Prevalence and characteristics of MAFLD in Chinese adults aged 40 years or older: A community-based study

Jing Zeng , Li Qin , Qin Jin , Rui-Xu Yng , Gung Ning , Qing Su , Zhen Yng , , Jin-Go Fn ,

a Department of Gastroenterology, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 20 0 092, China

b Department of Endocrinology, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 20 0 092, China

c Department of Endocrine and Metabolic Diseases, Shanghai Institute of Endocrinology and Metabolism, Shanghai Clinical Center for Endocrine and Metabolic Diseases, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200127, China

Keywords: Nonalcoholic fatty liver disease Metabolic dysfunction-associated fatty liver disease Prevalence Oral glucose tolerance test

ABSTRACT Background: Nonalcoholic fatty liver disease (NAFLD) was recently proposed to be renamed metabolic dysfunction-associated fatty liver disease (MAFLD) with the diagnostic criteria revised. We investigated the similarities and differences in the prevalence and clinical characteristics of MAFLD and NAFLD in Chinese adults. Methods: A cross-sectional study of 9980 Chinese individuals aged 40 years or older was performed be- tween 2011 and 2012 using randomized, stratified cluster sampling in Shanghai, China. A detailed ques- tionnaire and the results of abdominal ultrasonography, a standardized 2-h 75-g oral glucose tolerance test and blood biochemical examinations were collected. Results: A total of 9927 subjects were included in this study. The prevalence of MAFLD (40.3%) was signif- icantly higher than that of NAFLD (36.9%) ( P ＜ 0.05). MAFLD was highly prevalent in type 2 diabetes mel- litus (T2DM) (53.8%), impaired fasting glucose (35.7%) and impaired glucose tolerance (40.9%). High risk of advanced fibrosis based on fibrosis-4 was highly prevalent (14.7%) in lean MAFLD with T2DM. Among 9927 subjects, 3481 (35.1%) fulfilled the diagnostic criteria for MAFLD and NAFLD (MAFLD + NAFLD + ), 521 (5.2%) MAFLD + NAFLD-, and 181 (1.8%) MAFLD-NAFLD + . The MAFLD + NAFLD- group had more significant metabolic disorders than those in the MAFLD + NAFLD + group (all P ＜ 0.05). Among MAFLD-NAFLD + sub- jects, 82.9% had metabolic disorders. Conclusions: The new definition of MAFLD may better reflect the pathogenesis related to metabolism. Future research should focus on studying the natural history, pathogenesis and treatment effectivity of the overlap and non-overlap of NAFLD and MAFLD subjects.

Introduction

In 1980, Ludwig et al. proposed the term “nonalcoholic steato- hepatitis (NASH)” for a condition mimicking alcoholic hepatitis histologically but with no significant alcohol consumption [1] . In 1986, Schaffner et al. proposed the name “nonalcoholic fatty liver disease (NAFLD)” [2] . In the past several decades, a diagnosis of NAFLD is still made only after ruling out other causes of fatty liver, especially excessive alcohol intake. In fact, NAFLD is related not only to serious liver-related complications but also to metabolic disorders and cardiovascular and renal complications [3] .

Currently, NAFLD has become a worldwide health problem, af- fecting approximately 25% of the general adult population, and is high prevalent in subjects with co-existing risk factors such as obesity and diabetes [3] . The clinical characteristics and in- creasing number of NAFLD subjects suggest that both clinicians and subjects themselves should focus not only on “nonalco- holic” or other liver diseases but also on key metabolic disor- ders, including insulin resistance and other metabolic syndrome components [ 4 , 5 ]. Therefore, in 2020, an international expert group consensus statement proposed that NAFLD should be renamed “metabolic dysfunction-associated fatty liver disease (MAFLD)” to better reflect its metabolism-related etiology and disease hetero- geneity in pathological characteristics. The new definition and the revised diagnostic criteria of MAFLD are anticipated to better im- prove clinical care [ 6 , 7 ].

Therefore, in this article, we performed a cross-sectional study to investigate the prevalence and clinical characteristics of MAFLD in Chinese adults aged 40 years or older.

Patients and methods

Study population

This study is a part of the Risk Evaluation of cAncers in Chi- nese DiabeTic Individuals: A lONgitudinal (REACTION) study, which was a community-based cross-sectional study conducted among 259 657 Chinese individuals aged 40 years and older, from 2011 to 2012 [8] . The data presented in this study are based on the baseline survey of subsamples from Chongming District in Shang- hai, China. There were 9980 participants who were all required to complete a questionnaire including demographic information (date of birth, sex), history of chronic diseases (chronic liver disease, my- ocardial infarction, cerebral infarction, hypertension, diabetes, hy- perlipidemia, etc.), related medications, and smoking and alcohol consumption habits. To ensure its validity and response rates, the questionnaire was completed under the guidance of a physician.

The exclusion criteria for the study population were as follows: i) those with a history of severe chronic diseases, liver cirrhosis, or malignancy; ii) those with some important missing information; and iii) those who failed to complete an abdominal ultrasonogra- phy (US). Thus, a total of 9927 participants (3228 men and 6699 women) were eventually included in this analysis. The study proto- col was approved by the Ethics Committee of Xinhua Hospital Affil- iated to Shanghai Jiao Tong University School of Medicine (XHEC- C-2012-023), and all studies were carried out in accordance with the approved guidelines. Written informed consent was obtained from all participants.

Clinical and laboratory data collection

Participant characteristics and anthropometric indices, including body weight, height, waist circumference (WC), body mass index (BMI), systolic blood pressure (SBP), and diastolic blood pressure (DBP), were obtained. BMI was defined as weight in kilograms di- vided by the square of height in meters (kg/m2). SBP and DBP were measured in a sitting position after a five minutes rest with a mer- cury sphygmomanometer.

Peripheral blood samples were collected from the partici- pants in the morning after fasting for at least 12 h and were measured with a Hitachi 7600 series automatic analyzer (Hi- tachi, Tokyo, Japan). The variables included platelets (PLT), alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma- glutamyl transferase (GGT), total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipopro- tein cholesterol (LDL-C), serum creatinine (SCr) and partly high- sensitivity C-reactive protein (hsCRP).

A standardized 2-h 75-g oral glucose tolerance test (OGTT) was performed on all participants after an overnight fast. Fasting plasma glucose (FPG), plasma glucose after 2 h (2 h-PG), fasting plasma insulin and glycosylated hemoglobin Alc (HbAlc) were all obtained to assess the glucose metabolism status. Fasting plasma insulin and FPG were used to calculate a homeostasis model as- sessment of insulin resistance (HOMA-IR) index using the formula HOMA-IR = (FPG × fasting plasma insulin)/22.5 [9] .

The fibrosis-4 (FIB-4) index was calculated according to the following formula: FIB-4 = [age (years) × AST (U/L)] / [PLT

The estimated glomerular filtration rate (eGFR) was calcu- lated according to the MDRD equation: eGFR = 186 × SCr -1.154 × age - 0.203 ( × 0.742 if female) [10] .

Abdominal US was performed by a certified and experienced technician blinded to the clinical data of the study participants.

Work definitions

Obese, overweight and lean: a BMI greater than 25 kg/m2was used to define the obese population, 23-25 kg/m2, overweight and less than 23 kg/m2, lean [11] .

Type 2 diabetes mellitus (T2DM), impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) were defined according to the World Health Organization 2006 criteria [12] .

Metabolic syndrome (MetS) was diagnosed with WC ≥90 cm in men and ≥80 cm in women (central obesity) plus any two of the following metabolic disorders: i) blood pressure ≥130/85 mmHg or receiving anti-hypertensives, ii) TG ≥1.7 mmol/L or receiving specific drug treatment, iii) HDL-C＜1.0 mmol/L in men and＜1.3 mmol/L in women, and (iv) FPG ≥5.6 mmol/L or T2DM.

A subject was considered to have MAFLD if they had a fatty liver based on ultrasonographic features [13] , together with T2DM, overweight/obesity, or at least two metabolic disorders [6] : i) WC ≥90 cm in men and ≥80 cm in women (central obesity), ii) blood pressure ≥130/85 mmHg or receiving anti-hypertensives, iii) TG ≥1.7 mmol/L or receiving specific drug treatment, iv) HDL-C＜1.0 mmol/L in men and＜1.3 mmol/L in women, v) prediabetes (FPG 5.6-6.9 mmol/L or 2 h-PG 7.8-11.0 mmol/L or HbAlc 5.7%-6.4%), or vi) HOMA-IR ≥2.5.

A subject was considered to have NAFLD if they had fatty liver based on abdominal US, no or modest alcohol consumption, and no evidence of other concomitant liver diseases [ 14 , 15 ].

A subject was considered to have alcoholic fatty liver if they had fatty liver based on abdominal US, with excessive alcohol con- sumption ( ≥40 g/day in men, ≥20 g/day in women), usually for＞5 years, and/or ongoing daily alcohol consumption＞80 g over 2 weeks [ 14 , 15 ]. Alcohol intake was assessed by the questionnaire.

Chronic kidney disease (CKD) was defined as an eGFR＜60 mL/min per 1.73 m2.

The risk of advanced liver fibrosis (stage 3 or 4 according to Kleiner’s classification of MAFLD) was evaluated using the FIB-4 in- dex [16] : low risk, FIB-4 index＜1.3; intermediate risk, FIB-4 index 1.3-2.67; and high risk, FIB-4 index＞2.67.

Statistical analysis

Continuous variables are expressed as the mean ± standard de- viation (SD) for those with a normal distribution or median (in- terquartile range, IQR) for those with a skewed distribution. The Student’st-test and Chi-square test were used to test for differ- ences in continuous and categorical variables between the two dif- ferent groups. The significant level was set atP＜0.05. All statis- tical analyses were performed using SPSS version 22.0 (SPSS Inc., Chicago, IL, USA).

Results

General characteristics

A total of 9980 adults agreed to be included in this study. After applying the exclusion criteria, 9927 participants were included in the analysis ( Fig. 1 ). The mean age was 56.0 ± 7.8 (40-73) years, and 3228 (32.5%) were male. A total of 4170 (42.0%) participants were classified as obese, and 2417 (24.3%) as overweight. Among 9927 subjects, there were 2548 (25.7%) with T2DM, 916 (9.2%) with IFG and 2482 (25.0%) with IGT. A total of 3583 (36.1%) had hyper- tension, 5386 (54.3%) had MetS and 56 (0.6%) had CKD. Thirty-two (0.3%) had a history of myocardial infarction, 54 (0.5%) had a his- tory of cerebral infarction and 329 (3.3%) had a history of coronary heart disease.

Fatty liver (4185, 42.2%) was based on the abdominal US. A total of 3662 (36.9%), 4002 (40.3%) and 523 (5.3%) participants fulfilled the diagnostic criteria for NAFLD, MAFLD and alcoholic fatty liver, respectively, and the prevalence of MAFLD was signif- icantly higher than that of NAFLD (P＜0.05). A total of 3481 (35.1%) subjects fulfilled the diagnostic criteria for both NAFLD and MAFLD (MAFLD + NAFLD + ). A total of 181 (1.8%) subjects fulfilled the diagnostic criteria for NAFLD but not MAFLD (MAFLD-NAFLD + ), and 521 (5.2%) fulfilled the diagnostic criteria for MAFLD but not NAFLD (MAFLD + NAFLD-) ( Fig. 2 ).

Fig. 1. Flow chart of the study participants. US: ultrasonography; REACTION study: the Risk Evaluation of cAncers in Chinese DiabeTic Individuals: A lONgitudinal study.

Age-, sex-, and BMI-adjusted prevalence of MAFLD and advanced fibrosis

As shown in Table 1 and Fig. 3 , the prevalence of MAFLD was increased with the increase of BMI (P＜0.05). The prevalence of MAFLD in men had almost reached its highest level (42.3%) be- tween the ages of 45 to 49 years, but the peak prevalence (45.8%) in women was at the age of 60 years and older. MAFLD was sig- nificantly more prevalent in men under 50 years of age than in women under the same age, but the opposite was true among participants older than 50 years. The prevalence of MAFLD in the T2DM group (53.8%) and IGT group (40.9%) was significantly higher than that in the IFG group (35.7%) and normal group (32.4%) (P＜0.05).

Table 1 Prevalence of MAFLD in different sex and BMI groups.

The risk of advanced liver fibrosis was evaluated by the FIB-4 index. As shown in Fig. 4 , the prevalence of MAFLD subjects with a high risk of advanced fibrosis increased with age (P＜0.05). A high risk of advanced fibrosis in MAFLD was significantly more preva- lent in lean subjects (11.1%) than in overweight (10.2%) and obese groups (8.9%) (P＜0.05). The prevalence of advanced fibrosis in MAFLD was the highest in the lean T2DM group (14.7%).

Characteristics of MAFLD

The characteristics of the participants with MAFLD are shown in Table 2 . Compared to the control group without fatty liver and with normal liver enzyme levels, these subjects in the MAFLD group were older and had more body fat with a significantly larger WC and a higher BMI. The subjects in the MAFLD group also had significantly higher SBP, DBP, FPG, 2 h-PG, HbAlc, fasting plasma in- sulin, HOMA-IR, TG, TC, SCr, ALT, AST, GGT and FIB-4 score (allP＜0.05). The prevalence of other metabolic-related diseases, includ- ing myocardial infarction, cerebral infarction, coronary heart dis- ease and CKD, were significantly increased in the MAFLD group (allP＜0.05) ( Table 2 ).

Similarities and differences between MAFLD and NAFLD

As shown in Table 3 , there were no significant differences in the levels of WC and BMI between the MAFLD + NAFLD- and MAFLD + NAFLD + groups. The subjects in the MAFLD + NAFLD- group had significantly higher TG, TC, LDL-C, ALT, FPG, 2 h-PG, HbAlc, SBP, DBP, and SCr levels than those in the MAFLD + NAFLD + group. Compared to the MAFLD + NAFLD + group, the subjects in the MAFLD-NAFLD + group were younger and leaner and had sig- nificantly lower TG, LDL-C, ALT, FPG, 2h-PG, HbAlc, HOMA-IR, SBP, DBP, and SCr levels ( Table 3 ). The FIB-4 scores in these groups were not significantly different.

Table 2 Characteristics of MAFLD

Table 3 Characteristics of MAFLD + NAFLD + , MAFLD + NAFLD- and MAFLD-NAFLD + .

Metabolic disorders in MAFLD-NAFLD +

A total of 181 patients fulfilled the diagnostic criteria for MAFLD-NAFLD + , only 31 (17.1%) had no metabolic disorders, and 150 (82.9%) had one metabolic disorder ( Fig. 5 ).

Fig. 2. Concordance of MAFLD and NAFLD diagnostic criteria. MAFLD: metabolic dysfunction-associated fatty liver disease; NAFLD: nonalcoholic fatty liver disease; HCV: hepatitis C virus; BMI: body mass index.

Fig. 3. Prevalence of MAFLD in different ages and groups, including normal glucose tolerance, IFG, IGT and T2DM grouped by sex and BMI. A : Prevalence of MAFLD in different ages grouped by sex; B : prevalence of MAFLD in different groups including normal glucose tolerance, IFG, IGT and T2DM grouped by sex; C : prevalence of MAFLD in different ages grouped by BMI; D : prevalence of MAFLD in different groups including normal glucose tolerance, IFG, IGT and T2DM grouped by BMI. MAFLD: metabolic dysfunction-associated fatty liver disease; BMI: body mass index; IFG: impaired fasting glucose; IGT: impaired glucose tolerance; T2DM: type 2 diabetes mellitus.

Fig. 4. Prevalence of a high risk of advanced fibrosis in MAFLD according to FIB-4 in different age and groups including normal glucose tolerance, IFG, IGT and T2DM grouped by sex and BMI. A : Prevalence of a high risk of advanced fibrosis in MAFLD in different age groups by sex; B : prevalence of a high risk of advanced fibrosis in MAFLD in different groups including normal glucose tolerance, IFG, IGT and T2DM grouped by sex; C : prevalence of a high risk of advanced fibrosis in MAFLD in different ages grouped by BMI; D : prevalence of a high risk of advanced fibrosis in MAFLD in different groups including normal glucose tolerance, IFG, IGT and T2DM grouped by BMI. MAFLD: metabolic dysfunction-associated fatty liver disease; FIB-4: fibrosis-4; IFG: impaired fasting glucose; IGT: impaired glucose tolerance; T2DM: type 2 diabetes mellitus; BMI: body mass index.

Fig. 5. Metabolic disorders in MAFLD-NAFLD + . MAFLD: metabolic dysfunction-associated fatty liver disease; NAFLD: nonalcoholic fatty liver disease; TG: triglyceride; HDL-C: high-density lipoprotein cholesterol.

Discussion

We analyzed the prevalence, clinical features, and related metabolic disorders of MAFLD in Chinese adults aged 40 years or older based on a cross-sectional study. Our results revealed that the prevalence of MAFLD was higher than that of NAFLD. The presence of obesity, T2DM, IFG and IGT were all significantly predictive of MAFLD. The proportion of individuals with a high risk of advanced fibrosis was the highest in lean T2DM patients with MAFLD. Those MAFLD + NAFLD- had obvious metabolic disor- ders and a large percentage of MAFLD-NAFLD + patients also had metabolic disorders.

In the past 20 years, with the rapid development of the econ- omy and lifestyle changes, NAFLD has become the most common chronic liver disease in China. With the uncertainties of the mech- anisms [17] , there is no cure treatment. Because most of NAFLD is closely related to insulin resistance, central obesity, and T2DM [3] , the term “MAFLD” was proposed to rename “NAFLD” to bet- ter reflect its metabolism-related etiology and disease heterogene- ity in its pathological characteristics. Lin et al. in a “real world”data analysis revealed that the prevalence of MAFLD was 31.2% (4087/13 083), and the prevalence of NAFLD was 33.2% (4347/ 13 083) using the National Health and Nutrition Examination Sur- vey (NHANES)-1988-1994 database [18] . As expected, compared with patients with NAFLD, patients with MAFLD are more likely to have multiple metabolic comorbidities and have more cases of advanced fibrosis [18] . In our study, the prevalence of NAFLD was higher than that estimated before [3] . As the population in our study was mainly the middle-aged and elderly population with high proportion of obesity and MetS, the epidemic of obesity and metabolic disorders increased the prevalence of NAFLD. And the prevalence of MAFLD was even significantly higher than that of NAFLD. The main reason of the current high prevalence of MAFLD may be that other liver diseases also coexist with MAFLD because of coexistence of T2DM or obesity. Although most of the NAFLD fulfilled the diagnostic criteria for MAFLD, part of subjects with other liver diseases (not NAFLD) also developed fatty liver with in- sulin resistance and metabolic dysfunction, so we categorized these subjects as MAFLD + NAFLD-. In our study, a total of 521 (13.0% of MAFLD) fitted this category. We should not only treat MAFLD for these patients, but also other liver diseases, especially alcoholic fatty liver.

A crucial step in the management of patients with NAFLD is the identification of advanced fibrosis, which is directly associated with liver-related events [19] . It has been reported that subjects with IFG, T2DM or other metabolic abnormalities of MetS have an increased risk of advanced fibrosis [20] . Advanced fibrosis and liver complications are more prevalent in NAFLD patients with di- abetes [ 21 , 22 ]. Therefore, glucose metabolism abnormalities may also predict a worse long-term prognosis of MAFLD patients [23] . In our study, MAFLD was highly prevalent in the T2DM (53.8%), IFG (35.7%) and IGT (40.9%) groups. The prevalence of advanced fibro- sis in MAFLD was high in the T2DM group (11.4%) and achieved the highest proportion in the lean T2DM group (14.7%). Therefore, we may need to pay attention to screening for possible advanced fibrosis in lean MAFLD, especially those with T2DM. OGTT is very important for MAFLD.

In many studies, central obesity and insulin resistance have been suggested the most important risk factors for NAFLD patients [ 24 , 25 ]. On the other hand, the coexistence of fatty liver related to excessive alcohol intake or other causes and the pres- ence of metabolic disorders are common [26] . Therefore, the re- naming of NAFLD to MAFLD is not simply a renaming of a simple medical term, but to separate fatty liver associated with metabolic dysfunction from NAFLD. Lin et al. also indicated that MAFLD can better identify higher risk patients than NAFLD [18] . In our study, the subjects in the MAFLD group had significantly higher BMI, WC, SBP, DBP, FPG, 2 h-PG, HbAlc, fasting plasma insulin, HOMA-IR, TG, TC, SCr, ALT, AST, GGT and FIB-4 score. The term “MAFLD” re- minds us that overweight/obesity, T2DM, and metabolic syndrome are undoubtedly the main causes of fatty liver rather than just so- called risk factors. In our study, the MAFLD + NAFLD- had more se- vere metabolic disorders. These were individuals developing fatty liver coexisting with alcohol abuse, genotype-3 HCV infection or Wilson’s disease. The MAFLD diagnosis of these individuals em- phasizes the multiple causes of fatty liver, such as alcohol and metabolic dysfunction, and requires consideration of weight man- agement and controlling metabolic and cardiovascular risk factors to achieve effective prevention and treatment of these chronic liver diseases. There were also a considerable number of NAFLD patients without MAFLD, so-called MAFLD-NAFLD + . Such individuals had neither traditional liver damage factors nor overweight/obesity, T2DM nor two or more metabolic and cardiovascular risk factors. Lin et al. found that the MAFLD-NAFLD + had no metabolic disor- ders as defined by the MAFLD diagnostic criteria, who were leaner with a mean BMI of 21.7 ± 2.1 kg/m2[18] . In our study, 181 (1.8%) MAFLD-NAFLD + were younger and leaner than the subjects of MAFLD + NAFLD + . However, 82.9% of MAFLD-NAFLD + already had one metabolic disorder. Based on this, MAFLD-NAFLD + also needed to be paid attention. The cause of MAFLD-NAFLD + may be sus- ceptibility gene polymorphism changes, such asPNPLA3; an early manifestation of MAFLD or other rare ignored causes, such as Wil- son’s disease. Clinicians need to carefully follow up these subjects to find if they will become MAFLD as time goes by, as long as they develop overweight, obese, T2DM, or another metabolic disorder, they should be diagnosed as MAFLD.

There are some limitations of this study. Firstly, fatty liver sta- tus was not based on liver biopsy but on US, which cannot de- termine the severity of fatty liver. Nevertheless, US is still widely used in population-based studies with reasonable accuracy. Further research is needed to provide data on liver histological results to confirm our observations. Secondly, certain genes may be closely related to MAFLD-NAFLD + , but the initial research design did not include genetic tests. Thirdly, this is a cross-sectional study and may lack evidence of the incidence of MAFLD, the prognosis of MAFLD and advanced fibrosis, and related risk factors. These are also very important and should be covered in future studies. De- spite these limitations, our data provide important insights into the prevalence and characteristics of MAFLD in Chinese adults.

In conclusion, our results have shown that MAFLD can bet- ter reflect the importance of metabolic disorders in fatty liver disease. However, the individuals with MAFLD-NAFLD + and with MAFLD + NAFLD- had their own unique characteristics. It is neces- sary to closely follow them and strengthen prevention and treat- ment of metabolic disorders as soon as possible. Future studies are needed to discover the natural history, pathogenesis and treatment effectivity of the overlap and non-overlap of NAFLD and MAFLD subjects.

Acknowledgments

None.

CRediTauthorshipcontributionstatement

JingZeng:Data curation, Formal analysis, Writing - original draft.LiQin:Data curation, Formal analysis, Writing - original draft.QianJin:Data curation, Investigation.Rui-XuYang:Data cu- ration, Investigation.GuangNing:Conceptualization, Methodology.QingSu:Conceptualization, Methodology.ZhenYang:Conceptual- ization, Funding acquisition, Supervision, Writing - review & edit- ing.Jian-GaoFan:Conceptualization, Funding acquisition, Supervi- sion, Writing - review & editing.

Funding

This study was supported by grants from the Collaborative Innovation Program of Shanghai Municipal Health Commission (2020CXJQ01), the National Natural Science Foundation of China (81873565 and 82100605), Shanghai Jiao Tong University Trans- med Awards Research (20190104), Star Program of Shanghai Jiao Tong University (YG2021QN54), Hospital Funded Clinical Research, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine (17CSK04 and 15LC06).

Ethicalapproval

The study protocol was approved by the Ethics Committee of Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine (XHEC-C-2012-023). Written informed consent was obtained from all participants.

Competinginterest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the sub- ject of this article.