Wednesday, may 25, 2005 7: 00 am – 5: 00 pm registration


Download 1.65 Mb.
Size1.65 Mb.
1   ...   25   26   27   28   29   30   31   32   ...   53


DEMOGRAPHIC AND ATTITUDINAL CORRELATES OF ADOLESCENTS´ WEIGHT, PHYSICAL ACTIVITY, AND DIET FROM A NATIONAL SURVEY. Matthew Farrelly1, Kevin Davis1, James Nonnemaker1, Joanne Pais1, Doug Evans2, 1RTI International, Research Triangle Park, NC United States; 2RTI International, Washington DC, DC United States

From 1980-2000 the percentage of adolescents aged 12–19 who are overweight has tripled from 5% to 15% (Ogden et al, 2002). This rise is troubling because it is associated with increases in Type 2 diabetes among adolescents (Rosenbloom et al, 1999) and overweight adolescents are more likely to become overweight or obese adults (Ferraro, Thorpe, and Wilkinson 2003). While increases in the prevalence of obesity are well-documented among adolescents, less is known about demographic and attitudinal correlates of adolescents´ weight, physical activity, diet and nutrition.

In this paper we analyze data on from the Healthy Youth Panel Study (HYPS) which was designed to address this gap in knowledge. HYPS is a nationally representative survey of approximately 2,000 youth ages 12-18 and contains questions on adolescents´ body mass index (BMI), physical activity (e.g., weekly exercise), dieting (e.g., attempts and methods for losing weight), and nutrition habits (e.g., fast food consumption) and related attitudes, beliefs and perceptions.

We first calculated the percentage of youth who are overweight or obese based on their BMI and compared the respondent´s BMI classification to perceptions of their own weight, the prevalence of overweight adolescents, and their peers´ perception of the respondent´s weight.

We then calculated summary statistics overall and by BMI classification for weight loss attempts in the past year and 30 days; reasons and methods for losing weight; level of physical activity in the past week; and typical eating habits, including fast food, fruit, vegetable, soft drink, juice and milk consumption. Preliminary results indicate, for example, that 36% tried to lose weight in the past year and among those that tried, 38% reported losing weight to look better and 44% for better health. Seventy-four percent ate fast food at least once in the past week.

We also examine the correlation between BMI weight classification, physical activity, diet and nutrition outcomes and related attitudes, beliefs and perceptions. These attitudes include preferences for and perceived benefits of physical activity; health and social consequences diet and weight; perceptions of the causes of overweight.

Finally, we explore the parental influence on children´s physical activity and diet. Preliminary results indicate that over one-third of adolescents report that their parents rarely or never discuss poor eating habits. In addition, 73% of youth report that they always or often make their own decisions about what they eat.

CONCURRENT 9, METHODS, Grouped papers


Chair: Irene Hantman

  • Congressional B



A major consideration for many program evaluation analyses of public health programs is the use of certain geo-politcal boundaries as a means of defining populations and then analyzing the trends in interventions and outcomes for various populations. The determination of which geo-political boundaries to use, such as county, school district or neighborhood, may have significant implications for the validity or the usefulness of the program evaluation. This paper uses the experience of the Minnesota Adolescent Community Cohort (MACC) Study to explore these issues of using geo-political boundaries in defining interventions and populations in program evaluation.

The use of small geographic areas has been in the field of health services research analyzing differences in different metropolitan areas for hospital utilization and surgical procedure patterns. Another type of study has been in the field of social epidemiology, analyzing the differences by zip code or census tract for socio-economic status and health status indicators. Finally, some studies have assessed health behaviors by the residential area of the individual to draw conclusions about the impact of the “community” on the impact health promotion programs. These studies have required carefully constructed statistical methods to assure the validity of the results.

The MACC study used a hybrid of counties, municipalities, school districts and urban neighborhood districts as boundaries for the GPUs. To assess the effectiveness of various intervention programs and the overall state tobacco control effort, process data need to be collected from the intervention and the locality where the intervention is implemented. Four criteria were used to determine the boundaries for the local unit of observation when structuring these geographic-political units (GPUs): first, the boundary should maintain general relationship to patterns of local program activities; second, there should exist a similar degree of homogeneity in smoking behavior across GPUs; third, sufficient number of teens in each GPU could be selected for interview; and fourth, the boundary follows closely existing geographic and/or political limits, so that the study participants are able to self-report in which GPU they reside on the survey. The first criterion is the most crucial factor because specificity is higher and causation is more attributable when boundary agreement is stronger between outcome/process data and program implementation.

This paper will describe the GPU sampling methodology used for the MACC Study and its advantages/disadvantages, and provide an example along with quantitative results of a practical application of how this methodology is being adopted and used in the Minnesota State Synar compliance checks program.


COMPREHENSIVE GEOSPATIAL DATABASE OF RISKS AND RESOURCES AND ITS USE IN DETERMINING URBAN ADOLESCENT SUBSTANCE USE PROBABILITY.. Michael Mason1, Ivan Cheung2, J. Coatsworth3, Frank Lawrence3, Thomas Valente4, Leslie Walker1, 1Georgetown University, Washington, DC United States; 2George Washington University, Washington, DC United States; 3Pennsylvania State University, University Park, PA United States; 4University of Southern California, Alhambra, CA United States

By better understanding the interactions among individual, social, and environmental factors, prevention program effectiveness may be maximized for urban youth. In this paper we will use data collected from a Washington DC adolescent sample to construct a comprehensive geospatial database to evaluate urban youth´s social environmental risk and resource factors based upon the subjects' home and routine locations (specific geography of activity spaces). Therefore, we are engaged in interdisciplinary (mental and public health, geography, and adolescent medicine) research to further understand the relationships among urban settings, social behavior, cultural meanings, and health outcomes. A GIS approach is adopted to integrate a large array of variables at different levels of geography. For example, risk factors include proximity to crime hotspots, deterioring urban built environment, and other known potential establishments with negative influence (such as liquor stores). Furthermore, we will also use GIS to assess the subjects' accessibility to positive resources such as public library, recreational and religious opportunities. These data form `risk and protection exposure´ esitimates for each teen. To explore the usefulness of the database, a multivariate model will be developed. However we will use spatial filtering techniques to address the issue of spatial autocorrelation and principal component analysis to deal with collinearity problem among the variables.

Data are gathered from the Ecological Interview (Mason, Cheung, & Walker, 2003), an in-depth socio-spatial interview probes teens as to where they go, what they do, as well how they travel to and from these locations, the time and day of the week they go to these locations, and how long they stay in these locations, on a typical week. We ask the teens to tell us specific geographical information of their locations such as addresses, cross streets, names of parks, metro stops, coffee shops, and landmarks. We have dicotomized the teens sociospatial activities into protective and risky locations based upon our research and literature on neighborhood effects:

Protective locations:(1) health services {measured from home location}; (2) recreational facilities {measured from home location}; (3) social services {measured from home location}; (4) religious resources {measured from home and routine locations}; (5) police stations {measured from home and routine locations}; (6) entertainment opportunities {measured from home and routine locations}

Risk locations: (7) alcohol outlets {measured from home and routine locations}; (8) crime incidences {measured from home and routine locations}; (9) deteriorated housing {measured from home and routine locations}.


URBAN ADOLESCENTS´ SOCIAL NETWORKS, SOCIO-SPATIAL PROCESSES, AND THEIR HEALTH PROMOTING VERSUS HEALTH COMPROMISING BEHAVIORS.. Michael Mason1, Ivan Cheung2, J. Douglas Coatsworth3, Frank Lawrence3, Thomas Valente4, Leslie Walker1, 1Georgetown University, Washington, DC United States; 2George Washington University, Washington, DC United States; 3Pennsylvania State University, University Park, PA United States; 4University of Southern California, Alhambra, CA United States

Relatively little research focuses on urban adolescents´ social networks, socio-spatial processes (the social geography of their daily lives), and their health promoting versus health compromising behaviors. Such basic research would more accurately inform the development of targeted and appropriate multilevel community-based interventions. Consequently, leaders in the adolescent substance use field now are calling for an increased understanding of contextual influences and social development among urban, minority youth in high-risk settings. Our research fills a need by addressing multiple processes and levels that influence health behavior, including contextual and socio-spatial variables. These contextual factors are examined in detail for their impact on drug use, a high risk health behavior to which urban youth are particularly vulnerable.

The purpose of this study is to describe and analyze three domains: (1) social networks, (2) individual resources (mental health) and (3) community resources (macro and micro community level factors) of three groups of urban adolescents: (1) Non-substance using, (2) Substance using, and (3) Treatment seekers (enrolled in substance abuse treatment), as a first step toward developing a brief, contextually relevant prevention program within primary care settings. This innovative study relies extensively on Geographical Information Systems (GIS) to characterize the micro level community factors (routine locations of the urban adolescents) and the macro level community factors (objective social environmental data) hypothesized to influence urban adolescent drug use. These micro and macro level factors are used systematically to produce contextually-based risk and protection exposure estimates and to then forecast the odds of membership in one of three aforementioned substance use groups.

Few empirical studies have addressed the additive and interactive effects of these domains in the daily lives of urban adolescents, especially when measured in significant detail and at multiple levels. Measuring neighborhood health data is a point of contention in the neighborhood effects literature as census tract data are too large to meaningfully explain environmental influences and conversely, subjectively identified neighborhoods are highly malleable concepts full of personal meanings, varying locations, and interpretations. We evaluate both macro and micro data in order address this point of concern in the literature.

This poster presentation is broad and illustrative of our primary approach towards understanding complexities of urban substance use involvement and represents our unique methodological contribution of comparing the effects of macro community level data to micro level community data.

CONCURRENT 10, EFFICACY TRIALS, Organized symposium


Chair: J. David Hawkins

  • Congressional A


TAKING PREVENTION SCIENCE TO SCALE: THE COMMUNITY YOUTH DEVELOPMENT STUDY. J. David Hawkins1, Richard Catalano1, Michael Arthur1, 1University of Washington, Seattle, WA United States

Communities That Care (CTC) is an operating system grounded in prevention science, for helping communities build positive, healthy futures for their youth. The system helps communities organize and involve concerned community members to set clear priorities for action, identify gaps in policies and practices, and select, implement, and evaluate tested, effective programs and policies.

The Community Youth Development Study (CYDS) is a randomized controlled study of the Communities That Care operating system involving 24 communities across seven states. The study tests whether the CTC system significantly strengthens community protection, and reduces community levels of risk, adolescent drug use and abuse, and related problem behaviors. The study is built on the foundation of a five year descriptive study that included these 24 communities.

This symposia panel presentation presents on initial findings from the study. The first paper examines the baseline comparability of control and experimental communities on levels and trends of substance use across three waves of data over five years. The second paper analyzes the results of implementation of CTC in the experimental communities, including evidence regarding the fidelity of implementation of interventions, and the degree to which factors thought to be predictive of adoption of science-based approaches predict implementation progress. The third paper examines the power of the study to detect intervention effects, given alternative analytic models that might be applied to the data.


BASELINE COMPARABILITY BETWEEN CONDITIONS IN THE COMMUNITY YOUTH DEVELOPMENT STUDY. Beth Egan1, David Murray2, J. David Hawkins1, Michael Arthur1, 1University of Washington, Seattle, WA United States; 2University of Memphis, Memphis, TN United States

The Community Youth Development Study (CYDS) is a randomized, controlled community-level trial of the Communities that Care (CTC) operating system, a field-tested strategy for mobilizing communities to use prevention science to plan and implement community prevention services systems. Twenty-four communities have been matched on population size, percent white, crime rates, percent of students eligible for free and reduced lunch at the school district level, and unemployment rates, then randomized to either control or intervention conditions. Three waves of baseline data from the CTC Youth Survey were collected in 1998, 2000, and 2002 for 6th, 8th, and 10th grade students prior to the CTC intervention. The present study examined comparability between conditions on levels and trends of substance use across three waves of baseline data within this extended nested cross-sectional design. Missing data were addressed using multiple imputation on all items across the three waves of data. All substance use outcomes were dichotomized into either no use or any use. Dichotomized outcomes were analyzed with a generalized linear mixed model using the canonical logit link and random community-level intercepts and slopes in the GLIMMIX macro of SAS PROC MIXED. PROC MIANALYZE was used to combine the results across the 10 imputations. Two models were run: 1) an unadjusted model with fixed effects for Condition, Time, and Condition X Time; 2) the previous model adjusted for age, sex, race, family history of substance, frequency of attendance at religious services, and percent of students eligible for free and reduced lunch at the school level. Results for the unadjusted model revealed no baseline differences between conditions in levels at 2002 or slopes for demographic variables or substance use prevalence. In the adjusted model, there were no significant differences between conditions in levels at 2002. Differences between conditions in slopes in the adjusted model approached, but did not reach significance, for binge drinking and 30-day marijuana use, with prevalence rates higher in control communities in 1998 and these differences diminishing to similar prevalence levels in 2002. Findings indicate that random assignment at the community level resulted in comparable levels and trends in demographics and substance use prevalence across conditions.


ANALYSIS STRATEGIES FOR A COMMUNITY TRIAL TO REDUCE ADOLESCENT ATOD USE. David Murray1, M. Van Horn2, J. David Hawkins3, Michael Arthur3, 1University of Memphis, Memphis, TN United States; 2University of South Carolina, Columbia, SC United States; 3University of Washington, Seattle, WA United States

The Community Youth Development Study (CYDS) will evaluate the Communities That Care operating system for its effects on alcohol, tobacco, drug use, and other outcomes among adolescents resident in the 24 participating communities in seven states. The CYDS is a group-randomized trial which employs a combination of both cross-sectional and cohort designs. We use data from an earlier study that included the CYDS communities to estimate power for CYDS intervention effects given several analytic models that might be applied to the multiple baseline and follow-up surveys that define the CYDS cross-sectional design. We compare pre-post mixed-model analysis of covariance models against random coefficients models, both in one- and two-stage versions. The results indicate that the two-stage pre-post mixed-model analysis of covariance offers the best power for the primary outcomes and will provide adequate power for detection of modest but important intervention effects.


INSTALLING A SCIENCE-BASED PREVENTION SYSTEM: PREDICTING IMPLEMENTATION PROGRESS AND FIDELITY. Rose Quinby1, Abby Fagan1, J. David Hawkins1, Blair Brooke-Weiss1, Koren Hanson1, Michael Arthur1, 1University of Washington, Seattle, WA United States

Community coalitions are a popular strategy for achieving healthy youth development. And yet, there is limited empirical evidence that community mobilization produces positive youth outcomes. For example, the Robert Wood Johnson (RWJ) funded Fighting Back evaluation of 12 coalitions found that few of the coalitions reached the desired outcome of reduced substance use (D. Hallfors, et. al., 2002).

The Fighting Back evaluation offered recommendations for achieving desired effects. They recommend that if the coalitions incorporate a standardized structure, clearly defined goals and outcomes, an emphasis on implementing research-based programs, and measurement of both dose and quality of program implementation. The Communities That Care (CTC) system seeks to provide these structural elements recommended in the Fighting Back evaluation.

This paper reports early results of implementation of CTC in the experimental communities of the Community Youth Development Study. The following areas will be reviewed: progress achieving structured benchmarks and milestones specified in the CTC process and the degree to which tested, effective programs selected for implementation in experimental communities have been implemented with fidelity. The paper also investigates the degree to which factors thought to be predictive of adoption of science based approaches, such as key leader training and effective coalition functioning, predict progress in implementing tested, effective prevention programs (Arthur et. al. 2003).

3:00 PM – 3:15 PM


  • Regency Foyer

3:15 PM – 4:45 PM




Chair: Sheppard Kellam

  • Regency A


A COMBINED TRIAL OF EFFECTIVENESS AND INSTITUTIONALIZATION OF THE WHOLE DAY PROGRAM FOR FIRST GRADE CLASSROOMS (WD). Sheppard Kellam1, Jeanne Poduska2, C. Brown3, John Reid4, Carla Ford2, Natalie Keegan2, Amy Windham2, Linda Chinnia5, 1American Institutes of Research and Johns Hopkins University, Baltimore, MD United States; 2American Institutes for Research, Baltimore, MD United States; 3University of South Florida, Tampa, FL United States; 4Oregon Social Learning Center, Eugene, OR United States; 5Baltimore City Public School System, Baltimore, MD United States

This proposal is for a symposium on the dual test of 1) the effectiveness of a universal prevention program for drug abuse, psychopathology, academic failure, and the use of services, and 2) the utility of a multi-level structure for mentoring and monitoring for sustaining and institutionalizing WD, as the data warrant. Under the aegis of BCPSS, a partner for the last twenty years, we currently are testing a third generation preventive intervention, the classroom based WD, which integrates components previously tested separately into one program. According to life course /social field theory successful social adaptation to task demands in the main social fields, including learning and behavior in first grade classrooms, is preventive of later problem outcomes. WD adds enhanced reading instruction, classroom behavior management, and family-classroom partnering to the standard first grade program. We randomized schools, and then teachers and students within schools, to receive either the standard first grade program (SC) or the standard program plus WD. Implementation is monitored by intervention staff, using checklists and quality ratings reflecting the core elements of the intervention being delivered. Independent assessment staff measure impact on teacher practices, student achievement, behavior, and other aspects of social adaptation and psychological well-being in SC and WD classrooms. Three consecutive cohorts of 1st graders are involved from 24 classrooms in 12 schools. An initial cohort was followed during first grade and will be assessed in third. Two additional cohorts will serve as systematic replicates, with continuing training of WD teachers and their multi-level supervisors in the second cohort, and training of SC teachers, as the data warrant, in the third cohort. The 2nd and 3rd cohorts will test the utility of the mentoring and monitoring under increased ownership by BCPSS. The results in 1st grade of the first cohort will be presented. Our analyses will test for theoretically predicted mediating and moderating influences on intervention impact.

After an overview describing Life Course/Social Field Theory, prior results, and design, the proposed symposium includes three presentations:

1. Implementing the Whole Day Program for First Grade Classrooms

2. Measurement of Impact and Institutionalization in the Whole Day First Grade Program;

3. First Year Results of the Whole Day Program for First Grade Classrooms (WD): Effectiveness and Early Experience with Institutionalization

Share with your friends:
1   ...   25   26   27   28   29   30   31   32   ...   53

The database is protected by copyright © 2020
send message

    Main page