Background: Negative urgency (NegUrg) refers to impulsive behaviors aimed at avoiding negative emotions. It has been described as a predisposing factor in psychopathology, with a moderate degree of heritability. This study investigates the genetic, neuroanatomic, cognitive, and social-emotional underpinnings of NegUrg.
Methods: The present project examines data from 225 healthy participants, age 7 to 21, from the Pediatric Imaging, Neurocognition, and Genetics (PING) study. Study sample was split into subsets for model building and cross-validation (training sample, 80%) and external validation of the final model (test sample, 20%). Across three models, best subset or least absolute shrinkage and selection operator (LASSO) regression was performed to predict scores on the Negative Urgency subscale of the Urgency, Premeditation, Perseverance, Sensation Seeking, Positive Urgency (UPPS-P) Behavior Scale. Mean square error (MSE) and r2 values were evaluated as indices of model fit. Model I contained 14 variables derived from sociodemographic information, neuromedical history, cognitive, and socio-emotional data. Model II included cortical thickness measures of ten regions of interest (from five homologous pairs) in the cingulate and orbitofrontal cortices. Model III focused on genetic contributions to NegUrg, investigating 265 single nucleotide polymorphisms (SNPs) from nine candidate genes associated with dopaminergic and serotonergic transmission and metabolism, language development, and psychopathology.
Results: LASSO regression performed for Model I yielded seven significant predictors for NegUrg, including age, gender, household income, behavior problems, maternal alcohol use during pregnancy, negative affect, and anxiety. Fit indices for the training sample (MSE = 0.35, r2 = 0.25) and the test sample (MSE = 0.35, r2 = 0.30) were comparable, indicating appropriate model fit. Best subset regression results for Model II indicated that three right cingulate regions (i.e., rostral anterior, caudal anterior, and posterior cingulate cortices) were significant predictors of NegUrg. Model fit indices were, again, comparable across study samples (training sample: MSE = 0.41, r2 = 0.04; test sample: MSE = 0.47, r2 = 0.05). For Model III, the LASSO regression retained three SNPs from two candidate genes (CADM2 and SLC6A4) to predict NegUrg. Fit indices supported the validity of this model as well (training sample: MSE = 0.43, r2 = 0.16; test sample: MSE = 0.38, r2 = 0.11).
Conclusions: The results of this study provide a construct and criterion validation of the UPPS-P Negative Urgency Scale. The principal findings were that individual differences in sociodemographic factors, as well as psychological, genetic, and neuroanatomical variables related to emotional regulation (as opposed to cognitive control) contribute to NegUrg from childhood to early adulthood. Of particular interest, we provide evidence of associations between NegUrg and candidate genes CADM2 and SLC6A4. CADM2 is involved in synaptic organization, is robustly expressed in the anterior cingulate, and has been previously linked to hyperactivity and impulsivity in children. SLC6A4 is a serotonin transporter, particularly expressed in corticolimbic regions, with a role in emotion processing and emotion-driven behavior. Taken together, we provide support for a genetic basis to NegUrg that is in line with both the sociodemographic and neuroanatomical findings of the construct.
Cognition: Education, Intelligence, Memory, Attention , Gene Finding Strategies , Statistical Methods , Development
