <0.001 
LAP 82.96±8.81 43.89±2.82 <0.001 
FLI 64.94±3.32 42.17±1.87 <0.001 
TC, mmol/L 5.16±0.10 4.99±0.06 0.167 
TG, mmol/L 2.56±0.25 1.70±0.09 0.002 
LDL-C, mmol/L 3.22±0.08 3.13±0.04 0.337 
HDL-C, mmol/L 1.11±0.03 1.17±0.02 0.073 
ALT, U/L 36.64±3.48 32.20±1.48 0.201 
AST, U/L 28.68±2.68 26.16±0.74 0.208 
GGT, U/L 45.40±5.08 36.34±1.78 0.040 
ALP, U/L 99.15±3.18 92.12±1.40 0.031 
URIC, μmol/L 386.87±11.30 385.32±5.24 0.896 
INS, μIU/mL 15.67±0.96 10.97±0.38 <0.001 
FPG, mmol/L 5.45±0.09 5.48±0.09 0.909 
HbA1C, % 5.37±0.06 5.43±0.09 0.756 
HOMA-IR 3.89±0.30 2.69±0.10 <0.001 
FSH, mIU/mL 5.19±0.36 4.84±0.15 0.325 
LH, mIU/mL 4.54±0.29 5.14±0.13 0.050 
E2, pg/mL 28.61±1.13 31.75±0.69 0.037 
TT, ng/dL 323.75±6.83 570.36±11.72 <0.001 
SHBG, nmol/L 22.47±0.85 37.26±1.13 <0.001 
FTI, % 53.02±1.97 57.13±1.14 0.101 

Continuous variables are presented as mean ± standard error of the mean.

TT, total testosterone; BMI, body mass index; WC, waist circumference; HSI, hepatic steatosis index; LAP, lipid accumulation product; FLI, fatty liver index; TC, total cholesterol; TG, triglyceride; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; INS, insulin; FPG, fasting plasma glucose; HbA1c, glycosylated hemoglobin; HOMA-IR, homeostasis model assessment of insulin resistance; FSH, follicle-stimulating hormone; LH, luteinizing hormone; E2, estradiol; SHBG, sex hormone-binding globulin; FTI, free testosterone index.

The Association between Potential Risk Factors and Sex Hormones Levels

To investigate the association between potential risk factors and sex hormones levels, we performed a Pearson’s correlation analysis and drawn correlation matrix analysis in the studied populations. This revealed a significantly negative relationship between serum testosterone levels and FLI, BMI, WC, HSI, LAP, HDL-C, INS, and HOMA-IR (p < 0.001, respectively), as well as a significant positive relationship between serum testosterone levels and age, HDL-C (p < 0.001) in study A (Fig. 2a). FLI had the strongest negative correlation with serum TT levels (r = −0.436) compared to all the other risk factors.

Fig. 2.

Correlation matrix of potential risk factors and total testosterone levels. TT, total testosterone; BMI, body mass index; WC, waist circumference; HSI, hepatic steatosis index; LAP, lipid accumulation product; FLI, fatty liver index; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol; INS, insulin; HOMA-IR, homeostasis model assessment of insulin resistance. *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 2.

Correlation matrix of potential risk factors and total testosterone levels. TT, total testosterone; BMI, body mass index; WC, waist circumference; HSI, hepatic steatosis index; LAP, lipid accumulation product; FLI, fatty liver index; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol; INS, insulin; HOMA-IR, homeostasis model assessment of insulin resistance. *p < 0.05, **p < 0.01, ***p < 0.001.

Close modal

In study B, it was shown that TT levels were negatively correlated with age, FLI, BMI, WC, HSI, LAP, INS, HOMA-IR but positively correlated with HDL-C (Fig. 2b). Subsequently, we divided the FLI into three groups based on a previous study [20] and compared the sex hormone levels in three groups according to the FLI level, the participants with the FLI level higher than 60 had the lowest levels of TT (419.36 ± 110.75) and SHBG (26.79 ± 9.49). Moreover, there was no significant difference in FTI and E2 levels among the three groups (Fig. 3).

Fig. 3.

Sex hormone levels in different groups according to fatty liver index. Data were presented as mean ± standard deviation. #p < 0.05 versus the group of fatty liver index lower than 30, *p < 0.05 versus the group of fatty liver index between 30 and 60. TT, total testosterone; SHBG, sex hormone-binding globulin; FTI, free testosterone index; E2, estradiol.

Fig. 3.

Sex hormone levels in different groups according to fatty liver index. Data were presented as mean ± standard deviation. #p < 0.05 versus the group of fatty liver index lower than 30, *p < 0.05 versus the group of fatty liver index between 30 and 60. TT, total testosterone; SHBG, sex hormone-binding globulin; FTI, free testosterone index; E2, estradiol.

Close modal

The Independence of Hepatic Steatosis for Low Testosterone Level

To study the association between hepatic steatosis and low testosterone level, we used the FLI, a surrogate marker of hepatic steatosis, to distinguish potential patients with hepatic steatosis and to further evaluate the association by logistic regression. It is known that FLI levels of <30 rule out hepatic steatosis whereas a FLI level of ≥60 rules in hepatic steatosis [20]. In this manner, we divided subjects into three groups according to the FLI criterion. Table 3 summarizes that in each of the models, the odds ratio (OR) for low testosterone level increased when the FLI level increased based on study A. The risk of low testosterone level in male patients with potential hepatic steatosis increased 6.37-fold when compared to men in the lowest group of the FLI. After the data were adjusted for age, HOMA-IR, WC, BMI, current smoking, and alcohol consumption levels, the OR of low testosterone level in the patients with the FLI level higher than 60 decreased, but it was still significant (3.48, 95% CI: 1.87–6.50). Therefore, hepatic steatosis was shown to be a main risk factor for low testosterone level, independent of age, HOMA-IR, WC, BMI, smoking, and drinking status.

Table 3.

Odds ratio of fatty liver index for low testosterone level in study A

Model 1Model 2Model 3Model 4
FLI <30 Ref. Ref. Ref. Ref. 
30≤ FLI <60 2.14 (1.63–2.82) 2.12 (1.61–2.80) 2.04 (1.55–2.70) 1.62 (0.97–2.71) 
FLI ≥60 6.37 (4.94–8.21) 6.11 (4.72–7.91) 5.61 (4.32–7.29) 3.48 (1.87–6.50) 
Model 1Model 2Model 3Model 4
FLI <30 Ref. Ref. Ref. Ref. 
30≤ FLI <60 2.14 (1.63–2.82) 2.12 (1.61–2.80) 2.04 (1.55–2.70) 1.62 (0.97–2.71) 
FLI ≥60 6.37 (4.94–8.21) 6.11 (4.72–7.91) 5.61 (4.32–7.29) 3.48 (1.87–6.50) 

Model 1: unadjusted;

Model 2: adjusted for age;

Model 3: adjusted for age and HOMA-IR;

Model 4: adjusted for age, HOMA-IR, WC, BMI, current smoking, and alcohol consumption.

Ref., reference; FLI, fatty liver index.

Table 4 shows that the risk of low testosterone level in patients with potential hepatic steatosis increased 6.56-fold compared to men in the lowest group of the FLI. After adjusted for age and HOMA-IR, the OR of low testosterone level decreased but was still significant (4.26, 95% CI: 1.57–11.60). However, it was not significant in the highest group of the FLI after it was adjusted for SHBG (0.59, 95% CI: 0.14–2.60) (Table 4). This showed that the FLI was the main risk factor for low testosterone level and was independent of insulin resistance but not independent of SHBG levels.

Table 4.

Odds ratio of fatty liver index for low testosterone level in study B

Model 1Model 2Model 3Model 4
FLI <30 Ref. Ref. Ref. Ref. 
30≤ FLI <60 2.71 (1.02–7.18) 2.78 (1.04–7.49) 2.19 (0.79–6.05) 0.58 (0.16–2.06) 
FLI ≥60 6.56 (2.73–15.75) 6.82 (2.74–16.98) 4.26 (1.57–11.60) 0.59 (0.14–2.60) 
Model 1Model 2Model 3Model 4
FLI <30 Ref. Ref. Ref. Ref. 
30≤ FLI <60 2.71 (1.02–7.18) 2.78 (1.04–7.49) 2.19 (0.79–6.05) 0.58 (0.16–2.06) 
FLI ≥60 6.56 (2.73–15.75) 6.82 (2.74–16.98) 4.26 (1.57–11.60) 0.59 (0.14–2.60) 

Model 1: unadjusted;

Model 2: adjusted for age;

Model 3: adjusted for age and HOMA-IR;

Model 4: adjusted for age, HOMA-IR, WC, BMI, and SHBG.

Ref., reference; FLI, fatty liver index.

More people should be aware of low testosterone level because of its high prevalence rates and long-term detrimental effects on the health of the male population. The FLI, a clinical indicator of hepatic fat accumulation and in our study, we have provided evidence to the existence of a strong negative correlation between the FLI and serum TT levels in two large-scale populations of Chinese men of differing age ranges. There was also a negative correlation between the FLI and male low testosterone level which was independent of age, insulin resistance, WC, BMI, smoking, and drinking status, but the association was modulated by SHBG levels. In this way, our study has suggested that male subjects with excessive hepatic fat accumulation are more inclined to develop low testosterone level.

We observed a negative correlation between hepatic steatosis and serum testosterone levels, which are consistent with a few previous studies that have been conducted [14‒16]. A cross-sectional study conducted on 154 men with type 2 diabetes mellitus demonstrated that their serum TT levels were not associated with the presence of fatty liver disease [17]. However, discrepancies between these studies could possibly be explained by the varying differences in the size of the study populations and the selection of subjects chosen for the study. Nevertheless, these results still indicate that hepatic steatosis was negatively associated with serum testosterone levels within the male populations. It has been reported in several studies that SHBG levels were negatively associated with NAFLD within a general male population or with people who had type 2 diabetes mellitus [16‒18], which is consistent with our results in study B. Therefore, taking all the evidence into account, it is reasonable to speculate that hepatic production of SHBG and their subsequent serum levels can be influenced by hepatic steatosis. It has been indicated previously that older age was an independent risk factor for testosterone deficiency [2]. This finding was inconsistent with our results. In spite of this, a recent study has demonstrated that the prevalence of testosterone deficiency was not associated with age [30]. Therefore, it is still unclear whether older age contributes to low testosterone level, obesity, visceral obesity, chronic illness, or any other factors that become more prevalent as men age contributed greatly to low testosterone level. Therefore, it can be concluded that the association between age and male testosterone deficiency still needs to be further investigated.

We have shown that hepatic steatosis diagnosed based on the FLI was the main risk factor for male low testosterone level and this association was independent of age, insulin resistance, smoking, and drinking status. With the accumulation of both clinical and animal studies, it has increasingly been demonstrated that testosterone deficiency increased the risk of hepatic steatosis. In fact, the association between testosterone deficiency and hepatic steatosis may be more complex and multidirectional. Hepatic cells that accumulated excess fat secreted inflammatory cytokines in the liver, such as tumor necrosis factor and interleukin-6. These inflammatory adipocytokines have an inhibiting action which suppresses the secretion of hypothalamic GnRH and contributes to the decreased secretion of LH, which eventually leads to the reduction in the synthesis of testosterone in testis [31]. One study has demonstrated that testosterone deficiency was inversely associated with insulin resistance [32], which is consistent with the results found in study A and B of this paper. It should be noted that SHBG is a circulating glycoprotein secreted by the liver, which may be modulated by alterations in the structure and function of the liver. The results from study B indicate that SHBG levels may account for the association between hepatic steatosis and TT to a greater extent. Previous systematic review and meta-analysis showed similar results that men with higher TT levels had lower odds of NAFLD, meanwhile higher SHBG levels were associated with reduced odds of NAFLD [33].

The FLI, calculated by BMI, WC, TG, and GGT, was used as a clinical indicator for hepatic fat accumulation, which has been verified to align with SteatoTest and abdominal ultrasonography [20]. It was reasonable to demonstrate that the FLI was negatively associated with serum testosterone levels. In this study, we used the diagnostic standard value for hepatic steatosis based on the FLI to further evaluate the relationship between hepatic steatosis and TT levels. Hepatic steatosis was diagnosed using the methods of each study which included liver function, imaging, or liver biopsy. Employing the FLI may be more effective than employing hepatic imaging since testosterone deficiency is often associated with hepatic inflammation. In study A and B, it was noted that the FLI had the strongest negative correlation with serum TT levels when compared to other risk factors. This also indicated that elevated FLI levels were an important risk factor for male low testosterone level within two different age groups. However, the study B participants had lower ages with lower mean WC and higher mean BMI compared to the study A. The metabolic characteristics and visceral adiposity distribution caused by aging contributed to the difference of characteristics between the two studies. However, in two different age groups of studies A and B, elevated FLI levels were seen as an important risk factor for low testosterone level as well as having the strongest correlation with serum TT levels.

This is the first study with a relatively large sample size to investigate the association between the FLI and low testosterone level in two different age group population. The results have suggested that hepatic steatosis assessed by the FLI was the main risk factor for male low testosterone level, independent of age, insulin resistance, WC, BMI, smoking, and drinking status. However, the association between FLI and TT levels could be modulated by SHBG levels. Despite its strengths, our study still had some limitations. This is, due to the cross-sectional design of the study, we were unable to determine a direct causal relationship between hepatic steatosis and low testosterone level to avoid the effect of selection bias. A liver biopsy is regarded as a gold standard method for detecting hepatic steatosis, and the gold standard for steroid measurement is mass spectrometry; however, this method was not practicable to use in this population-based study, therefore, a further study is required based on these gold standards to verify the relationship between hepatic steatosis and testosterone levels.

The study A was as part of the Risk Evaluation of cAncers in Chinese diabetic Individuals: A lONgitudinal (REACTION) trial at Shanghai Jiao Tong University School of Medicine (clinical trial number: NCT01506869). The study A was approved by the Committee on Human Research at Ruijin Hospital, Shanghai Jiao Tong University School of Medicine (Clinical Trial Ethics Committee Approval Form [2011]14). All study participants provided written informed consent. The study B was approved by the Ethics Committee of Shandong Provincial Hospital affiliated to Shandong University. The procedures and purposes of this study were explained to the subjects, and all subjects provided written informed consent. The study was conducted in accordance with the Declaration of Helsinki.

The authors declare that there is no conflict of interest.

This work was supported in part by grants from the National Natural Science Foundation of China (82270839, 81770860) and Key Research and Development Plan of Shandong Province (2017CXGC1214).

Author contributions included the following: L.L. and M.L. analyzed the data and wrote the manuscript. P.C., Q.S., and L.F. contributed to clinical data collection. Y.L. and J.H. conducted research. C.Y. and Q.G. designed and performed the study and are responsible for the data. All authors read and approved the final manuscript.

Additional Information

Luna Liu and Man Li contributed equally to the manuscript.

The datasets generated during and analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

1.
Araujo
AB
,
O’Donnell
AB
,
Brambilla
DJ
,
Simpson
WB
,
Longcope
C
,
Matsumoto
AM
.
Prevalence and incidence of androgen deficiency in middle-aged and older men: estimates from the Massachusetts male aging study
.
J Clin Endocrinol Metab
.
2004
;
89
(
12
):
5920
6
.
2.
Liu
CC
,
Wu
WJ
,
Lee
YC
,
Wang
CJ
,
Ke
HL
,
Li
WM
.
The prevalence of and risk factors for androgen deficiency in aging Taiwanese men
.
J Sex Med
.
2009
;
6
(
4
):
936
46
.
3.
Giagulli
VA
,
Guastamacchia
E
,
De Pergola
G
,
Iacoviello
M
,
Triggiani
V
.
Testosterone deficiency in male: a risk factor for heart failure
.
Endocr Metab Immune Disord Drug Targets
.
2013
;
13
(
1
):
92
9
.
4.
Tsujimura
A
.
The relationship between testosterone deficiency and men’s health
.
World J Mens Health
.
2013
;
31
(
2
):
126
35
.
5.
Chalasani
N
,
Younossi
Z
,
Lavine
JE
,
Diehl
AM
,
Brunt
EM
,
Cusi
K
.
The diagnosis and management of non-alcoholic fatty liver disease: practice guideline by the american gastroenterological association, american association for the study of liver diseases, and american college of gastroenterology
.
Gastroenterology
.
2012
;
142
(
7
):
1592
609
.
6.
Machado
M
,
Cortez-Pinto
H
.
Non-alcoholic steatohepatitis and metabolic syndrome
.
Curr Opin Clin Nutr Metab Care
.
2006
;
9
(
5
):
637
42
.
7.
Choi
SY
,
Kim
D
,
Kim
HJ
,
Kang
JH
,
Chung
SJ
,
Park
MJ
.
The relation between non-alcoholic fatty liver disease and the risk of coronary heart disease in Koreans
.
Am J Gastroenterol
.
2009
;
104
(
8
):
1953
60
.
8.
Targher
G
,
Chonchol
M
,
Zoppini
G
,
Abaterusso
C
,
Bonora
E
.
Risk of chronic kidney disease in patients with non-alcoholic fatty liver disease: is there a link
.
J Hepatol
.
2011
;
54
(
5
):
1020
9
.
9.
Kalyani
RR
,
Dobs
AS
.
Androgen deficiency, diabetes, and the metabolic syndrome in men
.
Curr Opin Endocrinol Diabetes Obes
.
2007
;
14
(
3
):
226
34
.
10.
Rotter
I
,
Rył
A
,
Grzesiak
K
,
Szylińska
A
,
Pawlukowska
W
,
Lubkowska
A
.
Cross-sectional inverse associations of obesity and fat accumulation indicators with testosterone in non-diabetic aging men
.
Int J Environ Res Public Health
.
2018
;
15
(
6
):
1207
.
11.
Anderson
D
.
Sex-hormone-binding globulin
.
Clin Endocrinol
.
1974
;
3
(
1
):
69
96
.
12.
Grossmann
M
,
Hoermann
R
,
Gani
L
,
Chan
I
,
Cheung
A
,
Gow
PJ
.
Low testosterone levels as an independent predictor of mortality in men with chronic liver disease
.
Clin Endocrinol
.
2012
;
77
(
2
):
323
8
.
13.
Kley
HK
,
Nieschlag
E
,
Wiegelmann
W
,
Solbach
HG
,
Kruskemper
HL
.
Steroid hormones and their binding in plasma of male patients with fatty liver, chronic hepatitis and liver cirrhosis
.
Acta Endocrinol
.
1975
;
79
(
2
):
275
85
.
14.
Volzke
H
,
Aumann
N
,
Krebs
A
,
Nauck
M
,
Steveling
A
,
Lerch
MM
.
Hepatic steatosis is associated with low serum testosterone and high serum DHEAS levels in men
.
Int J Androl
.
2010
;
33
(
1
):
45
53
.
15.
Kim
S
,
Kwon
H
,
Park
JH
,
Cho
B
,
Kim
D
,
Oh
SW
.
A low level of serum total testosterone is independently associated with nonalcoholic fatty liver disease
.
BMC Gastroenterol
.
2012
;
12
:
69
.
16.
Lazo
M
,
Zeb
I
,
Nasir
K
,
Tracy
RP
,
Budoff
MJ
,
Ouyang
P
.
Association between endogenous sex hormones and liver fat in a multiethnic study of atherosclerosis
.
Clin Gastroenterol Hepatol
.
2015
;
13
(
9
):
1686
93.e2
.
17.
Shin
JY
,
Kim
SK
,
Lee
MY
,
Kim
HS
,
Ye
BI
,
Shin
YG
.
Serum sex hormone-binding globulin levels are independently associated with nonalcoholic fatty liver disease in people with type 2 diabetes
.
Diabetes Res Clin Pract
.
2011
;
94
(
1
):
156
62
.
18.
Hua
X
,
Sun
Y
,
Zhong
Y
,
Feng
W
,
Huang
H
,
Wang
W
.
Low serum sex hormone-binding globulin is associated with nonalcoholic fatty liver disease in type 2 diabetic patients
.
Clin Endocrinol
.
2014
;
80
(
6
):
877
83
.
19.
Seo
NK
,
Koo
HS
,
Haam
JH
,
Kim
HY
,
Kim
MJ
,
Park
KC
.
Prediction of prevalent but not incident non-alcoholic fatty liver disease by levels of serum testosterone
.
J Gastroenterol Hepatol
.
2015
;
30
(
7
):
1211
6
.
20.
Bedogni
G
,
Bellentani
S
,
Miglioli
L
,
Masutti
F
,
Passalacqua
M
,
Castiglione
A
.
The Fatty Liver Index: a simple and accurate predictor of hepatic steatosis in the general population
.
BMC Gastroenterol
.
2006
;
6
:
33
.
21.
Sun
K
,
Lin
D
,
Li
F
,
Qi
Y
,
Feng
W
,
Yan
L
.
Fatty liver index, albuminuria and the association with chronic kidney disease: a population-based study in China
.
BMJ Open
.
2018
;
8
(
1
):
e019097
.
22.
Khang
AR
,
Lee
HW
,
Yi
D
,
Kang
YH
,
Son
SM
.
The fatty liver index, a simple and useful predictor of metabolic syndrome: analysis of the Korea National Health and Nutrition Examination Survey 2010-2011
.
Diabetes Metab Syndr Obes
.
2019 Jan 24
12
181
90
.
23.
Ze
EY
,
Kim
BJ
,
Jun
DH
,
Kim
JG
,
Kang
H
,
Lee
DY
.
The fatty liver index: a simple and accurate predictor of colorectal adenoma in an average-risk population
.
Dis Colon Rectum
.
2018
;
61
(
1
):
36
42
.
24.
Huh
JH
,
Ahn
SV
,
Koh
SB
,
Choi
E
,
Kim
JY
,
Sung
KC
.
A prospective study of fatty liver index and incident hypertension: the KoGES-ARIRANG Study
.
PLoS One
.
2015
;
10
(
11
):
e0143560
.
25.
Yadav
D
,
Choi
E
,
Ahn
SV
,
Koh
SB
,
Sung
KC
,
Kim
JY
.
Fatty liver index as a simple predictor of incident diabetes from the KoGES-ARIRANG study
.
Medicine
.
2016
;
95
(
31
):
e4447
.
26.
The Lido Study Group
.
Fatty liver is an independent predictor of early carotid atherosclerosis
.
J Hepatol
.
2016
;
65
(
1
):
95
102
.
27.
Erenpreiss
J
,
Fodina
V
,
Pozarska
R
,
Zubkova
K
,
Dudorova
A
,
Pozarskis
A
.
Prevalence of testosterone deficiency among aging men with and without morbidities
.
Aging Male
.
2020
;
23
(
5
):
901
5
.
28.
Zhu
A
,
Andino
J
,
Daignault-Newton
S
,
Chopra
Z
,
Sarma
A
,
Dupree
JM
.
What is a normal testosterone level for young men? Rethinking the 300 ng/dL cutoff for testosterone deficiency in men 20-44 Years old
.
J Urol
.
2022
;
208
(
6
):
1295
302
.
29.
Lee
JH
,
Kim
D
,
Kim
HJ
,
Lee
CH
,
Yang
JI
,
Kim
W
.
Hepatic steatosis index: a simple screening tool reflecting nonalcoholic fatty liver disease
.
Dig Liver Dis
.
2010
;
42
(
7
):
503
8
.
30.
DeFina
LF
,
Radford
NB
,
Leonard
D
,
Wilson RK
,
Cooper TC
,
Clark SM
.
The association of cardiorespiratory fitness, body mass index, and age with testosterone levels at screening of healthy men undergoing preventive medical examinations: the cooper center Longitudinal Study
.
Maturitas
.
2018
;
118
:
1
6
.
31.
Herman
AP
,
Krawczyńska
A
,
Bochenek
J
,
Dobek
E
,
Herman
A
,
Tomaszewska-Zaremba
D
.
LPS-induced inflammation potentiates the IL-1β-mediated reduction of LH secretion from the anterior pituitary explants
.
Clin Dev Immunol
.
2013
;
2013
:
926937
.
32.
Souteiro
P
,
Belo
S
,
Oliveira
SC
,
Neves
JS
,
Magalhães
D
,
Pedro
J
.
Insulin resistance and sex hormone-binding globulin are independently correlated with low free testosterone levels in obese males
.
Andrologia
.
2018
;
50
(
7
):
e13035
.
33.
Jaruvongvanich
V
,
Sanguankeo
A
,
Riangwiwat
T
,
Upala
S
.
Testosterone, sex hormone-binding globulin and nonalcoholic fatty liver disease: a systematic review and meta-analysis
.
Ann Hepatol
.
2017
;
16
(
3
):
382
94
.