The Role of Behavior Genetics Research
in NIDA's Vulnerability to Drug Addiction Initiative
Conference Room A
6001 Executive Blvd., Rockville MD
May 27, 1999
- Remi Cadoret
- Andrew Heath
- John Hewitt
- Kenneth Kendler
- Matt McGue
- Ming Tsuang
Moderator: Naimah Weinberg
- What is currently covered in our funded portfolio, and what are the scientific gaps?
Our current portfolio was described as very high quality but somewhat limited in number and scope given the significance of the field for future genetics work.
Gap areas included: new investigators, training, pharmacogenetics and studies of endophenotypes, studies of specific drugs (other than nicotine), studies of minority populations to whom previously developed models may not apply, gender, clinical populations and high-risk paradigms. Some of these were characterized as "second generation" studies, building on the previous work on heritability in the general population and helping clarify the mechanisms by which drug use disorders develop. Comorbidity studies and truly developmental approaches are weak or lacking; these studies of progression and stages are key to understanding development of drug use disorders, and have major prevention and treatment implications. Prospective or sequential cohort designs are missing and felt valuable as well. Sufficient sample sizes were a concern. Replication of current studies is needed for validity and credibility in the field. Many of the above recommendations bear on the issue of phenotype definition, on which the whole gene-search enterprise will rest. A paucity of adoption studies limits understanding of gene-environment interaction.
NIDA is not funding any behavior genetic methodology studies, such as high risk sampling designs and two or three locus models being studied by other institutes. A lack of "nosologically-driven" research was noted.
The current portfolio is largely limited to domestic researchers and populations, which may limit the potential yield from studying population isolates.
Cross-fertilization among different researchers and approaches was felt to be absent.
- What can behavior genetic research contribute to the search for vulnerability genes?
Behavior genetic studies have much to offer beyond establishing heritability. Strong doubts were expressed about the readiness of the drug abuse field at this time to mount a successful candidate gene search. Many issues need to be further resolved first, to which behavior genetics, or genetic epidemiology, can contribute. In fact, the relationship between heritability and gene localization is termed "loose."
Behavior genetics can contribute to definition of preclinical phenotypes. In fact, a recommendation was made to study low exposure dependent individuals, and relatives with related deficits who don't meet full dependence criteria, rather than confining the studies to high exposure and heavily dependent individuals and family members. Genetically informed typological research is urged.
Behavior genetics can play two roles: to inform molecular studies about sample selection (phenotype issues), and to refine and guide what needs to be measured after the population is selected. The drug abuse field is felt to be still dealing with the first task, and not yet ready for the second.
Behavior genetic studies can also clarify issues about gene expression, as drug exposure may suppress or accelerate the expression of ("turn on" and "turn off") key genes.
- Recommendations for cost-effective approaches to behavior genetics research, encouraging further research, and addressing barriers to such research in the drug abuse field.
- Cost-effective approaches that may speed key findings
- Support secondary data analysis (e.g. COGA data on cocaine)
- Cross-institute collaboration in supporting grants; perhaps allowing NIH-funded centers to include R01's supported by more than one Institute
- Administrative supplements to on-going studies for marginal additional cost
- Assist currently-funded and future researchers
- Foster cross-fertilization between different approaches e.g. with animal and behavior genetics studies
- Offer guidance regarding validated instruments, without imposing requirements that limit the scientists' choice; foster common assessment tools that can promote cross-site collaborations while not generating untenable subject burden
- Educate and ensure balance by review committees
- Provide guidance regarding consent forms and ethical issues; hold meeting and generate materials to instruct IRB's
- Scientific and methodologic recommendations
- Studies of the equal environment assumption in drug abuse research, on which twin studies of heritability rest; a combination of approaches, such as twin and near-age sib studies, are needed
- Support both studies of drug specificity and of poly-drug users
- Combine cross-sectional and longitudinal approaches; consider sequential cohort designs to speed results and reduce costs
- General broadening of strategies funded in portfolio
- Support multiple threshold models; define samples by etiologic homogeneity, not just diagnostic homogeneity (by characteristics, not simply by end-points); consider sample definition on frequency and quantity, not only on dependence criteria
- Examine secular trends, cross-cultural studies to clarify environmental impacts on findings
- Consider pharmacogenetic studies
- Consider support for a national adoption registry, with broad support, to fund gene-environmental interaction studies using this increasingly rare but informative natural phenomenon
- Increase visibility of current research findings
- Special journal section
- Gordon conference or other scientific forum
- Public education re: utility of findings, not to predict individual risk but to help develop pharmacotherapies and target medication interventions
- Help speed publication