Serum DHEA Levels Distinguish Mild From Severe Nonalcoholic Fatty Liver Disease

May 22, 2006

May 22, 2006 (Los Angeles) — Patients with nonalcoholic fatty liver disease (NAFLD) with advanced fibrosis can be accurately distinguished from those with less severe disease using serum dehydroepiandrosterone (DHEA) levels as a biomarker, Michael R. Charlton, MD, chief of hepatology at the Mayo Clinic in Rochester, Minnesota, stated during a presentation here at Digestive Disease Week.

"More than 25 million Americans have this liver disease," Dr. Charlton told Medscape. DHEA modulates processes such as oxidative stress, insulin resistance, and fibrosis seen in severe NAFLD. The investigators conducted a prospective trial to test the hypothesis that serum DHEA levels in mild NAFLD would differ from DHEA levels in severe NAFLD.

Dr. Charlton and colleagues obtained serum samples from 78 patients with NAFLD, 53 of which were defined as mild (fibrosis levels 0 to 2), and 25 of which were defined as severe (fibrosis levels 3 and 4). The control group consisted of 44 individuals with cholestatic liver disease. The investigators measured serum DHEA concentrations with ELISA before liver biopsy, and they used logistic regression to explore the relation between DHEA levels and nonalcoholic steatohepatitis (NASH) outcomes.

All patients with severe NASH (n = 25) had low plasma DHEA concentrations (<.45 µg/dL). Participants with severe NAFLD also had higher C-reactive protein (CRP) levels (.7 mg/L) compared with mild NAFLD patients (.4 mg/L; P = .05). The investigators observed a dose-related effect when DHEA concentrations were graphed against fibrosis levels in patients with NASH. A dose-response relationship was not seen between fibrosis levels and DHEA concentrations when values from the control patients were graphed.

One regression model included as predictors 2 levels of DHEA (<0.45 µg/dL or >= 0.45 &#181;g/dL), patient age (<45 years; 45 years or older), and sex, and generated an operator curve with an area under the curve (AUC) value of 0.83. Another model, using only DHEA concentrations (0.45 µg/dL or more) as a predictor generated an AUC value of 0.87. This model had a sensitivity of 100%, but only a specificity of 58.5%.

Dr. Charlton cautioned, "I'm not sure that at a clinical level it has any implications at this point; it's a first look analysis.... But if we see the same thing in other populations then I think we'll be happy to say that this is something people should be thinking about."

If these data are borne out by planned analyses of similar data from Indiana University and the University of California, San Francisco, clinicians might someday be able to help some of their NAFLD patients avoid the pain and expense of a liver biopsy. "You could tell a patient with a normal DHEA or above this threshold 'you don't have liver disease, you can just relax,' and keep checking. It might be worth screening with a biopsy later," Dr. Charlton told Medscape.

"[DHEA is] not that difficult to measure, so it might be of big importance in practice," Wolfgang Tacke, MD, a gastroenterology specialist at St. Josef's Hospital in Konigstein, Germany, told Medscape in an interview for outside comment. Dr. Tacke was not involved with the study.

Funding for this study was provided by the National Institutes of Health. Dr. Charlton reports no relevant financial relationships.

DDW 2006: Abstract 115. Presented May 21, 2006.

Reviewed by Maria L. Gaiso, PhD


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.