According to the International Diabetic Federation diabetes atlas (10th edition, 2021), the global prevalence of diabetes has reached 10.5% and is expected to rise to 11.3% in 2030 and 12.2% in 2045. Approximately 6.7 million people (20–79 years old) died of diabetes or its complications in 2021, accounting for approximately 12.2% of all deaths worldwide (1, 2). According to an epidemiological survey, the prevalence of diabetes mellitus in mainland China was 12.8% in 2017, indicating that diabetes mellitus has become an important public health problem in the world, especially in China (3). Therefore, identifying the high-risk population of diabetes is important to solving the public diabetes burden.

Numerous observational studies indicated that liver enzymes such as aspartate aminotransferase (AST), alanine aminotransferase (ALT), and gamma-glutamyl transferase (GGT) were positively correlated with the risk of diabetes mellitus and metabolic syndrome, especially serum GGT (4–11). The liver plays an important role in maintaining the homeostasis of glucose metabolism, and liver enzyme concentrations are positively correlated with liver damage. Serum ALT and AST are released from the injured hepatocytes. GGT widely exists on the surface of cell membranes and is a key participant in the metabolism of glutathione. Glutathione is an important cellular antioxidant, which is closely related to inflammation and oxidative stress in tissues (12, 13), and oxidative stress and inflammation play an important role in the development of insulin resistance (14).

Obesity refers to a state that is obviously overweight, caused by the excessive accumulation of body fat (15). Obesity can promote global inflammation and peripheral insulin resistance, increasing the prevalence of various metabolic diseases, such as diabetes and non-alcoholic fatty liver disease (16, 17). According to a 2015–2017 national cross-sectional study in China, the incidence of overweight and obesity is 30.1% and 11.9%, respectively, and the proportion of the population with a body mass index (BMI) ≥30 kg/m² is 6.3%. The prevalence of diabetes mellitus (DM) in BMI <25 kg/m², 25 kg/m² ≤ BMI < 30 kg/m², and BMI ≥30 kg/m² is 8.8%, 13.8%, and 20.1%, respectively (3).

To better assess the correlation between liver enzymes and DM risk, obesity classification should be considered. However, limited studies showed whether obesity plays a role in the correlation between liver enzymes and the risk of diabetes mellitus. So, we conducted a large population cross-sectional study in Guangzhou, China, to clarify the relationship and provided guidance for identifying a high-risk population of diabetes mellitus.

In this manuscript, we focused on the relationship between liver enzymes and DM risk in different obesity subgroups in a cross-sectional study. We got the cutoff points of BMI and WC according to the two-slope linear regression model, and restricted cubic splines were used to analyze the correlation between liver enzymes and DM risk in different classes of the obese population divided according to cutoff points. Our results showed that only serum GGT level was correlated with DM risk instead of ALT and AST. It is worth noting that the correlation disappeared in the subgroup with WC ≥98.99 cm.

For obesity categories, previous studies conventionally grouped individuals according to the WHO criterion or percentile (20, 21). Our obesity classifications were not based on these criteria, considering that the criteria cannot reflect DM risk accurately. We conducted a two-slope regression model to find out the cutoff point where the DM risk increased faster. The cutoff points were 98.99 cm and 30.55 kg/m², respectively. Interestingly, the BMI cutoff point was similar to the WHO obesity classification (BMI ≥ 30 kg/m²); however, the Chinese always use the Asian obesity classification (BMI > 28 kg/m²). Moreover, the WC cutoff point was much higher than the abdominal obesity classification (WC > 90 cm for men and WC > 85 cm for women). We used the common criteria as the cutoff point to repeat the analysis, and the results showed a significant association between GGT and DM risk in all groups ( Table S1 ). The results implied that different regional populations with different ages may need more special criteria for DM risk and other obesity complications.

Our study showed that after grouping by BMI or WC, there was no significant correlation between ALT or AST and DM risk, and only GGT was correlated in some subgroups. The results were consistent with previous studies (4, 22–24). The results suggested that GGT levels were more significant for evaluating DM risk than other liver enzymes. However, there were other studies that showed conflicting results. A population study in Europe showed that ALT was significantly correlated with impaired glucose tolerance (IGT) but not with GGT or AST (25), suggesting that ALT and prediabetes are closely related. However, the study had limitations. The sample size of that study was 157, and it only adjusted for age and gender. A population study in China showed that GGT levels were not significantly associated with HOMA-IR in non-alcoholic fatty liver disease patients (NAFLD). The study only included 212 patients. The sample size may not be powerful and the HOMA-IR was not equal to DM risk (26). Another prospective study in Korea that included 548 patients showed that GGT levels were associated with DM risk only in women but not in men. Our study population mainly consisted of women, which may lead to conflicting results (4). A Mendelian randomization study showed that genetically higher ALT was associated with a higher DM risk but not GGT (27). However, the Mendelian randomization study was based on genes, not considering environment function and compensation function, and we cannot exclude the possibility that GGT was associated with DM risk. GGT is a protein that exists widely on the cell membrane and is closely related to the metabolism of glutamate in cells. GGT regulates the level of oxidative stress in cells and tissues, which is closely related to diabetes mellitus risk (14). In this respect, GGT has a more reasonable metabolic relevance, which partly explains why GGT in liver enzymes has a more significant diabetes risk association than ALT or AST.

We observed that GGT was associated with DM risk in different subgroups, but this association disappeared in the group with WC ≥98.99 cm. The result showed that the association will be disturbed for severe abdominal obesity individuals. A 15,792 middle-aged community-based prospective cohort study in the USA showed that GGT was associated with DM risk, after adjusting for body mass index, WC, and other confounding factors (28). The result was inconsistent with ours to some extent, and for this reason, we cannot get the association in the group with WC ≥98.99 cm. The contradiction may rely on WC stratification. The possible mechanism underlying the phenomenon was unclear and may be proposed as follows. Severe abdominal obesity may be a confounding factor for GGT and DM risk, which was similar to other points (29). Abdominal obesity with increased visceral fat is closely related to systemic inflammation and oxidative stress, and abnormal oxidative stress usually leads to increased serum GGT concentrations (12). So, patients with severe abdominal obesity, who are at high DM risk, may be accompanied by elevated serum GGT concentration raised from abnormal oxidative stress and led to the confounding associations between GGT and DM. The other explanation was that the risk factors for the development of diabetes mellitus in the population with severe abdominal obesity were not similar in other groups. Furthermore, the risk of DM reflected by GGT may not play a major role in a population with severe abdominal obesity, and GGT will also show different associations with DM risk in a population with different pathophysiological states.

Our research also has certain limitations. The major limitation is that the data of the ultrasonic diagnosis of fatty liver were incomplete, considering that the liver enzymes were highly associated with fatty liver and we could not exclude the confounding factor. Secondly, our study has a cross-sectional design. Causal inferences could not be drawn between serum GGT and DM risk among different subgroups, so more in-depth prospective studies may be needed to prove it. Thirdly, in the subgroup of people with WC ≥98.99 cm, the disappearance of the relationship may lie in the sample size of the subgroups (n = 144), and we should get more samples to ensure the associations. Lastly, the population in this study consists of middle-aged individuals aged >40 years old in South China and cannot represent the younger population and the North China population. Furthermore, the population mainly consists of women, partially because we invited residents over the age of 40 years and women are predominant in this age range in China. The pathophysiological mechanism of this research still needs to be further clarified.

In summary, in this cross-sectional study of a large population, we found that increased serum GGT levels are correlated with the risk of diabetes, and showed different effects in subgroups with different BMI or WC values. Serum GGT may be a better reference marker for predicting the risk of diabetes mellitus than AST or ALT in clinical practice. Individuals with elevated liver enzymes, especially GGT, should be alert to the risk of diabetes mellitus. However, GGT has limitations on DM risk in a population with severe abdominal obesity. For this population, even elevated GGT cannot reflect the risk of diabetes mellitus, and other biomarkers should be considered.

Our study suggests that serum GGT levels have greater reference significance than AST or ALT for the risk of diabetes in the middle-aged population. Moreover, GGT levels correlate with DM risk except for those with severe abdominal obesity. In clinical practice, GGT should be combined with WC to determine DM risk.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Institutional Review Committee of Sun Yat-sen Memorial Hospital affiliated to Sun Yat-Sen University. The patients/participants provided their written informed consent to participate in this study.

All authors listed have made a substantial, direct, and intellectual contribution to the work, and approved it for publication.

This study was supported by grants from the National Science Foundation of China (U20A20352) and Guang Dong Clinical Research Center for Metabolic Diseases (2020B1111170009). The funders were not involved in research design, data collection and analysis, preparation of the manuscript, or decision to publish.

We are indebted to the participants in this study for their continued and excellent support and our colleagues for their valuable assistance.

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.