This R24 application seeks to upgrade and further develop a resource (Family Research Center: FRC) that supports alcohol-related research by investigators at multiple institutions, generating data that can be widely utilized and should help foster larger-scale collaborative data analyses, and novel data collection and data- sharing (`Open Science') initiatives. Historically, utilizing state vital (e.g. birth) record data, the resurce has enabled implementation of a wide range of cohort studies, using varied sampling designs, leading to over 60 funded research projects (including secondary data-analyses; biomedical/translational projects) by 24 faculty (including many early career) investigators. We seek to extend this resource, generating the capacity to further link and jointly analyze relevant state and research data, and through illustrative analyses utilizing the resource, to improve prediction models for early-onset alcohol use, transitions in alcohol use, and alcohol- related outcomes (e.g. to model neighborhood effects and their interplay with family and individual including genetic factors); and in turn to use research data and derived prediction models to inform interpretation of state data-bases and their limitations (e.g. accuracy and predictive utility), and identify areas where state data might be utilized to guide the design and implementation of future data-collections (e.g. for studies of risk- processes in families at high risks of future parental separation and later child early alcohol use). Because the structures of state data-bases are rather similar across states, the data-linking capabilities that we develop will be scalable, and can be made available to others, enabling integration of research study and state vital record or other data in a secure computing environment. Secondary data analyses, which will be used to prioritize programming and data-base linkage tasks, will also function as demonstration projects (DPs), to establish the utility of working across multiple cohort study and state data-bases. They will be organized around the themes of (i) interplay of parental alcoholism and parental separation effects and their relationship to early- onset alcohol use; (ii) interplay of neighborhood, school environment and family and individual factors in predicting transitions in alcohol use; (iii) generalizing outcomes to high-risk families identifiabe through parental recurrent drunk driving convictions. These efforts will be supplemented by (iv) work on the feasibility of matching designs (e.g. sibling) for next-generation human studies of alcoholism; (v) reconsenting for additional data-sharing, and modest additional data-collection, from research participants from the clinical neuroscience Human Connectome Project, most of whom derive from our existing cohort studies, to enable release of a broader array of alcohol-relevent variables to the research community, and guide our move to an `Open Science' model. We hope these efforts will also motivate alcohol researchers to come together in collaborative analyses of many rich existing data-sets of alcohol misuse in young people and parents, to leverage these data and analyses, and link and better inform interpretation and utilization of state data-bases.
Alcoholism is a disorder which, when it occurs in parents, is associated with important environmental risks to children - for example, increased risks of child physical and sexual abuse, increased risk of parental separation - which risks in turn predict increased probability that a child will become an early-onset drinker (e.g. before age 15), or engage in other early risk behaviors (e.g. early sex, early illicit drug use, early tobacco use, early suicide attempt); and will have increased later risks of becoming alcohol dependent or of developing other addiction or mental health problems. Here we seek to upgrade and further develop a resource that allows investigators to combine data from prospective research studies, with limited statewide individual-level vital record data, to determine how state data can improve prediction in the research data (for example, by enabling improved characterization of neighborhood risk-factors), and how research data can in turn be used to improve interpretation of state data (e.g. by establishing potential connections between family structures in a neighborhood and risks of teen alcohol, tobacco and drug involvement, and establishing ways to better identify families with children at risk). Alcoholism is also a disorder where the research o biomedical scientists, including basic (e.g. genome) scientists, clinical neuroscientists, epidemiologists and psychosocial researchers can become better integrated, a process that we seek to facilitate through upgrading and further developing a resource utilized for a wide variety of research types, implementing innovative research approaches.
| Micalizzi, Lauren; Marceau, Kristine; Brick, Leslie A et al. (2018) Inhibitory control in siblings discordant for exposure to maternal smoking during pregnancy. Dev Psychol 54:199-208 |
