Construal Level as a Novel Pathway for Affect Regulation and Cancer Control
- Conditions
- Smoking ReductionCancerSmoking Cessation
- Interventions
- Behavioral: High-level construalBehavioral: Down-regulation of craving for cigarettesBehavioral: Up-regulation of goal energization
- Registration Number
- NCT04620915
- Lead Sponsor
- University of Oregon
- Brief Summary
The objective of the proposed research is to conduct a longitudinal experiment on the neurocognitive pathways and individual differences in high-level construal for affect regulation and smoking cessation. The population is adult smokers aged 25-55 who have tried and failed to quit and who are experiencing poverty. The primary endpoints are (a) the similarity in neural representation of high-level construal to one of two candidate pathways, (b) the presence of meaningful individual differences in the neural representation of high-level construal, and (c) as a secondary endpoint, the effect size of the high-level construal condition on smoking as measured by cigarettes per day.
Each of these endpoints corresponds to a specific null hypothesis. The null hypothesis for the first endpoint is that high-level construal is not significantly different in its neural representation from down-regulation of craving, which would suggest that high-level construal does not operate through distinct mechanisms from traditional treatments. The null hypothesis for the second endpoint is that the between-subjects variability in the neural representation of construal level does not significantly relate to relevant individual differences measures (e.g., traits, task behavior), which would suggest that individual differences are not meaningfully related to outcomes. Finally, the null hypothesis for the secondary endpoint is that the magnitude of the effect of high-level construal on smoking as measured by reductions in average cigarettes per day is not significantly greater than in the other conditions, which would suggest that the efficacy of the high-level construal condition is not significantly greater than a standard text-messaging intervention.
The primary endpoints will be assessed at baseline and change from pre-to-post training (8 weeks).
- Detailed Description
OVERVIEW:
The proposed work will achieve the three specific aims (two confirmatory, one exploratory) in the context of a 3-arm translational experiment. Assessments of neurocognitive mechanisms of high-level construal, as well as of two candidate pathways, will be completed at baseline and endpoint sessions. The 300 participants enrolled in the study will complete a multimodal battery that will assess neural, behavioral, and self-report indices relating to the three processes of interest - construal level, down-regulation of craving, and up-regulation of goal energization - and then are randomized to one of three experimental conditions related to those processes. This design is highly advantageous because it allows for the establishment of the mechanisms of the construal-level intervention, to compare them with the mechanisms of the other two processes, and to test whether, and to what extent, our potentially novel affect regulation strategy engages and alters those mechanisms (Aim 1); and to identify individual differences in the effects of that novel strategy (high-level construal) on patterns of brain activation, affect regulation, and cessation outcomes (Aim 2). Though the translational experiment design is not intended to be an intervention per se (because the evidence base does not yet support a full-scale trial and materials for such an intervention still need to be developed), the investigators will nonetheless quantify the effect size of high-level construal on changes in smoking so future RCTs have that information and can be adequately powered to detect an effect.
ASSIGNMENT OF PARTICIPANTS TO CONDITION:
Participants will be randomly assigned to a condition using the randomizer function in the RedCap participant tracking and management software. This assignment will happen only after participants are screened, consented, enrolled, and complete the baseline session. In other words, all 300 participants will be treated in exactly the same way through the end of the baseline session, and only at that point will RedCap be used to assign participants to their condition. All participants will have equal probability of being assigned to the conditions (i.e., 33.33% chance of assignment to each). The study coordinator will know which participant is enrolled in each condition, but researchers involved in data analysis (i.e., Drs. Berkman, Fujita, Chavez, and Weston, as well as the graduate students) as well as the research assistants who interact with the participants will be blind to condition during data gathering and analysis.
METHODS FOR SAMPLE SIZE CALCULATION AND DATA ANALYSIS:
The method for selecting the sample size was based on a power analysis of the least sensitive test - Aim 2. The key statistical tests for the three aims are as follows. For Aim 1, the test of neural similarity between conditions is essentially a paired-samples t-test on similarity scores (e.g., pattern correlations calculated within subjects). This is a high-powered comparison because the covariance within subjects of the scores is high. For Aim 2, the individual difference analyses are between-subjects correlations and regression (or partial regression or logistic regression) coefficients relating neural similarity to affect regulation scores, responses to trait and individual difference surveys, and smoking behavior as measured by cigarettes per day. This family of test has relatively low power because it is entirely between subjects. For Exploratory Aim 3, the key test is the condition-by-time interaction (i.e., differences among the groups in change in smoking behavior from pre-to-post training). This test is a within-between (sometimes called "mixed") ANOVA analysis that is more powerful that the individual difference analyses for Aim 2 but less powerful than the fully within-subjects tests for Aim 1.
INTERVENTION CONTENT:
The content for all three training arms is a set of brief (\<160 character) messages designed to enable participants to practice one specific affect regulation strategy (experimental arms). In brief, all messages in the experimental arms are generated by a large group of mTurk workers who are smokers; the process of message generation and validation has already begun. During the award period the investigators will add even more messages to the corpus and validate them. In the high-level construal condition, workers compose messages that encourage thinking about purpose of quitting and future goals (e.g., "What would quitting mean to you and your future?"). In the down-regulation of craving condition, workers compose messages that encourage effortful inhibitory or attentional control of cravings (e.g., "When you feel an urge to smoke, think about the health consequences"). And in the up-regulation of goal energization condition, workers compose messages that tie one specific core value (which the mTurk worker ranked as within his or her top three core values) to the goal of quitting (e.g., for the core value of "family", a message might be, "Quitting will help you model a healthy lifestyle for your family"). The content for the up-regulation of goal energization condition is matched to participants' own top three core values.
INTERVENTION DELIVERY:
All participants receive text messages and complete biweekly online "booster" sessions for 8 weeks.
Beginning on their quit date, participants will be sent messages via SMS text messaging 5 times each day for 28 consecutive days, then 4 times each day for the following 28 days. The investigators chose to increase the messaging frequency in the first month of cessation given the higher likelihood of relapse during this period. The order of the messages will be pseudo-randomized such that each message will be seen no more than three times across the 8-week training period, separated by at least 10 days. The timing of the messages will be adjusted for each participant to be evenly spaced starting at wake-up and ending 15 minutes before bedtime and adjusting for weekday/weekend differences. Participants will reply to each message and use a 5-point scale to rate its perceived helpfulness. Participants can also text "SOS" to the system at any point to initiate support via text if they feel tempted to smoke. Daily text messaging is an ideal delivery format for this training because it allows for in vivo participant contact at moments when smoking decisions are being made. Text messaging has very high adoption rates in the United States even among underserved communities that are typically difficult to sample densely for extended durations, and low SES users are comfortable receiving texts throughout the day. The investigators have used text messaging for experience sampling and for theory-based intervention in a community sample (see Preliminary Studies 5 and 6). Participants will provide their mobile phone number at the baseline session; those without a text message-enabled phone or those who do not wish to use their own, will be provided a prepaid phone at the end of the baseline session.
In addition to the text messaging, participants will complete biweekly online "booster" sessions using Qualtrics with a custom, personalized link sent to the participant via email and text. The booster sessions are designed to reinforce the active interventions (high-level construal, goal energization, down-regulation of craving) and provide opportunities to practice the assigned affect regulation strategy. In the sessions, participants will be reminded of their assigned strategy and write brief responses to specific prompts (e.g., "What are two specific ways that quitting now will change your future life?", "Why is quitting important to you today?", and "What are two ways you can change how you think to reduce your craving for cigarettes?"). Participants will also generate triggers to smoking (situations or cues) and practice using down-regulation of craving when they see those triggers. This exercise helps form the habit of deploying the strategy in everyday life. Not all participants have computer access, so, as in previous studies, participants will be provided access in the lab and the research team will work with the local library system to ensure all study sites are accessible from their computers.
BOOSTER COMPLETION AND RETENTION:
Dr. Berkman is currently using similar "booster" sessions, which are programmed in and are delivered via Qualtrics, in a current translational experiment on healthy eating (R01 CA211224). The automated reminders and links, which can be sent via email and text messaging depending on participant preference, greatly increase completion of the boosters. Together with the reminders, graded incentives (e.g., a monetary bonus for completing \>90% of the boosters during the study) have greatly increased completion of the at-home booster sessions.
Recruitment & Eligibility
- Status
- RECRUITING
- Sex
- All
- Target Recruitment
- 300
- Low-SES
- Persistent smokers: cigarette smokers (at least 10 cigarettes/day for at least 1 year)
- Want to quit but have tried and failed at least once
- Income-to-needs ratio (INR) is less than 2.0, meaning that their household income adjusted for household size is below 200% of the federal poverty line
- Ages 25-55
Scan
- Metal implants (e.g., braces, permanent retainers, pins)
- Metal fragments, pacemakers or other electronic medical implants
- Claustrophobia
- Weight ˃ 550 lbs.
- Women who are pregnant or believe they might be pregnant
People in this population are likely to have some comorbid psychiatric, substance use, and/or other health disorders that might pose a challenge to retention and intervention compliance. Such comorbidities are inherent to the population of interest (persistent smokers) so they will not be exclusionary criteria; instead, we will gather information about psychiatric, substance use, and medical comorbidities on intake so that we can monitor and report any associations with attrition, compliance, and effects of the experimental conditions.
E-cigarette use is acceptable - it is not an exclusionary criterion - but it will be recorded and covaried as appropriate in the analyses.
To increase the homogeneity of the sample in terms of cessation aids, we require that all participants use pharmacological cessation aids such as nicotine replacement therapy (NRT). This inclusion criterion also more realistically models how cessation happens in vivo, as medical care providers often recommend adding pharmacological assistance such as NRT to quit programs. We will provide patches or gum (e.g., Nicoderm) to participants who cannot afford. Participants who want or are able to provide their own NRT will be included as long as they agree to continue using NRT for the duration of the training period. We will monitor NRT use weekly to ensure compliance with this inclusion criterion.
No exclusions will be made on gender, race, or ethnicity, so the sample will reflect the demographic profile of the United States. Eligible participants will be scheduled for the Zoom pre-session.
Study & Design
- Study Type
- INTERVENTIONAL
- Study Design
- PARALLEL
- Arm && Interventions
Group Intervention Description High-level construal High-level construal Participants will be sent messages asking them to imagine what their lives will look like in the future if they succeed ("What would quitting mean to you and your family's future?"; Yeager et al., 2014). Effortful down-regulation of craving for cigarettes Down-regulation of craving for cigarettes Participants will be sent messages that encourage inhibitory control of cravings for cigarettes (e.g., using cognitive reappraisal or attentional control) and that provide strategies to do so (e.g., "When you feel an urge to smoke, think about the health consequences"). Up-regulation of goal energization Up-regulation of goal energization Participants will be sent messages that encourage them to consider the core values that drive their desire to quit smoking.
- Primary Outcome Measures
Name Time Method Aim 1: Neural similarity at baseline among the proposed psychological mechanisms At baseline Neural similarity as indexed by Pearson's correlations derived from the similarity matrices produced by Representational Similarity Analysis. The correlation is among the vectorized 3D images representing the patterns of BASELINE functional neural activity related to (a) high-level construal, (b) down-regulation, and (c) up-regulation of goal energization. There will be 3 correlations in total (a with b, a with c, and b with c).
Aim 1: Neural similarity in pre-post change among the proposed psychological mechanisms 56 days after the baseline session Neural similarity as indexed by Pearson's correlations derived from the similarity matrices produced by Representational Similarity Analysis. The correlation is among the vectorized 3D images representing the patterns of PRE-TO-POST CHANGE in the functional neural activity related to (a) high-level construal, (b) down-regulation, and (c) up-regulation of goal energization. There will be 3 correlations in total (a with b, a with c, and b with c).
Aim 2: Degree of prediction success of change in smoking from surveys 56 days after the baseline session Cross-validated machine learning (ML) prediction of endpoint (56-day) smoking quantity in terms of cigarettes per day based on responses to baseline responses to the Levels of Personal Agency Questionnaire. Degree of prediction will be expressed in Pearson's r correlation between (a) actual # of cigarettes per day at endpoint and (b) ML-predicted # of cigarettes per day.
Aim 2: Correlation of pattern representation of high-level construal with survey measure Within two weeks of enrollment Correlation between the similarity matrices produced by Representational Similarity Analysis and the self-report measures assessed at baseline. The measure is the Pearson's correlation between (a) the vectorized 3D image representing the patterns of baseline functional neural activity related to high-level construal and (b) the Levels of Personal Agency Questionnaire. The Outcome is the Pearson's r between (a) and (b).
Aim 2: Prediction success of change in smoking from task data 56 days after the baseline session Cross-validated machine learning prediction of endpoint (56-day) smoking quantity in terms of cigarettes per day based on responses to behavioral performance on the Construal Level Task as measured by the difference in response time in milliseconds between in the high- and low-level conditions. Degree of prediction will be expressed in Pearson's r correlation units.
Aim 2: Prediction of craving ratings from multivariate representations of high-level construal Within two weeks of enrollment Cross-validated machine learning prediction of baseline craving ratings during reactivity to personalized cigarette smoking cues based on multivariate neural representation of high-level construal. Ratings are on a 1 to 5 scale from "no craving" to "extreme craving". Degree of prediction will be expressed in Pearson's r units. Higher r values indicate better prediction of craving ratings.
- Secondary Outcome Measures
Name Time Method Aim 3: Effect size of high-level construal on smoking at endpoint 56 days after the baseline session Difference in baseline-to-endpoint change in cigarettes per day among the three conditions.
Aim 3: Time-series of the effect size of high-level construal on smoking across the training period Inclusive of days 1-56 of the training period Trajectory of the group difference of change in cigarettes per day from the baseline assessment.
Trial Locations
- Locations (2)
University of Oregon, Lewis Integrative Sciences Building
🇺🇸Eugene, Oregon, United States
University of Oregon, Social and Affective Neuroscience Laboratory
🇺🇸Eugene, Oregon, United States