The best magazine
Discussion
NAFLD has emerged as an important predictor of T2DM and thus it is of interest to investigate the relationship of this disorder with coffee consumption in the context of attempting to identify a potential mediating role on the inverse relationship of coffee with T2DM. Our findings regarding ALT, AST and NAFLD liver fat score shed light on a potential mediating relationship that may underlie this inverse association.
Results of the present cross-sectional examination of a multi-ethnic, non-diabetic cohort reveal significant inverse associations of caffeinated coffee consumption with ALT and AST, and in the case of ALT, there was a stronger magnitude of association across multivariate models and with the exclusion of individuals whose transaminase levels may be affected by alcohol consumption. Analyses also revealed an inverse relationship of caffeinated coffee with NAFLD liver fat score, a non-invasive validated tool for identifying NAFLD using routinely available clinical and laboratory data. These findings are consistent with accumulating evidence demonstrating a beneficial impact of coffee consumption on the liver. Specifically, a beneficial effect of coffee on the risk of NAFLD as defined by elevated liver transaminases or abdominal imaging, a reduction in risk of fibrosis associated with coffee consumption in patients with NAFLD and NASH, and inverse associations of coffee with ALT and GGT. Previous studies of associations of coffee with liver enzymes however have been conducted largely in Japanese male populations or among individuals with alcoholic and viral liver diseases. Investigations of associations of coffee with liver markers in liver diseases of metabolic origin (ie. NAFLD) are scarce. Our findings are consistent with the only other study of healthy individuals conducted in North America and extend these investigations to a large multi-ethnic population with more complete metabolic characterization, including directly-measured insulin resistance and diabetes status assessed by OGTT. Our assessment of usual, rather than recent, intake of caffeinated and decaffeinated coffee consumption allowed us to better characterize coffee drinking behavior in this population and assess the independent relationships of habitual intake of coffee types with markers of liver injury.
Our findings suggest that caffeinated coffee's beneficial impact on NAFLD may potentially mediate its inverse association with T2DM. The relationship of NAFLD with IR and oxidative stress is well established. It is possible that the antioxidant properties of coffee may help to offset reactive oxygen species, thereby contributing to a reduction in IR and subsequent fatty acid deposition in the liver. Results of our fourth mechanistic model support this notion as additional adjustment for S I attenuated associations of caffeinated coffee with markers of liver injury. Interestingly, for ALT, the more sensitive and liver-specific transaminase, the association with caffeinated coffee consumption remained significant in this model suggesting that this relationship is not entirely explained by S I.
Investigations of the relationship of coffee consumption with fetuin-A are extremely limited. To our knowledge, only one study has examined the specific relationship of caffeinated coffee with fetuin-A which similarly reported null associations.
Our analyses did not identify significant associations of decaffeinated coffee with markers of liver injury, which is notable considering the inverse relationship of decaffeinated coffee with T2DM was shown to be stronger than that of caffeinated coffee in a meta-analysis. It is likely that the lack of significant associations in the current analysis may be related to the generally modest consumption of decaffeinated coffee and narrow range of intakes in this population. The low level of decaffeinated coffee consumption in the current study may explain why our findings of a null association of decaffeinated coffee with fetuin-A are discordant with the only other study investigating this relationship. Wedick et al. (2011) reported an inverse association of decaffeinated coffee with fetuin-A in their randomized control trial in which individuals consumed 5 6 oz cups of decaffeinated coffee/day. It is possible that higher levels of consumption than reported in the current study are necessary to document this relationship. Alternatively, it is possible that null associations of decaffeinated coffee with liver markers may be explained by differing mechanistic pathways that are being impacted by these coffee types. It has been suggested for example that decaffeinated coffee may specifically benefit beta cell function rather than IR, which appears to be more strongly linked with caffeinated coffee in this cohort.
The current study has a number of strengths including its large sample size, multi-ethnic population and detailed measures used to assess participant characteristics including diabetes status, assessed using the OGTT, and S I determined by FSIGTT. Further, a validated FFQ was used to measure usual intake of coffee and nutritional covariates. Unlike the majority of previous studies, our FFQ differentiated between caffeinated and decaffeinated coffee intake offering insight into the relationship of usual intake of each coffee type with outcome variables. The examination of multiple markers of liver injury, including liver transaminases, fetuin-A, and NAFLD liver fat score, represents a further strength of this study.
Potential limitations of the current study include the cross-sectional design as this precludes conclusions being drawn regarding the causality or the temporal nature of associations. As an observational study, there is potential for residual confounding. However extensive data were collected on potential covariates and these variables were carefully considered in the development of regression models. The reliance on participant self-report and the semi-quantitative design of the FFQ may introduce a degree of misclassification error in estimates of coffee consumption due to imprecise definitions of portion sizes. Such error, however, would likely be non-differential thus resulting in conservative estimates of associations. Additionally, NAFLD was not assessed using the gold standard liver biopsy, as this technique is not suitable for use in epidemiologic studies. Liver transaminase levels have been shown to be well correlated with directly-measured liver fat by magnetic resonance spectroscopy, and the NAFLD liver fat score is a validated, non-invasive method found to be more robust than similar predictive scores for identifying NAFLD, making these techniques appropriate alternatives to liver biopsy.
Finally, it should be emphasized that the magnitude of associations of caffeinated coffee with liver markers in the present study was modest, a finding that may be due in part to misclassification of consumption (which would dilute effect estimates, as described above), as well as the relatively modest consumption of coffee in the population, particularly in comparison to other studies investigating associations of coffee with markers of liver injury; the majority of which reported average consumption of 1 or more cups of coffee/day. In the present study, the median (IQR) of caffeinated and decaffeinated coffee intake were 0.4 (0–0.8) servings/day and 0 (0–0.03) servings/day, respectively. Conducting this study in a population in which coffee types were being consumed at higher levels may have improved our ability to detect associations of coffee with outcome variables particularly if coffee's beneficial effects are exerted only at higher doses. This may be particularly important for decaffeinated coffee since consumption of this type was quite low in this cohort.
Source: ...