Liver Enzymes are Associated with Hyperglycemia in Diabetes: A Three-Year Retrospective Study

Abstract

Purpose

The correlation between liver enzymes and blood glucose in diabetes mellitus remains unclear. The purpose of present three-year retrospective study was to explore the association between liver enzymes and glycemia in type 2 diabetes mellitus (T2DM).

Patients and Methods

2046 participants (non-fatty liver, n=1253; fatty liver, n=793) with T2DM were included in our study. For all these patients, related basic clinical information, biochemical parameters and liver ultrasonic data were collected. The effects of liver enzymes on blood glucose levels were analysed by linear and binary logistic regression models.

Results

In the comparative analyses between groups, gamma-glutamyltransferase (GGT) and serum alkaline phosphatase (AKP) levels increased with glycosylated haemoglobin A1c (HbA1c) concentration, but alanine aminotransferase and aspartate aminotransferase did not show the same trend. Adjusted for possible confounding factors, the linear regression models indicated the strong positive correlations between Log₁₀ AKP, Log₁₀ GGT and three blood glucose parameters (fasting glucose, postprandial glucose, and HbA1c) (all P<0.01). And such correlations were more significant in the subgroups of fatty liver. Multivariate binary logistic regression showed that the ORs increased by 1.01 times higher per Log₁₀ AKP unit in the highest blood glucose tertile (P<0.01).

Conclusion

GGT and especially AKP can aggravate hyperglycemia in T2DM patients, which provides new clues for the research and treatment of T2DM.

Keywords:

• liver enzymes
• fasting blood glucose
• postprandial blood glucose
• HbA1c
• T2DM

Ethics Statement

The research is consistent with the Declaration of Helsinki. The ethics Committee of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine reviewed and approved the study. Informed consent is exempted, because the clinical data were collected only for a retrospective study and this study did not involve any personal privacy or commercial interest. We keep all participants anonymous anywhere in the article.

Acknowledgments

We appreciate Meng-Na Yu who participated in the early relevant data (about 200 cases, mainly for another study) collection.

Disclosure

The authors report no conflicts of interest in this work.