Using Mobile Technology to Improve Self-Regulation

Applying Novel Technologies and Methods to Inform the Ontology of Self-Regulation - Aim 4 Dartmouth Study: Using Mobile Technology to Improve Self-Regulation

This study will evaluate the degree to which engaging targets produces a desired change in medical regimen adherence (across 4-week interventions) and health behavior among smokers (n=50) and overweight/obese persons with binge eating disorder (n=50) (smoking in the former sample and binge eating in the latter sample). The investigators will employ a novel mobile behavioral assessment/intervention platform to engage targets in these samples, given that (1) it offers self-regulation assessment and behavior change tools via an integrated platform to a wide array of populations, and (2) content within the platform can be quickly modified as needed to better impact targets. This is the fourth and final phase of a study that aims to identify putative mechanisms of behavior change to develop an overarching "ontology" of self-regulatory processes. This trial builds on NCT03352713.

Health risk behavior, including poor diet, physical inactivity, tobacco and other substance use, causes as much as 40% of the illness, suffering, and early death related to chronic diseases. Non-adherence to medical regimens is an important exemplar of the challenges in changing health behavior and its associated impact on health outcomes. Although an array of interventions has been shown to be effective in promoting initiation and maintenance of health behavior change, the mechanisms by which they actually work are infrequently systematically examined. One promising domain of mechanisms to be examined across many populations and types of health behavior is self-regulation. Self-regulation involves identifying one's goals, and maintaining goal-directed behavior. A large scientific literature has identified the role of self-regulation as a potential causal mechanism in promoting health behavior. Advances in digital technologies have created unprecedented opportunities to assess and modify self-regulation and health behavior. In this project, the investigators plan to use a systematic, empirical process to integrate concepts across the divergent self-regulation literatures to identify putative mechanisms of behavior change to develop an overarching "ontology" of self-regulatory processes. This multi-year, multi-institution project aims to identify an array of putative psychological and behavioral targets within the self-regulation domain implicated in medical regimen adherence and health behavior. This is in service of developing an "ontology" of self-regulation that will provide structure and integrate concepts across diverse literatures. The investigators aim to examine the relationship between various constructs within the self-regulation domain, the relationship among measures and constructs across multiple levels of analysis, and the extent to which these patterns transcend population and context. The project consists of four primary aims across two phases of funding (UH2 and UH3 phases). Note that Aims 1-3 were conducted under our prior UH2 phase, and the investigators herein include the protocol for Aim 4 to be conducted in the UH3 phase: Aim 1. Identify an array of putative targets within the self-regulation domain implicated in medical regimen adherence and health behavior across these 3 levels of analysis. The investigators will build on Multiple PI Poldrack's pioneering "Cognitive Atlas" ontology to integrate concepts across divergent literatures to develop an "ontology" of self-regulatory processes. The expert team will catalog tasks in the self-regulation literature, implement tasks via online testing (Mechanical Turk) to rapidly obtain large datasets of self-regulatory function, assess the initial ontology via confirmatory factor analysis and structural equation modeling, and assess and revise the resulting ontology according to neural similarity patterns across tasks (to identify tasks for Aim 2). Aim 2. Evaluate the extent to which putative targets can be engaged and manipulated within the self-regulation domain both within and outside of laboratory settings. Fifty smokers and 50 overweight/obese persons with binge eating disorder will participate in a lab study (led by Poldrack) to complete the tasks identified under Aim 1. The investigators will experimentally modulate engagement of targets (e.g., stimulus set of highly palatable foods images or tobacco-related images as well as self-regulation interventions). A comparable sampling of 100 persons will participate in a non-lab study (led by Multiple PI Marsch) in which the investigators will leverage our novel mobile-based behavioral assessment/intervention platform to modulate target engagement and collect data in real-world conditions. Aim 3. Identify or develop measures and methods to permit verification of target engagement within the self-regulation domain. Led by Co-I MacKinnon, the investigators will examine cross-assay validity and cross-context and cross-sample reliability of assays. The investigators will employ discriminant and divergent validation methods and Bayesian modeling to refine an empirically-based ontology of self-regulatory targets (to be used in Aim 4). Aim 4. The investigators will evaluate the degree to which engaging targets produces a desired change in medical regimen adherence (across 4-week interventions) and health behavior among smokers (n=50 each at Dartmouth and Stanford) and overweight/obese persons with binge eating disorder (n=50 each at Dartmouth and Stanford) (smoking in the former sample and binge eating in the latter sample). The investigators will employ a novel mobile behavioral assessment/intervention platform to engage targets in these samples, given that (1) it offers self-regulation assessment and behavior change tools via an integrated platform to a wide array of populations, and (2) content within the platform can be quickly modified as needed to better impact targets. The proposed project is designed to identify valid and replicable assays of mechanisms of self-regulation across populations to inform an ontology of self-regulation that can ultimately inform development of health behavior interventions of maximal efficacy and potency. This protocol details the Aim 4 study at Dartmouth led by Multiple PI Marsch. This phase of the study takes what the investigators learned about self-regulation in the first three phases and applies it in two samples that are exemplary for "lapses" in self-regulation: individuals who smoke and overweight/obese individuals with binge eating disorder. The investigators learned in Aim 2 that many real-world conditions (e.g., temptation, negative affect) may decrease self-regulation, whereas training through the mobile intervention described below may increase self-regulation. The primary purpose of this Aim 4 study is to target self-regulation to impact health behaviors.

Laddr is an integrated, personalized, web-based self-regulation assessment and behavior change system. It integrates tools that have been shown to be effective for a wide array of behavioral phenomena ranging from substance use and abuse, mental health, risk-taking, chronic pain management, medication adherence, diet, exercise, diabetes and other chronic disease management, and smoking. The organizational structure, functionality and content within Laddr's system centrally embrace these fundamental aspects of behavior change; thus, the Laddr platform is not "diagnosis-specific" but rather enables integrated care for any combination of individuals' goals, needs, and preferences.

[Binge eating sample only] Self-reported binge eating episodes are assessed four times daily (morning, early afternoon, late afternoon/evening, and night) over a 28-day period. A binge eating episode is defined as self-reported overeating and loss of control. Overeating is assessed by the question "Since the last prompt, when you ate most recently, did you overeat?" and is scored as 0 (no) or 1 (yes). Loss of control is assessed by the question "When you ate most recently, did you lose control over your eating?" and is scored as 1 (not at all) to 5 (totally), where a 4 or 5 is considered loss of control. We are reporting the mean and standard deviations of the target behavior for each sample by week (i.e., Week 1, Week 2, Week 3, and Week 4). Data were averaged for all assessments each week to calculate the number of binge-eating episodes per week.

[Smoking sample only] Self-reported smoking is assessed four times daily (morning, early afternoon, late afternoon/evening, and night) over a 28-day period. A smoking episode is defined as self-reported smoking of more than zero cigarettes and is assessed by the question "Since the last prompt, how many cigarettes have you smoked?" Participants are asked to input a number into a number field. We are reporting the mean and standard deviations of the target behavior for each sample by week (i.e., Week 1, Week 2, Week 3, and Week 4). Data were averaged for all assessments each week to calculate the number of cigarettes smoked per week.

Inclusion criteria: Age 18-50 years Understand English sufficiently to provide informed consent Access to a computer in a setting in which the participant is comfortable providing sensitive information Use a smartphone operating system compatible with Laddr Additional inclusion criteria for binge eating sample: 27 ≤ BMI ≤ 45 kg/m2 Have binge eating disorder according to DSM-5 criteria Non-smoking (defined as no cigarettes in past 12 months-this includes former and never smokers) Confirmed interest in an eating intervention Use a smartphone compatible with Fitbit Additional inclusion criteria for smoking sample: Smoke 5 or more tobacco cigarettes/day for past year 17 ≤ BMI < 27 kg/m2 Confirmed interest in a smoking quit attempt Use a smartphone compatible with the iCO Smokerlyzer Exclusion criteria: Enrolled in Aim 2 study Any current substance use disorder o Will not exclude based on use of substances Currently pregnant or plans to become pregnant in next 3 months Lifetime history of mental disorder due to a medical condition Lifetime history of major psychotic disorders (including schizophrenia and bipolar disorder) Current use of prescription pain medications (e.g., Vicodin, oxycodone) Current use of any medication for smoking (e.g., Wellbutrin, varenicline) o Exceptions: will not screen out for nicotine replacement therapy (e.g., patch, gum, lozenge, nasal spray, inhaler) Current use of any medication for weight loss Have undergone weight-loss surgery (e.g., gastric bypass, lap band) Current nighttime shift work or obstructive sleep apnea Note: We will not exclude based on e-cigarette use. Additional exclusion criteria for binge eating sample: Compensatory behavior (e.g., purging, excessive exercise, fasting) o Already excluded as part of the DSM-5 binge eating disorder criteria Lost weight in recent past (>10 pounds in past 6 months) Currently in a weight-loss program (e.g., Weight Watchers, Jenny Craig) o Will ask about, but won't exclude on, online/mobile app weight-loss programs as part of the screener Currently on a special diet for a serious health condition Currently in therapy with a clinician for binge eating Nickel allergy (because Fitbit band contains nickel) Additional exclusion criteria for smoking sample: Currently in therapy with a clinician for smoking Binge eating behavior

After completion of the study, a de-identified dataset (i.e., stripped of all codes and other information that could be linked back to an individual participant) will be generated and made available to the research community as a whole. Informed consent procedures will ensure that participants are aware that consenting to participate in the study means consenting to inclusion in this open data set.