Jennifer B Unger, Ph.D. - US grants

The research projects described in this TTURC involve the assessment of tobacco-related psychosocial variables, such as cultural values relevant to smoking, exposure to pro- ant anti-tobacco media, and perceived social norms. Unfortunately, validated measures of these constructs often are not available. The Measurement Core will assist the other projects by researching, developing, and validating measures of tobacco-related psychosocial variables. The Measurement Core will perform the following functions: - Conduct literature reviews to identify existing measures of tobacco- related psychosocial variables. - Develop new measures of tobacco-related psychosocial variables, such as exposure to pro- and anti-tobacco media, perceived social norms about smoking, and cultural values relevant to smoking. - Help the investigators adapt existing measures for use with early adolescents and/or respondents with limited reading proficiency. - Assess the comparability of measures across cultures and languages. - Pilot test the existing and new measures in samples of adolescents from the target populations. - Conduct reliability and validity studies of these measures, including internal consistency reliability, test-retest reliability, construct validity, content validity, concurrent validity, predictive validity, and discriminant validity. - Monitor the research literature to stay current with new published measures of tobacco-related psychosocial variables. - Provide access to students and trainees to learn about measurement issues. - Maintain a library of publishes measures, existing datasets, and documentation. - Identify and investigate other measurement issues as they arise in our ongoing research. - Provide consultation and guidance to investigators in the use of qualitative and ethnographic research methods.

The United States is becoming a multicultural society. In culturally diverse areas such as Southern California, adolescents in varying stages of acculturation are exposed to the social influences of their peers, the U.S. society, and the cultures of origin. Cultural values from multiple societies interact and influence adolescents' attitudes, beliefs, values, and health behaviors. Some of these cultural values may be risk factors for adolescent smoking, and some of these values may be protective factors. As the U.S. becomes increasingly multicultural, other cultures throughout the world also are being increasingly influenced by the U.S. culture. Societies that were previous isolated from Western influences, such as Mainland China, now are exposed to movies, music, television, and other influences from the United States and other Western cultures. Unfortunately, U.S. the media includes frequent portrayals of adolescents engaging in risk behaviors, including smoking. Little is known about how cultural values influence adolescents' decisions about smoking. A better understanding of the role of cultural values in adolescent smoking could lead to the development of smoking prevention interventions to teach adolescents ways to remain true to the values of their cultures without resorting to unhealthy behaviors such as smoking. This cross-cultural study will investigate the role of cultural values in the smoking uptake process. This study will analyze longitudinal data from two ongoing studies: a multicultural sample of adolescents in California, and a sample of Chines adolescents in Wuhan, China. Both of these data sources will include longitudinal data from initial cohorts of 7th-grade students, and both datasets will include assessments of cultural values and smoking behavior, This study will determine which cultural values are associated with the risk of smoking initiation, occasional smoking, and/or progression to established smoking among adolescents in these two countries, and it will determine whether these associations are consistent across cultures or culture-specific. Specific goals of the study are to include a Multi-ethnic Cultural Values Scale, along with measures of smoking behavior, in ongoing longitudinal surveys of adolescents in California and Wuhan, China, and to conduct statistical analyses to determine how cultural values increase or decrease the risk of smoking in these tow groups of adolescents.

DESCRIPTION (provided by applicant): Among Hispanic-American adolescents, acculturation to the United States culture has been identified as a risk factor for substance use. However, the complexities of acculturation processes and their influences on health risk behaviors remain only partially understood. Previous studies of adolescent acculturation and its consequences have not considered the family and peer contexts in which acculturation and ethnic identity formation occur. Understanding the social contexts of acculturation is important because most psychosocial risk factors for adolescent substance use involve peer and family influences. In immigrant families, multiple generations of family members are acculturating, typically at differing rates and with different patterns. Parent-child discrepancies in acculturation patterns could affect parental authority and parent-child communication, which in turn might influence the adolescents' selection of peers and activities and ultimately the adolescents' involvement in substance use. This study will examine adolescent acculturation patterns in the family context, the effects of parent-child acculturation discrepancies on family functioning, the effects of family functioning on peer social networks, the effects of social networks on adolescent substance use, and the moderating effects of community SES. The study will consist of a year of qualitative and pilot research, followed by a three-year longitudinal study and a year of qualitative research to contextualize the quantitative results. Over a three-year period, high school students in 4 predominantly Hispanic high schools in Southern California representing a range of SES (N=2000) will complete questionnaires about their acculturation patterns and family functioning. Social network analysis will be used to describe the structure of the adolescents' social networks and the prevalence of substance use among social network members. We will geocode the students' home addresses and the addresses of their social network members and use U.S. Census data to estimate the socioeconomic characteristics of the neighborhoods in which the adolescents likely spend time. Analyses will be conducted to understand the psychosocial mechanisms by which family acculturation patterns influence the risk of adolescent substance use, and the role of the social network and neighborhood contexts on those processes. The results will provide important information that can be used to develop more effective substance use prevention programs for adolescents in acculturating families.

Effect Moderation and Mediation in Prevention Programs While school and community based tobacco and drug abuse prevention programs based on social influences have produced the greatest cumulative evidence for prevention effectiveness, their effects are inconsistent. One possible cause of this inconsistency is effect moderation such that factors either intrinsic or extrinsic to some young people make them refractory to the interventions. In the first 4 years of the USC TTURC we carried out 2 longitudinal trials testing social influences prevention programs in a wide variety of cultural contexts in southern California and central China.We found evidence in both countries for a dispositional attribute moderator effect such that youth scoring high on hostility or depression experienced reductions in smoking measured 1 and 2 years post program, but those scoring low did not. In California we also found a moderator effect for social context, with a progam containing collectivist content (act for the good of others) effectively reducing smoking in homogeneous Hispanic schools but not in multi-cultural schools, and a program with an individualist approach (look after yourself) vJorking in multi-cultural schools but not homogeneous Hispanic schools. We propose a model whereby dispositional phenotypes interact with social context (programs) to alter response to prevention programming, and do so by working through mediation pathways involving cognitive, affective, and motivational arousal. Likewise we propose environment X environment interactions such that features of environment render one more or less receptive to a prevention program. Socio-cultural and dispositionai variations in populations might account for many of the inconsistencies in prevention program effects. We propose 3 studies: 1) a quantitative analysis of effect mediation by dispositional attributes and effect mediation by cultural context in 4 trials previously carried out in diverse cultural, regional, and international settings, 2) rigorous development of better phenotype an socio-cultural measures to better assess effect moderation and mediation. 3) a pilot trial designed to replicate the dispositional moderation effect and, through high intensity interventions negate, the effect by reaching otherwise refractory youth. Project 3 will do program by genotype assessments on one of the previous trials for comparison with program by phenotype analyses in this project.

DESCRIPTION (provided by applicant): Among Hispanic adolescents, cultural factors such as acculturation, perceived discrimination, and acculturative stress are risk factors for drug use. Previous studies suggest that the effects of acculturation on drug use may be mediated by family functioning; when adolescents acculturate more rapidly than their parents, the parent- child acculturation discrepancy can lead to breakdowns in family communication and cohesion and increases in family conflict, which in turn can lead to risky behaviors such as drug use among the adolescents. Although there is considerable information about the role of family acculturation patterns in drug use among Hispanic adolescents, we do not know how family acculturation patterns affect adolescents as they transition into the next stage of life, emerging adulthood. Emerging adulthood, a provisional stage in which young adults try out adult roles but may not become firmly entrenched in them, is characterized by identity exploration, residential instability, self-focus, a subjective feeling of being between adolescence and true adulthood, and increased opportunities. In general, individuals who navigate this transition smoothly emerge with a more developed sense of identity, successful interpersonal relationships, new life roles, and minimal levels of substance use. However, the high level of independence and low level of social constraints that occur during emerging adulthood also make this a high-risk period for drug use and other problem behaviors. Understanding how factors in adolescence affect drug use in early adulthood can lead to better intervention efforts for adolescents. More research is needed to determine how cultural factors influence the transition to emerging adulthood among Hispanics and to identify the risk factors for escalation of drug use during this pivotal transition. This continuation application proposes to follow an established cohort of Hispanic adolescents (12th graders in the 2008-2009 school year) for an additional 4 years to identify the predictors of their drug use trajectories during the transition to emerging adulthood. The Specific Aims of the proposed research are the following: (1) Conduct an additional 4 annual surveys of the participants; (2) Construct individual growth curves to describe their use of tobacco, alcohol, marijuana, and other drugs; (3) Identify the predictors of the growth curve parameters, including cultural factors, peer factors, substance use during high school, psychological factors, and demographic factors; (4) Evaluate emerging adulthood variables as potential mediators of these influences, including identity exploration, experimentation / possibilities, instability (residential mobility, role transitions), self-focus, and subjective experience of being in-between adolescence and adulthood; and (5) Make recommendations for the development of improved health education messages to prevent drug use among acculturating Hispanics during the transition to emerging adulthood.

Abstract: Population Core The projects in the USC-TCORS will require access to population resources with sufficient representation of tobacco product exposures, population subgroups, and longitudinal follow-up. The USC-TCORS Population Core (PC) will provide the Center with a sufficiently large, diverse, and well-followed sample to address the proposed scientific goals. We leverage data from a highly successful 2013-2018 initial funding period, in which we developed and tracked two cohorts with excellent retention. For 2018-2023, PC proposes the following aims. Aim 1. Expand, follow, track, and assess a Center-wide combined cohort. The PC will leverage data collected during the current award from two samples of adolescents, now young adults, for additional follow-up: (i) the Children's Health Study (CHS), which has extensive smoking information collected through schools since early adolescence, and complete information on tobacco product use starting at age 16-18 years; CHS participants will be aged 20-22 in 2018 (N=1,366); and (ii) the Happiness and Health (H&H) study, which has comprehensive tobacco product use information and covariates since enrollment in 2013 at age 14; H&H par- ticipants will be aged 18-19 in 2018 (N=2,534). Using now well-established methods, we will also recruit a new younger sample, the Trends in Tobacco Use Survey (TITUS) study, from 9th grade classrooms (age 14) in schools where CHS and H&H were originally drawn from (N=2,500). We will integrate these three resources into combined Center-wide cohort of 6,400 youth and young adults. The PC will conduct annual surveys on the Center-wide cohort to serve P1 and P3 aims, utilizing a common electronic survey infrastructure developed in partnership with the Administrative Core and ensuring common survey items across projects. The PC will also collect and manage a database of cohort data to be made available for the other projects, Rapid Re- sponse Projects, collaborations with other TCORSs, and junior investigator pilots (from the Career Enhance- ment Core). The database will be used for secondary data analysis projects and to identify populations with selected characteristics for recruitment into P4. The database will also be drawn upon for shared aims regard- ing social media exposure and tobacco use for P1 and P3. Finally, the database will be a resource for addi- tional original data collection projects. Aim 2. Facilitate access to other population resources outside of the Center-wide cohort within the USC Department of Preventive Medicine. The PC will enhance existing connec- tions with and access to population cohorts and archival databases including numerous diverse samples. In sum, PC cohorts will provide an unmatched population resource in terms of diversity, follow-up length, meas- urement intensity, and richness of data. Our resource will complement and provide unique information to other large samples such as the Population Assessment of Tobacco and Health survey by providing population di- versity, product diversity assessment, and tobacco product exposure analysis that is not otherwise available.

ABSTRACT: PROJECT 1 (P1) Background: Social media is a ubiquitous channel of exposure to novel e-cigarette products, particularly for youth. Yet, there is little evidence characterizing e-cigarette product diversity portrayed on social media and the impact of social media exposure to products on tobacco product use. P1 aims to provide FDA with targets for regulation using social media as a platform to understand e-cigarette product diversity, and to inform regulation that benefits the ?population as a whole? by protecting vulnerable young non-users while also considering impact on adult smokers who have switched (or are interested in switching) to e-cigarettes. Method: Two complemen- tary studies will address two aims. Aim 1. To analyze continuously collected social media posts with e-cigarette and other tobacco product-related keywords to determine patterns of occurrences that indicate trends in product marketing, conversations about e-cigarette products, and their diverse product characteristics. To this end, Aim 1 will examine new keywords and themes that co-occur with youth-oriented posts aimed at non-tobacco-users, and health-oriented posts aimed at current smokers regarding switching to e-cigarettes. P1 tests the hypothesis that there will be postings on e-cigarette products with youth-oriented themes (i.e., about e-cigarette flavors, vape clouds, discounts, and cartoons) that tend to co-occur with one another, and that such themes will be unlikely to co-occur with postings on e-cigarette products with health-oriented themes (e.g., claims about reduced harm of e-cigarettes to self and others, or about smoking cessation). P1 also tests the hypothesis that messages with youth-oriented themes will produce more user engagement (likes) and dissemination (retweets, shares) than messages without them. Aim 2. To determine whether participation (e.g., posting, liking, sharing) in e- cigarette-related social media, especially posts that contain youth-oriented themes described above, is associ- ated with tobacco product susceptibility and use among youth and young adults. P1 will analyze publicly available e-cigarette-related social media postings generated by participants in the TCORS? Center-Wide cohort (recruited with support from the Population Core) and link the data to survey responses on tobacco product use. P1 tests the hypothesis that participants who engage with and disseminate youth-oriented tobacco messages about e- cigarettes will subsequently be more likely to report transitions to higher levels of tobacco product use. Integra- tion with ?Intersections of Products with Populations? theme: P1 will study domains of product diversity studied in other USC-TCORS projects. P1 also is likely to identify novel domains of product diversity in Aim 1. P1 will address population diversity from a thematic perspective in social media messages by contrasting post- ings with themes meant to engage populations interested in switching to a lower risk product vs. vulnerable youth who are not using tobacco products. FDA Scientific domains addressed: Behavior and Marketing.

The proposed study will examine the extent to which, among Hispanic youth, the effects of culturally related stressors on alcohol/drug use and conduct problems are mediated by physiological stress responses. Prior work has established the psychosocial and biological pathways through which physiological stress responses potentiate alcohol/drug use and conduct problems. What is not known is the extent to which cultural stressors, over and above other sources of stress, predict physiological stress responses, and whether cultural stressors exert an indirect effect on alcohol/drug use and conduct problems through physiological stress responses. This question is important for prevention science because few interventions have been developed to offset the effects of cultural stressors, and our results will provide essential information regarding whether such interventions are needed ? as well as the protective mechanisms on which such interventions should focus. We propose a 3-year accelerated longitudinal cohort study with two cohorts, one beginning in the seventh grade and one beginning in the ninth grade, to carry out the study aims and to test the study hypotheses. An accelerated longitudinal design includes multiple age cohorts, where each cohort starts at a different age and each is followed for the same amount of time. Such a design allows us to examine five years of development through only three years of data collection. We will recruit and follow 300 Hispanic 7th and 9th graders in Los Angeles and Miami-Dade counties. Adolescents will be followed for 3 years and assessed at both macro (longer measures administered every 6 months) and micro (daily measurement bursts using shorter measures and saliva sample collection for cortisol assays). We will examine the moderating effects of three evidence- based protective mechanisms ? family functioning, life skills, cultural assets, and ethnic socialization ? on the direct and mediated effects of cultural stress on physiological stress responses and on alcohol/drug use and conduct problems. These moderation and moderated mediation analyses will indicate the specific mechanisms that should be targeted within prevention programs. This information is critical to decreasing disparities in conduct problems and alcohol/drug use among Hispanic adolescents.

Abstract/Project summary E-cigarettes are a big problem among high school students ? with rates of daily use exploding from 12% in 2017 to 28% in 2019. Adolescents are exposed to e-cigarette endorsements from social media influencers who promote these harmful products on behalf of e-cigarette brands as a healthier alternative to conventional cigarettes. This project proposes to comprehensively describe e-cigarette influencer marketing on Instagram (the most widely used platform for e-cigarette influencers) and on other social media platforms popular among U.S. teenagers; to examine associations between exposure to social media marketing and patterns of initiation and change in e-cigarette use; and to explore youth attitudes and ideas related to e-cigarette influencers and e- cigarette prevention campaigns. For Aim 1, we will use innovative, validated deep-learning models - a form of artificial intelligence that automatically discovers patterns in data - to evaluate trends in marketing themes and compliance with federal requirements for warning labels and sponsorship disclosures in e-cigarette promotional images and videos on Instagram, YouTube and TikTok to be collected from 2021 to 2023 (over four million posts, in all). We will use ANOVA tests to examine temporal trends in user engagement with e-cigarette marketing posts and compare the proportion of underaged users among followers of e-cigarette influencers on these platforms over the two-year period. For Aim 2, we will measure the association of exposure to e-cigarette marketing (on Instagram, YouTube, TikTok and Snapchat) with subsequent initiation, escalation or cessation of e-cigarette use among a socioeconomically- and racially-diverse cohort of 5,000 high school students living in urban areas of Southern California via ongoing longitudinal cohort surveys (2021-2023). We will use three waves of survey data in multilevel multinomial logistic regression to measure changes in exposure to- and engagement with e-cigarette marketing on Instagram, YouTube, TikTok and Snapchat; we will use repeated-measures random effects logistic regression models to evaluate the association of exposure to marketing content on these platforms with subsequent changes in e-cigarette use. Ten focus groups among survey participants will complement survey findings and will be analyzed using the Grounded Theory method. Using the social media data, surveys and focus groups, we will also quantify and characterize the level of exposure to- and perception of anti-e-cigarette prevention campaigns on social media. This project will be the first to thoroughly evaluate changes in e-cigarette use habits among youth as a result of exposure to e-cigarette marketing and interaction with e-cigarette influencers on the most popular social media platforms among adolescents.