We appreciate the interest of Dr Johnston and colleagues in our recent publication and also the questions they present.

Histological evaluation of liver biopsy remains the gold standard for assesment of hepatic pathology, despite the recent development of other advanced imaging techniques. However, percutaneous liver biopsy poses a small, albeit inherent, risk even in the most experienced hands, which is why it should preferably be performed when the benefits of knowing the histology outweigh the risks to the patient. Therefore, serum aminotransferases are often used as surrogate markers to exclude or confirm suspicion of liver disease, even in scientific investigations, particularly in non-alcoholic fatty liver disease (NAFLD) studies.1 Recently, it was shown that severe histological findings can be present in NAFLD patients having serum alanine aminotransferase (ALT) levels within the reference limits.2 Thus, serum ALT is not a valuable criterion for histological assessment of disease severity in NAFLD, which is considered to be the hepatic counterpart of the metabolic syndrome. Because of this dilemma, a lower upper reference limit for serum ALT has …