This study was designed to examine whether proactive and reactive aggression are meaningful distinctions at the variable-and person-based level, and to determine their associated behavioral profiles. Data from 587 adolescents (mean age 15.6; 71.6 % male) from clinical samples of four different sites with differing levels of aggression problems were analyzed. A multi-level Latent Class Analysis (LCA) was conducted to identify classes of individuals (personbased) with similar aggression profiles based on factor scores (variable-based) of the Reactive Proactive Questionnaire (RPQ) scored by self-report. Associations were examined between aggression factors and classes, and externalizing and internalizing problem behavior scales by parent report (CBCL) and self-report (YSR). Factor-analyses yielded a three factor solution: 1) proactive aggression, 2) reactive aggression due to internal frustration, and 3) reactive aggression due to external provocation. All three factors showed moderate to high correlations. Four classes were detected that mainly differed quantitatively (no 'proactive-only' class present), yet also qualitatively when age was taken into account, with reactive aggression becoming more severe with age in the highest affected class yet diminishing with age in the other classes. Findings were robust across the four samples. Multiple regression analyses showed that 'reactive aggression due to internal frustration' was the strongest predictor of YSR and CBCL internalizing problems. However, results showed moderate to high overlap between all three factors. Aggressive behavior can be distinguished psychometrically into three factors in a clinical sample, with some differential associations. However, the clinical relevance of these findings is challenged by the person-based analysis showing proactive and reactive aggression are mainly driven by aggression severity.

It is unclear which aspects of empathy are shared and which are uniquely affected in autism spectrum disorder (ASD) and conduct disorder (CD) as are the neurobiological correlates of these empathy impairments. The aim of this systematic review is to describe the overlap and specificity of motor, emotional, and cognitive aspects of empathy in children and adolescents with ASD or CD. Motor and cognitive empathy impairments are found in both ASD and CD, yet the specificity seems to differ. In ASD facial mimicry and emotion recognition may be impaired for all basic emotions, whereas in CD this is only the case for negative emotions. Emotional empathy and the role of attention to the eyes therein need further investigation. We hypothesize that impaired motor and cognitive empathy in both disorders are a consequence of lack of attention to the eyes. However, we hypothesize major differences in emotional empathy deficits between ASD and CD, probably due to emotional autonomic and amygdala hyper-responsivity in ASD versus hypo-responsivity in CD, both resulting in lack of attention to the eyes.

Advances in commercial wearable devices are increasingly facilitating the collection and analysis of everyday physiological data. This paper discusses the theoretical and practical aspects of using such ambulatory devices for the detection of episodic changes in physiological signals as a marker for mental state in outdoor environments. A pilot study was conducted to evaluate the feasibility of utilizing commercial wearables in combination with location tracking technologies. The study measured physiological signals for 15 participants, including heart rate, heart-rate variability, and skin conductance. Participants' signals were recorded during an outdoor walk that was tracked using a GPS logger. The walk was designed to pass through various types of environments including green, blue, and urban spaces as well as a more stressful road crossing. The data that was obtained was used to demonstrate how biosensors information can be contextualized and enriched using location information. Significant episodic changes in physiological signals under real-world conditions were detectable in the stressful road crossing, but not in the other types of environments. The article concludes that despite challenges and limitations of current off-the-shelf wearables, the utilization of these devices offers novel opportunities for evaluating episodic changes in physiological signals as a marker for mental state during everyday activities including in outdoor environments.

Machine learning techniques are increasingly being applied to clinical text that is already captured in the Electronic Health Record for the sake of delivering quality care. Applications for example include predicting patient outcomes, assessing risks, or performing diagnosis. In the past, good results have been obtained using classical techniques, such as bag-of-words features, in combination with statistical models. Recently however Deep Learning techniques, such as Word Embeddings and Recurrent Neural Networks, have shown to possibly have even greater potential. In this work, we apply several Deep Learning and classical machine learning techniques to the task of predicting violence incidents during psychiatric admission using clinical text that is already registered at the start of admission. For this purpose, we use a novel and previously unexplored dataset from the Psychiatry Department of the University Medical Center Utrecht in The Netherlands. Results show that predicting violence incidents with state-of-the-art performance is possible, and that using Deep Learning techniques provides a relatively small but consistent improvement in performance. We finally discuss the potential implication of our findings for the psychiatric practice.

No abstract