With the development of urbanization and the improvement of living standards, the incidence of diseases related to metabolic disorders is steadily increasing, which makes them serious diseases threatening human health (1–3). Metabolic syndrome (MetS) is defined as a group of complex metabolic disorders characterized by insulin resistance, hypertension, atherogenic dyslipidemia and abdominal obesity, etc. Several societies harmonize that the MetS can be defined when any three or more of the following factors are met: (a) elevated waist circumference based on population-and country-specific definitions; (b) elevated triglycerides or being previously diagnosed as hypertriglyceridemia and taking antihypertriglyceridemia medication; (c) reduced high-density lipoprotein cholesterol (HDL-C) or being previously diagnosed as reduced HDL-C and taking medication for reduced HDL-C; (d) elevated blood pressure (BP) or being previously diagnosed as hypertension and taking antihypertensive medication; (e) raised fasting plasma glucose level or being previously diagnosed as type 2 diabetes and taking antiglycemic medication (4, 5). Non-alcoholic fatty liver disease (NAFLD), a kind of metabolic stress liver injury closely related to insulin resistance and genetic susceptibility, has been regarded as the leading chronic liver disease and primary cause of abnormal liver biochemical indexes found in physical examination (6). The prevalence of NAFLD is increasing worldwide, with an average prevalence of about 24% (7). The close correlation between NAFLD and MetS, and the reciprocal causality between them has been reported. Therefore, some of risk factors and serum diagnostic markers for both NAFLD and MetS may be consistent (8, 9). Current literature has shown there are no drugs available for the treatment of MetS and NAFLD. Owing to significant increased incidence of MetS and NAFLD, it’s essential to seek for new therapeutic agents or targets for those.

Serum bilirubin, mainly originating from the catabolism of hemoglobin in senescent erythrocyte, is commonly used as a biochemical index for the diagnosis of hepatobiliary and metabolic diseases. The conjugation between free bilirubin and UDP-glucuronosyltransferase (UGT) 1A1, which catalyzes the transfer of glucuronic acid, leads to the generation of conjugated bilirubin. For a long time, bilirubin has been deemed as a metabolic waste of iron porphyrin compounds, which means no beneficial effects can be provided by bilirubin. However, latest studies have shown that mildly elevated bilirubin, such as that found in Gilbert’s syndrome(GS), may serve as an important endogenous tissue protector. Meanwhile, it can act as a physiological modulator of oxidative stress and chronic inflammation in MetS (10, 11). In a meta-analysis including 9 observational studies, serum bilirubin levels are demonstrated to be inversely associated with adverse metabolic outcomes. Unfortunately, subgroup analysis was not performed considering that fewer studies were included. In addition, the lack of information about direct bilirubin (DBIL) and indirect bilirubin (IBIL) limits the value of that meta-analysis to evaluate which kind of bilirubin is associated with MetS (12). Both MetS and NAFLD seem to be associated with serum bilirubin, including total bilirubin (TBIL), DBIL, and IBIL. And the increase in bilirubin levels has been demonstrated to be negatively correlated with the prevalence of NAFLD (8, 13–16). Nevertheless, the alternative study based on Mendelian randomization analysis did not find a causal relationship between bilirubin levels and the risk of NAFLD (17, 18). Similarly, Bellarosa et al. reported that bilirubin does not provide protection against MetS and NAFLD in children population with severe obesity (8). What is noteworthy is that the association between bilirubin and NAFLD or MetS in normal-weight adults remains controversial. Importantly, there are no meta-analysis evaluating the association between serum bilirubin and NAFLD currently. In this context, we conducted this meta-analysis to clarify the relationship between serum bilirubin levels and the MetS or NAFLD.

This meta-analysis was designed and implemented according to the Preferred Reporting Items for Systematic Reviews and Meta−Analyses(MOOSE) guidelines [Supplementary Material 1], and the search strategy [Supplementary Material 2], eligibility criteria and outcomes had been registered in the PROSPERO database (CRD42021293349).

MetS is associated with an increased risk of cardiovascular disease and all-cause mortality (1). NAFLD, characterized by fat degeneration and accumulation in hepatocyte, is regarded as the “liver manifestation of metabolic syndrome”. Owing to significant increased incidence of MetS and NAFLD (1, 38), it’s essential to seek for new therapeutic agents or targets for MetS and NAFLD. At present, hyperbilirubinemia is considered to play a beneficial protective role in numerous oxidative stress and inflammation-related diseases, such as coronary heart disease, diabetes, and stroke (11, 39, 40). What relations have been existed between bilirubin and MetS or NAFLD, and whether hyperbilirubinemia could lower the risk of MetS or NAFLD, however, needs further discussion.

Except for MetS, this study is the first to systematically review and summarize published studies in order to assess the correlation between bilirubin subtypes and NAFLD through meta-analysis. Serum TBIL is inversely associated with MetS in male and gender-neutral group, but not in female. However, the inconsistency existed for the analysis results derived from the cross-sectional and cohort study when we evaluated the association between TBIL and MetS in male or female group There were 7 cross-sectional studies in both male and female group, while 5 cohort studies were included in male and 2 cohort studies in female, respectively. Thus, we speculated that the reason for the inconsistency may be ascribed to the differences in the number of cross-sectional and cohort studies. Fewer cohort studies might bring about the inconsistency. DBIL exhibits an inverse association with MetS, regardless of gender. IBIL displays a negative correlation with MetS in females but not in males. No stastistical correlation is found between TBIL and NAFLD. DBIL is negatively correlated with NAFLD in male subgroup. IBIL manifests an inverse association with NASH in NAFLD patients.

Bilirubin and its precursor biliverdin can increase the antioxidant activity of vascular endothelial cells (41). Bilirubin can also increase insulin sensitivity by regulating levels of cholesterol metabolism, adipokines and peroxisome proliferator-activated receptor γ (PPARγ) (39). What’s more, it can selectively bind to proliferator-activated receptor α (PPARα), causing the decrease in lipid accumulation by increasing the number and function of mitochondria (40, 42, 43). In this analyses, an inverse association was found between TBIL levels and MetS among male and gender-neutral group. In females, TBIL tends to be negatively associated with MetS, although no statistical correlation is found between them(P=0.06). In addition, an inverse association between TBIL and MetS exists in cross-sectional subgroup, but not in cohort subgroup. Insufficient number of included studies in cohort subgroup could affect the accuracy of overall results. Therefore, in line with previous reports, we confirm a protective role of TBIL in MetS (12).

An obvious negative correlation between DBIL and MetS was revealed in all subgroups, regardless of gender and study type. Moreover, correlation between DBIL and MetS is closer than that between IBIL and MetS. As we know, DBIL tends to build looser bound with albumin than IBIL. Hence, it is easier for DBIL to separate from albumin than IBIL. As a result, DBIL might directly act on target organs and molecules (44). Collectively, DBIL may possess better prognostic value than IBIL according to the analysis conducted by we and other investigators (32, 33).

In general, average bilirubin levels in males are slightly higher than those in females. GS, an inherited metabolic liver disease characterized by IBIL elevation, is also most frequently occurs in young males. This discrepancy may be explained by a gap in life style and the effect of sex hormines on the glucuronic acid (45, 46). Hwang et al. demonstrated that all 3 types bilirubin levels are inversely relevant with MetS in females, while DBIL exhibits significant inverse association with MetS in males, after adjusting for the confounding variables including lifestyle (32). Another investigation indicated that all bilirubin subtypes are negatively correlated with MetS in Korean men and women. Nevertheless, this significant inverse correlation between TBIL or IBIL and MetS vanishes according to the adjusted multivariate analysis model (adjusting for age, smoking status, alcohol consumption and so on) (33). In this analysis, we adjusted for the potential confounding factors such as age, gender, life styles, etc. Our data showed that all the studies type of DBIL were inversely related with the MetS in male whereas the IBIL was only inversely related with the MetS in female. These results likely indicate that DBIL is more related to the MetS than the other subtypes of bilirubin in male, and the protective effect of IBIL against MetS is more pronounced in female group. This is partially consistent with Hwang’s findings (32). Apart from uridine diphosphate-glucuronosyltransferase (UGT1A1), the protective effect of bilirubin may be also related to other metabolic enzymes that can regulate the bilirubin metabolism, such as heme oxygenase (HO). Bilirubin is produced under the action of HO, the rate-limiting enzyme of heme catabolism. Downregulation of HO activity inhibits bilirubin production (47). Notably, the gender difference exists in HO activities, which is related to oxidative stress (48) or high iron storage (49). This finding can be used, at least in part, to explain the reasons for the different effects of bilirubin on men and women. However, the exact mechanism is still needed to be explored. Besides, limited number of included studies may reduce the analysis accuracy.

NAFLD does not only refer to the fat accumulation in the liver caused by excessive free fatty acids, but also involves numerous metabolic problems such as oxidative stress, insulin resistance, and mitochondrial dysfunction. Furthermore, a wide spectrum of histological lesions ranging from pure hepatic steatosis to NASH are included in NAFLD. Various laboratory and clinical studies have demonstrated that bilirubin most likely reduces hepatic lipid accumulation by increasing PPARα activation and inhibiting PPARγ transcriptional activity in humanized mice with the Gilbert’s polymorphism (HuUGT*28) or humans with GS (39, 40, 50, 51). According to our analysis, the negative correlation between DBIL and NAFLD only exists in male subgroup, while no significant associations between all bilirubin subtypes with NAFLD is found in gender-neutral group. Furthermore, DBIL rather than other bilirubin subtypes manifests significiant inverse association with MetS or NAFLD in male population. Although no significant relation between IBIL and NAFLD is noticed in health screening population, an inverse association is reported between IBIL and NASH occurring in NAFLD patientsis. These findings are not compatible with the results provided by Luo et al. We think this inconsistence can be ascribed to the following factors. In Luo’s study, NAFLD was diagnosed based on liver ultrasonography rather than liver biopsy which is regarded as,the gold standard for NAFLD diagnosis (52, 53). As a result, mild fatty liver might be missed considering the insensitiveness originating from ultrasonography. In addition, Luo’s study mainly enrolled mid-aged adults with lower incidence of NAFLD, therefore bringing about the risk to draw an biased conclusion (17). In this context, three separate investigation diagnosed NAFLD based on liver biopsy results, and the authors found that IBIL is inversely associated with the severity of liver damage in NASH patients (15, 16, 37). Oxidative stress has been documented to promote the progression from hepatic steatosis to NASH (38, 53). In view of the potent oxidation resistance, it is conceivable that IBIL may provide a protective effect through antioxidant activity in lipotoxic diseases such as NAFLD. Even through, the association between serum IBIL and NAFLD and the underlying mechanism behind this association still need to be explored. Nowadays, studies on association between TBIL or DBIL and NASH in NAFLD patients is insufficient.

Women at reproductive period have a different metabolic status from those at post-menopause. It may reflect the effects of decreased estrogen levels, which have a certain impact on lipid metabolism and insulin resistance. In addition, estrogen deficiency hastens the development of hepatic steatosis and the progression of hepatic fibrosis (53, 54). Serum bilirubin levels in post-menopause may be higher than pre-menopause, which is also related to estrogens deficiency (45, 46). Elevated bilirubin levels have been reported to be closely related to decreased prevalence rate of MetS or NAFLD (12, 13). And our results partially support this finding. Nevertheless, elevated bilirubin levels may not counteract the effect of estrogen deficiency in postmenopausal women. Thus, the prevalence rates of MetS and NAFLD are significantly higher in postmenopausal women than pre-menopausal women (1, 2, 7). For premenopausal women, not only bilirubin but also estrogen can protect from developing MetS and NAFLD. Because of the lack of detailed information for determining whether a female participant is in post-menopause or not, it is unfeasible to conduct subgroup analysis based on this factor. Further research is needed to be carried out to clarify this issue.

Why the association is inconsistent between serum bilirubin and MetS or NAFLD? The possible reasons are as follows: Primarily, except for the liver, other factors such as other tissues, gene, et. are also involved important role in the pathogenesis and development of MetS or NAFLD (55). For instance, steatosis in PNPLA3-associated NAFLD is not accompanied by features of MetS, while PNPLA3-uncorrelated NAFLD closely resembles MetS with regards to its causes and consequences (56). There is a basic research showed that bilirubin deficiency renders mice susceptible to hepatic steatosis in the absence of insulin resistance. It adopted a kind of contrarian strategy to prove that the pathogenesis of MetS and NAFLD is not exactly identically (57). In addition, the variety of definitions of MetS and the variety of diagnostic methods for NAFLD in different studies is also make it challenging that assess the consistent-association of serum bilirubin with MetS and NAFLD.

The inverse correlation between serum bilirubin and MetS or NAFLD suggests that bilirubin might be utilized as a potential and promising strategy to assist in lowering the risk of developing MetS and NAFLD. Exogenous bilirubin supplement is the most direct way to prevent the occurrence of MetS and NAFLD. So far, several studies have attempted to carry out targeted therapy for cancer, inflammation and vascular diseases utilizing bilirubin nanoparticles (BRNP) or bilirubin coated stents (58). And the preliminary efficacy is promising. Moreover, basic research shows that BRNP reduces diet-induced hepatic steatosis (59). On the other hand, increasing endogenous bilirubin production is also a feasible treatment strategy. In this regard, inducing “iatrogenic Gilbert syndrome” by uricosuric drug-probenecid has been demonstrated to be capable of reducing the liver gluconaldehyde acidification activity followed by the increase in serum bilirubin (60). Curcumin supplemented by diet has been documented to increase bilirubin levels through targeting HO-1. These strategies have been proven to be effective and safe in vitro and in vivo. Therefore, increasing bilirubin levels may be an advantageous treatment strategy for MetS and NAFLD (47).

Although the stratification has been executed as far as possible, the shortcomings exist in this meta-analysis. Firstly, the absence of RCTs or prospective cohort studies brings down the credibility of analysis results. Secondly, the classification criteria for serum bilirubin are not identical among different studies, which may partially account for the existence of heterogeneity. Thirdly, insufficient studies on the association between TBIL or DBIL and NASH in NAFLD patients that impairs the credibility and clinical value of this analysis. Fourthly, the studies on MetS were all conducted in Asian countries included China, Korea and Japan. There are no obvious difference in geographical location, dietary patterns, figures, and life styles among these coutryies. Thus, the dietary patterns may have little impact on the results related to MetS in this meta analysis. In terms of NAFLD, four studies derived from Western countries and five studies derived from Eastern countries. Nevertheless, it is unfeasible to conduct subgroup analysis in view of lacking detailed information on diet. Fifthly, in terms of age, most studies included in our meta analysis refer to populations at all ages. However, most original studies had adjusted age as a covariate. For the remaining studies, the lack of detailed information on age makes it infeasible to carry out subgroup analysis on age. Meanwhile, lacking of detailed information on menopause, it is infeasible to conduct subgroup analysis on pre or post-menopause. Finally, follow-up time for each included study is inconsistent or unclear, therefore underscoring the incidence of MetS or NAFLD. Consequently, more high-quality, large-scale, prospective and long-term follow-up studies are urgently needed.

In brief, our meta-analysis indicates that serum TBIL and DBIL levels, especially serum DBIL levels, supporting an inverse connection with MetS, Moreover, serum IBIL could decrease the onset of NASH in NAFLD patients. Therefore, appropriately elevated serum bilirubin levels seem to reduce the risk of MetS and NAFLD. Regulation of bilirubin metabolic pathways may be a potential strategy and exogenous bilirubin supplement may be a medicine to assist in lowering the risk of developing MetS and NAFLD. Bilirubin is still far from being used in the clinic at present. Large-scale prospective and high-quality animal or clinical studies are required to establish to investigate the association and potential prevention of bilirubin on MetS or NAFLD.

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

CL and ZY wrote and amended the original draft. LB and WH participated in extracting and analyzing the data. ST and WZ searched literature and produced the tables, figures. XC, ZH and ZD gave critical revisions, and final approval of the article. SZ contributed to design the study, interpretation of data, and the final approval of the article. All authors contributed to the article and approved the submitted version.

The current study was supported by a research grant from Beijing Municipal Administration of Hospitals Clinical medicine Development of Special Funding Support (ZYLX202125), Capital Health development Scientific Research project (2022–1–2182), Natural Science Foundation of Beijing Municipality (7202068), Beijing Advanced Innovation Center for Big Data-Based Precision Medicine (1212040205).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.