Individual-focused approaches to the prevention of college student drinking.

1As detailed by, the majority of young adults, in particular college students, consume alcohol. Moreover, a substantial proportion of those who consume alcohol misuse it, engaging in heavy episodic drinking,1 which directly and indirectly contributes to a host of harmful consequences. The rates of heavy drinking peak at ages 21 or 22, suggesting that most college students mature out of heavy drinking. Nevertheless, the harm they experience as a result of heavy drinking, such as poor academic and work performance or serious physical injury, may irrevocably alter students’ natural developmental trajectories. In an effort to prevent or mitigate such long-term harm, myriad prevention programs have been developed to reduce college student drinking by targeting individual factors associated with alcohol use and misuse, including alcohol expectancies, drinking motives, perceived norms, and natural ambivalence regarding behavior. A wealth of research has been devoted to evaluating the efficacy of these preventive interventions. The purpose of this article is to provide a comprehensive summary of the current state of the science with regard to individual-focused preventive interventions whose efficacy in reducing alcohol use and alcohol-related problems has been evaluated in the college student population using randomized controlled trials. Conclusions from earlier reviews in this area are described briefly, with greater focus given to summarizing evidence accumulated in the past 3 years (2007–2010).

2Larimer and Cronce conducted qualitative reviews of research published between 1984 and early 2007 that evaluated the efficacy of individual preventive interventions aimed at college students. Both reviews noted a dearth of support for educational or awareness models, including information-based and values-clarification approaches, whereas there was evidence of efficacy for skills-based interventions, including self-monitoring/assessment, alcohol expectancy challenge (AEC), and multicomponent skills training. Moreover, both reviews documented strong empirical support for brief motivational interventions (BMIs) delivered via mail, online, or in person. As the name implies, in-person BMIs are brief (i.e., typically delivered over one or two sessions) and focus on enhancing motivation and commitment to change problematic behavior. To this end, BMIs often provide personalized feedback regarding the client’s drinking and related consequences, alcohol expectancies, and drinking motives; when delivered alone in the absence of a trained facilitator, this personalized feedback component is referred to as a personalized feedback intervention (PFI). BMIs and PFIs often additionally include general alcohol information (i.e., alcohol education) and alcohol-specific coping and harm-reduction skills. PFIs typically include personalized normative feedback (PNF), which compares the client’s self-reported drinking behavior to the average drinking behavior of a specific reference group (e.g., typical student, typical female). PNF encourages clients to explore and enhance discrepancies between their perception of their own drinking as “typical” and the actual drinking behaviors of their peers—that is, that the majority of students drink moderately, often significantly less than the individual receiving the intervention. Like PFIs, PNF can be delivered as a stand-alone intervention in the absence of in-person contact. independently detailed empirical evidence supporting normative reeducation interventions, in particular computer-administered or in-person PNF interventions, that produced reductions in drinking and/or consequences mediated through changes in normative perceptions.

3Complementing the qualitative reviews by Larimer and Cronce, conducted a quantitative review evaluating 62 randomized clinical trials of 98 alcohol interventions for college students published during roughly the same time period (i.e., 1985 to early 2007). This meta-analysis similarly supported the efficacy of individual-focused alcohol interventions in reducing the quantity and frequency of alcohol use and alcohol-related negative consequences. The investigators further noted that significant intervention effects on indices of alcohol consumption peaked before the 6-month followup and that subsequently emerging effects on alcohol-related negative consequences lasted through long-term followup (ranging from 1 to 3.75 years). Specifically, concluded that individual interventions that used motivational interviewing techniques, included personalized feedback on alcohol expectancies and drinking motives with normative reeducation components, and included decisional balance exercises demonstrated greater efficacy in reducing alcohol-related consequences than did various comparison groups. This combination of intervention components is common to intervention approaches patterned after the Brief Alcohol Screening and Intervention for College Students (BASICS) program.

4In the years since the publication of the reviews by and Larimer and Cronce, numerous studies of individual-focused preventive interventions for college student drinking have been published. Of these, 36 studies evaluating 56 unique interventions, met criteria for inclusion in this review (see the tables for details). Studies were identified via a comprehensive search of electronic databases, including PsycINFO and MEDLINE (for a list of search terms used, see ), covering the period from late 2007 to early 2010. Additional studies were identified indirectly (e.g., they were referenced in the introduction section of one of the identified studies), and as-yet-unpublished studies were provided directly by authors. Studies were included if they used a randomized controlled trial approach—that is, if they randomly assigned individual participants (or intact groups) to one of two or more experimental conditions, including at least one active intervention and an ostensibly inert control (e.g., assessment only) group. Although the number of studies meeting inclusion criteria suggests that a meta-analysis may be warranted, a qualitative approach was selected for this review to facilitate more rapid communication with key stakeholders concerning the current state of alcohol prevention.2 However, intervention effect sizes are reported for relevant outcomes in all studies that included effect size estimates in the original report or provided sufficient postintervention data to calculate between-group estimates (see tables). Within-group effect size estimates also are provided for studies wherein significant within-person reductions in alcohol use or consequences were evident.

5Many of the studies included in this review evaluate the efficacy of multicomponent BMIs, many of which were adapted from the BASICS program. Most of these BMIs incorporated a PFI with PNF. Some studies evaluated one or more PFI/PNF interventions delivered alone, without the benefit of a trained intervention facilitator. Interventions were delivered via various modalities, including in-person group and individual sessions, mailed printed material, and Web-based content. In addition, some interventions were conducted in special settings (i.e., primary care, in the student’s home before entering college) or targeted high-risk student subpopulations (i.e., mandated/ sanctioned students, freshmen, or athletes).

6A total of 17 studies evaluated the impact of 14 unique PFIs/PNF and 4 PNF-only interventions implemented via written material, mail, computer, Web, or electronic diary on college student drinking (see table 1). Of 14 PFI/PNF interventions evaluated, 6 were associated with reductions in drinking but not drinking-related consequences relative to the comparison condition at followup. One PFI/PNF intervention was associated with reduced drinking-related consequences as well as alcohol use. Four additional PFI/PNF interventions were associated with significant within-person reductions in alcohol use and/or consequences across assessment periods, but between-group differences were not evident. Of four PNF-only interventions evaluated, three resulted in reductions in drinking outcomes at followup. The remaining PNF-only intervention had no effects on these outcomes but was associated with reductions in perceived drinking norms and increased readiness/ preparation for behavior change.

7The literature review also identified 17 studies evaluating 20 unique in-person BMIs (individual and group), most of which incorporated PFI and/or PNF (see table 2). Of these interventions, 13 were associated with reductions in drinking, alcohol-related negative consequences, and/or associated psychopathology, and three interventions exhibited a protective effect against the onset of or increase in alcohol use and/or related consequences. One of these studies demonstrated a sleeper effect of the intervention, with short-term reductions in drinking and subsequently emerging reductions in consequences. Also note that another of these studies was not specifically aimed at college students but targeted young adults (ages 18 to 24) who were employed; however, 75 percent of the sample concurrently was enrolled in school. This study found that BMI combined with PFI was equivalent to PFI alone in reducing drinking-related variables. Finally, one of these studies specifically evaluated intervention gains between the 12-month and 24-month followup and found an advantage for a BMI combined with coping skills over either component alone. A quantitative comparison of changes from baseline to the 12-month followup was not presented. However, figures displaying group means suggest a potential short- term effect of the BMI-only condition in reducing estimated blood alcohol concentrations (BACs), which, if counted, would bring the above total support for BMI conditions from 13 to 14.

8Other conclusions that can be drawn from the analysis of these studies include the following: Findings of studies evaluating BMI in specialized settings and high-risk subpopulations suggest that primary care is an effective venue for delivery of this type of intervention.Group BMI or BMI enhanced with parental coaching is effective in reducing drinking among college freshmen.BMI is effective for nonmandated high-risk drinkers.Studies involving students who had been mandated to participate in the interventions documented benefits of BMIs, in particular for females and those who received additional services, including coping skills, problem solving, and stress management training, in the context of a student assistance program. Another study (Carey et al. 2010) additionally found greater benefit of BMI participation in reducing alcohol consumption among female mandated students compared with two separate multicomponent educational programs; however, reductions in the BMI were similar to assessment only. Participation in any of the three interventions was associated with short-term reductions in alcohol consumption among male mandated students.

9Findings of studies evaluating BMI in specialized settings and high-risk subpopulations suggest that primary care is an effective venue for delivery of this type of intervention.

10Group BMI or BMI enhanced with parental coaching is effective in reducing drinking among college freshmen.

11BMI is effective for nonmandated high-risk drinkers.

12Studies involving students who had been mandated to participate in the interventions documented benefits of BMIs, in particular for females and those who received additional services, including coping skills, problem solving, and stress management training, in the context of a student assistance program. Another study (Carey et al. 2010) additionally found greater benefit of BMI participation in reducing alcohol consumption among female mandated students compared with two separate multicomponent educational programs; however, reductions in the BMI were similar to assessment only. Participation in any of the three interventions was associated with short-term reductions in alcohol consumption among male mandated students.

13Additional studies evaluated other specific alcohol interventions, in most cases comparing these approaches to other active interventions (e.g., BMI or PFI/PNF) (see table 3). Two studies published in the time period evaluated included alcohol expectancy challenge (AEC) protocols, which generally are considered to be more skills based than motivational in nature. evaluated a single-session, gender-specific in vivo (experiential) AEC. In contrast, assessed a two-session mixed-gender in vivo AEC, both alone and in combination with a BMI involving a PFI/PNF component. Both AEC interventions resulted in reductions in alcohol use but not alcohol consequences.

14Two other studies investigated the efficacy of BAC feedback, another cognitive–behavioral skills-based approach used to intervene with college students. One of these studies demonstrated a positive effect of the intervention (i.e., reductions in BACs), whereas the other reported a potential inadvertent opposite (i.e., iatrogenic) effect—that is, an increase in BACs. These mixed findings may be related to differences between the two studies in terms of the timing of the feedback (i.e., immediate versus delayed) and use of incentives to promote lower BACs (i.e., a $100 cash raffle for participants with BACs lower than 0.05 percent in the study by Glindemann and colleagues).

15Four studies evaluated alcohol education either as a stand-alone intervention (see ) or as a comparison intervention for PFI/PNF interventions with or without BMI. These studies generally found increases in alcohol knowledge among the students receiving the intervention. However, the interventions generated equivocal or negative effects on alcohol use and related consequences because they detected no group differences and/or lacked an assessment-only control group.

16Finally, eight studies tested nine unique multicomponent, education-focused programs, which included general alcohol information as well as elements typically associated with efficacious BMI and PFI/PNF interventions, such as personalized feedback, normative reeducation, challenge of positive drinking expectancies, and tips for harm reduction. Just over one-half of these programs were associated with reductions in drinking and/or alcohol consequences, whereas the remainder (i.e., Alcohol 101 Plus; an in-person, facilitator-led program; AlcoholEdu for College, 2006 version [Croom et al. 2008]; and Alcohol 101) produced equivocal results. Of note, because the effective multicomponent education programs (e.g., AlcoholEdu, 2007 version; AlcoholEdu for College; AlcoholEdu for Sanctions; and College Alc) included BMI and PFI/PNF elements, it is impossible to disentangle the effect of education alone from the effects of these efficacious components.

17In summary, studies published between 2007 and early 2010 provide consistent support for the efficacy of brief, personalized, individual motivational feedback (i.e., BMI with PFI/PNF) interventions and stand-alone PFI/PNF interventions. These studies also provide support for the efficacy of AEC interventions, although less consistent, and offer mixed support for BAC feedback. These conclusions are in line with previous reviews. Also consistent with previous reviews, there was an absence of support for programs solely including alcohol education, although multicomponent alcohol education–focused programs, which combine educational elements with BMI, PFI, and PNF components, had greater, albeit mixed, support.

18Although the balance of the evidence supports the efficacy of PFI/PNF-only interventions, additional research on these interventions is necessary to identify the elements and/or modalities that are associated with behavior change and to determine for whom in-person BMI is more (or less) efficacious compared with PFI/PNF-only interventions. The lack of intervention effects in a few of the BMI and PFI/PNF studies may reflect the potential absence (or ineffective delivery) of necessary intervention components or the presence of potential moderators of intervention effects (e.g., mandated student status). Additional research also needs to establish the efficacy of these brief interventions in reducing long-term risk. Thus, it may be necessary to modify and evaluate existing interventions and/or evaluate the effects of supplemental interventions in order to extend their short-term effects and enhance or prolong their impact on negative drinking consequences. Recent findings suggesting longer-term emergent effects on alcohol-related consequences, particularly in response to in-person BMIs, indicate that the addition of longer-term follow-up assessments will be necessary to achieve this. Finally, additional research is needed to evaluate the efficacy of BMIs in combination with other interventions, including interventions targeting environmental change, parenting practices, or psychiatric comorbidity. Ultimately, multiple intervention strategies may be necessary to produce lasting effects on college student drinking and related harm.

19Unfortunately, key stakeholders (e.g., college administrators, campus health professionals) face numerous barriers when trying to implement efficacious individual-focused alcohol interventions. For example, with the exception of commercially available programs, such as e-Chug or AlcoholEdu, the measures and feedback programs used in most intervention protocols are not easily accessible or not immediately useable. For those seeking to implement the BASICS approach, a published manual and measures are available. However, campus personnel may not have adequate resources (e.g., the expertise to train and supervise therapists, access to programs that can generate personalized feedback, or access to campus specific normative drinking data) to implement the program with sufficient fidelity.

20Many of these barriers can be overcome by pairing health and counseling personnel with faculty in academic departments who may have experience with program evaluation and implementation. Word processing and spreadsheet/database programs generally available to campus personnel can be used to generate basic personalized feedback. Distance-learning methods currently used to disseminate some evidence-based public health interventions (e.g., video- or Web-based conferencing of initial training and ongoing clinical supervision) could be adapted to support implementation of BMI protocols. Implementation of routine alcohol screening in campus health centers could be used to gather normative data for use in PFI/PNF and to identify students appropriate for intervention.

21Barriers to intervention implementation also necessitate additional research into increasing the reach of evidence-based approaches. This includes research related to training of providers and assessment of fidelity for in-person interventions, methods to improve impact and portability of Web-based or mailed/written interventions, and research on adaptation of efficacious interventions so they are appropriate for young adults from different cultural backgrounds and in contexts outside the traditional, mainstream college setting. To date, young adults in the workplace, community-college settings, tribal colleges and universities, historically Black colleges and universities, and other minority-serving institutions have been substantially underrepresented in efficacy trials of BMIs and related interventions. Careful consideration and the development of meaningful community partnerships to support the bidirectional learning necessary to adapt and implement efficacious brief prevention approaches in these settings are needed.