College Student Drinking Research From the 1940s to the Future: Where We Have Been and Where We Are Going

Abstract

Objective:

College student drinking is not a new phenomenon, yet the field of research studying college student drinking is relatively young. In recognition of the 75th anniversary of what is now the Journal of Studies on Alcohol and Drugs, this article reviews the path from the first article to focus exclusively on college student drinking as the topic (published in 1945) to the current state of the science and attempts to look forward to the next steps in the field’s research agenda.

Method:

Articles were selected by consensus of the authors from incarnations of the journal and other academic journals based on their relevance to the genesis of current best practices regarding college student drinking prevention.

Results:

Major eras and themes include (a) early efforts to describe and understand college student drinking; (b) building foundations for prevention and intervention efforts in response to growing concerns about high-risk drinking; (c) the emergence of harm-reduction efforts, normative interventions, and efforts to document campus strategies; (d) efficacious prevention efforts and high-risk drinking; (e) the “Call to Action” Task Force Report from the National Institute on Alcohol Abuse and Alcoholism; and (f) updates to the science (including emerging technology).

Conclusions:

Understanding the rich history of science related to college drinking prevention should prepare and guide our field for the next 75 years of scientific advances, leading to even greater understanding of the etiology and topology of college student drinking as well as more effective methods to reduce alcohol-related harms.

In 2012, a record 21.6 million people were enrolled in U.S. colleges and universities (Hussar and Bailey, 2013). Understanding threats to students’ health and well-being is paramount. These threats include alcohol consumption, which is associated with consequences ranging from academic impairment (Pascarella et al., 2007; Porter and Pryor, 2007; Singleton and Wolfson, 2009) to assaults and fatalities (Hingson et al., 2009).

Neither college student drinking nor its associated harms are new phenomena. Decades of research have documented drinking behavior and etiological factors and more recently identified successful strategies for reducing alcohol-related harm. Looking back at how the field has developed helps highlight the progress made as well as goals for the future.

Method

In recognition and celebration of the 75th anniversary of what is now the Journal of Studies on Alcohol and Drugs (JSAD), a review of college drinking research across the various incarnations of the journal was conducted. The journal has been and continues to be instrumental in advancing this field; however, to provide a cohesive picture of how the science has advanced to its current state, it was necessary to incorporate literature from other academic publications. Given the pivotal role of the journal in advancing prevention science related to college drinking, articles were selected by consensus of the authors from a comprehensive literature search of PsycINFO on the basis of their relevance to the genesis of current best practices, beginning with seminal articles establishing the nature of and problems related to college drinking. In this report, the articles are organized according to major eras and themes. Given the brevity of this review and its focus on prevention science, many other important studies could not be included, in particular, numerous studies examining the epidemiology and correlates of college drinking; studies exploring drinking and developmental trajectories; and studies examining event-specific drinking, including drinking games; however, their absence here does not diminish their invaluable contributions.

Results

Early efforts to describe and understand college student drinking

In 1945, Dr. Clements Collard Fry published in the Quarterly Journal of Studies on Alcohol the first article exclusively focusing on college student drinking as the topic. Almost 70 years later, it is striking how much has changed and how much remains relevant. Fry observed that college drinking is an accepted symbol of good fellowship, described “beer parties” often attended by faculty members, and indicated that many colleges have alcohol-themed songs. Fry noted that fraternities often served wine at dinner “in the hope that members will learn to appreciate proper wines with food” (p. 244); however, Fry warned a “state of intoxication” (p. 245) could be the primary purpose of some events. He also discussed the opportunity for World War II veterans to attend college and speculated on the role of alcohol in coping with stressors. Fry concluded with a plea that alcohol disorders be seen as a health concern:

It is the obligation of the university, which should be concerned with the health of its students, to view alcoholism as a possible symptom of emotional disturbance, in need of psychiatric attention, rather than as a form of behavior calling for customary disciplinary action. (p. 248)

Fry’s commentary opened the door for decades of subsequent research on cultural, social, and environmental influences on drinking, including factors like drinking motives and social norms. In addition, as more colleges move to offer outreach services and/or provide nonjudicial or conduct-based follow-up services to students after alcohol-related incidents (Kilmer and Bailie, 2012), it is interesting that this recommendation was made as early as 1945.

Hecht et al. (1948) extended Fry’s (1945) research to the study of college women’s drinking. In a statement that sheds light on a very different era of curfews and social norms, the authors explained that questionnaires were distributed every night, and “most of the interviewing was done after 11 o’clock, when students were home from dates” (p. 253). Drinking frequency and frequency of dating were related, with serious relationships less common among regular drinkers; however, no relationship was found between drinking frequency and grades. This study spurred additional research to understand factors contributing to and surrounding drinking among women.

Following the work of Hecht and colleagues (1948), Berezin and Roth (1950) explored other social factors affecting college women’s drinking, including religion, sorority membership status, and drinking context. The most significant contribution of Berezin and Roth was inclusion of sorority membership status. Results indicated that sorority women consumed more alcohol than non-sorority women on both dating and nondating occasions, highlighting the need for further research with this population.

Straus and Bacon (1953) conducted the first wide-scale study of drinking at 27 colleges, described as “the most definitive study of college drinking” by Gusfield (1961, p. 428), and later as a “classic” by O’Malley and Johnston (2002, p. 23), facilitating further research on college student drinking in the 1960s. For example, Gusfield (1961) documented heavy drinking by fraternity men, examined the impact of parental drinking and religious affiliation, and demonstrated the importance of peer influence on heavy alcohol use. The importance of parental messages about abstinence, perceived parental knowledge and approval of drinking, and attitudes about drinking also appeared in the literature at this time (Shaw and Campbell, 1962). Studies in this era also broadened diversity of research samples, including studies of alcohol use among African American college students (Maddox and Borinski, 1964; Maddox and Williams, 1968).

Investigators also began examining psychological constructs related to excessive drinking. In a study comparing “problem drinkers” in college with “alcoholics,” Williams (1965) found that problem drinking fraternity men were more likely to check each of 24 unfavorable adjectives (e.g., cynical) to describe their “real self” than were non-problem drinkers. A follow-up study of personality characteristics of college students using the Heilbrun Need Scales demonstrated significant differences between problem and non-problem drinkers on 9 of 15 variables, differences that were considered in the context of theories of the development of “alcoholism” (Williams, 1967).

Williams (1968) took this research a step further, examining how college students described themselves in actual drinking situations. Data were collected before, during, and at the end of parties in which participants could order as many drinks as they chose containing 2 oz. of distilled spirits. The more participants drank, the more their self-ratings of confidence, disinhibition, and impulsiveness increased, and those who drank 6 oz. or more also reported being more aggressive. Williams (1968) concluded that because those with high scores on a problem drinking scale (i.e., on a scale from -1 to +12, scores ranging between +7 and +12) believed they could engage in behavior while drinking that was not allowed when sober, they were at risk for drinking more frequently and becoming dependent as their levels of use continued. Williams emphasized factors that put heavy drinking students at risk for dependence and set the stage for alcohol expectancy research in the 1970s and 1980s.

Subsequent studies shed light on enforcement of policies related to college student drinking. LeMay (1968) described “alcohol misconduct” as the most frequent charge against students referred for disciplinary action. LeMay observed that the academic problems of these primarily first-year students “suggest strongly that most will not become upper-classmen” (p. 942) and raised questions regarding why freshmen were overrepresented, including their age, inexperience with alcohol, and other possible factors causing them “both to underachieve and to act-out in such a manner that they were apprehended” (p. 942). Congruent with current thinking, LeMay acknowledged that alcohol use does not occur in a vacuum, emphasizing the importance of considering the context of high-risk drinking.

Part of considering this “context” was exploring individual differences that may affect drinking, such as anxiety and personal goals (e.g., social lubrication, coping with disappointment). Smart (1968) found a nonlinear association between anxiety and drinking, and Jessor et al. (1968) found that low need for academic achievement and peer affection were associated with more alcohol consequences for both men and women.

Descriptions of college student drinking patterns became more global as studies of college students in England (Einstein et al., 1975; Orford et al., 1974), Ireland (Parfrey, 1974), and multiple countries (Sargent, 1971) garnered more attention. These studies looked further at the context of alcohol use by considering rates of marijuana and tobacco use in addition to alcohol (Einstein et al., 1975) and documented major differences in cultural attitudes toward drinking (Sargent, 1971). College undergraduates were no longer simply readily accessible research participants but rather had become the focus of research.

In the process, early glimpses of what would later represent major breakthroughs in college student drinking research appeared. Well before heavy episodic drinking was commonly addressed in the literature, Orford and colleagues (1974) talked about “fast drinking” and concluded that there were significant sex differences between men and women, with men not only drinking more than women, “but also knocking back relatively small quantities very fast” (p. 1364). Orford also examined expected tolerance, drinking styles, concerns about drinking, “amnesia for the night before” (p. 1358) (i.e., blackouts), and “morning-after” (p. 1324) effects.

Responding to college student drinking: Building foundations for prevention and intervention efforts

Born out of funding provided by the Hughes Act of 1970, the creation of the National Institute on Alcohol Abuse and Alcoholism (NIAAA) marked a key turning point for prevention science (NIAAA, n.d., para. 1). It provided the funding necessary to allow the field to grow and gave credibility to the view that alcohol misuse was a public health concern that could be addressed through intervention and treatment versus a personal moral failing that could only be cured through willpower or religion (Warren and Hewitt, 2010).

For college drinking prevention approaches to truly gain ground, however, it was necessary to show that alcohol dependence was not fully mediated by physiological processes. Marlatt et al. (1973) hypothesized that loss-of-control drinking was the result of learning to expect positive effects from alcohol, leading to loss of control when greater levels of consumption were required to achieve the same effects because of tolerance. Using a regular vodka and tonic mixture (alcohol) and a vodka-beverage recipe that could not be distinguished in taste from plain tonic water (placebo), Marlatt and colleagues manipulated actual and anticipated beverage content in a fully crossed design (told alcohol/got alcohol; told alcohol/got placebo; told placebo/got alcohol; told placebo/got placebo) in a sample of alcohol-dependent and nondependent individuals. This balanced placebo design (see Marlatt and Rohsenow, 1981, and Rohsenow and Marlatt, 1981, for details on the history of the creation of this design) effectively disentangled physiological and psychological aspects of drinking and demonstrated that what people thought or expected they were drinking was what determined subsequent drinking, not the actual alcohol content of the beverage. This expectancy effect was subsequently demonstrated among male college students with respect to aggression (Lang et al., 1975). This work ultimately provided the foundation for additional research on expectancies, including brief interventions that incorporate expectancy challenge.

In addition to basic science on expectancies, potential responses to college student alcohol and substance use began to appear in the literature at this time. Biggs et al. (1974) examined the relationship between students’ alcohol and marijuana use, parental attitudes, and “collegiate norms.” Their final paragraph addressed campus administrators and staff:

Student personnel workers might well discuss the ramifications of this research for designing drug education projects on their campuses. Orientation programs, activities programs, and counseling programs should be concerned with intervening in the socialization process whereby students learn to be drinkers and smokers. Parents should be involved in all of these programs so that they learn more about the role of their norms in student use of the two drugs. If drinking alcohol represents conformity to parental norms whereas smoking marijuana represents nonconformity, student personnel workers should help parents and students understand the dynamics of this situation (p. 29).

Biggs and colleagues’ (1974) recommendations presaged much of the work on peer- and norms-based preventive interventions, as well as current research on parental interventions to reduce drinking on college campuses.

Overall, this era saw increased focus on developing prevention approaches for college drinking. For example, Filstead et al. (1976) reported on an Illinois statewide conference of campus stakeholders charged with identifying problems and solutions for alcohol-related issues on campus. Many reported suggestions bear a striking resemblance to evidence-based practices recommended today, including developing and consistently enforcing clear alcohol policies, training residence hall staff and faculty to identify and refer students exhibiting alcohol problems, incorporating peers into alcohol prevention activities, and coordinating campus and community alcohol treatment services.

Subsequent research (Engs, 1977) highlighted the importance of evaluating program effectiveness: An educational film and peer-led values-clarification exercises changed knowledge but did not change drinking behavior of college students. Engs concluded that it is important to distinguish between programs that change knowledge and those that change behavior, and argued we should “look before we leap” (p. 39), only implementing approaches shown to change behavior. This recommendation remains relevant in the current climate.

In addition to intervention development, researchers continued and expanded their efforts to document trends in drinking on college campuses. Hanson (1977a, 1977b) noted a shift from parental to peer influences on drinking and a reduction in drinking for sedative effects among students at 17 colleges from 1971 to 1975. There was a decrease in the amount of alcohol consumed over this period but no change in likelihood of drinking. Hanson (1977a, 1977b) hypothesized that other substances were being used to achieve sedation, supporting increased focus on comorbid use of alcohol and other substances.

Although college students did not appear until the 1985 survey (Johnston et al., 1986), the launch of the Monitoring the Future study in 1975 allowed the trends, trajectories, and relationships among different substances (including alcohol) to be examined. The wealth of information on adolescent and young adult substance use within the Monitoring the Future database continues to inform prevention efforts and responses to emerging needs to this day (Johnston et al., 2013).

Research on college students’ alcohol expectancies begun in the 1970s continued and accelerated into the 1980s (Brown et al., 1980). Southwick et al. (1981) were among the first to document a relationship between students’ drinking habits and self-reported positive and negative expectancies for a “moderate” dose and “too much” alcohol. Students reported greater behavioral impairment at heavier doses, with heavier drinkers expecting more pleasurable disinhibition and stimulation/dominance effects at moderate doses than lighter drinkers, suggesting that these expectancies may serve a motivating function in decisions to drink more.

Development of prevention approaches also continued, mostly advocating educational or disease-model approaches (Kazalunas, 1982; Kleinot and Rogers, 1982; Ramsey, 1982), although there was a move toward matching interventions to specific etiologic models of college drinking. For example, Kleinot and Rogers tested three components of alcohol prevention based on protection motivation theory (Rogers and Mewborn, 1976) as well as the health belief model (Becker, 1974). They found that interventions increasing perceived noxiousness of alcohol consequences, personal vulnerability, and effectiveness of moderate drinking strategies were most associated with increased behavioral intentions to drink moderately rather than excessively.

Emergence of harm-reduction efforts, normative interventions, and attempts to document campus strategies

Given prior research demonstrating how expectancies influence drinking, and the potential to modify such expectancies through intervention, Fromme et al. (1986) evaluated an 8-week skills training program (STP) versus traditional alcohol education (Alcohol Information School; AIS) and self-monitoring/assessment only (AOC) among high-risk college drinkers. The STP was a stark departure from Reaganera “just say no” programs in that abstinence was endorsed as the least risky option, but students were encouraged to set their own goals with regard to drinking. Skills covered in the STP were designed to help students who made the choice to drink do so in moderation and reduce alcohol-related harms. STP sessions included discussion of blood alcohol level (BAL) monitoring, managing high-risk drinking situations, discussion-based alcohol-expectancy challenge, and stress-management training.

In addition to seeing if STP would influence alcohol-related expectancies, the potential mediating effect of changing expectancies on mean drinks per week, hours per week, and BAL was tested. Significant group differences were not evident; however, within-group analyses showed significant reductions on all drinking outcomes from baseline to postintervention and 4-month follow-up in STP, with no reductions evident in AOC and reductions in only hours and BAL evident in AIS at 4-month follow-up (but not postintervention). Although modest, the success of this approach spurred evaluation of successive variants of the STP, later termed the Alcohol Skills Training Program (ASTP; Kivlahan et al., 1990).

By this time, the impact of peer influences on college drinking was clear. A growing movement involving peer health education became more visible. Researchers also began documenting the significant role played by misperceptions of the norms surrounding college student drinking. Perkins and Berkowitz (1986) showed that students tended to overestimate permissiveness of attitudes and drinking behaviors, and that these perceptions correlated with their own drinking. This research laid the groundwork for later interventions targeting individual normative reeducation and social norms mass marketing.

In addition to documenting drinking rates and consequences experienced by college students, national surveys also attempted to detail services and educational efforts offered on college campuses. After three administrations of the College Alcohol Survey in 1979, 1982, and 1985, a majority (64%) of schools in 1985 (compared with 37% in 1979) reported having a task force or committee focused on alcohol education/prevention, and 48% of campuses in 1985 had a dedicated alcohol education coordinator/specialist (compared with 14% in 1979) (Gadaleto and Anderson, 1986). However, in 1985, college alcohol programs primarily consisted of information and articles in campus publications (76%), films shown on campus (63%), speakers (63%), workshops focused on drinking attitudes (61%), poster and slogan campaigns (60%), educational handouts (51%), and discussion groups (50%). There was recognition of the need to address college student drinking, yet no clear guidelines on how to best do this.

At the same time, the minimum legal drinking age (MLDA) started to universally move to 21 in the United States (a transition accomplished in 1988). Although later research eventually showed that increases to the MLDA decreased both traffic crashes and alcohol consumption (Wagenaar and Toomey, 2002), initial evaluations of college students showed everything from shifts in where students did their drinking (George et al., 1989) to students’ efforts to avoid getting caught by enforcement, often associated with increases in risk taking (Brittain and Roberge, 1988). One article written by student affairs staff (Roberts and Nowak, 1986) concluded:

Another approach that may help during and after the transition to the [MLDA] of 21 would be to make funds available to institutions of higher education to develop, test, and disseminate information about model alcohol education programs. Approaches to alcohol education are already in use. These approaches need to undergo rigorous evaluation and then be made available for application throughout college campuses, (p. 489)

Although widespread dissemination of best practices was still 16 years away, it was time for the college-drinking field to identify efficacious programs.

Efficacious prevention efforts and high-risk drinking

In 1990, Kivlahan and colleagues conducted a study following up on Fromme et al. (1986), testing the same three conditions (ASTP, AIS, and AOC). Promising results led to a subsequent study by Baer and colleagues (1992) in which the ASTP was shortened to six sessions, altered to include in vivo expectancy challenge, and compared against self-help versus a 1-hour individualized feedback and advice session using motivational interviewing (Miller and Rollnick, 1991). Results suggested that both the ASTP and 1-hour brief motivational intervention (BMI) were effective in reducing alcohol use (by about 40%) over a 2-year period. Given the brevity of the BMI and its equivalence with the ASTP, this model (which formed the foundation of the Brief Alcohol Screening and Intervention for College Students [BASICS] program; Dimeff et al., 1999) represented a quantum leap, potentially reducing the resources needed by colleges to implement effective individual-focused indicated prevention.

In late 1994, Wechsler and his team from the Harvard School of Public Health College Alcohol Study used the term binge drinking to describe consumption of five or more drinks in a row for men and four or more drinks in a row for women in the past 2 weeks. Wechsler and colleagues deemed the behavior to be “widespread” (p. 1676) and stated that “the scope of the problem makes immediate results of any interventions highly unlikely” (p. 1677). They went on to warn that students who engage in this practice are “similar to other alcohol abusers elsewhere in their tendency to deny that they have a problem” (p. 1677). Although the phenomenon described by Wechsler and colleagues was promulgated in the mainstream media, the academic community raised questions about what “in a row” meant, the absence of the focus on blood alcohol concentration attained as a function of the weight of the person drinking and over how much time drinks were consumed, and the impact of viewing students as “in denial” (e.g., Dimeff et al., 1995). In time, academic journals (led by the Journal of Studies on Alcohol [JSA]) requested different terminology and generated policies requiring authors to make a distinction between a binge being several days of extended intoxication at the expense of other activities and heavy episodic drinking (or other, similar phrases) referring to massed consumption on isolated occasions (a policy maintained by JSAD; Alcohol Research Documentation, Inc., n.d., para. 2). (The journal changed names from JSA to JSAD in 2007.) Regardless of the controversy surrounding the definition, Wechsler’s team helped to highlight the concerns surrounding heavy drinking by some students on college campuses. Unfortunately, although successfully advancing the recognition of a potential concern, this work did not identify a solution.

One potential solution came through subsequent work by Marlatt and colleagues, which more definitively established the success of BASICS in reducing alcohol use and negative consequences among high-risk drinkers (Baer et al., 2001; Marlatt et al., 1998). Comparing BASICS with AOC longitudinally across 2 years, Marlatt and colleagues documented significant reductions in consumption and consequences among students who had received BASICS. Baer and colleagues reported that students who received BASICS as part of the Marlatt et al. trial reported significantly less frequent alcohol use, fewer drinks per drinking occasion, and fewer negative consequences than students who received no intervention at 4-year follow-up, thus establishing the potential long-term benefits of BASICS. Although the scientific base for BASICS was still growing, the program was manualized for broad dissemination (Dimeff et al., 1999), which facilitated implementation of BASICS on other campuses and promoted further efficacy research using BASICS.

“A Call to Action ”—The NIAAA Task Force Report

In 1999, a task force of experts in the epidemiology, etiology, prevention, and treatment of alcohol misuse by college students was convened by the National Institutes of Health’s National Advisory Council on Alcohol Abuse and Alcoholism. The Task Force on College Drinking was charged with characterizing the phenomenon of college student drinking, along with its harms, and synthesizing the research literature to provide key stakeholders with science-based guidance on how to address drinking on their campuses, as well as set the research agenda for future college drinking prevention studies. In 2002, NIAAA published the findings of the Task Force in A Call to Action: Changing the Culture of Drinking at U.S. Colleges, which described a four-tier system delineating the efficacy and effectiveness of different approaches in reducing alcohol use by college students, with Tier 1 having the greatest scientific support and Tier 4 having the lowest. Individual-focused prevention approaches geared toward changing the drinking behaviors of individual students, including BMIs and expectancy challenge, predominantly occupied Tier 1, whereas environmental approaches designed to alter the context in which drinking occurred, such as increased enforcement of the MLDA, mostly occupied Tier 2. Purely educational approaches were assigned to Tier 4. These recommendations were largely based on two literature reviews compiled by Larimer and Cronce (2002) and Toomey and Wagenaar (2002), which along with reports by other members of the task force were published in a special supplement of JSA made available to the public via www.collegedrinkingprevention.gov. Along with dissemination of the report and supporting scientific review articles on the website, copies of the task force report were mailed to college presidents at institutions across the United States.

Also part of the task force report, and published in a separate issue of JSA in 2002, was a study by Hingson et al., quantifying the morbidity and mortality of alcohol use among college students ages 18–24. This report estimated that approximately 1,400 college students lost their life to alcohol-related causes each year, and another 500,000 were injured while under the influence of alcohol. These numbers garnered the attention of the popular media, appearing in articles published beginning the second week of April 2002, in outlets including The New York Times, CNN, the San Francisco Chronicle, USA Today, and Fox News. The national spotlight on college drinking cast by the NIAAA Task Force report (and the accompanying articles in JSA) was instantly magnified by the media coverage of Hingson and colleagues’ statistics.

Building on the report of the task force, NIAAA implemented the Rapid Response to College Drinking Problems initiative in 2002. This unique mechanism paired teams of experienced alcohol research scientists with campuses experiencing an urgent alcohol-related problem, allowing for the swift implementation and evaluation of evidence-based approaches to reduce alcohol use and consequences. Ultimately, 5 teams of scientists and 15 campus intervention sites were funded through this mechanism. Results of eight of these interventions were reported in a special issue of JSAD (DeJong et al., 2009). Results provided support for in-person individual BMIs in a campus health setting (Schaus et al., 2009) alone and in combination with other strategies as part of a student assistance program (Amaro et al., 2009); provided preliminary support for a parent handbook in reducing drinking initiation, as well as noted growth in drinking for women across the course of the freshman year (Ichiyama et al., 2009); supported short-term efficacy of a group BMI for reducing alcohol use and related consequences for freshman women (LaBrie et al., 2009); suggested that reductions in perceived norms and increases in protective behavioral strategies may mediate efficacy of peer-based alcohol prevention approaches (Cimini et al., 2009); and demonstrated reduced growth in drinking for participants in residential learning communities compared with non-residential learning community participants through 18-month follow-up (Cranford et al., 2009). Studies reported in the JSAD supplement also provided initial quasi-experimental support for two comprehensive campus-community partnerships (Saltz et al., 2009; Wood et al., 2009) and provided an update to the estimates of alcohol-related harms by Hingson and colleagues (2009).

In addition to these tangible outcomes, the Rapid Response initiative (NIAAA, 2003) in combination with the 2002 NIAAA Task Force report resulted in a steep increase in high-quality research on college drinking prevention over the subsequent decade.

Updates to the science, moving beyond main effects, and emerging technology

Five years following publication of the NIAAA Task Force report, NIAAA commissioned a second review of the literature to provide an update of the findings on individual and environmental prevention strategies evaluated by Larimer and Cronce (2002) and Toomey and Wagenaar (2002). Published in Addictive Behaviors and JSAD, respectively, Larimer and Cronce (2007) and Toomey et al. (2007) revealed the exponential growth of research on college drinking prevention that had occurred since 2002. Specifically, although only 44 unique intervention conditions had been reviewed by Larimer and Cronce in 2002, covering a 15-year period from 1984 to 1999, a total of 60 intervention conditions targeting college student drinking had been tested using a randomized controlled design in the intervening 7 years (1999–2006). A similar trend was noted by Toomey and colleagues, with few studies available at the time of the 2002 review and evaluations of 110 environmental approaches published (of which 36 specifically targeted college students) at the time of the 2007 review. Although the broad recommendations remained the same, important insights were gained and new research directions determined through these reviews, most of which were summarized in the 12-page report released by NIAAA: What Colleges Need to Know Now: An Update on College Drinking Research (2007). Of note, additional reviews of the literature emerged around the same time, focusing on web-based stand-alone intervention approaches (e.g., Walters and Neighbors, 2005; Walters et al., 2005) and using meta-analytic techniques (e.g., Carey et al., 2007, 2009) that likely would have been underpowered in 2002 for all but the most prominent approaches.

As the evidence mounted for the efficacy of specific individual-focused interventions, exploration of moderators and mediators of treatment effect gained speed. A host of individual difference variables have been tested as potential moderators, including gender, high-risk drinking status, drinking motives, and personality differences, although no individual moderator has consistently emerged as significant. In terms of potential mediators, normative reeducation was a key component in two of the three Tier I programs listed in the 2002 NIAAA Task Force report and emerged as an efficacious stand-alone approach (i.e., personalized normative feedback) in the update by Larimer and Cronce in 2007. As such, changes in normative perceptions were the focus of a wealth of research, which overwhelmingly demonstrated the mediating role of correcting normative perceptions in decreasing alcohol use (e.g., Borsari and Carey, 2000; Doumas et al., 2009; Neighbors et al., 2006; Turrisi et al., 2009; Walters et al., 2007; Wood et al., 2007). The specificity of the normative referent group was also found to have a mediating role (e.g., Lewis and Neighbors, 2007). Motivational enhancement and alcohol-specific skill building were also critical components of the Tier I approaches, and there was substantial evidence suggesting the relevance of readiness to change and use of protective behavioral strategies to drinking. Some studies have shown support for the mediating role of these variables (e.g., readiness to change: Barnett et al., 2010; protective behavioral strategies: Barnett et al., 2007; Larimer et al., 2007) and others have not (e.g., readiness to change: Borsari et al., 2009; Wood et al., 2007).

In 2011, NIAAA commissioned yet another update of the literature on individual college drinking prevention for Alcohol Research and Health. Cronce and Larimer (2011) identified 36 randomized controlled trials evaluating 56 unique interventions designed to decrease college drinking and/or related negative consequences. Overall findings were consistent with Larimer and Cronce (2002, 2007); however, the again exponential expansion of research allowed for greater exploration of web-based (and other technology-based) programs versus in-person programs. Overall, personalized feedback interventions, patterned after the feedback offered as part of BASICS, were effective in reducing drinking; however, the changing nature of web-based materials and programs was noted as an area for concern, especially for commercial-based programs for which frequent version updates are the norm and little is known (or tested empirically) about variation in efficacy between versions.

Discussion

Collectively, the research on college student drinking over the past 75 years has helped reduce individual and societal harms, but much work remains. The prevalence of drinking among college students remains high, and increased co-occurring use of both legal (e.g., caffeine) and illicit substances (e.g., misused prescription drugs), and an apparent increase in heavy episodic drinking (≥10 drinks) in the form of pre-gaming, has made the need to understand the topology of college drinking in order to adequately intervene all the more urgent. The past decade has also seen unprecedented changes in terms of how college students interact with the world and receive information. Social networking mediums, including Facebook and Twitter, have virtually replaced face-to-face contact, phone calls, and email as primary means of communication. More research is needed to see how interacting via these mediums influences drinking and how they may be used for intervention. Most importantly, the speed with which technology is changing, and ever-tighter financial resources at the college, state, and federal level, requires that we attempt to develop intervention methods whose efficacy is not bound to a particular medium and that can be rapidly adapted to accommodate the needs of colleges and college students.