This is the first report on the high order interaction effects of steatohepatitis, constructed with a nested case-control design for the healthy population. We clearly demonstrated obesity, metabolic syndrome, and high coronary risk had interaction effects on steatohepatitis status.
The close relationship between insulin resistance and nonalcoholic fatty liver disease was demonstrated in a spectrum of abnormalities, from fatty liver only to nonalcoholic steatohepatitis. The nonalcoholic steatohepatitis is a critical link in the chain of metabolic fatty liver disorders that spans steatosis to cryptogenic cirrhosis. It is the hepatic manifestation of the insulin resistance (or metabolic) syndrome, and provides a clue to understanding fibrotic progression of other chronic liver diseases, particularly hepatitis C. Nonalcoholic steatohepatitis is often the first clinical indication of insulin resistance, with its complications of high blood pressure, coronary heart disease and type 2 diabetes[20
]. Our study clearly demonstrated the atherosclerotic risk burdens in nonalcoholic steatohepatitis.
Abdominal ultrasound is a useful tool to detect fatty liver disease, although with limited sensitivity and specificity. In this nested case-control design, non-invasive ultrasound as the diagnostic tool provided valid and reproducible instruments, which was more effective than the "gold standard" of liver biopsy. Also, the consistency of information collection and blinding in exposure and disease ascertainment made the results valid.
There have been abundant reports in the literature on the association of steatohepatitis and metabolic syndrome [21
]. For example, a cross-sectional study based on adults demonstrated clearly that metabolic syndrome is independently associated with fatty liver [21
]. In Taiwan, obesity and atherosclerotic risk factors play important roles in metabolic syndrome and cardiovascular burdens [23
]. Our study results imply that the burdens of obesity to the liver are also significant. In the severely obese patients, the best therapeutic modality is bariatric surgery, which is safe and has been successful in producing a considerable weight loss. This treatment can improve the control of diabetes mellitus, the metabolic syndrome, and presumably its sequelae. The bariatric surgery can reduce the fat, inflammation, and even the fibrosis in well-documented nonalcoholic steatohepatitis[25
Inflammation risk worsens steatohepatitis status and has incremental interactions with metabolic syndrome and risk profiles. Possible mechanisms of inflammatory cytokine increases might play roles in the steatohepatitis process. Our study demonstrates that CRP and lymphocytes were associated with steatohepatitis and fatty liver status. Anti-inflammatory effects of pharmacotherapy, such as insulin sensitizers, might play a role in the reduction of fatty liver and steatohepatitis severity.
The limitations of this study are as follows: First, we cannot specify the accuracy of ultrasound or biopsy evidence for fatty liver. Histopathological features of steatohepatitis have been the gold standard for diagnosis of steatohepatitis. Also, the misclassification of disease and control status might underestimate the true association effects. Due to large study populations and reliable abdominal sonography, the diagnosis of cases and careful selection of controls could validate the results. Second, the cases of steatohepatitis were prevalent cases, not incident cases. The associated factors might have been beneficial for survival status. We included the close population in fixed duration to improve the comparability of case and control status, and the strategy was able to decrease selection bias. The prospective cohort or randomized control trial studies would provide more evidence for risk factor reduction for steatohepatitis control. Finally, the study population was from the specified health examination center, and possible self-selection characteristics among healthy subjects could make external generalization of the results to the general population limited.