Introduction
Nonalcoholic fatty liver disease (NAFLD) is a spectrum of liver disease, including simple hepatic steatosis, nonalcoholic steatohepatitis (NASH), liver cirrhosis and even hepatocellular carcinoma (HCC). NAFLD has become the most common chronic liver disease in the world, with a global comprehensive prevalence of 25% and an increasing trend year by year [
1]. It is closely associated with metabolic syndrome (MetS) including obesity, hypertension, hyperglycemia and dyslipidemia [
2]. NAFLD may be both the liver manifestation and the cause of MetS [
3], and there is an interplay between these two diseases. Recently, NAFLD with metabolic disorders has been redefined as metabolic-associated fatty liver disease (MAFLD) [
4]. NASH is an inflammatory progressive form of NAFLD and can lead to cirrhosis and HCC without timely intervention. Recognition and diagnosis of different stages, particularly NASH, are very important for the management of NAFLD. Although liver biopsy is the gold standard for the diagnosis and the monitoring of NAFLD, the limitation is not only that it is an invasive procedure due to trauma and complications such as bleeding and bile fistula risks, but also its expensive nature of liver biopsy-based diagnosis, sampling and inter pathologist variations [
5]. Therefore, there is an unmet need for seeking brand-new noninvasive techniques or biomarkers that can be used not only to detect NASH among patients with metabolic risk factors or MetS, but also to monitor the effect of treatment.
Bone morphogenetic protein 9 (BMP9), a member of the transforming growth factor-β (TGF-β) superfamily [
6], is mainly synthesized by hepatic mesenchymal cells and released into the blood through autocrine or paracrine secretion [
7]. Preclinical and clinical studies highlight that BMP9 influences the factors in MetS such as type 2 diabetes (T2DM) and NAFLD [
8]. Luo et al. found that compared to healthy subjects, serum BMP9 levels were significantly lower in patients with T2DM [
9]. Similarly, another study revealed that BMP9 levels decreased significantly in MetS patients than in healthy controls. Hao et al. found circulating BMP9 decreased in patients with T2DM and NAFLD [
10]. Our previous study [
11] showed BMP9 was significantly reduced in NAFLD mice, and supplement of BMP9 improved NAFLD by downregulating the expression of genes related to glucometabolism and lipometabolism, thereby reducing liver inflammation. All these findings hint BMP9 has the potential as a serological biomarker and new treatment for NAFLD and MetS. This trial intends to explore whether BMP9 can be used as a novel biomarker for severity of NAFLD and MetS using a population-based cohort study, to provide evidence for early diagnosis and intervention of high-risk individuals for NASH, thereby reducing the adverse outcomes of NAFLD.
Methods
Study population
A total of 263 adults who underwent FibroTouch testing in Shanghai East Hospital from February 2020 to December 2021 were included in this study after obtaining informed consent. This trial was approved by the Ethics Committee of Shanghai East Hospital ([2020] Research Approval No. 009). The inclusion criteria were as follows: (1) age: 18–75 years; (2) sufficient information such as history, clinical manifestations, physical signs and laboratory examination was available from the medical record database. The exclusion criteria were as follows: (1) alcohol consumption > 30 g/d for men and > 20 g/d for women; (2) the presence of conditions that may lead to hepatic steatosis, such as the use of special drugs, total parenteral nutrition, Inflammatory Bowel Disease (IBD), hypothyroidism, celiac disease and hypercortisolism; (3) other liver diseases, malignant tumors, active infection and severe mental or cognitive impairments; and (4) pregnant or lactating women.
MetS is a constellation of risk factors, and the main components of MetS were selected as follows in our study: (1) overweight and obesity: body mass index (BMI) > 25.0 kg/m
2; (2) hyperglycemia: fasting plasma glucose (FPG) ≥ 6.1 mmol/L and/or 2 h plasma glucose (PG) ≥ 7.8 mmol/L or previously diagnosed T2DM; (3) dyslipidemia: triacylglycerol (TG) ≥ 1.7 mmol/L and high-density lipoprotein (HDL) ≤ 0.9 mmol/L (male), ≤ 1.0 mmol/L (female); and (4) hypertension: blood pressure ≥ 140/90 mmHg or use of medication for hypertension [
12]. Since NAFLD has become the most common chronic liver disease in metabolic syndrome characterized by steatosis associated with inflammation, fibrosis and current researches demonstrated hyperuricemia promotes the development of steatosis [
13,
14], it plays a more important role than blood pressure in NAFLD, and therefore, in our study, we used uric acid (UA) instead of blood pressure as one component of MetS.
All the participants were divided into NAFLD group and healthy controls according to the controlled attenuation parameter (CAP) of liver fat (CAP ≥ 240 dB/m indicated liver steatosis) by the FibroTouch detected diagnostic system and ultrasound [
15,
16]. In NAFLD group, the patients were subsequently re-divided into NASH at-risk and NAFL group according to the results of blood liver function (alanine aminotransferase (ALT) > 40 U/L or aspartate aminotransferase (AST) > 35 U/L or γ-glutamyl transferase (GGT) > 45 U/L) [
17].
Anthropometric and hematological parameters
The following general details were collected: age, gender, height, weight, past medical, drinking and medication history and BMI was calculated using the equation: [BMI (kg/m2) = weight/height2]. Blood samples were obtained from the antecubital vein of each participant after 8 h of overnight fasting. Routine blood test was measured by LH-750 Blood Cell Analyzer (Beckman Coulter, Fullerton, CA, USA). Serum biochemistry, lipid profile, FPG and glycated hemoglobin (HbA1c) were measured by automated biochemical analyzer (AU5400; Olympus Corporation, Tokyo, Japan). According to the manufacturer’s instructions, serum BMP9 level was detected by enzyme-linked immunosorbent assay (ELISA) kit (Hengyuan Biotechnology co., Ltd, Shanghai, China) and calculated using a standard curve. Three separate plates (3 technical replicates) were run for each sample.
Liver fat content and liver elasticity test
The liver stiffness measurement (LSM) values expressed in KPa, detected by FibroTouch (FIBROTOUCH-FT5000, HISKY MED, WX, China), were used to determine the degree of liver fibrosis, and the CAP values were expressed in dB/m to determine the degree of steatosis. FibroTouch was operated by an experienced ultrasound physician and applied to detect the liver stiffness and steatosis of all participants. An ultrasonic probe was placed in the area covering the seventh to the ninth intercostal spaces from the right anterior axillary line to the midaxillary line of the participants. Marks were left on uniform hepatic tissues of proper thickness that were free from either artery, bile ducts or cysts. The device was then switched to elastography mode, and participants were asked to hold their breath for 3 s. When the image became stabilized, LSM was conducted. The measurement was taken 10 times, and the median was considered the final value and was expressed as the value of elasticity (KPa). A reliable LSM was defined as more than 10 valid shots, a success rate of at least 60% and an interquartile range (IQR) < 33%. The fat attenuation parameter (FAP) value was also acquired. The results were defined as follows: LSM > 8.0 kPa implied the presence of abnormal liver elasticity; and CAP ≥ 240 dB/m implied hepatic steatosis: mild: 240 ≤ CAP < 265 dB/m; moderate: 265 ≤ CAP < 295 dB/m; and severe: CAP ≥ 295 dB/m.
Statistical analysis
Continuous variables were described as mean ± standard deviation (SD) according to their data distribution. Independent samples t-test was selected for comparison between two groups. Multiple groups of independent samples were compared by one-way analysis of variance, and based on the results of one-way statistical analysis, a logistic regression analysis model was constructed to explore relevant risk factors, and p < 0.05 was considered to be statistically significant. All data analysis and graph drawing were performed by Statistical product and service solutions (SPSS) Statistics 20, R 3.6.3 (IBM, Armonk, NY) and GraphPad Prism 8 (GraphPad Software, San Diego, USA).
Discussion
NAFLD is a major chronic liver disease closely associated with obesity, insulin resistance (IR), T2DM, dyslipidemia and other MetS components, and it is the hepatic manifestation of MetS. The pathogenesis of NAFLD is complex whereby the historical two-hit hypothesis is now replaced by a multiple parallel hits hypothesis also considering the impact of an altered adipokine secretory pattern, inflammation, gut microbiota, nutritional factors, genetic and epigenetic factors, and IR on the development and progression of NAFLD [
20]. NAFLD and MetS are closely related in both phenotype and pathogenesis. Coexistence of NAFLD and MetS should not be expected in every patient, one may precede the other, and common therapeutic modalities are beneficial for both entities.
BMP9, a hepatic cytokine that regulates blood glucose and lipid metabolism, is closely related to the pathogenesis of NAFLD and has a certain protective effect on NAFLD. It can inhibit hepatic lipid deposition, transform white adipose tissue (WAT) to brown adipose tissue and play a leptin-like role such as increasing energy expenditure to exert lipid-lowering effect [
21]. Furthermore, BMP9 can play a hypoglycemic role in regulating glucose metabolism by increasing insulin sensitivity and stimulating the synthesis and secretion of insulin release [
8]. Kuo et al. found that BMP9 induced WAT browning, inhibited high-fat diet-induced obesity and improved obesity-mediated IR and NAFLD in mice [
22]. Our previous study also found the expression of BMP9 was decreased in NAFLD mice and exogenous supplementation of BMP9 improved the phenotype of NAFLD [
11]. However, the relationship between serum BMP9 and the severity of MetS, NAFLD has not been explored thoroughly so far.
In the present study, we observed a negative correlation between the concentration of BMP9 and several indicators of glycolipid metabolism, including TG, FPG, HbA1c and BMI, while it was positively associated with HDL. This was consistent with other studies. Xu et al. found in middle-aged and older populations [
23] as well as patients with T2DM and NAFLD [
10], circulating BMP9 levels positively correlated with HDL and negatively correlated with markers of adiposity such as waist-to-hip ratio (WHR) or BMI, parameters of glucose metabolism (FPG and HbA1c) and TG. Animal studies also showed that upregulation of BMP9 levels inhibited the expression of lipogenic genes, ameliorated triglyceride accumulation in liver, decreased serum TC and TG [
24], improved glucose tolerance, decreased FPG and alleviate IR in mice [
11].
NAFLD is closely associated with obesity, and BMI is currently the most widely used anthropometric measure for assessing obesity worldwide and the most critical risk factor for NAFLD [
25]. In our study, we discovered the risk of developing NAFLD increased along with elevated serum BMP9 and BMI, which is consistent with Younossi et al [
26]. They also found that the prevalence of NAFLD was positively correlated with an increase in BMI. It is noteworthy that there was a significantly higher risk of NAFLD observed in men compared to women in our study. This has been confirmed in other studies that the prevalence of NAFLD is higher in men compared to women, and in women, the prevalence tends to increase after menopause, potentially due to the possible protective role of estrogen [
27].
NAFLD has insidious progression and lacks specific symptoms. Noninvasive assessment of the severity of NAFLD is a general trend. Serum transaminase is often used as an index to detect NASH clinically, however, it does not reflect the course and prognosis of NAFLD accurately [
28]. Another frequently used biomarker for NASH is cytokeratin 18 (CK18), which exhibits an overall sensitivity of 66% and specificity of 82% [
29]. Nevertheless, inflammatory markers such as ferritin, high-sensitivity C-reactive protein (CRP), tumor necrosis factor and interleukins were demonstrated poor accuracy and specificity in recognizing NASH [
30]. FibroTouch is now widely used in clinical practice as a noninvasive assessment method with excellent diagnostic performance for hepatic steatosis and fibrosis. The sensitivity of CAP values in diagnosing hepatic steatosis at stages S1, S2 and S3 was 81.5%, 86.7% and 100.0%, respectively, and the specificity was 80.6%, 91.7% and 96.5%, respectively [
31]. Additionally, LSM outperformed other noninvasive fibrosis indices, namely fibrosis 4 (FIB-4), aspartate aminotransferase-to-platelet ratio index (APRI) and γ-glutamyl transferase-to-platelet ratio index (GPRI), demonstrating significant superiority, albeit second to liver biopsy [
32]. Therefore, we employed FibroTouch in place of liver puncture to measure the degree of steatosis and fibrosis. In our research, we observed an upward trend in transaminases, CAP and LSM in patients with NAFLD compared to the healthy population. Moreover, we identified significant differences in BMP9 concentration among different liver enzyme levels, CAP and LSM. The current study also showed serum BMP9 levels differentiated NASH at-risk (58.13 ± 2.82 ng/L) from the other groups: healthy control (70.32 ± 3.70) and NAFL (64.34 ± 4.76 ng/L,
p < 0.0001). Notably, BMP9 is widely involved in the function of various receptors associated with liver fibrosis promotion, downstream signaling molecules and the expression of target genes in hepatocytes [
33], thus making it a valuable serum diagnostic indicator [
34]. All these findings suggest that BMP9 has the potential to serve as a biomarker in identifying at-risk NASH patients.
Uric acid levels may play a role in MetS, and high uric acid level is independently associated with all components of MetS such as hypertension, obesity, high triglycerides, low HDL and elevated FPG [
23]. In addition, another meta-analysis of prospective studies concluded that there is an independent linear relationship between increased UA levels and the occurrence of MetS [
35,
36]. Therefore, in our study we used UA instead of hypertension as one of the MetS parameters, which are in line with previous studies [
37]. Our results showed circulating BMP9 levels reduced with the number of MetS components. Xu et al. [
23] also found that plasma BMP9 concentrations were significantly associated with MetS and circulating BMP9 levels reduced progressively with an increasing number of MetS components.
Although our study observed close association of serum BMP9 with the severity of MetS and NAFLD, there are some limitations including the small sample size, and the cross-sectional study design did not allow us to establish causality between serum BMP9 and NAFLD development. Future longitudinal studies with larger and more diverse participants are needed. Additionally, the absence of a liver biopsy-proven NAFLD cohort limits our comprehension for the connection of NASH and BMP9. Nonetheless, our findings suggest that BMP9 is highly has the potential to be used as a noninvasive biomarker for monitoring the severity and progression of NAFLD and MetS.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.