Association of Chronic Hepatitis C and Diabetes: Is There a Gender Difference?

Abstract

To the Editor: Montenegro et al. (1) recently reported in the Journal an association between chronic hepatitis C (HCV) infection and development of diabetes mellitus (DM) type 2 in a cohort of patients in a Southern Italian town. We have three comments on their excellent article. First, we similarly evaluated the association between HCV antibodies and DM in the United States, using the National Ambulatory Medical Care Survey (NAMCS), a large nationally representative cohort, which provides outpatient data from physician offices in the United States for 2000–2009. All patients with complete data on HCV and DM status were analyzed. Among 252,450 patients initially analyzed, 220 patients had HCV and 18 of them were diabetic. A multivariate logistic regression model was used to adjust for age, hypertension, obesity, and any lipid therapy. On multivariate analysis, HCV was independently associated with DM (odds ratio (OR)=1.97; 95% confidence interval (CI) 1.29–3.07, P=0.02). These data support the authors’ conclusions by showing a similar association between HCV and DM in a different population: United States outpatients. However, unlike the authors, we did not analyze the relation of serum ALT (alanine aminotransferase) values to DM, and did not separate DM type 1 from DM type 2. However, the latter difference is unlikely to affect the conclusions because DM type 2 represents the overwhelming majority of DM in America. Second, 13 of 18 patients with DM and HCV were male, and only 5 were female. This represents a quantitatively large, statistically significant effect (OR for being male=6.67; 95% CI 2.05–21.79). Montenegro et al. (1) report the same phenomenon, incidentally without commenting about its significance (OR for being female=0.44, 95% CI 0.31–0.63, P<0.001, Table 3). Our data in conjunction with the authors’ data suggest there may be a gender difference in the proposed association. This gender difference is not surprising because many toxins affect the liver according to gender, e.g., different minimum levels of alcohol consumption/day to develop alcoholic cirrhosis for males vs. females (2); and difference in mortality from hepatitis E infection between females in third trimester of pregnancy vs. other females or males (3). This large difference between genders might offer clues to the pathophysiology of association between HCV and DM. Third, the authors describe taking blood samples from each participant in 1985 (as well as 1992 and 2005) for anti-HCV antibodies, assessed by immunoenzymatic enzyme-linked immunosorbent assay (Methods (1)). The authors may want to clarify that this blood test was performed subsequently (retrospectively) on the bloods drawn in 1985 because hepatitis C was only discovered in 1988–1989 (4) and the assays for it first became commercially available in 1992 (5).