Diabetes mellitus, an enormous public health concern, imposes staggering health and financial burdens worldwide (1, 2). There are two primary types of diabetes: Type 1 Diabetes Mellitus (T1DM) and Type 2 Diabetes Mellitus (T2DM) (3). Approximately 10% of adults worldwide suffer from T2DM, making it the most prevalent form of diabetes (4).T2DM, characterized by hyperglycemia resulting from disturbed glycemic homeostasis (5), is precipitated by defective insulin secretory responses and action. The liver, which is essential to glucose homeostasis by supplying endogenous sugar, is the first organ to receive insulin following its release from pancreatic islets (6). Moreover, hepatic glucose metabolic disorders could contribute to fast hyperglycemia in diabetic patients (7, 8).

Considering the important role of the liver in the pathogenesis and treatment of diabetes, various liver diseases have been associated with T2DM (9, 10). Previous studies have demonstrated a connection between T2DM and chronic liver diseases (CLDs), such as Non-alcoholic fatty liver disease (NAFLD), hepatocellular carcinoma (HCC) and viral hepatitis (11–13). A large meta-analysis from 156 studies estimated that the global prevalence of NAFLD among patients with T2DM was 65.04% (14). It was reported that coexistence of NAFLD and T2DM is common in everyday outpatient practice (15). There is a bidirectional relationship between NAFLD and T2DM that is confirmed by epidemiological data, clinical picture, diagnosis and pathomechanisms (16). However, the complex causal link between NAFLD and T2DM is still controversial, and it is important to bring the attention of clinicians and researchers to the relationship between these two metabolic diseases in order to prevent adverse clinical outcomes (16). Viral hepatitisis an acute or chronic inflammation of the liver parenchyma caused by viruses. There are 5 common types of viral hepatitis: A, B, C, D, and E, and most deaths from viral hepatitis are due to hepatitis B and hepatitis C (17). It is worth noting that hepatitis B Virus (HBV) and hepatitis C Virus (HCV) infections were reportedly higher among T2DM patients (18). However, other studies reported that the risk of HCV infection was rather low in patients with T2DM (19, 20). The discrepancies may be attributable to the sample sizes, target populations, control sources, potential biases from residual confounding, and reverse causation of the respective studies. These highlight how crucial it is to manage T2DM in CLDs patients and to make a firm determination of the causal relationships between T2DM and CLDs.

Mendelian Randomization (MR), a developing epidemiological technique, examines the effect of genotypic variation (exposure) on a phenotype (outcome) from a genetic standpoint. Using germline genetic variance as instrumental variables (IVs), MR analysis assesses causation between exposure and outcomes, thereby avoiding biases caused by confounding variables or reverse causation (21). Previous MR research assessed the relationship between NAFLD and T2DM in Europeans but skipped East Asians (22). To comprehensively evaluate the role of T2DM in CLDs across both demographics, we conducted a comprehensive two-sample MR study.

To the extent of our knowledge, this is the first study leveraging MR to comprehensively investigate the association of genetic predictors of T2DM with CLDs across various ethnic backgrounds. Our analysis indicates that T2DM was probably linked with NAFLD and viral hepatitis, and possibly associated with HCC, and HBV in Europeans. In East Asians, our findings suggest a significant inverse correlation between T2DM and ALT, AST, and HCV. However, we did not find any causal association between T2DM and HCC, viral hepatitis, and HBV. Figure 4 .

Epidemiological data has revealed that the age-adjusted relative risk of NAFLD is approximately 5.36 times greater in individuals with T2DM as opposed to the general non-diabetic population (45). Prior MR studies also presented evidence supporting correlations between genetic predisposition to T2DM and an escalated risk of NAFLD, particularly within European populations (46). These findings align with our MR analysis, albeit our study was conducted on a larger sample size. Numerous potential mechanisms may underpin the associations such as insulin resistance (IR), altered lipid metabolism, and inflammation (47). T2DM is classified as a chronic low-grade inflammatory disease characterized by the elevated levels of interleukin-6 and tumor necrosis factor (TNF)-α (48, 49). By stimulating cellular kinase and inhibiting kappa B kinase, these inflammatory cytokines can induce IR (50). IR may deliver substrates and precursors (e.g., free fatty acids, glucose, and glycerlol) for de novo lipogenesis and mitochondrial β-oxidation, processes that induce hepatic steatosis and increase the susceptibility to NAFLD (47, 51, 52). Moreover, higher blood glucose levels can activate carbohydrate response element-binding protein (ChREBP) signaling pathway, which stimulates the expression of several glycolytic genes, and ChREBP overexpression induced stearoyl-CoA desaturase 1 expression, increasing liver fat content (50, 53, 54).

NAFLD patients usually have abnormal liver function (55). Previous studies have shown that ALT, AST and AST/ALT ratio can predict NAFLD, and AST/ALT are closely related to IR and T2DM (55, 56). High circulating ALT and AST are widely used proxies of NAFLD, although they are not specific markers of NAFLD (57). Interestingly, our MR studies showed there was an inverse association between T2DM and NAFLD Proxies (ALT, AST) in East Asians. A retrospective investigation conducted in coastal Eastern India revealed that NAFLD patients of Indian descent exhibited a lower body mass index (BMI) and a lower prevalence of diabetes when compared to those in Western populations (58). Hence, we formulated the hypothesis that, with the exception of genetic predisposition and environmental exposures, the lifestyle and body weight in the West may account for a portion of these variations among populations and ethnicities but it remains to be validated.

Regarding the relation between T2DM with HCC, a recent meta-analysis reported that T2DM is associated with a significantly higher risk of HCC (12). Another larger cohort study also showed that the incidence of HCC was three times higher than the general population compared to patients with T2DM (59). Several factors could explain this observation, firstly, long-term T2DM can cause oxidative stress and telomere shortening, which induces DNA damage, apoptosis and chromosomal instability in hepatocytes, increasing the risk of HCC (60, 61). Secondly, IR is highly suspected of being a cancerogenic condition, which may attribute to hyperinsulinemia and increased bioavailable insulin-like growth factor I (IGF-I) (59, 62). In Europeans, our MR study identified a causal effect of T2DM on HCC that was only suggestive; in East Asians, no causal effect was observed. The potential explanation for this association specific to ethnicity is still unknown. The observed discrepancies in outcomes across ethnic groups may potentially be attributed to variations in residual confounding and selection bias, factors that are unavoidable (26, 63). Additionally, environmental, genetic, dietary, and lifestyle factors may account for the inverse result. Thus, additional research is required to examine the ethnic variations in disease risk profiles in order to develop more effective treatment approaches.

Additionally, numerous studies have investigated the correlation between T2DM and viral hepatitis, such as HBV and HCV. HBV infection is the most prevalent chronic viral infection on a global scale, affecting roughly 30% of the world’s population in their lives; over 350 million individuals are chronic carriers of the virus (64). Also, it is estimated that around 71.1 million people are chronically infected with HCV, and that 1.75 million new cases of HCV infection were identified in 2015 (65). Previous study shows that patients with T2DM are more likely to be infected with HBV and HCV (18). This could be attributed to a multitude of possible factors. Firstly, HCV replication may be favored by hyperinsulinemia and/or the increased serum levels of free fatty acids observed in patients with T2DM (13). Then, T2DM has been associated with an immunocompromised condition, resulting in disruptions of immune function that potentially heighten the vulnerability to infection with HBV and HCV (66, 67). Also, it is worth noting that the process may be significantly influenced by interleukin-6, tumor necrosis factor-alpha, and various other immune-mediated pathways (9, 66, 68). Within the European population, this MR study revealed a suggestive association between T2DM and HBV and a positive correlation between T2DM and viral hepatitis. On the other hand, we discovered an inverse relationship between T2DM and HCV and absent relation between T2DM and viral hepatitis, HBV in East Asians. Considering the divergent outcomes observed in Eurasian populations, we posit that the relationship between T2DM and CLDs may indeed exhibit ethnic variations. Certainly, we consider the need for further investigation into potential molecular mechanisms and pathways involved in these ethnic disparities. Future studies, particularly those conducted in diverse ethnic populations, should delve deeper into understanding the intricate interplay between T2DM and CLDs. Our MR study also has certain shortcomings. First, comparatively small numbers of cases, such as NAFLD and viral hepatitis in East Asians, may result in low precision and potentially false negative results, which could influence the different MR results in different ancestral origins. Nevertheless, we have implemented the most extensive database of these diseases available, and additional research encompassing larger case sizes and ethnic backgrounds is necessary to definitively ascertain this correlation. Next, because there were zero SNPs as IVs when we set the threshold at 5e-8 in the GWAS of CLDs, we were unable to perform reverse MR analysis to estimate associations between CLDs and T2DM. Furthermore, the implementation of summarized GWAS precludes the ability to perform subgroup analyses according to gender and age. To mitigate such biases, future studies may employ within-family genome-wide association studies (GWASs) (69).

In summary, this research employed MR analysis to deduce a causal relationship between T2DM and CLDs by utilizing GWAS data from various ethnic populations. Our results showed the positive association between T2DM and NAFLD, viral hepatitis, HCC and HBV infection but not with HCV infection in Europeans. In contrast, in East Asians, T2DM was negatively associated with NAFLD and HCV but was not associated with HCC, viral hepatitis or HBV. The result revealed varying causal associations between T2DM and CLDs in East Asians and Europeans. Further clinical trials should be conducted in individuals of various ancestral origins to investigate the interaction between T2DM and CLDs in order to generate novel concepts for early screening and prevention of CLDs in patients with T2DM, as suggested by these findings.

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

YZ: Conceptualization, Formal Analysis, Writing – original draft, Writing – review & editing. DL: Data curation, Methodology, Software, Writing – original draft. HS: Data curation, Methodology, Writing – original draft. WLu: Data curation, Software, Writing – original draft. JQ: Formal Analysis, Writing – original draft. SW: Project administration, Resources, Writing – review & editing. YG: Methodology, Writing – original draft. RL: Project administration, Writing – review & editing. FH: Resources, Writing – review & editing. JL: Data curation, Formal Analysis, Writing – review & editing. WLi: Conceptualization, Project administration, Validation, Writing – review & editing. FW: Conceptualization, Project administration, Supervision, Validation, Writing – review & editing.