MEASUREMENT OF IMPACT AND INSTITUTIONALIZATION IN THE WHOLE DAY FIRST GRADE PROGRAM. Jeanne Poduska1, Sheppard Kellam2, C. Brown3, Amy Windham1, John Reid4, Carla Ford1, Natalie Keegan1, 1American Institutes for Research, Baltimore, MD United States; 2American Institutes for Research, Baltimore, MA United States; 3University of South Florida, Tampa, FL United States; 4Oregon Social Learning Center, Eugene, OR United States
The measurement frame for the ongoing field trial of WD Program effectiveness and institutionalization derives explicitly from our analytic model which is based on life course/social field theory. The three components of the WD Program – effective academic instruction, family classroom partnering, and classroom behavior management – aim to reduce academic failure and behavior problems in the short term, and drug abuse, school drop-out, delinquency, and psychopathology in the long term. The measurement model incorporates indicators of each of these intermediate and outcome variables, measures of intervention fidelity, as well as individual, family, classroom, and community level variables that will help us understand mediation and moderation of program impact, e.g., classroom level of aggression, family structure, and poverty. Following from our analytic model, this presentation will describe the measures and methods used for testing the effectiveness of the WD Program as well as the institutionalization of WD. Constructs measured in the WD effectiveness trial include: 1) teacher instructional practices related to academic instruction, classroom behavior management, and family-classroom partnerships; 2) the children´s acquisition of reading-related skills, and other measures of social adaptational status; 3) psychological well being; and 4) measures of the multilevel contexts in which the interventions are set. Data collection methods involving both WD and standard setting classrooms include classroom observations using time sampling methodology, individual teacher and student interviews, and school records review. Additionally, structured implementation measures are used to assess WD fidelity and serve as the foundation for a system for continued monitoring of fidelity during WD institutionalization, an important basis for mentoring teachers, principals, and higher level school leaders.
IMPLEMENTING THE WHOLE DAY PROGRAM FOR FIRST GRADE CLASSROOMS. Carla Ford1, Jeanne Poduska1, Sheppard Kellam1, C. Brown2, Natalie Keegan1, John Reid3, Amy Windham1, 1American Institutes for Research, Baltimore, MD United States; 2University of South Florida, Tampa, FL United States; 3Oregon Social Learning Center, Eugene, OR United States
The focus of this paper is to describe the WD Program, including its theoretical underpinning, core elements, implementation fidelity measures, and structures for mentoring, modeling, and monitoring and for sustaining practices. Based on our analytic model, WD is an all-day, every day classroom-based program aimed at increasing student mastery of reading and other academic subjects and mastery of appropriate student behavior with the teacher and classmates/peers. There are three integrated components: 1) improving classroom instruction, particularly reading; 2) improving family/classroom partnering around behavior and homework; and 3) improving classroom behavior management. The first component involves enhancing teachers´ delivery of the standard reading curriculum, integrating reading instruction in all subject areas, and strengthening reading instruction for individual students. The second component involves strategies to engage and involve parents/caregivers in children´s learning – with a focus on increasing mastery of first grade reading skills. The third component, classroom behavior management, entails helping children learn how to be students and how to work together in the classroom to create a positive learning environment by monitoring their own behavior as well as their classmates.
Each intervention component addresses a specific set of observable elements or behavioral indicators that form the core of the intervention itself. These observables are the basis of the Fidelity of WD Program Checklist, a tool used to measure the extent to which defined intervention practices occur, and with what quality. WD facilitators use the checklist as a formative mentoring tool to match professional development, including intensity and duration, with individual teacher needs, current knowledge and skills, and learning goals. The professional development design includes, but is not limited to, demonstrations, guided practice, reflection, collaboration and problem solving, and peer coaching. Both the Fidelity of WD Program Checklist and professional development approaches and practices are important for sustaining and scaling up WD practices in first grade classrooms in Baltimore City in the next stages of our work, as the data warrant. The transition for sustaining and going to scale involves the roles of mentoring, modeling, and monitoring being assumed by BCPSS professional development under the aegis of the Chief Academic Officer. The multiple levels required for institutionalization will be described.
FIRST YEAR RESULTS OF THE WHOLE DAY PROGRAM FOR FIRST GRADE CLASSROOMS (WD): EFFECTIVENESS AND EARLY EXPERIENCE WITH INSTITUTIONALIZATION. Jeanne Poduska1, Sheppard Kellam1, C. Brown2, John Reid3, Linda Chinnia4, Carla Ford1, Natalie Keegan1, Amy Windham1, 1American Institutes for Research, Baltimore, MD United States; 2University of South Florida, Tampa, FL United States; 3Oregon Social Learning Center, Eugene, OR United States; 4Baltimore City Public School System, Baltimore, MD United States
In this paper, we present results on the effectiveness of WD during first grade and our experience in developing a foundation for its institutionalization. WD is directed at two correlated and confirmed early antecedents of drug abuse, co-morbid mental and behavioral disorders, and school failure, early aggressive, disruptive behavior and poor achievement. WD consists of three integrated classroom intervention components, each previously tested separately: teachers´ classroom behavior management; family/classroom partnerships regarding homework and discipline; and teachers´ instructional practices, particularly around reading. The trial is being carried out in 24 first grade classrooms in 12 elementary schools involving three consecutive cohorts of first grade children. Within each school, teachers and classrooms are randomized to intervention condition or to a standard program classroom (SC). We hypothesize that improving teachers´ practices will improve the classroom environment relative to behavior and learning which, in turn, will have an impact on children´s behavior, psychological well-being, and academic achievement. We further hypothesize that mastery in these areas will continue and will reduce later substance use, behavioral, and psychological problems. For the first cohort of students, we found by the end of first grade evidence of a reduction in off-task behavior for boys in less structured settings. We collected minute-by-minute observation of teachers´ instruction, or non-instructional practices in the classroom. After only four months of intervention, trends indicate that SC classes experienced the same amount of time of teaching in the morning as WD classes, they had less instruction in the afternoons. We found consistent trends favoring WD classes over SC on teacher practices related to behavior management and on reading, spelling, and overall quality of instruction. We will discuss the function of the three cohorts in regard to power and as systematic replicates. We also will discuss early experience in developing a multi-level structure of mentoring and monitoring as a foundation for institutionalization. The dual design employed in this trial appears to have considerable utility: 1) it provides data on WD effectiveness with two systematic replications; 2) it provides, through the first two cohorts, power for testing mediation and moderation; and 3) it aids building a foundation for institutionalization, as the data warrant.
CONCURRENT 2, INTEGRATING BIOLOGICAL AND SOCIAL FACTORS IN PREVENTION RESEARCH, Organized symposia
TRANSLATIONAL RESEARCH ON DRUG ABUSE: LINKAGES BETWEEN GENETICS AND PREVENTION.
Chair: Kevin Conway
TRANSLATIONAL RESEARCH ON DRUG ABUSE: LINKAGES BETWEEN GENETICS AND PREVENTION. Kevin Conway1, Ralph Tarter2, 1National Institutes of Health, Bethesda, MD United States; 2University of Pittsburgh, pittsburgh, PA United States
Several decades of research have shown that the etiology of drug abuse involves, at least in part, a sizeable genetic influence. Genetic epidemiologic studies of drug abuse have yielded results that are compelling in terms of consistency, magnitude of relative risk, and coherence of the message that drug abuse has genetic and environmental underpinnings in need of further explication. Like many other relatively common human disorders, drug abuse is now believed to arise from multiple genes exerting small effects, gene-by-gene interactions, gene-by-environment interactions, and a host of environmental factors and risk-conferring behaviors. Because the identification of gene-by-environment interactions is likely to prove key to understanding the development and course of drug abuse, research on the etiology of drug abuse has increasingly incorporated molecular genetic approaches. This level of translation has been far less remarkable in the prevention field, and it is the purpose of this symposium to discuss the potential (and actual) ways of integrating genetics and prevention research on drug abuse.
Dr. Leve´s paper identifies three ways in which behavior-genetic research can inform prevention science. The paper not only provides a conceptual perspective to bridge these research perspectives, it also provides several key examples to specify how this can occur. Dr. Neiderhiser´s paper addresses several ways in which two relatively separate subfields of drug abuse research – prevention and molecular genetics – represent mutually informative approaches to understanding drug abuse. Discussion will focus on the association between specific genes and specific behaviors, with considerable emphasis on environmental conditions that interact with genetic polymorphisms. Dr. Old´s paper will discuss the beginning stages of a study that integrates genetics into a 27-year follow-up of young adults (N=340) whose mothers participated in a trial of prenatal and infancy home visiting by nurses. The purpose of the new study is to examine the interaction between intervention condition, specific genetic polymorphisms, and adverse environmental exposures in preventing dysregulated behavior. Consistent with a an expectation of gene-by-environment interaction, it is hypothesized that the group with the highest rates of dysregulated behaviors will be those individuals in the control group who are both genetically vulnerable and who experience adverse environmental experiences. Dr. Tarter´s discussion will integrate the findings across papers and provide additional comments for dialogue among presenters and audience members. Dr. Tarter has a longstanding commitment to integrating research on drug abuse etiology, genetics, and prevention.
THE INTEGRATION OF BEHAVIOR GENETIC METHODS AND PREVENTION SCIENCE. Leslie Leve1, Daniel Shaw2, David Reiss3, 1Oregon Social Learning Center, Eugene, OR United States; 2University of Pittsburgh, Pittsburgh, PA United States; 3George Washington University, Washington, DC United States
Prevention studies have demonstrated that the environment has a critical influence on child outcomes, and that environmental interventions can lead to reduced problem behavior. At the same time, behavioral genetic (BG) research has shown that one third to one half of the variance in many problems behaviors is due to genetic influences. Theoretical models and research studies that bridge the findings from prevention research and BG research are greatly lacking, leaving many to question the value of BG work in prevention settings. In this presentation, we describe several ways that BG research can inform prevention science. We frame our discussion of BG-prevention integration within the context of specific examples from prevention science and behavior genetic research. First, BG research enhances our understanding of the etiology of problem behavior by providing information about genetic, shared, and nonshared environmental influences on behavior. For example, Rhee et al. (2003) found that the shared environmental effects on drug use initiation were greater than the shared environmental effects on problem drug use. This BG finding could bolster prevention research by highlighting the potential importance of the family system (and other shared environmental systems) in interventions targeting drug use initiation, as opposed to interventions targeting problem drug usage. Second, BG research can inform prevention research by identifying specific developmental periods where there is extensive change in the genetic expression of behavior. BG studies have shown that there is substantial change in some behaviors during the toddler and adolescent periods, and that much of this change can be explained by genetic influences (Reiss et al., 2000). Such behavior changes typically present parenting challenges, as caregivers adapt their parenting strategies in order to be effective given the new set of child behaviors. BG research can thus be useful in identifying opportune behaviors and times to intervene. Third, BG research can provide greater specification of the mechanisms driving parent–child interactions by identifying aspects of family interactions that reflect specific parental responses to genetically-influenced child characteristics as opposed to environmentally-mediated processes. This information can be used to inform preventionists about behavioral cycles that might be best suited to parent-based or to child-based interventions. For example, harsh parenting in response to genetically-influenced child inhibition may be more effectively served by a parent-based intervention, whereas parental intrusiveness as part of a coercive family cycle perpetuated by environmentally-influenced child behaviors may be better served by a child/family-based intervention.
PREVENTION RESEARCH AND MOLECULAR GENETICS: CAN ONE INFORM THE OTHER?. Jenae Neiderhiser1, 1George Washington University, Washington, DC United States
Understanding the role of genes in development across the lifespan has moved from a focus on twin studies to include molecular genetics. This has enabled us to identify specific genes that can help to explain specific disorders. For example, DRD2, a gene involved in dopamine reception, has been associated with novelty seeking and drug abuse (e.g. Benjamin et al, 1996; Sery et al., 2001) and the Serotonin Transporter gene has been associated with depression, anxiety and suicidal disorder (e.g. Anguelova et al., 2003). Identifying genes that contribute to specific disorders enables a better understanding of the mechanisms involved in the development of such disorders and has the potential to help focus prevention efforts.
Studies that have even greater potential in advancing our understanding of mechanisms and, thereby, are of more use for prevention, are those that focus on the interaction and correlation of specific genes and environments. There are a rapidly growing number of studies that have found evidence of interaction between a specific gene and environment (e.g. Caspi et al., 2002). Evidence of evocative genotype-environment (GE) correlation and genotype by environment interaction underscores the critical role of the environment in the development of disorders. In other words, although there may be a genetic risk for developing a disorder it is only under certain environmental circumstances that the disorder develops.
Studies that have examined the dual role of genotype and environment will be discussed with a focus on those that have examined relationships. These studies will be placed in the context of particular prevention strategies suggested by such findings. Finally, there will be a brief discussion of how prevention research can be used to better understand genetic mechanisms.
GENE X ENVIRONMENT INTERACTIONS EXAMINED IN A TRIAL OF THE NURSE FAMILY PARTNERSHIP. David Olds1, Sherry Leonard1, 1University of Colorado Health Sciences Center, Denver, CO United States
We are beginning a 27-year follow up of young adults (N=340) whose mothers participated in a trial of prenatal and infancy home visiting by nurses. This study will examine the interaction between specific polymorphisms and environmental exposures and the role the intervention may have played in preventing dysregulated behavior that results from these gene x environment interactions. In earlier phases of this trial, the program was found to reduce the number of cigarettes women smoked during pregnancy, children´s experience of child abuse and neglect, and the number of children´s arrests by child age 15.
Recent evidence suggests that prenatal tobacco exposure and child abuse and neglect each interact with specific genetic vulnerabilities that increase the likelihood of early impulsivity and oppositionality, later antisocial behavior and depression on the part of the young adult. A functional polymorphism in the gene that encodes the neurotransmitter-metabolizing enzyme monoamine oxidase A (MAOA) leads to increased rates of violent behavior when children have been abused. Children with low levels of MAOA have excess levels of neurotransmitters (e.g., norepinephrine, serotonin, and dopamine), which make them vulnerable to hyper-reactive responses to stressful events. A functional polymorphism in the promoter region of the serotonin transporter (5-HTT) gene has been found to increase the likelihood that the experience of both child abuse and neglect and stressful life events will lead to depression in adulthood. Moreover, a recent study suggests that children with a polymorphism in the Dopamine Transporter (DAT) who are exposed to tobacco during pregnancy are more likely to become impulsive and oppositional by age 3 than children with only the polymorphism or tobacco exposure by themselves.
In the current phase of follow up, we will conduct genotyping of the participants, and look at arrests, depression, substance abuse, and antisocial personality disorder as outcomes in a design matrix created by the cross-classification of Treatment conditions (Intervention versus Control), presence or absence of the polymorphism, and presence of the environmental risk factor (i.e., prenatal tobacco exposure, child abuse and neglect). We expect that the group with the highest rates of adverse outcomes will be those individuals in the control group who are both genetically vulnerable and who experience the environmental exposure.
The statistical power to detect these interactive effects will be low, but we will be able to gain preliminary insight into these relationships that can be used to guide future research.
CONCURRENT 3, METHODS, Poster forum
BINGE DRINING MODEL OF ALCOHOL USE, SUBSTANCE ABUSE AND DELINQUENCY, GENDER AND ADOLESCENT RISK TAKING, MODELING RECIPRICAL RELATIONS IN LONGITUDINAL DEVELOPMENTAL DATA
Chair: Ron Prinz
A FREQUENCY/QUANTITY/BINGE DRINKING MODEL OF ALCOHOL USE DURING THE TRANSITION TO ADULTHOOD. Karen Auerbach1, Linda Collins1, Susan Gore2, Robert Aseltine3, Mary Ellen Colten2, 1Pennsylvania State University, University Park, PA United States; 2University of Massachusetts at Boston, Boston, MA United States; 3University of Connecticut, Farmington, CT United States
Alcohol use and binge drinking rates have been found to be particularly high during the transition to adulthood (~ age 18-25), peaking around age 21 (Johnston, O'Malley, & Bachman, 2003). Alcohol use behavior has been often been measured in terms of recent use, the frequency of recent use, the quantity of recent consumption, and the frequency of recent binge drinking. Individual differences in these alcohol use behaviors have typically been analyzed using univariate methods (e.g., regression, trajectory analysis, and growth modeling). However, such methods are unable to fully capture the complexity of alcohol use behavior, which is made up of these four and other dimensions (e.g., frequency of drunkenness). Stage-sequential latent transition analysis (Guo, Collins, Hill, & Hawkins, 2000; Jackson, Sher, Gotham, & Wood, 2001) is capable of identifying individual patterns of alcohol use that are multidimensional and developmental. The present study uses latent transition analysis to identify a developmental model of alcohol use that combines these four dimensions (i.e. recent use, frequency, quantity, frequency of binge drinking) in order to better understand individual differences in the development of alcohol use over the transition to adulthood. We examined the alcohol use behavior of 1,325 emerging adults who participated in Gore, Aseltine, and Colten´s longitudinal Reducing Risk in Young Adult Transitions study from age 18 to age 22 (Gore, Kadish, & Aseltine, 2003). To determine the alcohol use model that best described these data, eight alternative multidimensional, stage-sequential models of alcohol use were identified and fit using latent transition analysis. From among these eight models, a five-stage model was identified as the best-fitting model. This model consisted of the following stages: (1) no use, (2) infrequent, low quantity, non-binge drinking, (3) frequent, low quantity, non-binge drinking, (4) frequent, high quantity, non-binge drinking, and (5) frequent, high quantity, binge drinking. Analysis of this model revealed that most participants showed either increases or stability in alcohol use over the studied age period. More notably, our analysis indicated that frequent use, high quantity of consumption, and binge drinking did not always co-occur. Rather, alcohol use is a complex behavior made up of unique patterns of various dimensions of alcohol use. Examining the relation between unique patterns of alcohol use and alcohol use and dependence and will increase our understanding of the relation between dimensions of alcohol use and problematic alcohol use. This study demonstrates the unique contributions of using latent transition analysis to model the development of alcohol use over the transition to adulthood.
THE RELATION BETWEEN THE WAXING AND WANING OF ADOLESCENT SUBSTANCE USE AND DELINQUENCY. Bethany Cara Bray1, Linda M. Collins1, 1The Pennsylvania State University, University Park, PA United States
Understanding the simultaneous relation between the development of adolescent substance use and delinquency is important for informing researchers about how changes in substance use are related to changes in delinquency and how preventive intervention programs for each of these behaviors may or may not influence the other. Research has shown that substance use and delinquent behavior can simultaneously wax and wane during adolescence. What has been difficult to determine, however, is why the two processes are occurring at the same time; for example, are substance use and delinquency independent functions simply occurring during the same age range or is one process truly influencing the other? Although traditional methods like correlation, transition matrices, regression, and ANOVA have been used to examine substance use and delinquency, these methods cannot simultaneously examine the development of these two processes over time. Associative latent transition analysis (ALTA; Flaherty, Tang, & Collins, 2003) is a new approach that models two longitudinal stage-sequential processes simultaneously. Using the ALTA methodology, a series of models was fit to a sample (N=3,225) of male participants aged 12 to 16, assessed annually from 1997 to 2001, from the National Longitudinal Survey of Youth 1997. Given that 81% of the sample initiated alcohol use, 51% initiated marijuana use, 49% destroyed property, and 38% attacked someone by the last time period, there is ample opportunity to explore the development of these processes. Four ALTA models relating substance use and delinquency cross-sectionally and longitudinally in a variety of ways are compared in order to determine the relation between adolescent substance use and delinquency. The results of these model comparisons, the relation between substance use and delinquency, and the implications related to preventive interventions will be discussed.