No abstract

Alcohol misuse during adolescence (AAM) has been associated with disruptive development of adolescent brains. In this longitudinal machine learning (ML) study, we could predict AAM significantly from brain structure (T1-weighted imaging and DTI) with accuracies of 73 - 78% in the IMAGEN dataset (n ~1182). Our results not only show that structural differences in brain can predict AAM, but also suggests that such differences might precede AAM behavior in the data. We predicted ten phenotypes of AAM at age 22 using brain MRI features at ages 14, 19, and 22. Binge drinking was found to be the most predictable phenotype. The most informative brain features were located in the ventricular CSF, and in white matter tracts of the corpus callosum, internal capsule, and brain stem. In the cortex, they were spread across the occipital, frontal, and temporal lobes and in the cingulate cortex. We also experimented with four different ML models and several confound control techniques. Support Vector Machine (SVM) with rbf kernel and Gradient Boosting consistently performed better than the linear models, linear SVM and Logistic Regression. Our study also demonstrates how the choice of the predicted phenotype, ML model, and confound correction technique are all crucial decisions in an explorative ML study analyzing psychiatric disorders with small effect sizes such as AAM.

Depression and coronary heart disease (CHD) are prevalent and often co-occurring disorders. Both have been associated with a dysregulated stress system. As a central element of the stress system, the FKBP5 gene has been shown to be associated with depression. In a prospective design, this study aims to investigate the association of FKBP5 with depressive symptoms in CHD patients. N = 268 hospitalized CHD patients were included. Depressive symptoms were measured using the Hospital Anxiety and Depression Scale (HADS-D) at four time points (baseline, and after 1 month, 6 months, and 12 months). The functional FKBP5 single-nucleotide polymorphism (SNP) rs1360780 was selected for genotyping. Linear regression models showed that a higher number of FKBP5 C alleles was associated with more depressive symptoms in CHD patients both at baseline (p = 0.015) and at 12-months follow-up (p = 0.025) after adjustment for confounders. Further analyses revealed that this effect was driven by an interaction of FKBP5 genotype with patients' prior CHD course. Specifically, only in patients with a prior myocardial infarction or coronary revascularization, more depressive symptoms were associated with a higher number of C alleles (baseline: p = 0.046; 1-month: p = 0.026; 6-months: p = 0.028). Moreover, a higher number of C alleles was significantly related to a greater risk for dyslipidemia (p = .016). Our results point to a relevance of FKBP5 in the association of the two stress-related diseases depression and CHD.