Non-alcoholic Fatty Liver Disease and Depression: Evidence for Genotype × Environment Interaction in Mexican Americans

Abstract

This study examines the impact of G × E interaction effects on non-alcoholic fatty liver disease (NAFLD) among Mexican Americans in the Rio Grande Valley (RGV) of South Texas. We examined potential G × E interaction using variance components models and likelihood-based statistical inference in the phenotypic expression of NAFLD, including hepatic steatosis and hepatic fibrosis (identified using vibration controlled transient elastography and controlled attenuation parameter measured by the FibroScan Device). We screened for depression using the Beck Depression Inventory-II (BDI-II). We identified significant G × E interactions for hepatic fibrosis × BDI-II. These findings provide evidence that genetic factors interact with depression to influence the expression of hepatic fibrosis.

Genotype-by-socioeconomic status interaction influences heart disease risk scores and carotid artery thickness in Mexican Americans: the predominant role of education in comparison to household income and socioeconomic index

“…There is GxE evidence if either null hypothesis is rejected ( Diego et al, 2003 ; Diego et al, 2007 ; Santos et al, 2014 ; Arya et al, 2018 ; Manusov et al, 2022 ). Rejection of either or both is evidence that the phenotypic response to the environment has a genetic basis.…”

“…We now extend this theory to a spectrum of such measures to model GxE for continuous SES environments as opposed to two levels of the SES variable. To this end, we employ variance and correlation functions ( Diego et al, 2003 ; Diego et al, 2007 ; Santos et al, 2014 ; Arya et al, 2018 ; Manusov et al, 2022 ), which we now define as: where the additive genetic variance is re-parameterized as an exponential function of the value of the environmental variable for the individual, , scaled against the sample mean, , and where the genetic correlation is re-parameterized as an exponential decay function of the difference of environmental variables for any pair of individuals and , and where , , and are parameters to be estimated. These functions can be interpreted as the variance and correlation functions of a Gaussian stationary stochastic process, ( Kirkpatrick and Heckman, 1989 ; Pletcher and Geyer, 1999 ; Jaffrézic and Pletcher, 2000 ; Pletcher and Jaffrézic, 2002 ; Diego et al, 2003 ; Meyer and Kirkpatrick, 2005 ) where the index variable of the process is one of the three quantitative SES variables.…”

“…BLUP accounts for additive genetic and environmental covariances among relatives based on known pedigree structure ( Diego et al, 2015a ). We then subtracted the BLUP genetic values from the original SES variable to get a BLUP-computed version of that variable that reflects primarily environmental effects ( Diego et al, 2015a ; Manusov et al, 2022 ). This lattermost variable is our GXE models’ focal (genetically corrected) environment.…”

“…We adopted a two-stage hypothesis testing approach that our team has implemented to analyze GxE models of biomedically significant trait-environment dyads ( Santos et al, 2014 ; Arya et al, 2018 ; Manusov et al, 2022 ). In the first stage, we examined if the overall GxE model provided a better fit to the data when compared with the polygenic model by the LRT.…”

Editorial: The biology and management of chronic diseases in Mexican Americans

“…The topic review includes data from the Mexican-ancestry population living on the U.S-Mexico border in the RGV. This population experiences high rates of complex diseases such as T2D (prevalence of 32%), obesity/overweight (60%), fatty liver disease (60%), and depression (19%) [ Zhang et al ; ( 2 , 5 )].…”