Article Text

Download PDFPDF
Liver enzymes and incident diabetes in China: a prospective analysis of 10 764 participants in the Guangzhou Biobank Cohort Study
  1. L Xu1,
  2. C Q Jiang2,
  3. C M Schooling1,3,
  4. W S Zhang2,
  5. K K Cheng4,
  6. T H Lam1
  1. 1School of Public Health, The University of Hong Kong, Hong Kong, Hong Kong
  2. 2Guangzhou No 12 Hospital, Guangzhou, China
  3. 3CUNY School of Public Health and Hunter College, New York, USA
  4. 4Department of Public Health, Epidemiology and Biostatistics, University of Birmingham, Birmingham, UK
  1. Correspondence to Dr C M Schooling, School of Public Health, The University of Hong Kong, Patrick Manson Building, 7 Sassoon Road, Hong Kong, Hong Kong; cms1{at}


Background Impaired liver function has been shown to be associated with incident diabetes. The independent role of the different liver enzymes, including γ-glutamyltransferase (GGT), alanine transaminase (ALT) and aspartate transaminase (AST), has not been addressed properly, taking into account their high collinearity. We used partial least squares (PLS) regression to identify the contribution of ALT, AST and GGT, which appears causally associated with diabetes as a validation factor, to incident diabetes in a South China population where liver impairment and diabetes are common.

Methods Participants were from the Guangzhou Biobank Cohort Study recruited in 2003–2008, with follow-up re-examination up to the end of 2012. Multivariable generalised linear models and PLS were used to examine the adjusted associations of ALT, AST and GGT with diabetes. Incident diabetes was defined as self-reported diabetes, and/or initiation of hypoglycaemia medication or insulin during follow-up, or fasting glucose ≥7.0 mmol/L, or 2 h oral glucose tolerance test, glucose ≥11.1 mmol/L at follow-up examination.

Results In 10 764 Chinese participants aged ≥50 years with no diabetes at baseline, 1228 (11.4%) developed diabetes during the median 4 years of follow-up. Using PLS, the risk for incident diabetes was higher by 18% (95% CI 8% to 27%) per 1 SD increment in log-ALT, and expectedly higher by 36% (95% CI 26% to 52%) for log-GGT, adjusted for age, sex, education, smoking, alcohol, physical activity, waist circumference and body mass index. Similarly adjusted, no association for log-AST (relative risk 0.92, 95% CI 0.85 to 1.01) was found.

Conclusions ALT but not AST was associated with incident diabetes. Further experimental studies are needed to confirm the causal association and clarify the underlying mechanisms.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles