This study revealed that the prevalence rate of both IFG and DM was 27.27% in patients with HBV infection, and the abnormal glucose metabolism rate was higher in CHB patients with LC or HBeAg (-) status.
Some of the findings in this study were similar to those in previous literature, which reported that the prevalence of DM was significantly higher in the HBV-infected population, 4–11 particularly in those with high viral load, with a long duration of CHB, with cirrhosis, 4, 6,7−10 or of Asian American race.8 A meta-analysis reported that the summary OR of the risk of DM for HBV patients was 1.99 (95% CI, 1.08–3.65) when compared with non-HBV individuals. 4 The prevalence rates of both IFG and DM in this study were higher than the 12.5% for DM and 7.8% for IFG in adults with CHB previously reported in a large HBV-infected multiethnic cohort study. 8 This might be attributed to the differences in the study population: the populations of the two former studies were Chinese, while the population of the latter study was American.
This study also reported a higher abnormal glucose metabolism (including IGR and DM) rate in CHB patients with LC. The prevalence rates of IGR and DM were 43.90% and 36.59% in CHB patients with LC in comparison to 18.42% and 13.16% in CHB patients without LC. This finding is consistent with those from a study that reported7–10, 17 that the odds ratios for DM in chronic hepatitis B cirrhosis patients were 1.74, 1.76 and 2.317 (95% confidence interval: 1.43–2.13, 1.44–2.14, 1.528–3.513, respectively) when compared with noncirrhotic chronic hepatitis B patients. 7–10, 17 However, the DM prevalence was higher than in other cross-sectional studies reporting a prevalence of 22.2% of DM among CHB patients with liver cirrhosis.17 The development of cirrhosis may increase the incidence of DM.7–10, 17
In addition, it is proposed that this is the first study to indicate that the abnormal glucose metabolism rate was higher in CHB patients who were HBeAg negative, with IGR and DM prevalence rates of 31.94% and 34.72% in patients who were HBeAg negative in comparison to 17.39% and 21.74% in patients who were HBeAg positive. Previous studies suggested that the HBsAg status could influence glucose metabolism, and maternal HBsAg carriers were an independent risk factor for gestational diabetes mellitus (GDM). 22 The incidence of GDM in pregnant women who were HBsAg positive was 6.48%, which was higher than the 3.41% incidence rate in those who were HBsAg negative.23 However, there was no significant association between the incidence of DM and viral load, HBeAg carrier status, or other HBV markers in pregnancy, 22–23 and there is no literature reporting glucose metabolism in HBeAg carriers.
Further analyses demonstrated that abnormal glucose metabolism manifested as elevated FPG levels and significantly decreased islet β cell function, as indicated by the HOMA-β values, but it did not manifest as insulin resistance, as indicated by the HOMA-IR values. We found that under the same glucose metabolism conditions, the FPG level of the CHB group was continuously higher than that of the non-HBV group; it was more than 6.0 mmol/L under IGR conditions and more than 7.0 mmol/L under DM conditions, while in the non-HBV group, it was lower than 6.0 mmol/L under all three glucose metabolism conditions. Simultaneously, the HOMA-IR value of the CHB group was consistently lower than that of the non-HBV group. However, under NGT conditions, the HOMA-β values of the HBV group was 47.53 mIU/mmol, only half of the reference value (100.00 mIU/mmol), and only one third of the non-HBV group 124.19 mIU/mmol, and these values continuously decreased with the deterioration of glucose metabolism.
A multicenter randomized parallel-group trial showed that the HOMA-β values in patients newly diagnosed with DM was only half the reference value (100 mmol*mIU/L2); it decreased progressively at a rate of 4.5% annually and deteriorated with the course of the disease.24 A new staging method for NGT, IGR and DM was proposed according to the function of β cells: normal phase of β cell function, compensatory phase of β cell function, decompensatory phase of β cell function, and failure phase of β cell function in the general population. The compensatory secretion of β cell function occurs in individuals with NGT and IR and reaches the peak of compensatory secretion. The decompensatory phase of β cell function has occurred in individuals with prediabetes or IGR.25 In recent years, most studies have confirmed that not all individuals with NGT were healthy, and some presented with IR.26 The risk of developing prediabetes and/or DM significantly increased in individuals with NGT but IR and dysfunction of β cells. 26 Therefore, the HOMA-β values of CHB patients under NGT conditions was even lower than that of those newly diagnosed with DM. The β cell function of the CHB population deteriorates directly to the decompensatory and failure phases, without undergoing normal and compensatory phases, even under NGT conditions, and this change leads to higher FPG levels and a high prevalence of IGR and DM in the CHB population. From this, we could conclude that the evident increase in the FPG levels in patients with CHB was associated with worsening β cell function but not insulin resistance.
In this study, it was also demonstrated that most of the HBV serological and virological indicators had negative effects, while LC, HBeAb levels and markers of liver inflammation and fibrosis had positive effects on both glucose metabolism and FPG levels. The main contributing factors for glucose metabolism and FPG levels were HBeAg (-) and HBeAb levels. However, HBV serological and virological indicators had no direct effects on islet β cell function, as indicated by the HOMA-β values. Therefore, we speculated that HBV indirectly affected islet β cell function through certain mechanisms.
Fundamental studies have found that hepatitis B virus infection could increase the production of tumor necrosis factor (TNF), especially in HBeAg-negative patients. 27 The overproduction of TNF could decrease the phosphorylation of insulin receptor substrates 1 and 2, inhibit phosphoinositol 3-kinase and protein kinase B, block the phosphorylation of glucose transporter 4, prevent the cell uptake of glucose28 and increase plasma glucose levels. Prostate six-transmembrane protein 2 (STAMP2) is a factor associated with inflammation and dietary adipocyte function and system metabolism. It can be induced by nutrition, feeding, and cytokines, such as TNF alpha, interleukin (IL)-1β, and IL-6, which can inhibit IR in rats. IR and visceral and hepatic insulin signaling disorders were observed in mice lacking STAMP2. In the presence of inflammation and obesity, the increased expression of STAMP2 has protective effects against insulin signaling in the liver.29 Moreover, hepatitis B virus X protein induces liver fat accumulation and IR by reducing the expression of STAMP2. STAMP2 downregulates the insulin-induced phosphorylation of the P3K p85 subunit, protein kinase and the expression of insulin receptor substrate 1, and the posttranscriptional level of insulin receptor substrate 1 plays a role;30 this leads to the increase in blood glucose levels and high abnormal glucose metabolism incidence.
Although those fundamental science studies have confirmed that hepatitis B virus infection could lead to increased hepatic glucose output and IR, they could not explain the decrease in HOMA-β values and FINS levels. Further fundamental science studies are needed to investigate the mechanisms of the hepatitis B virus infection effect on islet β cell function in CHB patients.
To our knowledge, this cross-sectional study was the first to compare the differences in HOMA-β values and FPG levels between CHB patients and non-HBV patients matched according to sex, age and BMI. This is the first study to compare the differences in abnormal glucose metabolism rates between CHB patients with HBeAg (+) and HBeAg (-) status, and it is also the first study to analyze the contributing factors of HBV virological and serological indicators on abnormal glucose metabolism. The results showed that in CHB patients, the FPG levels was higher, while the HOMA-β values was significantly lower. Additionally, the HOMA-IR values was also lower than that of non-HBV patients under the same glucose metabolism conditions. The β cell function of the CHB population deteriorated directly to the decompensatory and failure phases, without undergoing normal and compensatory phases, even under NGT conditions. Furthermore, CHB patients with LC or who were HBeAg negative had a higher abnormal glucose metabolism rate. HBV serological and virological indicators, markers of liver inflammation and fibrosis, and LC could directly affect glucose metabolism. However, the effects on islet β cell function were indirect. Therefore, the increased FPG levels of CHB patients was accompanied by significantly decreased islet β cell function but not insulin resistance. In addition, HBV could directly affect glucose metabolism and could indirectly affect islet β cell function through certain mechanisms.
Our study also has some limitations. The sample size was small, and it was a single-center, retrospective study. A further multicenter, prospective study with a large sample size is needed.
The findings of this study provide a reference to allow clinicians to monitor abnormal glucose metabolism in CHB patients, especially those with LC or HBeAg (-), focus on the protection of islet β-cell function, and avoid the application of insulin secretagogues in CHB patients with abnormal glucose metabolism.