Alcohol Availability
Drinking Contexts
Community and Policy
Economics
Program Portfolio

Drinking and Driving
Violence
Program Portfolio
Worksite
Program Portfolio

Youth
Program Portfolio
Advertising/Media
Program Portfolio

Minorities
Program Portfolio
AIDS
Program Portfolio
Basic Behavioral Research
Program Portfolio

A: Subcommittee for Review of Prevention Portfolio
B: Experts in Prevention
C: NIAAA Program Staff
D: NIAAA Staff, Representatives from other NIH Institutes, and Guests
E: Research Priorities from Scientists Who Prepared Reviews
F: Recommendations by Participants

Sixty-five percent of the U.S. population consumed alcoholic beverages in 1990, with 29 percent consuming alcohol weekly and 3.9 percent consuming > five drinks at least once per week (Midanik and Clark, 1994). There are a number of social, health, safety, and economic problems in the U.S. that are associated with the consumption of alcohol (see NIAAA report: Alcohol and Health 9, 1997). In most instances, alcohol is rarely the single and only cause of this problem. Of course, there are exceptions, e.g., alcoholic liver cirrhosis. In general, alcohol is a contributing factor that increases the risk beyond a base level. For example, traffic crashes are caused by a number of factors including speed, road and vehicle conditions, weather, and driver skills and experience. If the driver is drinking and impaired, the risk of a crash is increased. Another example involves the risk of birth defects. If the mother has been drinking heavily during the prenatal period, the risk of an alcohol-related birth defect (Fetal Alcohol Syndrome) is greatly increased.

It is useful to consider alcohol problems as acute and chronic. Acute problems are those which arise from drinking events usually involving heavy drinking such that the person is impaired. Acute alcohol problems may include traffic crashes (injuries and death for the driver and others); non-traffic injuries and fatalities (falls, fires, or drowning, as well as violent events resulting from domestic conflict or public assaults in which either the perpetrator or the victim has been drinking); and unprotected sex.

In 1993, 7 percent of all automobile crashes and 44 percent of fatal crashes involved alcohol use, resulting in nearly 17,500 deaths and 289,000 injuries, with a disproportionate representation of 16-24 year olds (NHTSA, 1994). Alcohol use has been implicated in 15-63 percent of fall fatalities, 13-37 percent of nonfatal injuries from falls, and in 33-61 percent of burn fatalities (Hingson and Howland, 1993). Available data suggest that alcohol use may be a major risk factor for drowning and other fatal and nonfatal injuries that occur on or near the water (Hingson and Howland, 1993).

It has been estimated that 50 percent of both victims and perpetrators of violence use alcohol, and this finding is consistent across countries, demographic subgroups, and types of violence (homicides, suicides, assaults) (Martin, 1992). Alcohol consumption represents a significant risk factor for husband-to-wife violence (Collins and Messerschmidt, 1993), and approximately 30 percent of child abuse cases may involve alcohol (Murdoch et al., 1990).

Associations have been reported between alcohol use and high-risk sexual behavior, such as failure to use condoms or engage in safe sex practices in heterosexual adults (Ericksen and Trocki, 1992) and homosexual men (Paul et al., 1991). Individuals who satisfy diagnostic criteria for alcohol abuse/dependence, consume > two drinks per day, or consume > 9 drinks on one occasion at least once per year are at a much higher risk of exposure to HIV and of developing AIDS (Stinson et al., 1992). It has also been reported that adolescents are more likely to engage in unprotected sex when they consume alcohol (Strunin and Hingson, 1992).

Examples of chronic problems resulting from long-term and often heavy drinking include alcohol dependency, alcohol abuse, alcohol poisoning, Fetal Alcohol Syndrome, and health/medical consequences.

Prevalence rate of alcohol abuse or dependence has been estimated at 7.4 percent (Grant et al., 1994). Alcohol-related morbidity is a significant factor in non-Federal, short-stay hospital discharge surveys. In 1993, an alcohol-related diagnosis was listed first in the medical records for approximately 429,000 (1.5 percent) of the discharge episodes from short-stay hospitals for persons aged 15 years and older (Caces et al., 1995).

Chronic, excessive alcohol consumption is a significant risk factor for liver cirrhosis, which accounted for 25,407 deaths in the U. S. in 1992 (DeBakey et al., 1995).

The incidence of Fetal Alcohol Syndrome (FAS) has been estimated to range from 0.97 to 1.90/1000 live births (Alcohol and Health 9, 1997). Moreover, most experts agree that the incidence of partial presentations of FAS, including neurobehavioral anomalies, is much higher.

Alcohol abuse also can contribute to a variety of social, legal, and occupational difficulties, including problems with friends, family, and spouses; fighting; problems at work; difficulties with police; financial problems; loss of control over drinking; binge drinking; and alcohol-related health problems or accidents. Two or more social consequences were reported by 12.8 percent of current drinkers in 1990 (Midanik and Clark, 1995).

There is evidence that the younger the age of drinking initiation, the greater the risk that the drinker at some time in life will develop an alcohol disorder (alcohol abuse or dependence). Young persons who began drinking before age 15 were four times as likely to develop alcohol dependence and twice as likely to develop alcohol abuse as those who began drinking at age 21. The risk for lifetime alcohol abuse decreased by 8 percent with each increasing year of age of drinking onset (Grant and Dawson, 1998). This finding is based upon cross-sectional data and needs to be confirmed with longitudinal analyses.

The estimated cost of alcohol abuse in 1992 in the U.S. was $148 billion, and it is estimated that alcohol is involved in about 100,000 deaths annually. Even though some 14 million Americans (about 7 percent of the adult population) satisfy the diagnostic criteria for alcohol abuse and/or dependence, most Americans consume alcoholic beverages without adverse effects. Moreover, many individuals who experience alcohol-related problems do not satisfy the criteria for alcohol abuse/dependence. Consequently, the prevention of alcohol-related problems needs to include these individuals, with specific interventions for specific populations.

Within each type of alcohol-involved problem, it is possible to consider levels and rates of such problems across age groups (children, youth/adolescents, young adults, adults, and elderly), gender, and racial/ethnic groups.

The prevention of excessive alcohol consumption has to take into account what has been called the "prevention paradox". This paradox was first observed in the prevention of coronary heart disease (Rose, 1981), wherein a preventive effort to reduce the cholesterol levels of only high-risk individuals had less of an overall effect in reducing the incidence of heart disease than reducing the cholesterol levels of an entire population. This same perspective was extended to alcohol consumption by Krietman (1986), who found a similar phenomenon with reference to many alcohol problems. Alcohol-dependent drinkers are much more at risk individually for a wide variety of problems. Yet, since they are a relatively small percentage of total drinkers, they do not account for the majority of alcohol-involved problems, particularly acute problems. This had been confirmed earlier by Moore and Gerstein (1981), who found that most event-based alcohol problems, such as injuries, are accounted for by moderate and heavy, non-dependent drinkers. This means that even so-called "moderate" drinkers, on the average, can incur acute problems during a heavy drinking event or while drunk.

Heavy drinking has been defined as 5 or more drinks per occasion for men and 4 or more drinkers per occasion for women, following the research of Cahalan et al. (1969), Johnston et al. (1996), and Midanik et al. (1996). This level of consumption on a single drinking occasion has often been called "binge" drinking. Wechsler et al. (1994) found that college students who over the previous year binge drank one to two times (infrequent binge drinkers) or more than two times (frequent binge drinkers), had five times and 10 times greater risk of drinking and driving when compared to those drinking college students who did not binge. These college binge drinkers were at substantially greater risk of a variety of alcohol problems during the academic year than non-binge drinking students. Duncan (1997) used the Brief Risk Factor Surveillance System from 47 states to examine the relationship between binge drinking, chronic regular drinking, and DWIs and found that DWI rates were significantly associated with binge drinking but not with chronic heavy drinking. For example, it is estimated that dependent drinkers only account for 7-15% of alcohol-involved traffic crashes.

The challenge for the prevention of alcohol problems is not to shift the target of prevention away from dependent drinkers to moderate drinkers. Rather, the challenge is to find those strategies that can reduce specific problem events, even those which result from long-term drinking. There is consistent evidence that heavy drinking, whether in a specific event or overtime, greatly increases the risk of social, health, and economic problems. Drinking while driving can increase the risk of a traffic crash for any type of drinker, and regular heavy consumption can substantially increase the risk of health problems, especially liver disease. In addition, there is evidence from studies of alcohol price and changes in alcohol availability, caused by strikes or changes in control measures, that heavy and dependent drinkers are affected by population-level prevention strategies (Mäkelä, 1992).

Thus, the challenge for the prevention of alcohol problems is formidable, because the domain for prevention is much wider than simply identifying and targeting dependent or likely dependent drinkers. This has been described as a shift from "high-risk drinkers to high-risk drinking". A much wider public health perspective is essential. The change is not trivial and has notable implications for the future investments in prevention research.

Given the significant morbidity, mortality, and social consequences associated with excessive alcohol consumption, a review of the National Institute on Alcohol Abuse and Alcoholism's (NIAAA) extramural grant portfolio seemed appropriate. Hence, NIAAA's Subcommittee for the Review of the Extramural Research Portfolio for Prevention met on 21-22 October 1998. The charge to the Subcommittee was to examine the appropriateness of the breadth, coverage, and balance of the prevention portfolio, identifying research areas that are well covered and others which are either under-investigated or which otherwise warrant significantly increased attention. The Subcommittee was asked also to provide specific advice and guidance on the scope and direction of the Institute's extramural research activities in the prevention area.

The Subcommittee for the Review of the Extramural Research Portfolio for Prevention consisted of a NIAAA Advisory Council chair and an advisory group of seven individuals. Four of these individuals have demonstrated expertise in alcohol-related areas, and three individuals have demonstrated expertise in non-alcohol-related areas (see Appendix A).

The review process was initiated by having experts (see Appendix B) in prevention prepare written assessments of the state of knowledge, gaps in knowledge, and research opportunities. NIAAA program staff (see Appendix C) presented the current extramural portfolio, categorized into the areas of policy, problem areas, youth and media, special populations and basic research and methodology, and training and career development. All information was shared with experts, selected NIAAA staff, and the chair and advisory group before the meeting.

A summary of FY 98 prevention awards is detailed below.

¹ includes SBIR awards and reimbursable funds

NIAAA Prevention Studies

On 21-22 October 1998, experts and NIAAA program staff made abbreviated presentations of their material followed by discussion among all of the participants, including representatives from other NIH Institutes and guests (see Appendix D).

Prevention can occur within a variety of sites, settings, and situations, including the family, workplace, schools, colleges and universities, medical care system, roadways/ transportation, waterways, communities, states, nation, and internationally. Prevention of alcohol-associated problems is a national priority in itself, and any set of research priorities should reflect both a consideration of scientific opportunities and findings as well as the needs of prevention policy and program decision makers at the national, state, and local levels. There is a considerable demand now for scientific-based prevention strategies by all states and communities, and there is an increasing requirement for "evidence-based" prevention by governments at all levels. The NIAAA prevention research portfolio should be at the forefront of providing research findings of value to practitioners.

After completing this process, the chair and advisory group, with input from the experts, delineated the following list of research priorities, in order of importance. The challenge to NIAAA is how to select from among the large number of potential target problems, groups, and settings for those that are sufficiently important to dedicate scare research resources or that have a sufficient research base from which to test potential prevention strategies to reduce such problems over the next 10 years. Following are the research priorities that the Subcommittee recommends.

The Subcommittee concludes that NIAAA should give emphasis in prevention research to the design and testing of prevention interventions. This emphasis is based upon two conclusions. First, there is sufficient basic and developmental research in some areas of prevention to support the actual testing of interventions in actual field settings, such as communities, colleges, and organizations. Second, even when the complete understanding of the causal mechanisms and risk factors involved in a specific set of alcohol problems is lacking, intervention testing can actually increase the scientific understanding of causal relationships. In short, the Subcommittee concludes that alcohol prevention research can best confirm its understanding of a problem by actually testing interventions based upon the best existing basic research. Prevention interventions can both test for reductions in alcohol problems as well as increase the understanding of causal mechanisms.

There are a number of possible approaches to alcohol-problem prevention, including studies of individuals, family, and group risks; specific types of prevention interventions or approaches; location or site of the prevention intervention, e.g., family, school, or workplace; and the risk factors surrounding a specific alcohol problem. This complexity has stimulated a wide range of prevention studies within the NIAAA portfolio that cover the spectrum of alcohol problems. With the exception of a few large-scale community intervention studies and state-level alcohol policy studies, the prevention portfolio can be characterized as a considerable number of highly focused but separate research studies. This has provided a rich and diverse research portfolio. Missing, however, is an emphasis on comprehensive, research-based interventions that have as their goal the actual reduction of specific alcohol problems. There is no question that the investment of research funds in a number of smaller risk studies provides greater coverage of possible topics and that large, comprehensive prevention interventions reduce the funds for such a large number of diverse studies. However, the pressing societal need for prevention strategies that actually reduce problems and the public call for more science-based prevention in communities and states, support new priorities that emphasize studies of large scale and comprehensive prevention interventions that actually reduce problems.

Studies of alcohol problems have shown that they are the result of interaction of many factors acting together and rarely result from a single cause. In like fashion, single prevention interventions appear to have limited effectiveness in isolation, but multiple strategies have the potential to be mutually reinforcing, that is, enhancing to each other. Priority should be given to studies of the effects of integration of prevention strategies, e.g., changes in hours of alcohol sale, reductions in the densities of alcohol outlets, or efforts to teach parents and school children about drinking linked with purposeful reductions in the sales of alcohol to youth.

Thus, the following priorities emphasize testing prevention interventions with recognition of the need for basic research in key areas as well as the importance of targeted specific studies.

Fund community-level (cities and towns, neighborhoods, and colleges) prevention interventions (both program and policy studies) which show potential effectiveness to: (a) reduce alcohol-involved injuries and death, and (b) delay initiation of youth drinking and reduce frequency for youth drinking, binge drinking, and alcohol-related problems for youth.

The most promising evidence of effectiveness for prevention interventions (both program and policy interventions at the local and state levels) has shown reductions in acute alcohol problems, including alcohol-involved traffic crashes, injuries, as well as assaults and deaths (across all ages). Such studies should be expanded and replicated in a variety of locations and conditions to test their generalizability as well to seek the most effective prevention interventions and mix.

In addition, there is evidence of the effects of school and community programs and minimum age policies (at the state level and enforcement at the local level) on reducing the age of drinking initiation by youth and their drinking levels. Unfortunately, the specific delay of initiation based only upon school and parent programs appears to decay over time. The clear potential, based upon existing research, is that comprehensive community prevention interventions to reduce acute alcohol-associated problems can also delay youth drinking initiation. Studies of community alcohol policy in conjunction with school/parent programs are needed to determine the most effective means to sustain this effect and to investigate the potential reductions in youth drinking levels and binge drinking as associated alcohol-related harm for youth, including violence, injuries, and unprotected sex.

Building upon the community-based prevention research to date, emphasis should be given to determining the most cost-effective combination of policy, education, and media strategies at the local level that can reduce acute alcohol problems. Priority should be given to proposed studies that build upon the developmental and efficacy research findings concerning effective interventions over the past five years. By investing in such a priority, NIAAA can support the development, testing, implementation, and evaluation of important prevention initiatives that can address both long-term public health and safety issues as well as the potential to reduce alcohol dependency.

Thus, based upon the evidence to date, this is the most promising area of alcohol prevention research and should be given priority over the next 10 years by NIAAA. Some supporting priorities as described below:

1. Establish better surveillance systems, based on development of actual and surrogate

   outcomes, measures, and indicators of alcohol involvement in non-fatal injuries and in fatalities as well as better linked survey items concerning drinking and self-reported alcohol problems.

   The development of a monitoring system to accurately measure alcohol involvement in non-traffic injuries and death is essential to any long-term assessment of the effectiveness of community/state-level interventions to reduce alcohol impairment leading to increased risk of trauma and reductions in age of drinking initiation. A useful model for the development of such a monitoring system for alcohol-involved traffic crashes is the Fatality Analysis Reporting System (FARS) of the National Highway Traffic Safety Administration. It is recommended that NIAAA take the lead in developing a Federal agency collaboration in the development of such a monitoring system that can be used to augment the data available via FARS. There is a clear need for such a database to go beyond traffic crashes. More specifically, it is recommended that a working relationship with such agencies as the U.S. Department of Justice and the Center for Disease Control be established by NIAAA with the express purpose of creating a system like FARS to monitor alcohol-involved assaults, rapes, theft, suicides, other intentional injuries, alcohol and drug overdose, and homicides. In addition, this database should develop a means to collect data on non-traffic but alcohol-related injuries and deaths, including falls, burns, and drowning.

   In addition, there are a number of state and national surveys sponsored by the federal government which include questions concerning personal drinking, high-risk drinking, and personal problems potentially related to drinking. Examples include the national household survey, the individual state needs assessment surveys, the Center for Disease Control Behavioral Risk Factor Surveillance surveys, etc. Frequently these surveys are designed and undertaken in very separate and uncoordinated ways that often limits the generalizability of results concerning alcohol across surveys at the state and national levels. The field of alcohol epidemiological research has developed the measurement of drinking and alcohol-related problems in surveys to a very high level. This has enabled prevention studies supported by NIAAA to make use of a robust range of survey items. However, such robust and field tested survey questions are often not utilized by other institutes and agencies. As a result, it is recommended that NIAAA stimulate a need among National Institute of Health as well as other centers, institutes, and departments of government to better coordinate survey efforts and to make more extensive use of tested survey questions from the alcohol research field. This will be invaluable to the advancement of prevention research.

2. NIAAA should assume the lead in developing a series of interagency cooperative initiatives and associated funding that support research concerning the prevention of alcohol-related problems that are within the responsibility of more than one Federal agency.

   Prevention of alcohol-associated problems is within the domain of a large number of Federal Agencies and Departments, and the careful and scientific study of alternative strategies for preventing such problems requires considerable funds. There is an opportunity and need for NIAAA, as the major Federal health research agency concerned with alcohol abuse and alcoholism, to assume the lead in interagency research initiatives to develop and test strategies for preventing alcohol-associated problems. These cooperative initiatives/programs can also provide the types of funding required for careful and comprehensive prevention research.

3. Confirm, through longitudinal research, that any delay in the initiation of drinking by youth contributes to reductions in future alcohol problems, including alcohol dependency, binge drinking, and alcohol-associated traffic crashes, injuries, fatalities, and violence.

   There is promising evidence that age of initiation of drinking is related to alcohol dependency. This evidence is largely based upon cross-section studies. There is research evidence based upon longitudinal studies that the legal drinking age results in a lower age of drinking initiation, lower incidence of alcohol-involved traffic crashes, and delayed drinking into young adulthood. Priority should be given to longitudinal research to determine if actual delay in youthful drinking initiation results in reduced incidence of alcohol dependency, associated health problems, and alcohol-involved acute problems such as traffic crashes, injuries, unplanned pregnancy, HIV infections, fatalities, and violence. Such tests can be a part of comprehensive community prevention interventions.

4. Determine the best means to restrict social and retail access of alcohol by youth at the local level and determine the most effective combination of local/state policy, youth sales and serving enforcement, school-based education, family/community training and involvement, and local news coverage (media advocacy) that can achieve the lowest access.

There is evidence of the relationship between frequency and volume of drinking by youth and the convenience and availability of alcohol to drink either in social or family situations as well as retail sales. Therefore, lower social and retail access to alcohol by youth has a real potential to reduce initiation of drinking, frequency and high-volume drinking, as well as reduce alcohol problems.

Monitor the implementation of state policies concerning alcohol, the administration and legal support of this implementation, and determine the effectiveness of state policies on reducing alcohol-involved problems such as traffic crashes, injuries, fatalities, and violence at the state and local level.

Every year there are a number of state policies that are directed at reducing alcohol problems. These policies provide the opportunity to undertake single and multi-policy multi-year studies and stimulate important methodological advances that capitalize on natural variation in intervention and implementation overtime. There is need to establish and maintain a research database that can identify/track state and local changes in alcohol-related policies and programs and enforcement efforts in order to evaluate effects. Pre-planned set of designs would facilitate rigorous evaluation. Such research should be used to define underlying effects in order to learn what (a) policies and (b) their administration actually results in decreased excessive consumption and related consequences.

Determine the generalizability as well as the need for custom-designed (targeted) prevention strategies for gender groups, racial/ethnic minorities, social class, elderly, and age groups.

There is clear evidence of differences in rates and frequencies of drinking as well as associated alcohol problems according to gender, age, racial/ethnic, education, and occupational factors. There are many reasons why custom-designed prevention programs may be required. However, there are insufficient research resources to design, field test, implement, and evaluate a full range of prevention programs and policies that can address the numerous combinations of individual characteristics and special groups. Therefore, this priority recommends emphasis on research to determine the generalizability of more universal prevention strategies across gender, age, and racial/ethnic groups. For example, how effective are programs/policies to reduce alcohol-involved traffic crashes, in general, with women, especially young women or how effective are community programs to increase the age of drinking initiation across all youth on drinking initiation by young African Americans, Native Americans, or Hispanics/Latinos? Attention should be given to the college-age group and testing the effectiveness in reducing drinking, especially high-risk drinking with college youth, that may have been developed for a larger age group as well as developing and testing college-age specific strategies. Also, the prevention of the Fetal Alcohol Syndrome is a good example of a gender-specific intervention directed at pregnant women who drink.

The second aspect of this recommendation, based upon the first, is the design and testing of prevention strategies which are targeted to unique age, gender, racial/ethnic, educational, and occupational groups in order to be effective or achieve more effects than strategies designed for general populations. This emphasis, based upon the results of tests of generalizability, is to design and test targeted prevention strategies when it has been shown that general population programs/policies are not effective with the special group under consideration. Such prevention designs should emphasize cultural relevance to the target group(s).

Develop and test interventions for reducing alcohol consumption in workplace that are generalizable across occupations, age groups, gender, ethnicity, and that are acceptable to management and union and is appropriate for workplaces of varying sizes and types.

There has been an historical emphasis on employee assistance programs as a means to assist existing problem drinking (often dependent) employees and family members with such problems and to assist the employee to continue to maintain productive employment. This priority is to give emphasis to workplace prevention strategies to actually reduce the level of drinking on the job, especially in high-risk occupations such as transportation, law enforcement, and heavy construction. Such prevention strategies should test educational programs only, workplace policies only, and a combination of educational approaches and workplace policies.

 Priority 5

Determine the causal relationship between alcohol and violence and explicate the risk curve between violent and non-violent incidents and levels of alcohol consumption.

There is considerable evidence that drinking is associated with violence. For example, approximately half of all murder victims had been drinking prior to their death and drinking is frequently reported in conjunction with domestic violence. Research is needed which goes beyond this simple epidemiological finding to study the factors that yield this result. We simply do not understand the causal or contributory effects of drinking by the perpetrator or the victim and/or the setting that may involve drinking on resulting violent events. Emphasis should be given to studies that seek to determine the causal or contributory relationships among drinking; settings; and violence both at the home (domestic violence) and outside of the home in public settings, as well as prevention interventions that seek to reduce violent events in such settings. Studies of prevention interventions can actually contribute to our understanding of the causal mechanisms involved in the relationship of violence and drinking.

Understand effects of changes in alcohol price alone and in interactions with alcohol control policies on a variety of alcohol problems, including traffic and non-traffic injuries and fatalities, violence/crime, and health consequences.

There is considerable evidence of the responsiveness (sensitivity in economic terms) of drinking to the price of alcohol. The higher the price of alcohol the lower the purchase and consumption of alcohol, and vice versa. There is further evidence that aggregate price levels, e.g., at a state or national level, are related to the levels of alcohol-associated problems, such as traffic crashes, homicide, cirrhosis of the liver, and other health problems. However, there is a need for studies of actual changes in price and their effects on drinking levels of youth and the general population, as well as on alcohol-associated problems.

In addition, econometric studies of alcohol price have most often been conducted without a consideration for the potential interactive effects of changes in alcohol prices with other public policies. Therefore, a priority should be given to studies that examine the interaction or synergistic effect of public policies that both change alcohol prices as well as seek to reduce alcohol problems though the enactment and implementation of non-price policies.

It should be noted that an important research opportunity exists in the study of the interactions among price, outlet density, local policies that affect alcohol availability, and other factors, such as family history of alcohol dependency and abuse, personality variables, peer influences, and demographics with alcohol problems and alcohol dependency.

Test the effects of media and product marketing/advertising on consumption and alcohol-related problems. This is an important priority which addresses efforts to determine the effects of media and alcohol product marketing on consumption especially youth drinking.

1. Determine the effect of the entertainment media and promotional campaigns on expectancies, especially for youth, related to alcohol consumption.
2. Determine the relationships between new forms of alcohol packaging, products, and marketing and alcohol-related problems.
3. Design and test the purposeful use of media to reduce heavy and high-risk consumption. This supporting priority is separate from the above other priorities which address effects of specific marketing and entertainment on drinking. This supporting priority address the potential effects of both "counter advertising" as well as planned and purposeful efforts to utilize the new media in support of public health prevention programs, especially at the local level, to reduce alcohol-involved problems.

Development of well-defined prevention strategies to reduce HIV risk by disentangling alcohol/ drug use and sexual behavior in clinical and general populations.

The relationships among heavy drinking, high-risk sexual behavior, and HIV infection are not well understood, and there is often an implicit assumption that illicit drugs are the major substances for increasing the risk of HIV. However, studies of drinking and risky sex suggest that the ability of alcohol to impair judgment and cognitive processes increase the risk of unprotected sex. Therefore, the design and testing of prevention strategies to reduce the contribution of drinking to risk of HIV exposure are needed. Such prevention strategies would both provide a means to better determine the contributory relationship of drinking to HIV infection as well as test means to reduce such exposure.

The following general observations and recommendations are also advanced by the Subcommittee for consideration by NIAAA:

Observation: The current prevention research portfolio is extensive but is not organized along any particular progression or logic. Current prevention intervention projects are not evaluated in terms of the level of research evidence that exists to support the potential value for this intervention. Further, demonstrated effective prevention interventions are not seen within a progression which tests the robustness of these findings in situations with less and less researcher control, in other words, the real world outside of the laboratory.

Recommendation: Adopt a phases of prevention research model or paradigm that identifies natural and progressive steps in the development of prevention research, beginning with foundational or pre-intervention research and progressing through diffusion testing. The Subcommittee recommends consideration of the following phases of research model by NIAAA as defined in Holder et al. (1999) and summarized below:

Phases of Research in Developing and Testing Alcohol-Problem Prevention Interventions

Phase I - Foundational Research: Basic studies to define and determine the prevalence of specific alcohol-associated problems; establish working causal models of factors and processes that yield the specific problems or increase the risk of a problem; and provide the foundations for the development of effective prevention interventions.

Phase II - Developmental (preliminary effectiveness) Studies: Preliminary studies to develop and test the likely effectiveness, safety, and early cost estimates of new interventions or an existing intervention.

Phase III - Efficacy Studies: Rigorous studies (of maximized internal validity) of intervention effects, safety, and costs under optimal conditions with maximal implementation (or availability or enforcement) and acceptance (or adoption at the community, organizational, or group level; or participation, compliance, or adherence at the individual level).

Phase IV - Effectiveness Studies: Studies of real-world effectiveness of preventive interventions with purposeful or natural variation in implementation and acceptance.

Phase V - Dissemination/Diffusion Research/Evaluation: Studies of the effects of different levels, or types of implementation, or acceptance on effectiveness, safety or costs.

Since research opportunities for the prevention of alcohol problems can be stimulated by investigators as well as naturally occurring situations and settings, this model accounts for both investigator-initiated research and program-initiated research or so called "natural experiments". Natural experimental studies sponsored by NIAAA have yielded

considerable scientific knowledge about prevention strategies that can actually reduce alcohol-involved problems. Since the timing for such opportunities is outside of the control of the scientist, a phases research program must allow for the potential to respond in a timely fashion to the opportunities to study these natural experiments. Examples of prior prevention research studies that NIAAA has supported include: minimum drinking age laws, changing the BAC limit for drunk driving, zero tolerance laws aimed at reducing youth drinking and drinking initiation, and an alcohol price change caused by a tax increase in California.

Observation: The field of alcohol-problem prevention research has had some important and remarkable findings over the past 10 years. Much is now known about potentially effective prevention strategies. Often the importance of NIAAA sponsored prevention research is not recognized or known. Also there is a diverse audience for the prevention research sponsored by NIAAA including the U.S. Congress, the National Prevention Network, National Association of State Alcohol and Drug Abuse Directors, alcohol and drug authorities in each state, etc. This is a much wider audience than the National Institute of Health. Better working communication with these diverse audiences both provides a means to inform them about the prevention research program of NIAAA but also provides a means to learn about the needs for scientific-based results in support of prevention interventions.

Recommendation: Prepare a dissemination plan for sharing the scientific findings from alcohol- problem prevention research in a way that is useful to prevention practitioners. This plan should provide for the dissemination of accomplishments of NIAAA prevention research to prevention constituencies throughout the U.S. For example, NIAAA sponsored much of the major research concerning the effects of changes in the minimum drinking age and this research was used to support national legislation which had important public health and safety implications. Subsequent research, much of it sponsored by NIAAA, further demonstrated the actual effects in terms of saved lives and reduce traffic crashes for young people.

NIAAA should develop a report immediately on "State of the Science in Alcohol-Associated Problem Prevention" that summarizes the specific general conclusions, findings, and principles resulting from prevention research sponsored by NIAAA. Such a publication would have importance to the U.S. Congress in demonstrating the value of NIAAA sponsored prevention research as well as to prevention specialists and community activists at the state and local levels.

Concurrently, it is essential that NIAAA seek a means for communications with a diverse audience of organizations to which scientific-based prevention is increasingly important. Such communication should be expected to be two-way, i.e., a means to communicate with various constituencies and to learn from them and about their needs. Further, NIAAA should include members of the hospitality industry which produces, distributes, and sells alcohol in regular communications in order to better disseminate the results of prevention research as well as learn about the perspectives of this industry.

Observation: There is concern about the visibility of prevention research within NIAAA and a question of whether the co-location with treatment and health services research within the Division of Clinical and Prevention Research is the most effective.

Recommendation: Create a prominent and highly visible organizational location for prevention research that demonstrates the strong commitment of NIAAA to prevention and provides a means to "show case" the activities and leadership of prevention research as well as stimulates even greater leadership within the field of alcohol-problem prevention.

The scientists who prepared reviews of current alcohol-prevention research efforts (Appendix E) proposed a number of suggestions for research priorities. Moreover, candidate recommendations were discussed by participants at the meeting and are listed in Appendix F.

Back to Top

(Jan Howard, Ph.D.)

The Prevention Research Branch (PRB) at NIAAA was formed in the Fall of 1987. I was appointed its Chief in November of that year. The professional staff already assigned to the PRB consisted of four persons. Two of them had Ph.D.'s (in psychology), and both of the others had substantial graduate work. The portfolio of grants assigned to the PRB consisted of a small potpourri of studies in the general area of social science. There was confusion in the minds of senior staff at NIAAA regarding the meaning of prevention research, because in the recent past there had been a Division of Prevention and Research Dissemination at the Institute which had been involved in prevention program activities. Those activities along with the Division Director (Robert Denniston) had been transferred to the newly created Office of Substance Abuse Prevention (OSAP) at ADAMHA.

In essence, the task at hand was clear: Define the meaning of prevention research, frame and implement a research agenda, attract established researchers to the field, improve the scientific quality of grant applications, and prove that prevention research in the alcohol area deserved to be recognized as a legitimate field of inquiry in its own right. Ideally, all that would be accomplished without denigrating the concept of "prevention activity" or "prevention programs", because at a higher theoretical level prevention researchers understand that the proof of the value of their research lies in its impact on the focal problem that society is trying to prevent. In the National Cancer Institute from whence I had come, the bottom line was the reduction of cancer mortality and morbidity. And the same was true for the National Heart, Lung, and Blood Institute where I had previously helped conduct multicenter clinical trials.

While the PRB was embarking on its research mission, the newly created Office of Substance Abuse Prevention was heavily engaged in its more applied mission of stimulating prevention programs throughout the United States. With help from OSAP, the National Association of State Alcohol and Drug Abuse Directors, which also includes all the state prevention coordinators, launched the first in a series of annual conferences to apprise prevention activists of the latest research findings regarding effective (and ineffective) preventive intervention strategies. Over the ensuing 10 years, there has been increasing pressure on the PRB's of NIAAA and NIDA to participate in the process of science-based technology transfer so that prevention activists supported by federal and state agencies can make informed choices. To the extent possible, we have tried to be partners in this endeavor, which can be a humbling experience when salient research is yet to be undertaken or when relevant research findings are in conflict.

With the exception of myself, none of the original PRB professional staff are still among us, although they played critical roles in implementing the PRB agenda. Early on, we were helped in our mission by several senior sociologists who jointed the PRB on yearly assignments. These scientists, included the late Larry Ross, internationally known for his research on drinking and driving, who helped mentor staff and write literature reviews; Frank Camilleri, who had strong methodological credentials and essentially wrote the RFA on Community-Based Prevention Research; and Phyllis Langton, an expert on alcohol policy research, who chaired a working group and edited an NIAAA/CSAP monograph on the Challenge of Participatory Research: Preventing Alcohol-Related Problems in Ethnic Communities.

The present staff, all but one of whom have been with the PRB more than six years, are all well published health scientist administrators with Ph.D.'s in a variety of social/ behavioral science disciplines and with special interests in selected segments of the grant portfolio. They and I are responsible for providing technical assistance to applicants and re-applicants; for writing and circulating RFA's and program announcements; for publishing monographs, guides, chapters and papers; for reviewing and editing massive amounts of material to be published by NIH, NIAAA, and other government agencies; for interfacing and collaborating with dozens of public and private organizations responsible for research and/or prevention activity; for presenting papers at national conferences; and for identifying prospective researchers and stimulating research in a variety of understudied areas.

Initially, our RFA's and program announcements were so detailed in their literature reviews and descriptions of projected research themes that it was an easy step to convert them into articles for publication. This was our way of saying: "Prevention research is a legitimate area of scientific inquiry, and here is our proof." We no longer feel the need for that degree of documentation in our program announcements. But we still believe it important to review and evaluate the state of the science. Over the past eight years, PRB staff have served as senior editors or co-editors of 14 books, monographs, guides, and special issues of journals on a wide range of prevention-research topics including, methodological issues, community trials, effects of alcohol advertising and media-based prevention strategies, alcohol-related violence, rural alcohol problems, economic issues, alcohol problems in racial/ethnic communities, women and alcohol, alcohol abuse among youth, the bio-psycho-social matrix of risk, family-based prevention research, and guides for sentencing DWI offenders and disposing of alcohol use/abuse offenses by youth. The two guides were developed in collaboration with the National Highway Traffic Safety Administration.

Over the ten years since its inception, the PRB has issued a substantial number of program announcements and requests for applications. These announcements were designed to publicize our research agenda, to stimulate applications in key subfields of prevention research (e.g., underage drinking, worksite alcohol problems, high-risk sexual behavior, alcohol abuse among the elderly, alcohol-related violence, economic issues in prevention research, racial/ethnic minorities) and to identify cross-cutting intervention and methodological issues. In the case of the RFA's, designated funds were set aside to support high-priority studies (e.g., research on community-based interventions, effects of alcohol warning labels, alcohol-related problems among racial/ethnic minorities, prevention of alcohol-related HIV/AIDS, economic issues, and underdeveloped areas of prevention research). Co-funding from OSAP (later CSAP) made it possible to issue two of these RFA's and recently a third that focused on the effects of alcohol advertising. Two new RFA's have just been issued for funding in FY 1999 concerned with preventing alcohol abuse among college students and preventing fetal alcohol syndrome. The Department of Education and CSAP collaborated in the development of the college-focused RFA and will help co-fund grants that evaluate preventive interventions.

As our research agenda was being developed, the PRB sponsored a large number of Working Group meetings that included extramural researchers and PRB staff. In many instances revised drafts of the solicited papers for these Working Groups resulted in the monographs described above. In one unique case, four working groups were held over an eight-year period in an attempt to articulate a research agenda for preventive intervention studies in work settings. The United States Coast Guard funded one of these meetings. In spite of these efforts, all current worksite studies supported by the PRB can be considered pre-intervention studies. Attention has shifted away from Employee Assistance Programs that were the core of PRB-funded worksite research when the Branch was initially formed; but with rare exceptions, proposals for other types of intervention research have either not been forthcoming or have not been funded.

Until June 1996, grant applications that were coded AP (Alcohol Prevention) were reviewed by standing and ad hoc Initial Review Groups that were organized and managed by NIAAA. The usual review committee for AP applications was known as ALCP 2, which focused primarily on prevention and epidemiological studies. For the first several years, the PRB closely monitored the fate of AP applications, subdivided by type of grant (R01, R03, R29) and the grant's Type 1/ Type 2 and new or amended status. These data clearly indicated that AP grants had a relatively steep hill to climb in the review process, particularly if they proposed to test preventive interventions.

Our response and that of the prevention research constituency involved multiple actions. Gradually, guided by PRB suggestions, the Office of Scientific Affairs added more experienced prevention researchers to the review committee. Second, two members of ALCP 2 raised the issue for discussion at the beginning of a review meeting and expressed concern that prevention research was being stigmatized as not being scientifically based, when in other fields of health research it was considered a very important area of scientific study and discourse. There was also concern that prevention intervention grants were being evaluated in terms of inapplicable methodological criteria, inconsistent with the state of the science. Third, program staff brought to the attention of the review committee that R03 and R29 applications, which were mechanisms used relatively frequently by prevention researchers for pilot studies, appeared to be evaluated by the same criteria as R01's, contrary to the rules of the review process. Committee members immediately implemented a new procedure to call attention to R03's and R29's so that they would be treated more appropriately. Simultaneously, the review outcomes for these types of grants changed appreciably, permitting more of them to be funded. Fourth, the PRB intensified its efforts to obtain set-aside funds for RFA's wherein prevention researchers would be competing solely against prevention researchers.

Fifth, and perhaps most important, the PRB initiated an extramural technical assistance program consisting of annual workshops and one-on-one mentoring for applicants who had failed to obtain a fundable score. The workshops were also used as a mechanism for recruiting new investigators to the field of alcohol-problem prevention research. This approach has since been adopted by other Branches within NIAAA to improve the scientific quality of grant submissions and to stimulate interest in various subfields of alcohol research. Attempts to assess the value of the PRB program suggested that both types of technical assistance were useful, depending on the experience and credentials of the persons being mentored. A definitive answer to the benefit question proved difficult because the participants were not randomly assigned to intervention and control categories.

Since June 1996, most grants that are coded AP have been reviewed by IRG's within the Division of Research Grants, now called the Center for Scientific Review (CSR). A particular IRG concerned with prevention studies has been reviewing AP applications from NIAAA as well as prevention applications from other institutes such as NCI and NHLBI. So far, the AP grants appear to have received their appropriate share of fundable scores. However, further organizational changes are underway within CSR that have important implications for the future review of AP prevention grants.

Finally, it is important to mention the expedited review system for natural experiments that has been implemented through the collaborative efforts of the PRB, the NIAAA Office of Scientific Affairs, and the Center for Scientific Review. In several situations, it has been necessary to expedite the review of applications that propose to study naturally occurring preventive interventions to enable the investigators to collect uncontaminated baseline data. With the permission of NIAAA and DRG (or CSR), the applicant was allowed to submit his or her application as soon as possible, off cycle; and the application was either reviewed in an upcoming regularly scheduled IRG meeting or through a specially convened telephone review. In three cases that resulted in funding, the telescoped review process even permitted the principal investigators to amend their original applications. In several other situations, the investigator applied for and received an administrative supplement to an existing grant to enable the collection of baseline data preceding the implementation of a new naturally occurring preventive intervention. Then, he or she submitted a regular grant application to study the impact of the new intervention.

This Fall, the PRB negotiated a contract with CDM Group, Inc. in Bethesda, Maryland to design and pilot test a Distance Learning and Mentoring Program for potential grant applicants. The program is intended to establish a system of mentoring and learning for investigators new to the field of alcohol prevention research, particularly investigators from underrepresented racial and ethnic minority groups. Funds to support the project come primarily from the NIH Office of Research on Minority Health, with co-funding from NIAAA. PRB staff concluded that the training workshops and one-on-one technical assistance visits that the Branch has been utilizing for many years could not meet the needs of new investigators who are associated with colleges and universities that are geographically distant from the main stream of alcohol research. These new researchers much have access to mentors who can interact with them on a more sustained basis. The proposed innovative system will utilize a variety of electronic communication technologies (such as the internet, electronic libraries, and e-mail) to pair new researchers with experienced mentors despite long geographic distances. It should also permit mentoring and guidance over a long enough period for the prospective applicants to frame and fine tune viable research proposals.

According to the generally accepted definition of prevention research used by the Office of Disease Prevention at NIH, and established by the U.S. Public Health Service in 1984, prevention research includes "only that research designed to yield results directly applicable to interventions to prevent occurrences of disease or disability, or the progression of detectable or asymptomatic disease." In addition to intervention research per se, prevention research includes so-called pre-intervention studies, the ingredients of which appear to vary somewhat from institute to institute. In the 1984 definition, pre-intervention research includes the identification of risk factors for disease and disability; development of methods for identification of disease controllable in the asymptomatic state; and refinement of methodological and statistical procedures for quantitatively assessing risk and measuring the effects of preventive interventions. A more all-inclusive category of prevention studies (prevention-related research) has also been identified. It includes "research which has a high probability of yielding results which will likely be applicable to disease prevention," namely, studies aimed at elucidating the chain of causation of acute and chronic diseases. Such basic research is perceived as generating the fundamental knowledge that contributes to the development of future preventive interventions.

This broad conception of prevention research has guided the orientation of the Prevention Research Branch over the course of its existence, permitting it to claim as relevant a fairly wide range of studies concerned with risk and protective factors, mediating processes, and methodological approaches. Thus, many of the studies supported by the PRB have focused on identifying and dissecting a variety of alcohol-related problems (such as violence, driving under the influence, absenteeism, underage drinking, and high-risk sexual behavior) as a prelude to the design, development, and testing of appropriate preventive interventions.

Within the category of preventive intervention research, the PRB distinguishes between tests (or evaluations) of interventions that occur naturally in society (natural experiments) and tests of interventions designed and developed by the investigators themselves, which we refer to as studies of investigator-initiated interventions.

Generally speaking, studies of naturally occurring prevention strategies have focused on policy interventions, such as changes in laws and regulations relevant to the reduction of alcohol abuse and related problems. And, clearly, these types of studies have made very important contributions to policy science in alcohol-problem prevention research. Using quasi-experimental techniques such as time series analysis, investigators have proved the effectiveness of laws that raised the legal minimum drinking age, which lowered legally acceptable BAC limits to "zero tolerance" for youthful drivers, that permitted administrative license revocation for drivers who violate drinking/driving laws, and that limited the availability of beverage alcohol through constraints on its sale and distribution.

Much of the research on natural experiments in the alcohol area has been funded by NHTSA; but NIAAA has broadened the sphere of such research to include studies of the effectiveness of alcohol warning labels, the effects of raising the price of beverage alcohol on consumption levels and related problems, the impact of a variety of laws and regulations on alcohol-related violence (such as suicide), the effects of bans on alcohol advertising, and the effectiveness of naturally occurring community-based prevention programs in terms of reducing deaths from alcohol-related crashes. In several RFA's and program announcements, the PRB has encouraged investigators to nest studies of investigator-initiated interventions within the context of larger natural experiments - for example, having a subset of primary-care physicians call attention to the risk of birth defects from drinking during pregnancy within the context of the warning label that mentions such risks. An ongoing study of the overall effects of the zero tolerance law in California includes a substudy of the effectiveness of relevant "investigator-initiated" preventive interventions that have been implemented in selected communities (e.g., media advocacy). The technology of natural experiments has also been used to evaluate the possible deleterious effects on drinking behavior of naturally occurring events such as the revocation by the spirits industry of their self-imposed ban against television advertising.

In at least two situations involving proposed natural experiments, the principal investigator received a fundable priority score even though the naturally occurring intervention to be studied rested on a foundation of uncertainty. This was true for the so-called "nickel a drink" referendum in California, which would have raised the tax on beverage alcohol and for the announcement by the spirits industry that they intended to advertise their products on television and radio. In the case of the pending referendum, which later failed to win voter approval, the principal investigator proved to the IRG that the baseline survey of store prices would be a worthy research project in its own right. When the referendum failed, a tax increase by the state legislature combined with a federal tax increase on beverage alcohol made it possible to continue the study past the baseline year. In the case of spirits advertising, the timetable was obviously in the hands of the industry; but shortly after baseline data were collected, the advertising began in earnest. Clearly, investigators who respond in a timely manner to naturally occurring policy changes, and agencies that fund such studies, must be willing to prepare for and adjust to somewhat unique forms of research uncertainty.

Over time, an increasing number of researchers supported by the PRB have launched studies of preventive interventions that they themselves have designed, developed, and implemented - frequently in collaboration with representatives of the system, community, or organization in which those strategies are being tested. Moreover, it is anticipated that the number of such investigators will increase appreciably in the near future, because all the recent RFA's issued by the PRB (concerned with FAS, college binge drinking, and alcohol-related sexual risk-taking) focus exclusively on intervention rather than pre-intervention research. Within that rubric, investigators may still elect to conduct natural experiments, but the option of testing investigator-initiated (i.e., "research-driven") interventions generally provide more choices. Currently, a broad range of investigator-initiated strategies is being tested. They can be categorized in terms of the alcohol problem being addressed (e.g., underage drinking, alcohol-related trauma), the settings in which research is being conducted (e.g., health-care systems, communities, worksites, schools), the populations or groups being targeted, and characteristics of the interventions themselves. It is traditional in alcohol-relevant prevention research to distinguish between interventions that focus on the environment and those that focus on individuals. This is true for naturally occurring prevention strategies as well as investigator-initiated interventions. But, as noted, natural experiments have usually focused on changes in laws and policies, which are generally interpreted to be changes in the environment. Investigator-initiated interventions are as likely to focus on targeting and changing individuals as they are on changing the environment.

Environmental interventions: The concept of "environment' in this context is not necessarily used in a consistent fashion, but there is a general understanding by people in the field that it stresses the importance of social/structural factors and forces that tend to be beyond the influence of individuals as individuals. Thus, several community-based studies stress the importance of implementing specific policies that control the availability of alcohol (e.g., conditional use permits, responsible beverage service) and/or enhance the enforcement of sanctions against sales of alcohol to minors and drinking and driving. As noted, media advocacy is another environmentally oriented strategy that is being implemented and tested at the community level. In addition, media strategies being studied include specially-designed TV and radio prevention messages for youth; a guide for parents to use in confronting the problem of drinking and driving; and a series of visits and discussions with parents that involve prepared materials about the problem of underage drinking. Depending on how media strategies are designed and implemented, they may be considered to be focusing on the environment, the individual, or both. Where parents or families are the targets of the intervention, as is occurring more frequently in alcohol prevention research, the concept of environment may be extended to include families as micro-systems, which have their own potential for norm setting, norm enforcement, and social control, as well as socialization.

In the grant portfolios of PRB staff members, the environmentally oriented interventions being tested also include anticipatory guidance protocols implemented by primary-care physicians, condom distribution procedures and safe-sex advertisements in gay bars, prevention-focused norm-setting strategies implemented by fraternities and sororities, and attempts to determine appropriate outcome measures for policy interventions in college settings.

Interventions focused on individuals: It is becoming more and more difficult to distinguish between environmental interventions and those that focus on individuals, because there is increasing appreciation of the need to combine these approaches. For example, in the current Northland studies, one of the two interactive grants involves classroom interventions that would historically be considered to have an "individual" focus, while the other grant is concerned with community task forces and policy change. Similarly, the interactive studies of media messages in Vermont involve the development of messages for TV and radio, on the one hand, and a community action component to bolster the effects of the messages. In addition to these composite studies, the PRB supports several projects that are entirely school- or curriculum-based. But here again, it is unclear whether they should be labeled as having an "individual focus," given that the schools have endorsed and are implementing the prevention curricula.

It is perhaps more appropriate to pin the "individual-focus" label on the group of studies that in one form or another involve one-on-one motivational counseling. This is true for Marlatt's successful preventive interventions with college students and for some of the HIV/AIDS intervention studies. Moreover, Goldman's "expectancy challenge" interventions with college students are concerned with changing the expectancies or beliefs of individuals rather than the large environment that may generate those beliefs.

Since its inception, the PRB has emphasize the need for research that develops and/or tests interventions that have the potential of preventing or reducing alcohol abuse and related problems. Yet, over the ten-year period of the PRB's existence, the majority of its grants have been characterized as "pre-intervention" rather than "intervention" studies. Currently, the proportion of pre-intervention studies is 55%, which is probably lower than at any previous time in the history of the Branch. To ensure that the proportion of intervention studies will continue to increase, the PRB has recently been restricting its RFA's to requests for intervention research.

The titles of the pre-intervention studies currently supported by the PRB suggest the variety and range of these research efforts. They consist of studies that explore risk and protective factors, family processes as mediating variables, environmental cognitions of drinking and of alcohol-related violence, development of alcohol cognition's and expectancies, predictors of alcohol use and abuse in rural or urban settings and among different racial and ethnic groups, methodological tools for prevention research, relationships between alcohol abuse and labor market outcomes, mediators and moderators of alcoholism inheritance, methods for detecting underage drinking, follow-up behaviors of DWI offenders, worksite influences on problem drinking, gang behavior and alcohol use, psychosocial correlates of adolescent driving behavior, and relationships between alcohol use/abuse and risky sexual behavior or various forms of violence. Sometimes the studies are longitudinal and involve competitive renewals; in other instances, time is not a key variable of interest, except as a manifestation of process.

It has gradually become apparent to us that investigators who conduct pre-intervention studies do not necessarily take the next step of moving into the intervention phase, even after they have identified critical parameters of the focal alcohol problem. Understanding the science of cause does not necessarily prepare one for understanding the ingredients of behavioral change. In some cases, however, principal investigators have successfully moved from pre-intervention to full-blown intervention research. Illustrative is Mark Goldman's transition from basic studies of expectancy theory to more applied expectancy challenge studies, or Ron Stall's transition from studies of relationships between alcohol use or abuse and risky sexual behavior to tests of safer-sex interventions for alcohol abusers in treatment. In other cases, investigators with a history of pre-intervention research have joined a team of investigators engaged in intervention studies.

Intervention research: Two new RFA's have just been released requesting applications for intervention studies that respectively address the problem of fetal alcohol syndrome and the problem of alcohol abuse on college campuses. The PRB anticipates that the responses to these RFA's will be of sufficient quantity and quality to use the existing set-aside funds. However, we would like to encourage further research in these two areas and have requested that future NIAAA funds also be committed to expanding these subfields of study.

Other area of intervention research that are considered high priority include:

Pre-intervention research: Clearly, preventive intervention research must rest on a firm foundation of more basic pre-intervention studies. But important questions concern the ingredients of that foundation. What tends to be missing from the pre-intervention portfolio of the PRB is basic research on processes of behavioral and social change as opposed to studies of causal processes or risk and protective factors. It is also clear that in certain areas of prevention research, none of the studies supported by the PRB are testing investigator-initiated interventions or even naturally occurring prevention strategies. In these situations, it is important to consider what further pre-intervention research may be necessary to move the process forward. Alternatively, it may be necessary to attract more intervention researchers to the field of alcohol-focused prevention research.

Back to Top

State of Knowledgez (Paul J. Gruenewald, Ph.D.)

There is a rapidly growing research literature on theoretical models and empirical relationships between the physical and social availability of alcohol, alcohol use and alcohol-related problems. Theoretical contributions include the development of (1) demographic-ecological models relating the geographic distribution of alcohol outlets to population growth and alcohol use, (2) geographic-ecological models relating the distribution of outlets to the distribution of alcohol-related traffic crashes, and (3) social-ecological models relating the geographic distribution of population characteristics and outlets to availability and alcohol-related violence. Empirical contributions include (1) conclusive demonstration of the effects of large-scale changes in availability on alcohol use and problems, (2) successful assessment of the effects of outlet densities on alcohol sales independent of reciprocal causation, (3) extraordinary growth in social area analyses relating patterns of violence to patterns of availability, (4) the development and application of advanced geostatistical methods for the analysis of community-level data, and (5) the preliminary development of multi-level approaches to the analysis of the relationships between environmental measures of availability and individual alcohol problems. Although important future research agendas can be identified in every area of study, greatest emphasis should be placed upon the assessment of neighborhood alcohol availability in relation to community-based alcohol-related problems.

Back to Top

State of knowledge (A. James McKnight, Ph.D.)

Historically, efforts to prevent injury from alcohol impairment have been largely directed to modification of long-term drinking behavior through education and information, counseling and, in the case of alcohol dependence, therapy. Over the past two decades, prevention efforts have been increasingly focused upon the context in which drinking occurs, attempting to prevent episodes of impairment from drinking and harm to the impaired. The nature of prevention varies greatly as a function of drinking context. From the viewpoint of prevention, drinking contexts are more readily classified in terms of the resources available to provide prevention than in the physical setting. Dividing prevention resources into the following three categories will accommodate the most important differences among settings.

Servers - who provide alcohol in licensed on-premise establishments.
Hosts - who provide alcohol in social gatherings.
Peers - who accompany drinkers where alcohol is served.

Servers of alcohol in licensed establishments are most clearly distinguished from other resources by the extent to which their prevention activity is prescribed by law and regulation. The sale of alcohol by licensed on-premise establishments has been the primary focal point of efforts to prevent harm from the over-consumption of alcohol. These prevention efforts include direct-intervention bartenders and waitpersons, as well as policy-level prevention by managers and owners. The strongest incentive to prevention of alcohol service to the already impaired appears to be rigorous enforcement of laws in all but one state prohibiting such service. Even a modicum of enforcement yields substantial increases in compliance, offering potential cost-benefit relationships exceeding those of almost any available alcohol countermeasure.

Most of the drinking that occurs in the company of others occurs at gatherings hosted by some individual or organization. While not subject to the same degree of outside control as licensed establishments, hosts at least provide a focal point for organized prevention efforts. There is a relative lack of research or programs addressing involvement of social hosts in preventing harm from overdrinking among guests. Obstacles include general lack of public support for legal measures holding hosts liable for harm, lack of any good route of access for formal programs, and difficulty that hosts have in monitoring the drinking of guests. Access to hosts is also limited, with public information being the only way of reaching the large numbers of social hosts. The objectives of such efforts address any and all who are present when drinking occurs, and do not differentiate hosts from peers.

Often referred to as "peer intervention", the participation of the drinker's associates in preventive efforts has gained widespread public attention. It is of particular value in drinking that occurs outside of licensed establishments and hosted gatherings and which lack any one person with a legal or social obligation to intervene. While the concept of peer intervention applies to all that drink, it has been most frequently advanced and studied within the under-age population. Attempts to induce people to intervene directly in the drinking and driving of intoxicated associates have been most successful with youth, both because they seem more receptive to the idea of intervention and because they can be reached through high school and college with the kind of group-interactive training believed necessary to build skill and confidence in carrying out intervention. Among adults, the proposition that friends don't let friends drive drunk seems best advanced through non-confrontational approaches.

Back to Top

State of Knowledge (Alexander C. Wagenaar, Ph.D. and Traci L. Toomey, Ph.D.)

There are 319 research evaluation studies for 35 alcohol control policies. By far the most studied alcohol policy is the minimum legal drinking age, with 102 studies. The second most studied policy is alcohol excise tax, with 50 studies. Two other policies, warning labels on products and density of alcohol outlets had more than 20 studies each and 26 studies were identified that assessed amount and effects of general exposure to alcohol advertising. Other than these five alcohol policies, all other alcohol policies have received little, if any, scientific research attention.

Alcohol control policies can be categorized by how they affect drinking behavior: (1) how, when, and where alcohol is sold, (2) where and when alcohol is consumed, (3) price of alcohol, (4) broader social environment surrounding alcohol use, (5) how existing policies are enforced, and (6) how underage youth obtain alcohol. Most of the extant research literature on alcohol policy is in 1, 3, 4, and 6. The dearth of studies of policy implementation and enforcement is notable.

The 35 policies may also be differentiated by whether they can be implemented at national, state, local, or institutional levels.

Alcohol consumption is the most common outcome measure (129 studies). The second most commonly analyzed outcome measure is traffic crashes and related drink-driving measures (93 studies). Only four policy studies used measures of unintentional injuries other than traffic crashes, and only nine studies examined intentional injury outcomes. Thirty-one policy studies examined other alcohol-related problems, including health, crime, school, family, and other problem indicators. Thus, one factor that impedes the growth of research literature on alcohol policies is limited data on multiple outcomes.

The overwhelming majority of studies are focused on state-level policies, with national policies receiving the second most research attention. Local and institutional policies have received little research attention. There were at least 463 alcohol control bills introduced during the 1997 state legislative sessions and 148 were enacted. Most legislative activity occurred for alcohol policies that affect how, when, and where alcohol is sold, with 229 bills introduced and 86 passed. Another large amount of legislative activity occurred for policies affecting the price of alcohol and the social environment.

For only a few alcohol policy topics are the research evidence extensive - excise taxes/price, warning labels, and legal drinking age. Even for these topics, many research questions remain unanswered, most notably what are their effects on alcohol-related problems beyond drinking behavior and traffic crashes? There is a moderate amount of evidence in the literature for privatization of distribution systems, hours of sale, days of sale, density of outlets, server training, and advertising. For all other alcohol policy issues, there is little or no research in the peer-reviewed scientific literature.

Most programmatic interventions to reduce alcohol problems and its consequences in recent years have focused on reducing the demand for alcohol by youth, traditionally through school-based programs. Programs from the 1960s to mid-1980s used information-based and affective-change strategies and were found to be ineffective (Moskowitz, 1989). More recently, programs based on the social influence model have emerged, teaching specific drug use resistance skills or more general life skills. Some of these programs have shown beneficial effects, although the effects typically decay quite rapidly after program implementation ends (Botvin et al., 1995).

As the limitations of school-based interventions, such as DARE (Clayton et al., 1996), have become apparent, interest in community-based interventions has increased. Results from three large-scale community-based alcohol intervention projects were recently published. Project Northland was a 28-community, randomized trial with social-influences-model school curricula implemented in sixth through eight grade, supplemented with peer leadership, parent education, and community task forces (Perry et al., 1996). Results showed significantly lower prevalence of alcohol use after three years of intervention, with the effects most notable among those who were nonusers of alcohol at baseline. The effects decayed, however, as these adolescents moved into high school (Komro, 1998).

A five-year, quasi-experimental prevention trial implemented in three communities was designed to specifically reduce alcohol-related injuries. Results indicate significant reductions in alcohol sales to minors and alcohol-involved traffic crashes, and no demonstrated effects on sales to intoxicated patrons or other broader measures of alcohol availability (Holder, 1997).

The Communities Mobilizing for Change on Alcohol (CMCA) project was a randomized, 15-community trial of a community organizing intervention designed to reduce the accessibility of alcoholic beverages to youth under the legal drinking age (Wagenaar et al., 1994). Results show that the CMCA interventions significantly and favorably affected both the behavior of 18-20 year olds and the practices of on-sale alcohol establishments may have favorably affected the practices of off-sale alcohol establishments, but had little effect on younger adolescents (Wagenaar et al., in press).

Back to Top

State of Knowledge (Frank J. Chaloupka, Ph.D.)

Numerous econometric studies of the demand for alcoholic beverages have concluded that increases in the prices of beer, wine, and distilled spirits will reduce alcohol consumption. However, estimates of the magnitude of the effects of price on various measures of alcohol use and abuse vary significantly across studies. Moreover, there appears to be important differences in the effects of price on alcohol use by various population subgroups, including those defined by age, gender, race/ethnicity, and drinking behavior, with particularly mixed evidence on the impact of price on the heaviest drinkers.

Econometric research on the impact of price on outcomes related to alcohol use and abuse produces generally consistent evidence that increases in the prices of alcoholic beverages lead to reductions in drinking and driving and related accidents, other accidents, delinquency, violence and other crime, liver cirrhosis and other health consequences of alcohol use, and other negative consequences of alcohol consumption, while also improving educational attainment.

Back to Top

(Susan E. Martin, Ph.D. and Jan Howard, Ph.D.)

This grouping is not a single program within PRB but 2 ongoing programs focused, respectively, on community and policy studies and on economics. The issue of alcohol availability permeates each of the community/policy studies, and it is also a focus in research concerned with DUI and violence, which has been summarized in a separate paper.

The community and policy portfolio includes four grants: The Community Prevention Trial to Reduce Alcohol-Involved Trauma (CT) which is a single study that currently involves two grants; Complying with Minimum Drinking Age (CMDA), a new study that continues the work initiated in the Communities Mobilizing for Change on Alcohol (CMCA) study; and a newly-funded Evaluation of Efforts to Reduce Border Binge Drinking. CT and CMCA were funded in response to a 1990 RFA; the former was awarded a competing continuation, while the latter has been completed. CMDA, like CMCA, focuses on the problem of youth access to alcohol. It examines the combined effects of compliance interventions and deterrence interventions to reduce the sale of alcohol to underage persons. The interventions, aimed at on- and off-premise alcohol outlets, include a compliance intervention designed to reduce underage alcohol sales, followed by a deterrence-based intervention. Nine intervention communities will be compared with 14 comparison sites using a two-group time series design.

The Community Prevention Trial to Reduce Alcohol-Involved Trauma involves three intervention communities that were paired with three matched control sites in California and South Carolina in a quasi-experimental prevention trial of a comprehensive community intervention to reduce alcohol-involved trauma. In the three intervention communities (two in California and one in South Carolina), a multi-faceted package of environmental interventions has been implemented, and the prevalence of alcohol-related trauma is the primary dependent variable. The five intervention components are: (1) a community knowledge, values, and mobilization component which involves working with existing community coalitions to develop an integrated public awareness and education program; (2) a responsible beverage service component; (3) an underage drinking component that includes normative curriculum programs in schools and community programs for parents; (4) a drinking and driving component aimed at increasing the actual and perceived risk of apprehension for DUI; and (5) an access to alcohol component that includes use of local zoning powers to reduce outlet density. Preliminary findings suggest that the project has significantly reduced alcohol-involved traffic crashes and alcohol sales to minors and that an environmentally-directed approach to prevention, using policies as the form of intervention, can reduce alcohol problems at the local level and produce substantial cost savings.

The border binge drinking study is examining the implementation and impact of the Safe Border Project, a coordinated community effort to enhance enforcement of drinking laws at the (San Diego-Tijuana) border and to establish a media advocacy program designed to alter expectancies regarding the risks of being arrested for public drunkenness and DUI.

Drs. Howard and Heurtin-Roberts are currently managing the small "Economics" portfolio described below. Four of the principal investigators are economists, and one (the PI of the gin epidemic study) is an historian. In addition, at least two grants being managed by Dr. Martin have economists as principal investigators. Thus, Dr. Michael Grossman is studying the impact of alcohol regulations on violence, and Dr. Frank Sloan is examining the impact of various deterrents on drunk driving.

Two of the studies in the Howard/Heurtin-Roberts portfolio are closely related companion projects and are funded together as an IRPG (Interactive Research Project Grant). The overall IRPG addresses alcohol demand and labor market outcomes by means of econometric analyses of existing national survey data (NLAES). The first of these studies examines the price responsiveness of beverage specific demand for alcohol and whether heavy and/or frequent drinkers are less sensitive to price. In addition to NLAES data on drinking behavior, the study utilizes data on the prices for beer, wine, and spirits reported by the American Chamber of Commerce Researchers Association. In his final report to NIAAA, Dr. Manning (the PI) concludes that: "The point estimates of the various overall price coefficients are of a magnitude that suggests that price could be an effective tool for reducing alcohol consumption." However, the results also suggest that "price may not be well targeted at the problem end of drinking behavior."

The second study in the IRPG investigates relationships between alcohol use and labor market outcomes such as labor supply, employment, earnings, and wages, again using the NLAES data for analysis. The two studies together employ an innovative synthesis of methodologies and results to assess the social costs of alcohol misuse as a means of better targeting prevention efforts.

Another econometric analysis of existing data sets seeks to understand 1) the effects of alcohol abuse and dependence on labor market success and marital status and stability; and 2) how parents' alcoholism, labor market success, and marital status affect their children. The study is composed of several sub-studies that are being incorporated to analyze complex interlinkages between parents' and grandparents' alcoholism and children's behavioral and other problems.

A panel analysis of young workers entering the job market, using data from the National Longitudinal Survey of Youth (NLSY), examines young adults' drinking patterns in relation to labor market outcomes. It investigates the effect of heavy alcohol consumption on job choices, job advancement, and investments in education following departure from high school.

The portfolio also contains an interesting historical case analysis of the London Gin Epidemic (1720-1751) being conducted by means of archival research. The study seeks to investigate the factors (social, cultural, political and economic) that influenced both the formulations of alcohol control policy and its actual impact on the consumption of distilled spirits and beer. Qualitative methods are being used to identify cultural constraints influencing policy formulation and the impact of policy on consumption. Quantitative methods being used employ an econometric time series analysis charting the relative impact of policy and market forces on consumption.

While the community-based studies portfolio is small, it strongly suggests that environmentally directed approaches are effective in addressing alcohol-related behavioral risks. Further research priorities include the following:

With respect to the economics portfolio, important areas of pre-intervention and intervention research need to be expanded or launched anew.

Back to Top

DRINKING AND DRIVING, VIOLENCE, AND WORKSITE PROBLEMS
DRINKING AND DRIVING: NEEDED RESEARCH

State of Knowledge (Robert B. Voas, Ph.D.)

Since DOT was founded in 1967, four programs have had a significant impact on alcohol research and policy, gradually resulting in the integration of drinking and driving into the mainstream of public health programs: (1) development of practical breath-testing devices, which resulted in the blood alcohol concentration (BAC) measure being written into the nation's drunk-driving laws; (2) development of reliable traffic record systems with standardized definitions of crash types at both state and federal levels; (3) demonstration of the impact of drinking age on alcohol-related fatalities among underage drinkers; and (4) development of the citizen activist movement. Unfortunately, all to often drinking and driving research is driven by current policy and program developments rather than by theory.

Although the national concern with drinking and driving has led to the implementation of tougher, more consistent drunk-driving legislation, such legislation is not always based on research. Research supporting lower BAC limits is convincing, with some alcohol impairment beginning at .02 BAC (Moskowitz et al., 1985), and roadside survey and crash data showing that there is a significant increase in the risk of involvement in an alcohol-related fatal crash at a BAC of > .05 (Zador, 1991). Currently, 39 states have laws that suspend the licenses of offenders who have a breath sample over the legal limit. There are studies that have demonstrated that this law has a general deterrent effect (Klein, 1989) and reduces recidivism among DUI offenders (Voas et al., 1998a). In contrast, states have passed open-container laws principally based on their face validity.

Most drivers in fatal crashes do not have a prior DWI, but those convicted of a previous DWI have an increased risk of being involved in fatal crashes (Simpson et al., 1996). Reducing DWI recidivism has been shown to accompany license suspension (Ross, 1992), vehicle impoundment (Voas et al., 1998b), vehicle license plate impoundment (Rodgers, 1994), alcohol-safety interlocks (Voas et al, in press), treatment for alcohol-related problems (Wells-Parker et al., 1995), and vehicle license plate tagging (Voas et al., 1997). There is little evidence that jail sentences have a special deterrent effect on offenders who are actually incarcerated (Simpson et al., 1996).

While state legislation has driven much of the progress in the reduction of drinking and driving to date, it is clear that action at the community level is critical to the effective implementation of impaired driving laws. To have an impact, new legislation must be enforced and publicized and combined public information/community action programs have been found to be particularly effective. Evidence that public information alone can impact drinking and driving is generally lacking (Atkin, 1989). There is considerable evidence, however, that publicizing new laws and enforcement programs can increase their effectiveness (Blomberg, 1992). For some years, the National Safety Council supported a p