Risk and Prevalence of Fatty Liver Assessed in A Scientic Research Institute in Beijing

Background: To investigate the prevalence and distribution of employees with fatty liver disease (FLD) and assess the risk factors associated with diseases and biochemical markers. Methods: Disease risk markers, and the levels of biochemical markers of 2541 participants from an Institute of Scientic Research in Beijing, were assessed according to standard methods and FLD was diagnosed by abdominal ultrasonography. Chi-square test was used to compare the results between participants with and without FLD. Results: The prevalence of FLD, based on abdominal ultrasonography examination, was 43.5% (1105/2541) in this population. The overall prevalence increased with age, and it was substantially higher in males (50.0% vs females: 34.6%, P < 0.001). The gender and age distributions were signicantly different between the two groups (χ 2 =59.8 for gender and t= -15.2 for age or χ 2 =205.6 for age group, P < 0.05). Chi-square test results revealed statistically signicant differences in the BMI, prehypertension, hypertension, prediabetes, diabetes, hyperlipidemia, mixed type hyperlipidemia, hypertriglyceridemia, and the rise of GGT, ALT, AST, total bilirubin and indirect bilirubin between populations with and without FLD (P < 0.05). The risk of disease and biochemical markers contributing to the development of FLD was assessed by calculating the odd ratio (OR) and 95% condence interval (95% CI), and the results are ordered as follows: overweight and obesity [6.85 (5.54-8.48)], diabetes [6.00 (3.66-9.82)], prediabetes [5.06 (3.10-8.26)], ALT rise [4.67 (3.58-6.10)], GGT rise [4.01 (3.03-5.30)]. Multiple factors analysis was conducted to further compare those identied risk factors of FLD, which showed statistically signicant differences in age, overweight and obesity, prediabetes, diabetes, mixed type hyperlipidemia, hypertriglyceridemia and ALT rise by Logistic regression analysis (P < 0.05). OR=(positive/negative in FLD group)/(positive/negative in Non-FLD group). Regression models was used to evaluate the relations between risk factors and FLD morbidity from specic causes by adjusting the confounding factors. In this study, multivariate analysis with statistically signicant risk factors by Chi-square test at univariate model was performed by Logistic regression analysis. All reported P-values were two-sided, and differences with P < 0.05 were considered statistically signicant. transpeptidase aminotransferase aspartate aminotransferase (AST), bilirubin (TBIL), (DBIL), bilirubin (IBIL), cholesterol (TC), triglyceride (TG), density (LDL-C),


Background
Fatty liver disease (FLD) has become one of the most common disorders of liver since the improvement of people's living conditions and adoption of a more westernized diet. The prevalence of FLD has consistently risen over the recent years in China, especially for non-alcoholic fatty liver disease (NAFLD) which is associated with a higher energy intake and normal metabolic pro le in patients with a genetic predisposition. In western countries, the prevalence of FLD in the adult population is estimated to be 20% − 33% [1] , increasing the importance for its clinical diagnosis [2,3] . FLD is typically diagnosed by imaging and ultrasonography at present, and is highly prevalent (27%) in the urban population in China, with 7.6% − 34.2% being young people. It is mainly related to obesity and metabolic syndrome (MetS) [4,5] . The hepatitis disease, initially found to be induced by hepatitis virus, is still a leading cause of chronic liver disease in China, even though viral hepatitis infection has decreased from 10-7% between 1992 and 2006 in mainland China [6] , and infectious hepatitis is an important factor leading to hepatocellular carcinoma. While as a visible and most frequently diagnosed liver disease in Chinese clinics, FLD represents a particularly alarming threat to human health and public healthcare system, especially because it can readily progress to steatohepatitis, brosis or cirrhosis, even liver cancer. Approximately 10% − 25% of patients with silent liver disease develop non-alcoholic steatohepatitis (NASH), and 5% − 8% may develop liver cirrhosis within 5 years from diagnosis [7,8] , as well as having an increased risk to develop cardiovascular diseases and diabetes. Furthermore, 12.8% of patients with liver cirrhosis may develop hepatocellular carcinoma (HCC) within 3 years from diagnosis [9,10] . NASH is considered to be a major inducer of cryptogenic cirrhosis as patients with NASH have 70% risk to develop NAFLD [11,12] .
Because of the youth oriented tendency of FLD and the increased concern of scienti c researchers for their health due to the high technology mastering demands in their working environment, it was imperative to assess the health of employees in an institute of scienti c research in China. To gain further information on the prevalence and risk of fatty liver, the current study was designed to analyze the retrospective data of adult participants in a research institute in Beijing, based on physical examinations.

Study participants
A total number of 2760 consecutive adult workers who underwent routine physical examinations at the Fangshan Branch of Beijing Nuclear Industry Hospital from September 1, 2018 to December 31, 2018 were retrospectively enrolled in the study.

Physical examination, biochemical and radiological analysis
The physical examination was performed by a professional personnel doctor who had the title of attending or above. The routine examinations of internal medicine, surgery and ophthalmology were carried out according to the uni ed standards. Patients were presented to the hospital for blood sampling in the morning after 10 h of fasting. All serological measurements were performed on-site at the certi ed laboratory. Unicel DxC 800 automatic biochemistry analyzer (Beckman Coulter Laboratories, United States) was used for the determination of plasma concentration of glucose, total cholesterol (TC), triglyceride (TG), low density lipoprotein cholesterol (LDL-C), gamma glutamyl transpeptidase (GGT), alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TBIL), direct bilirubin (DBIL), indirect bilirubin (IBIL) and other biochemical markers. Abdominal ultrasonography was conducted to detect FLD with standard imaging criteria [13] . Electrocardiography and chest Xray were performed to rule out serious heart and lung diseases. Hepatitis B test in serum was performed to excludes HBV/HCV chronic infections. All employee were excluded from acute infection during physical examination.

Diagnostic criteria and de nitions
The participants were allocated into 4 groups according to their Body mass index (BMI): thin (≤ 18.4 kg/m 2 ), normal (18.5-23.9 kg/m 2 ), overweight (24.0-27.9 kg/m 2 ) and obese (≥ 28.0 kg/m 2 ) [14,15] . According to the Chinese and national hypertension guidelines for the management of hypertension [16,17] , hypertension was diagnosed when the systolic pressure was ≥ 140 mmHg and/or the diastolic pressure was ≥ 90 mmHg; prehypertension was diagnosed when the systolic pressure was 120-139 mmHg and/or the diastolic pressure was 80-89 mmHg under no medication and no self-reported hypertension. Hyperglycemia was diagnosed when the fasting plasma glucose levels were 6.1-6.95 mmol/L for prediabetes, and ≥ 6.95 mmol/L for diabetes when patients were receiving medication or had reported diabetes [18,19] . Dyslipidemia was de ned when the cholesterol TC levels were ≥ 5.7 mmol/L and/or TG ≥ 1.8 mmol/L and/or LDL-C ≥ 3.6 mmol/L and/or HDL-C < 0.8 mmol/L [20] . The various types of hyperlipidemia were diagnosed when TC content was ≥ 5.7 mmol/L and TG ≥ 1.8 mmol/L for mixed type hyperlipidemia, when TC levels were ≥ 5.7 mmol/L and TG < 1.8 mmol/L, and when LDL-C < 3.6 mmol/L and HDL-C ≥ 0.8 mmol/L for hypercholesterolemia, when the TC content was < 5.7 mmol/L and TG ≥ 1.8 mmol/L, and when LDL-C levels were < 3.6 mmol/L and HDL-C ≥ 0.8 mmol/L for hypertriglyceridemia, and when TC concentration was < 5.7 mmol/L and TG ≥ 1.8 mmol/L and LDL-C ≥ 3.6 mmol/L and HDL-C < 0.8 mmol/L for low density lipoprotein cholesterolemia. Abnormal elevation in the levels of gamma glutamyl transpeptidase (GGT), alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TBIL), direct bilirubin (DBIL) and indirect bilirubin (IBIL) were de ned when these were ≥ 50 U/L, ≥ 40 U/L, ≥ 40 U/L, ≥ 20 µmol/L, ≥ 6 µmol/L and ≥ 15 µmol/L, respectively.

Statistical analysis
The participants` physical examination results were stored in the database of the health center management system (Peking University Medical Information Technology Co., Ltd), and exported as an excel spreadsheet to extract the relevant data. All statistical analyses were performed using the SPSS statistical software suite (version 20.0, Chicago, IL, United States). Measurement data were expressed as mean ± SD. Comparisons between quantitative data and categorical variables were carried out by Student's t-test and χ 2 test, respectively.
Odds ratio (OR) was used to express ratio between positive to negative in FLD group and positive to negative in Non-FLD group. i.e. OR=(positive/negative in FLD group)/(positive/negative in Non-FLD group). Regression models was used to evaluate the relations between risk factors and FLD morbidity from speci c causes by adjusting the confounding factors. In this study, multivariate analysis with statistically signi cant risk factors by Chi-square test at univariate model was performed by Logistic regression analysis. All reported P-values were twosided, and differences with P < 0.05 were considered statistically signi cant.

Demographics and distribution of fatty liver
The study population of physical examination of employees in the institute of scienti c research of Beijing comprised 1466 males and 1075 females besides 19 of having HBV/HCV in ections, 12 of lung shadow, 32 of broid, 9 of in uenza and 147 of missing. The average age of participants was 39.4 ± 9.4 years old (range: 19-60 years) and they were divided to four groups, each with a ten years coverage. The percentage of FLD was 43.5% based on abdominal ultrasonography (1105/2541), with an obvious higher prevalence in males than in females (50.0% vs 34.6%, P < 0.001). As presented in Table 1, the prevalence of FLD was much higher in males than in females (65.7% vs 34.3% in the 30 to 39 year-old group, and 57.1% vs 41.7% in the 40 to 49 year-old group), and this trend did not exist in the other two age groups (for both P < 0.001, χ 2 = 205.6) (shown in Table 1).  Table 2).  The age prevalence distribution of diseases and biochemical markers in the 1105 participants with FLD is displayed in Table 4. The overall prevalence increased with age, which was statistically signi cant different for all disease markers and for two of the biochemical markers (hyperlipidemia and ALT rise), and there was no age bias.

Comparison of prevalence between populations with fatty liver and non-fatty liver
Comparison of the fatty liver to the non-fatty liver control group revealed gender and age distribution differences (χ 2 = 59.8 for gender and t= -15.2 for age or χ 2 = 205.6 for age group, P < 0.05), therefore we analyzed the odds ratio (OR) to identify the risk factors associated with FLD. The Chi-square test was performed to assess factors such as BMI, prehypertension, hypertension, prediabetes, diabetes, hyperlipidemia, mixed type hyperlipidemia, hypertriglyceridemia, and the rise of GGT, ALT, AST, total bilirubin and indirect bilirubin between populations with and without FLD (P < 0.05). The values of OR and 95% con dence interval (95% CI) are listed in a descending order in Table 5, starting with overweight and obesity, diabetes, prediabetes, ALT rise, GGT rise and so on.  Table 6. Logistic regression analysis revealed that there were statistically signi cant differences in age, overweight and obesity, prediabetes, diabetes, mixed type hyperlipidemia, hypertriglyceridemia and ALT rise (P < 0.05).

Discussion
FLD is a common clinical pathologic syndrome that manifests from hepatic steatosis and excessive fat accumulation induced by a variety of factors. Despite the increase in alcohol sales in China, the intake of alcohol in public institutions has been signi cantly reduced by government policies. However, NAFLD is still prevalent and is strongly associated with metabolic disorders. Not surprisingly, FLD has been frequently reported in patients with type 2 diabetes mellitus (T2DM) and obesity, accounting for 40-80% and 30-90%, respectively [21] . Thus it is necessary to track FLD progress as it may develop to steatohepatitis, fatty liver cirrhosis, and associated virus hepatocellular carcinoma. Although liver biopsy is the gold standard for the diagnosis of FLD, abdomen ultrasonography is still a general approach to screen FLD especially in adult populations. The prevalence of NAFLD has increased globally and recently became the predominant cause of chronic liver disease around the world. The epidemiology and demographic characteristics vary worldwide [3] . The incidence rate of FLD, diagnosed by ultrasonography, was reported to be more than 46% in developed countries [7,22,23] and 39.5% in developing countries [4,24] , and it is still on the rise.
In the current study in Fangshan Branch of the Beijing Nuclear Industry Hospital, the prevalence of FLD was 43.5%, which was higher than the results reported by another survey study in Beijing [24] . However, only 19 to 60-year-old subjects were employed in the research and 17.9% of patients below 40-year-old were diagnosed with FLD (16.1% is the standard). The prevalence increased with age, which was dramatically higher than previously estimated values in other Chinese cities [25,26] . Overall, the age group with the highest FLD prevalence was the over 50-yearold group [27] , at which age the adiponectin levels in women also generally decreased. According to the results shown in Table 1, there was a negative correlation between FLD prevalence and adiponectin levels, which could explain the lower incidence of FLD in women over 50-year-old as at this age they undergo menopause and as a result have lower estrogen and higher androgen levels. As a whole, the prevalence of FLD, overweight and obove, prehypertension (hypertension) and hyperlipidemia was relatively high in the total population (Table 2). Actually, age comes with experience and higher responsibility in the working environment. Intense competition for title assessment and promotion and the prolonged working hours in this institute has resulted in the personnel to decrease physical activity time, leading to the prevalence of overweight and obesity and hypertension, which are risk factors for developing FLD with diabetes [3,11,20,24] . The westernized diet, irregular diet, reduced physical exercise, hypertension and diabetes are major determinants of FLD among Chinese scienti c researchers. The incidence of overweight and obesity, hypertension and diabetes was considerably high in the 1105 participants diagnosed with FLD. As seen in the Table 2, the prevalence of the other biochemical markers was consistent with that of disease markers, and both resembled that of hyperlipidemia and FLD.
Previous reports have indicated that males have more risk to develop FLD than females after the age of 50 [28] .
However, alcohol-related FLD (AFLD) in males is more prominent than in females under excessive alcohol consumption. In this research, AFLD and NAFLD were not separately investigated. The rate of rise of the levels of biochemical markers for MetS in males exceeded that of females, with the exception of AST. The prevalence of FLD was 50% in males and 34.6% in females, which was relatively higher than what has been reported in a previous Chinese study [29] . The results of gender distribution of prevalence and categorization in the 1105 participants with FLD (Table 3), were in agreement with the ndings of a Japanese study [30] . Besides, males usually consume high fat diet and as a consequence store fat in their abdomen, whereas females that look after their weight tend to utilize fat in the subcutaneous tissue by weight loss plans such as physical exercise. The reason why the prevalence of overweight and obove was still high remains uncertain. The differential fat accumulation in males and females may be a consequence of their different lipid metabolism pathways. The comparative values and distribution of different genders are listed in Table 3. Brie y, our study, along with previously published ndings, highlights the importance of prevention and screening of FLD in speci c age populations. It has been reported that a high protein and hypocaloric diet are associated with improved lipid pro le, glucose homeostasis and liver enzymes function in NAFLD, independent to BMI decrease or body fat mass reduction [31] . However, our results showed statistically signi cant differences in all disease markers and two of the biochemical markers (hyperlipidemia and ALT rise) between the age groups in the FLD group. As shown in Table 4, the prevalence of prehypertension and ALT rise decreased by aging, and this rise was surprisingly higher in the 19 to 29 year-old group than in the other three groups. On the other hand, although the prevalence of other markers increased with age, there were no signi cant differences between the age groups.
In the current study most FLDs were diagnosed in participants without any clinical symptoms. Over 50% of cases with FLD had mildly increased contents of biochemical markers, most of which were also accompanied by a rise in cholesterol, triglyceride, LDL cholesterol, GGT, AST levels. There were signi cant differences in hyperlipidemia, mixed type hyperlipidemia, hypertriglyceridemia, GGT, AST and AST between the FLD and non-FLD group (Table 5). These alterations were found to be associated with certain disease markers. However, in this cohort study analysis of risk factors contributing to the prevalence of FLD was not performed. These results suggest that FLD, with NAFLD being the most common type, may be correlated with those disorders. Additionally, age and gender are basic risk factors as they play important roles in the development of NAFLD [32] . Multiple Logistic analysis was carried out to identify risk factors for the analysis of the prevalence of FLD. The vital risk factors determined in this study were: age, overweight and obesity, prediabetes, diabetes, mixed type hyperlipidemia, hypertriglyceridemia and ALT (Table 6). However, our study included a single-site population and diagnosis of FLD was based on ordinary physical examination, which may limit the application of the identi ed markers for screening larger populations. Therefore, further studies with higher number of participants from various institutes of scienti c research, including more demographic and biochemical data, are required in the future.
Recently, FLD has also been attributed to NAFLD, which is characterized by a heightened systemic proin ammatory state. NAFLD accelerates the risks of arteriosclerosis and other cardiovascular diseases or associated events [33] . However, the previously reported correlation relationship between total cholesterol levels and cancer risk [34] is controversial because in the current study we found that FLD prevalence was associated with other risk factors (e.g., gender, age, overweight and obesity, hypertension, diabetes and MetS), and NAFLD was closely related to the incidence of cardiovascular disease, a clear risk factor for hyperlipidemia [35] . The incidence and mortality rates of cardiovascular diseases were the highest amongst chronic diseases worldwide [36] . The most common causes of death by NAFLD are atherosclerotic cardiovascular diseases and hepatic cirrhosis [37] , hence it is important to regularly screen for FLD to prevent development of disease. The screening and prevention of obesity and hypertension are also of great importance, as they are risk factors of cardiovascular diseases and their prevalence was 39.1% and 7.9%, respectively, in the FLD group. Therefore, more attention should be paid to systemic diseases of the cardiovascular, urinary, digestive and respiratory system during physical examination, instead of the incidence of overweight and obove, diabetes, hypertension and dyslipidemia.
Conclusion high prevalence of FLD was discovered in employees of the institute of scienti c research by physical examination, thus targeted intervention should be carried out as soon as possible. However, this study also has some shortcomings. First, the demographic information was relatively small; second,the diagnosed FLD had not reganding severity with mild, moderate and severe; besides, the level of AST and ALT is parallel, while in this study, the rise of ALT and AST was not similar; lastly, the incidence of pre-hypertension was too high in this study, more relative studies are needed to gain more detailed information for FLD in the general populations from other graduate schools and colleges to determine the contribution of the above risk factors to FLD development and to set up better guidelines for prevention and control of chronic diseases.

Declarations
There study was approvaled by the ethics committee of Soochow University.
Ethics approval and consent to participate Not applicable.

Funding
The present study was supported by the Nuclear Energy Development Project (No. 2016-1295), and Postgraduate Research & Practice Innovation Program of Jiangsu Province (KYCX17-2017).

Availability of data and materials
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.
Author contributions SGQ conceived and designed the study. SGQ and HQC performed this study, HQC and DML extracted the data, SGQ analyzed the data. SGQ, HQC and DML participated in the writing of this paper. All authors have read and approved this manuscript.