Insulin-related biomarkers and their relationship with hepatic fibrosis risk in patients with psoriasis, metabolic syndrome and non-alcoholic fatty liver disease

Abstract

I read with interest the clinical study by Campanati et al [1]. I appreciate their study on the preventive effect of etanercept on the development of hepatic fibrosis in patients with psoriasis, metabolic syndrome and nonalcoholic fatty liver disease (NAFLD). They speculated that the pathological basis was dependent on insulin resistance. Among several insulin-related biomarkers, they used indices of insulin sensitivity and resistance, respectively known as homeostasis model assessment-insulin resistance (HOMA) and quantitative insulin sensitivity check index (QUICKI), in combination with serum insulin. HOMA and QUICKI were easily calculated using fasting glucose (FPG) and insulin, and they have been used in the past [2]. I have two queries on their study outcome. First, patterns of significant difference resulting from use of different insulin-related biomarkers among four groups stratified by two different medications were the same in their Table 1, Fig. 1 and Fig. 2. This statistical outcome could be explained by the mathematical viewpoint. The definition is presented as follows: HOMA = (FPG 9 insulin)/405; QUICKI = 1/(common logarithms(FPG 9 insulin)). QUICKI is calculated as 1/(common logarithms (405 9 HOMA-IR)), and when predicting QUICKI by HOMA-IR, it becomes a monotone decreasing function. Mathematically, HOMA-IR and QUICKI are the same indicators, except that their distribution can be characterized as either logarithmic-normal or normal. Furthermore, I previously reported that serum insulin and HOMA showed no difference when they were used as insulin-related biomarkers [3]. When the two scatterplots in Fig. 4 by Campanati et al. were checked, the explanation rate presented by R square and the level of significance were almost the same. Second, they recruited 89 patients with psoriasis, metabolic syndrome and NAFLD to participate in their study. Since the application of QUICKI and HOMA could be applied to subjects with FPG under 140 mg/dl [4, 5], I speculate that Campanati et al. excluded patients according to these criteria (please see data presented in Table 1 by them). As serum insulin was not regulated by these criteria, I recommend their use of serum insulin by including patients with diabetes or pre-diabetes. I suppose that selection bias can be avoided by this process. Their study includes information that is valuable to the prevention of hepatic fibrosis risk for patients with psoriasis, metabolic syndrome and NAFLD. Although insulinassociated biomarkers were the main contributors to the target pathological status, they used the aspartate transaminase/alanine transaminase ratio as an indirect indicator of hepatic fibrosis risk. Direct risk assessment of hepatic fibrosis should also be conducted as an advanced study.