In our current study, approximately 40% of AIH patients had severe liver fibrosis, with MADs present in 87.3% and NAFLD in 23.0% of AIH patients, respectively. In the severe fibrosis cohort, the proportion of AIH patients with DM was higher than that in the mild or moderate fibrosis cohort. Compared with AIH patients without NAFLD, those with NAFLD were older (median age), had higher BMI, and exhibited more severe fibrosis. Interestingly, only male, γ-GT, the presence with MADs and the presence with DM were identified as independent factors associated with severe fibrosis, using multivariable analysis.

The precise underlying mechanism of AIH remains to be explored, but it is speculated that it is due to autoimmune-mediated attacks against hepatocytes with unknown cause(s) [1]. AIH commonly affects middle-aged women [28], as supported by our current study, which showed that the median age of all patients was 54.9 years and the majority were female (82.3%). In addition, we found that approximately 94% of AIH patients had at least moderate liver fibrosis at the time of definitive diagnosis by liver biopsy, and around half of them presenting with severe fibrosis.

In our cohort, we found that the prevalence of NAFLD was 23.0%, but other study has shown a prevalence of 17% [16]. Interestingly, we observed a significant difference in median age between AIH patients with and without NAFLD. Our study utilized a combination of liver histology, ultrasound and CT to detect hepatic steatosis, which may explain the higher prevalence of NAFLD compared to comorbidity rate based solely on liver histopathology [16]. Diagnosing AIH with NAFLD can be challenging because some patients with ANA-positive NAFLD may be misdiagnosed, leading to delayed diagnosis and treatment. In addition, liver biopsy may have certain limitations due to potential sampling errors. Therefore, to diagnose NAFLD in our study, we utilized a variety of examination methods, including pathology, ultrasound, and CT. This approach may have increased the likelihood of detecting NAFLD in AIH patients. Our findings of older age in AIH patients with NAFLD are consistent with previous studies that have identified older age as a characteristic of AIH patients with concurrent NAFLD. Moreover, the distribution of both AIH and NAFLD in adults shows a single peak in patients in their 60s [28, 29], which further supports our observations of older age in AIH patients with NAFLD.

Unhealthy lifestyle habits, such as irregular eating and low physical activity[4], contribute to obesity [6] and are a major cause of the increased prevalence of MADs worldwide. In our study, we observed that AIH patients with NAFLD had higher BMI and a higher proportion of obesity than patients without NAFLD, consistent with previous research linking metabolic syndrome and NAFLD [29]. Previous studies have demonstrated a close association between liver fibrosis and metabolic syndrome in NAFLD [30], with MADs being independent risk factors for liver fibrosis [14]. In our study, we reported a high prevalence of MADs in AIH patients. Although there was no significant difference in the prevalence of MADs between severe and mild/moderate liver fibrosis, we found that the combination of MADs could be an independent factor for severe fibrosis in AIH patients in multivariate analysis. This finding aligns with previous research demonstrating an association between MADs and liver fibrosis in other chronic liver diseases [8]. In addition, we attempted to determine which single factor might be more important in promoting liver fibrosis, further analysis demonstrated that DM is also an independent risk factor. This supports the concept that MADs promote liver fibrosis, and aligns with previous study in other chronic liver diseases, showing that the presence of DM in CHB patients is associated with an increased risk of liver fibrosis, and cirrhosis occurrence [31]. In our study group, the prevalence of MADs in AIH patients was high, and 75.9% of patients were over 50 years old. We have primarily described patients diagnosed with AIH, who, in addition to being older, also have a significant proportion of comorbid metabolic-associated diseases. Therefore, we aim to investigate whether these comorbid conditions affect inflammation and fibrosis in AIH patients. However, this does not allow us to infer that patients with metabolic-associated diseases are at risk for AIH, as the study population differs. We did not assess the incidence of AIH among patients with metabolic-associated diseases, but this is an important research point. In future studies, we could explore whether the prevalence of AIH varies among specific patient populations.

In our cohort, fibrosis and/or cirrhosis is a common feature when a high proportion of patients are diagnosed with AIH. Therefore, it appears to be impossible to distinguish fibrosis caused by AIH or MADs, given the high comorbidity rate of MADs. However, MADs, including hypertension, diabetes, dyslipidemia, and obesity, do not themselves constitute causes of cirrhosis. Cirrhosis, as an outcome of end-stage chronic liver disease, is more commonly associated with metabolic disorders causing NAFLD, which can serve as a foundation for chronic liver disease leading to cirrhosis. In our current retrospective cohort study, relatively few of AIH patients had NAFLD. In addition, our retrospective study primarily focused on patients with AIH to explore the impact of concurrent MADs on inflammation or fibrosis in AIH patients. Due to the specific characteristics of patient medical history data in liver disease at the time, we were unable to collect sufficient data on risk factors related to MADs, such as high-sugar or high-salt diets. However, we plan to address this issue in future studies.

Previous research reports that AIH patients with NAFLD often exhibit mild elevations in hepatobiliary enzymes compared to those without NAFLD [16]. However, we did not find a significant difference in hepatobiliary enzymes between AIH patients with and without NAFLD. Such discrepancy could be attributed to the high degree of liver fibrosis in our cohort, leading to a presentation of chronic hepatitis rather than acute hepatitis. In addition, we observed a significantly higher percentage of patients with severe fibrosis in the NAFLD subgroup than in the non-NAFLD subgroup, which may have contributed to the lack of difference in liver-enzyme indicators between the two groups.

In addition, we found that AIH patients with NAFLD had a higher percentage of severe fibrosis and were more likely to present with cirrhosis. This is consistent with previous research, demonstrating that the coincidence of AIH and NAFLD is more likely to lead to cirrhosis [15]. However, concurrent NAFLD was not identified as an independent risk factor for severe liver fibrosis or liver inflammation in univariate analysis. This finding is consistent with previous research, which found that hepatic steatosis is not associated with liver fibrosis in AIH patients, using non-invasive vibration-controlled transient elastography [32]. However, this may also be due to the relatively small sample size. NAFLD from the current retrospective study was not identified as an independent factor for severe liver fibrosis using restricted multivariable statistical analysis. We will verify this issue in future studies from multiple centers with different genetic as well as environmental backgrounds.

Most cases of AIH present with one or more significant titers of autoantibodies. ANA, smooth muscle antibodies (SMA), and antibodies to kidney microsome-1 (anti-LKM1) are standard autoantibodies. According to the pattern of autoantibodies detected, AIH is classified into two types: 1. ANA and/or SMA antibodies characterize type 1 AIH (AIH-1), which accounts for almost 90% of cases; 2. anti-LKM1 and antibodies to liver cytosol type 1 (anti-LC1) characterize type-2 AIH (AIH-2) [33]. In our current study, the positivity for anti-SMA was very low, i.e., 2.7%, which is much lower than the recommended diagnostic criteria for AIH, where positivity for anti-SMA is in the range of 80% [34]. The reasons may be that some practitioners focus more on ANA antibodies during initial screening, possibly overlooking SMA. In fact, upon reviewing the original data from our current study, the detection rate for ANA was 100%, while for SMA it was only around 84%. Approximately 46 patients lacked SMA detection data, which might have contributed to the lower observed positivity rate in statistics. In addition, all these participating patients were confirmed as AIH based on comprehensive scoring systems, including autoantibodies and typical AIH liver histopathological features in our current retrospective study. Some patients were still diagnosed with AIH based on pathology, IgG levels, and overall scoring, even though ANA antibodies were negative. Thus, we believe that the AIH diagnosis is reliable. The positivity rates of ANA and SMA varied among different populations with AIH in different disease states, which is supported by a previous study on AIH-related acute liver failure patients in the US, where the positivity rate for ANA was 70% and for SMA was 43% [35]. Therefore, different populations may have varying rates of autoantibody positivity. Given that our study is retrospective with a relatively small sample size and liver biopsy being an invasive procedure, patients' consent for liver biopsy may contribute to selection bias to some degree. We will minimize the selection-bias in future studies by including multi-center data from different genetic and environmental backgrounds.

We acknowledge the limitations of our current study. First, the sample size was relatively small, and the cross-sectional study design limits our ability to establish a firm causal relationship. Second, we did not investigate the correlation between MADs or NAFLD and prognosis, which will be explored in future studies. Third, we chose NAFLD, rather than metabolic-associated fatty liver disease (MAFLD), as one of MADs exposure in the study. Because of the concept of MAFLD was defined in 2020 [36], while our current retrospective study focused on the period from 2016 to 2020. Due to lacked data on abdominal circumference, we are unable to test the concept of MAFLD in the current study, but it will be determined in future study.

In summary, our study provides novel insights suggesting that MADs, particularly when combined with DM, are significant risk factors for promoting hepatic fibrosis in AIH patients. These findings underscore the importance of considering and managing metabolic factors in the management of AIH.