ABSTRACT: Emotion regulation (ER) is a key transdiagnostic process and treatment target, with particular
relevance to distress disorders (depression, generalized anxiety disorder). Distress disorders, which are
characterized by heightened negative emotions, are prevalent conditions with considerable public health burden
and comparatively lower response to treatment. The efficacy of interventions for distress disorders may be
improved by investigating components of effective ER, both as instructed in the lab (ER capacity) and as
measured naturalistically in daily life (ER tendency). Effective ER is based on key components such as accurate
perceptions of one's emotions (i.e., emotional awareness) as well sensitivity to the environmental context (i.e.,
contextual sensitivity), which facilitate the selection and implementation of the appropriate ER strategy. For
individuals with distress disorders, elevated levels of perseverative negative thinking (PNT) interferes with these
processes, contributing to ER deficits and increased symptoms. This contextualized and integrative model of ER
in distress disorders has not been tested, and very little is known about the predictors or outcomes of effective
ER tendency—a critical gap given that ER capacity is irrelevant if ER is not employed skillfully in daily life.
Furthermore, examinations of ER tendency can show how individuals, considering their particular characteristics
and abilities, can optimally match particular ER strategies to the specific situations they encounter. The goal of
this application is to evaluate components of ineffective ER— both in the lab and in daily life— that influence the
severity and course of distress disorders and functioning. Consistent with NIMH strategic objectives and the
RDoC framework, this project incorporates multimodal assessment, dimensional symptom measurement, and
machine learning approaches. The proposed study will be comprised of 300 adults, oversampled for elevated
PNT (50%). Participants will complete a lab assessment to measure predictors of ER capacity, followed by a 10-
day ecological momentary assessment study examining predictors of ER tendency in daily life. To better capture
negative emotions as they occur, reports in daily life will be physiologically-triggered with algorithms that detect
potential episodes of psychological stress. Effective ER will be operationalized in multiple ways in the lab and
in daily life, using self-reported changes in affect, physiological indices, and perceived ER success. Effective ER
capacity and tendency will then be examined as predictors of distress symptom, functioning, and well-being
trajectories assessed monthly for 12 months. Additionally, an exploratory aim is to create predictive models from
a multifaceted battery of theoretically motivated variables that impact ER tendency and subsequent clinical
outcomes. To this end, machine learning will be used to build a clinically-relevant framework for how person-
level, situation-level, and ER strategy use interact to predict optimal ER. Overall, this project contributes to the
long-term goal of identifying and refining targets for personalized interventions, by more precisely isolating key
mechanisms of effective ER for specific individuals and the contexts they encounter in their daily lives.
