Adolescence is characterized by rapid neurobehavioral development. The endocannabinoid (eCB) system modulates several co-developing neurotransmitter systems and is well-implicated in preclinical substance use (SU) models. Few studies, however, have investigated associations between the eCB system and SU in adolescence. The present study evaluated associations between circulating eCB lipids and SU, impulsivity, and behavioral approach endophenotypes. Concentrations of eCBs [2-arachidonoylglycerol (2-AG) and N-arachidonoylethanolamine (AEA)] and related small bioactive lipids were quantified in serum samples from a substudy of youth (n = 432) enrolled in the Adolescent Brain Cognitive Development Study using liquid chromatography-mass spectrometry. Youth completed the UPPS-P Impulsive Behavior and Behavioral Inhibition/Behavioral Approach System Scales, a SU interview, and questionnaire on factors that impact circulating eCBs. Generalized additive models assessed for associations and sex interactions between eCB concentrations and behaviors. 2-AG and 2-oleoylglycerol were broadly associated with behavioral approach and urgency, while AEA and other N-acylethanolamines were associated with higher-order functions like perseverance and planning, with several sex interactions observed. 2-AG and AEA were negatively associated with low-level substance use and full-unit substance use, respectively. Although relationships between serum eCBs and concentrations in brain are not known, the associations observed between peripheral eCB concentrations and substance use behaviors could reflect differences in eCB homeostasis in the brain. Observed sex differences may suggest hormonal or sex-specific eCB influences on brain development. Further work using direct measures of eCB concentrations in the brain are needed to validate these speculations. Future studies should investigate these mechanisms longitudinally.
Within-disorder heterogeneity complicates mapping neurobiological features of psychopathology to DSM conceptualizations. The present study explored patterns of diagnostic classification errors among disorders with commonly co-occurring features to examine this heterogeneity. We conducted classification analyses with the UCLA Phenomics Study database using a support vector classifier to differentiate disorders via whole-brain task-based functional connectivity, predicting that model misclassifications would provide insight about brain connectivity characteristics shared across disorders. We also explored whether symptoms and specific brain networks could account for misclassification rates. The classification model performed better than chance (44% accuracy, p = .01) and revealed that misclassification of schizophrenia (SCZ) as bipolar disorder (BD; 38%) and BD as SCZ (36%) was symmetrical. Attention-deficit/hyperactivity disorder (ADHD) was misclassified as BD at the highest rate (46%) and higher than the converse (17%). SCZ and ADHD were misclassified least (SCZ as ADHD 15%; ADHD as SCZ 22%). Considerable variance in misclassification of SCZ as BD (R2 = .83) and BD as SCZ (R2 = .71) could be accounted for by symptoms of both SCZ and BD. Permutation testing revealed disorder- and network-specific effects, with certain networks improving classification accuracy and others hindering it for specific disorders. An approach focused on classification errors replicated known disorder overlap, producing errors in the expected configuration. Further, it identified clinical and neural features within and across diagnostic categories that contribute to disorder misclassification and within-disorder heterogeneity. This approach may facilitate neurobiologically informed phenotypic differentiation within diagnostic groups.
The endocannabinoid (eCB) system underlies many biological processes and is thought to be critical in adolescent brain development. However, the eCB system’s contributions to brain morphometry during adolescence remains understudied. Here we investigated cross-sectional associations between brain structure and circulating eCBs and related lipids in a subsample of adolescents ( N = 294, ages 10–14, 44.5% female, 65.3% White) from the Adolescent Brain and Cognitive Development (ABCD) study utilizing a structural equation modeling approach. After accounting for sex and age, significant covariance between circulating eCBs and related lipids and cortical thickness, surface area, and volume in brain regions throughout the cortex were revealed. While the mechanisms underlying these associations need further research, these results suggest that the eCB system likely plays a critical role in early-adolescent brain development. Future work should determine underlying mechanisms, evaluate these associations longitudinally, and consider sex differences and the modulatory effects of pubertal hormones.
Background: Advances in functional magnetic resonance imaging (fMRI) have led to the ability to study the brain across many contexts. However, the large number of features generated by functional connectivity approaches may overfit the data. These problems can be overcome with either feature selection (FS) or dimensionality reduction (DR), which can be applied to less complex models. We utilize two open-source datasets to compare the performance of DR/FS methods on cognitive task decoding using a suite of ML classifiers. New Method: While DR and FS methods have been used previously in decoding research, no systematic comparison of their performance has been undertaken. Here, we compare available methods using commonly utilized machine learning libraries to establish which methods provide the best predictive performance. We then conduct statistical tests to examine the relative contributions of DR and FS methods and classifiers on decoding accuracy. Results: Neither DR or FS was found to be superior. However, differences were identified across datasets and tasks. In the majority of methods and datasets, a peak in predictive performance was found using a small percentage of the total number of original features. Comparison with existing methods: Some methods perform better than the baseline method of prediction with all available features or selecting features randomly. Decoding performance utilizing the HCP datasets with certain DR/FS methods exceeds that of deep learning approaches. Conclusions: Simple machine learning models with DR/FS have competitive decoding performance. These results suggest a “sweet spot” for the tradeoff between the retention of features and predictive accuracy.