To the Editor:
In a paper published in the September 2005 issue of HEPATOLOGY,1 Kunde et al. described the spectrum of nonalcoholic fatty liver disease (NAFLD) among morbidly obese women and examined the possible implications of adopting the healthy ranges for alanine aminotransferase (ALT) activity, as recently proposed by our group,2 for the diagnosis of chronic liver disease. We are very grateful to the authors for their interest in our work.
The study population is very peculiar (women with morbid obesity undergoing gastric bypass surgery). From a clinical perspective, it is should be clarified what could be the benefit of identifying individuals with biochemically silent liver disease—of any severity—among morbidly obese patients. The frequency of chronic liver damage in these individuals is approximately 100% (the present series confirms these findings), and the management approach to preventing long-term liver disease complications coincides with strategies (counseling, medical, and/or surgical therapies) aimed at curing obesity. From a methodologically point of view, any cost-effectiveness extrapolation of data from this highly selected series to the general population is inappropriate. It should also be considered that for a pretest probability near 100, no test can be used to rule a disease in or out. What is more, no statistical approach is capable of widening the diagnostic meaning of a laboratory test. Measuring ALT to discriminate patients with and without fibrosis progression (i.e., isolated portal fibrosis and nonalcoholic steatohepatitis [NASH] from “simple” fatty liver disease) seems inappropriate, because its increase only indicates cytolysis. Thus, ALT can be used as a marker to identify those with disease activity, but not for staging fibrosis. Finally, the experience accumulated with hepatitis B virus (HBV) and hepatitis C virus (HCV) carriers indicates that a single ALT determination—the approach adopted in the present study—is insufficient for the workup of patients at risk for liver disease, and repeated determinations are necessary.
We fully agree with the authors' opinion that it would not be opportune to prescribe hepatitis serology tests and abdominal ultrasound in any obese individual with ALT levels exceeding the healthy ranges. In this regard, we recently recommended both the abandonment of the traditional concept of a universal “normal limit” for ALT, and a flexible interpretation of results that is performed according to the clinical context.2 Diagnostic and workup algorithms and thresholds should be diversified on the basis of many factors, including patient characteristics, pretest probabilities, and treatment opportunities. This approach is currently being adopted in the management of patients with chronic liver disease.3 This does not mean, however, that overweight and obese individuals with ALT values between the “healthy” and “traditional” threshold should not be appropriately counseled solely due to their high frequency. As documented by Kim et al.,4 these subjects are at an increased risk for death from chronic liver disease. The authors' observation that the healthy thresholds mainly detect the early stages of chronic liver disease—particularly amenable for behavior therapy—should be interpreted as a support for this policy.
What is “statistically normal” is not necessarily “healthy.” Any hesitation in recognizing this simple fact will ultimately contribute to aggravating the obesity epidemic in Western countries.