Redefining the alanine aminotransferase upper limit of normal improves the prediction of metabolic syndrome risk.

Abstract

Multiple studies have recently proposed the lowering of upper limit of normal (ULN) for alanine aminotransferase (ALT) to improve the diagnostic sensitivity for viral hepatitides and metabolic syndrome (MS). We have tried to validate some of the proposed ULNs in the diagnosis of MS. We used data from the HepaMeta Study conducted in 2011 in Slovakia, which explored the prevalence of MS in eastern Slovakia. Patients were tested for the criteria of MS and ALT. Different, previously published, ALT cutoffs were then used to calculate odds' ratios, sensitivity, specificity, and accuracy of MS and its components. Manufacturers' recommended ULN used in our institution (0.8 µkat/l, 47 U/l for men and 0.6 µkat/l, 35 U/l for women) failed to predict any significant risk of MS. Lowered cutoff (72% of the original ULN) identified the patients with the highest age-adjusted probability of MS (odds ratio 3.194, 95% confidence interval 1.398-7.295). ALT was significantly associated with elevated levels of triacylglycerols, hyperglycemia, and obesity. In patients with MS, one must consider liver involvement if the patient has ALT levels in the upper third of the reference range. There is the need for discussion about the feasibility of lower ALT ULN in clinical practice.