Search

PurposeFluid cognitive functioning (FCF), or the capacity to learn, solve problems, and adapt to novel situations, is instrumental for academic success, psychological well-being, and adoption of healthy behaviors. Our knowledge concerning factors associated with FCF, including those that may be targeted with interventions to improve outcomes, remains limited. MethodsWe used a machine learning (ML) framework in conjunction with a large battery of measures from 9,718 youth from the Adolescent Brain Cognitive Development (ABCD) study to identify factors associated with the observed variability in FCF performance. Youth age-corrected composite FCF score was derived from the National Institutes for Health Toolbox Neurocognitive Battery. A ML pipeline using a stack ensemble of multiple ML algorithms and nested cross-validation to avoid overfitting was conducted to examine factors associated with FCF. Results The identified ML algorithm explained 14.74% of variance (95%CI: 14.53-14.88%) in FCF. Among the most important factors were those that replicated previous research (e.g., socioeconomic factors), as well as novel, potentially modifiable factors, including extracurricular involvement, screen media activity, and sleepConclusionsPragmatic and scalable interventions targeting these behaviors may not only enhance cognitive performance but may also protect against the negative impact of socioeconomic and mental health factors on cognitive performance in at-risk youth. The longitudinal data from ABCD will be able to begin to assess causality by examining how changes in these factors affect subsequent cognitive performance.

Obesity is a global epidemic, yet successful interventions are rare. Up to 60% of people fail to achieve clinically meaningful, short-term weight loss (5–10% of start weight), whereas up to 72% are unsuccessful at achieving long-term weight loss (5–10% loss for ≥ 5 years). Understanding how biological, cognitive, and self-regulatory factors work together to promote or to impede weight loss is clearly needed to optimize obesity treatment. This paper describes the methodology of the Cognitive and Self-regulatory Mechanisms of Obesity Study (the COSMOS trial). COSMOS is the first randomized controlled trial to investigate how changes in multiple biopsychosocial and cognitive factors relate to weight loss and one another across two weight loss treatments. The specific aims are to: 1) Confirm that baseline obesity-related physiological dysregulation is linked to cognitive deficits and poorer self-regulation, 2) Evaluate pre- to post-treatment change across time to assess individual differences in biomarkers, cognition, and self-regulation, and 3) Evaluate whether the acceptance-based treatment (ABT) group has greater improvements in outcomes (e.g., greater weight loss and less weight regain, improvements in biomarkers, cognition, and self-regulation), than the standard behavioral treatment group (SBT) from pre- to post-treatment and 1-year follow-up. The results of COSMOS will provide critical information about how dysregulation in biomarkers, cognition, and/or self-regulation is related to weight loss and whether weight loss treatments are differentially associated with these factors. This information will be used to identify promising treatment targets that are informed by biological, cognitive, and self-regulatory factors in order to advance obesity treatment.