Reply:

Authors


  • Potential conflict of interest: Nothing to report.

Reply

We thank Dr. Tsochatzis and his colleagues for their comments and appreciate the opportunity to respond. We would like to point out that their results demonstrating an association between low adiponectin levels and insulin resistance in a “mixed” chronic hepatitis C (CHC) cohort are in agreement with the results reported in our article and in extensive published literature on insulin resistance unrelated to CHC. However, we wish to emphasize that the only robust method to determine if this is a virus-specific effect in CHC is to undertake a comparison with healthy controls not infected with hepatitis C virus, as was done in our publication. With regard to the statistical methodology, because the homeostasis model assessment of insulin resistance (HOMA-IR) is a continuous outcome, the power to detect a statistically significant linear dependence of HOMA-IR on leptin and adiponectin levels is largest if one fits a linear regression model to the continuous outcome. The authors suggest dichotomizing the outcome into, for example, HOMA-IR ≥ 3 versus HOMA-IR < 3, but this is well known to be a statistically less robust and powerful method of detecting a linear association.

Despite the above considerations, we did undertake the additional analyses as suggested by Tsochatzis and his colleagues. As expected, leptin (P < 0.001) and adiponectin (P < 0.02) were independently associated with insulin resistance in the control group with both linear and logistic regression analyses. Moreover, body mass index (P = 0.004), leptin (P < 0.001), and adiponectin (P = 0.04) remained as predictors for HOMA-IR ≥ 3 in the CHC group with logistic regression analysis. In our study, logistic regression analysis was not performed because of the statistical considerations mentioned earlier and because there is no universally accepted cutoff point in HOMA-IR that defines insulin resistance.

We would like to add that the data presented by Dr. Tsochatzis and his colleagues in their letter is confounded by the inclusion of female patients (49%). In women, insulin sensitivity is influenced by the menstrual cycle, age, and intake of estrogen-containing preparations. Our study demonstrating that CHC infection is not, in itself, associated with specific changes in either leptin or adiponectin levels implies that these adipocytokines cannot be the mediators of the virus-specific changes in insulin sensitivity that are evident among hepatitis C virus–infected persons. We agree that more studies to further characterize the mechanisms involved in the development of insulin resistance in CHC are warranted.

Ian Homer Y. Cua M.D.*, Jacob George M.D., Ph.D.*, * Storr Liver Unit, Department of Medicine, University of Sydney at Westmead Hospital, Westmead, New South Wales, Australia.

Ancillary