ORIGINAL ARTICLE The population distributions, upper normal limits and correlations between liver tests among Australian adolescents Jacob George, 1 Elizabeth Denney-Wilson, 2 Anthony D Okely, 3 Louise L Hardy 2 and Robert Aitken 4 1 Storr Liver Unit, Westmead Millennium Institute, University of Sydney and Westmead Hospital, 2 New South Wales Centre for Overweight and Obesity, School of Public Health, University of Sydney, 3 Child Obesity Research Centre, University of Wollongong, Wollongong and 4 New South Wales Department of Health, New South Wales, Australia Aim: Relatively little is known about the results of liver tests among adolescents. The purpose of this study was to describe the distributions of liver tests, to identify the upper normal limits (UNLs) and to describe the correlations among liver tests. Methods: Overnight fasting blood samples were collected from a representative population sample of 500 Grade 10 students (15 years old) attending schools in Sydney, Australia. Weight, height and waist girth were measured. UNLs were calculated for each enzyme as the 95th percentiles of the healthy body mass index category, after excluding those with high blood pressure, elevated low-density lipoprotein cholesterol (3.4 mmol/L) or triglyceride concentration (2.25 mmol/L) (n = 246). The distributions of alanine aminotransferase (ALT), g-glutamyltransferase (GGT), aspartate aminotransferase (AST) and alkaline phosphatase (ALP) were described, UNLs were identified and the correlations between enzyme concentrations were described. Results: The UNLs for ALT were 32 and 20 U/L; for GGT, 24 and 19 U/L; for AST, 33 and 26 U/L; and for ALP, 385 and 183 U/L, for boys and girls, respectively. ALT concentration was strongly correlated with GGT and AST (r = 0.5–0.6). Neither ALT nor GGT concentrations were correlated with ALP concentration, but AST concentration was moderately correlated with ALP concentration. Conclusions: These data are valuable in defining the distributions of liver tests, normal liver test ranges and relationships between liver tests among adolescents. Key words: adolescents; liver tests; upper normal limits; UNLs. Overweight and obese children and adolescents are at increased risk for the development of a range of chronic health problems, including non-alcoholic fatty liver disease (NAFLD). 1 NAFLD is characterised by the accumulation of triglyceride droplets that displace the hepatocyte nucleus peripherally and varying degrees of inflammation and fibrosis of the liver. NAFLD ranges in severity from a relatively benign disorder (non-alcoholic fatty liver or simple steatosis) to steatohepatitis (non-alcoholic ste- atohepatitis), cirrhosis and liver failure. 2 The natural history and rates of progression of this disease in adults is the subject of considerable debate, while that of paediatric and adolescent NAFLD is unknown. The biological risk factors for NAFLD among adults include obesity, type 2 diabetes and dyslipidaemia, all components of the metabolic syndrome. Further, it is now apparent that the concomitant presence of metabolic risk factors worsens the progression of liver disease of any aetiology. Although the risk factors for NAFLD among children and ado- lescents have not yet been thoroughly characterised, over- weight or obesity and insulin resistance appear to be risk factors; male gender may also increase the risk. 3,4 In the wake of a high incidence of overweight among young people and the likely association between NAFLD and excess adiposity, the prevalence of NAFLD is also likely to increase among young people. 4 Fundamental to clinical practice associ- ated with obesity and liver function is an understanding of the distributions of liver enzymes, the development of upper normal limits (UNLs) based on healthy population data and the evidence as to the need to take account of the stage of pubertal development. However, none of these pieces of information appear to be currently available. Key Points 1 Upper normal limits for the liver enzymes alanine aminotrans- ferase (ALT), g-glutamyltransferase (GGT), aspartate aminotrans- ferase (AST) and alkaline phosphatase (ALP) are identified, allowing risk stratification of adolescents and supporting clinical research. 2 The distributions of ALT and GGT concentrations were similar both within and between the genders whereas the distributions of AST and ALP were markedly different for boys and girls. 3 ALT concentration was strongly correlated with GGT and AST (r = 0.5–0.6), but AST and GGT concentrations were only very weakly correlated. Neither ALT nor GGT concentrations were correlated with ALP concentration, but AST concentration was moderately correlated with ALP concentration. Correspondence: Michael L Booth, Senior Research Fellow, School of Public Health, Level 2, Medical Foundation Building K25, University of Sydney, NSW 2006, Australia. Fax: +61 2 9363 4907; email: michael_ booth@health.usyd.edu.au Accepted for publication 9 April 2008. doi:10.1111/j.1440-1754.2008.01386.x Journal of Paediatrics and Child Health 44 (2008) 579–585 © 2008 The Authors Journal compilation © 2008 Paediatrics and Child Health Division (Royal Australasian College of Physicians) 579