Reciprocal relations of protective behavioral strategies and risk-amplifying behaviors with alcohol-related consequences: targets for intervention with female college students.
Drinking of alcoholic beverages
Public television (Social aspects)
College students (Surveys)
College students (Behavior)
College students (Social aspects)
Alcohol and youth (Social aspects)
Traffic safety (Social aspects)
Monte Carlo method (Social aspects)
Luebbe, Aaron M.
|Publication:||Name: Journal of Alcohol & Drug Education Publisher: American Alcohol & Drug Information Foundation Audience: Academic; Professional Format: Magazine/Journal Subject: Health; Psychology and mental health; Social sciences Copyright: COPYRIGHT 2009 American Alcohol & Drug Information Foundation ISSN: 0090-1482|
|Issue:||Date: August, 2009 Source Volume: 53 Source Issue: 2|
|Topic:||Event Code: 290 Public affairs|
|Product:||Product Code: E197500 Students, College; 9108218 Highway Safety Programs NAICS Code: 92612 Regulation and Administration of Transportation Programs|
Transactional associations of protective behavioral strategies (PBS) and risk-amplifying behaviors (RAB) to alcohol-related negative consequences were tested. A sample of 138 undergraduate women was assessed with self-report measures at two time points four months apart. Over and above quantity and frequency of alcohol consumption, engagement in risk-amplifying behaviors, but not protective behavioral strategies, predicted increased negative consequences concurrently. However, use of PBS but not RAB, predicted changes in experiencing negative consequences longitudinally. Frequency of negative consequences did not predict changes in either protective behavioral strategies or risk-amplifying behaviors over time. Results suggest that PBS and RAB may both be important but independent targets for intervention and prevention with college-aged women. Specifically, short-term intervention might target RAB, whereas prevention efforts might focus on RBS.
Keywords: Protective behavioral strategies, risk-amplifying behaviors, alcohol, college women
College students' alcohol use has long been implicated in experiences of academic difficulties, health-related problems, and engagement in high-risk behaviors (Naimi et al., 2003; NIAAA, 2004; Wechsler, Dowdall, Davenport, & Castillo, 1995). Traditionally, alcohol research has focused on men's heavier consumption and greater risk for experiencing alcohol-related consequences. Yet as consumption increases, women experience more negative consequences than men (Presley & Pimentel, 2006), and women's slower metabolism of alcohol leaves them with a smaller margin of error in underestimating alcohol effects (Nolen-Hoeksema, 2004; Mallett, Lee, Neighbors, Larimer, & Turrisi, 2006). While consumption levels certainly impact the number and severity of negative alcohol-related consequences, drinking quantity and frequency fail to provide a complete understanding of their occurrences (Gruenwald, Johnson, Light, Lipton, & Saltz, 2003). Besides drinking, college students' engagement in protective behavioral strategies may buffer resulting negative outcomes (Martens et al., 2004), whereas engagement in risk-amplifying behaviors when drinking may magnify subsequent consequences (Park & Grant, 2005). In particular, women may be more likely than men to engage in certain protective behaviors or be at unique risk if they do not utilize them (e.g., keeping possession of a drink). Also, certain risky behaviors may amplify the intensity of negative consequences particularly for women (e.g., unprotected sex). In light of the National Institute on Drug Abuse's (2003) call for additional research on women's alcohol use and corresponding women-specific interventions, the current study tested the transactional relations of protective behavioral strategies (PBS) and risk- amplifying behaviors (RAB) to negative alcohol-related outcomes, both concurrently and longitudinally, in college-aged women.
What Might Buffer Harmful Drinking Consequences for Women: Protective Behavioral Strategies (PBS)
College students report using several protective behavioral strategies (PBS) while drinking such as eating food (e.g. Clapp, Shillington, & Segars 2000), keeping track of number of drinks consumed (Delva et al., 2004), and alternating alcoholic and non-alcoholic drinks (Martens et al., 2005), to reduce subsequent adverse events. PBS are inversely related to several indices of drinking including weekly number of drinks, heavy episodic drinking, highest number of drinks, and 30-day frequency of drinking (Martens et al., 2005). Despite operationalizing negative consequences differently and measuring slightly dissimilar protective strategies, five studies to date have found that use of more PBS was associated with students experiencing fewer negative consequences (Benton, Benton, & Downey, 2006; Benton et al., 2004; Delva et al., 2004., Martens et al., 2004; Martens et al., 2005).
Despite the promise of these studies in identifying potential targets for intervention and prevention, limitations in this literature still exist. First, most studies have not controlled for consumption, and the three that did only controlled either for number of drinks (Benton et al., 2006; Benton et al., 2004) or frequency of 2 week binge-drinking episodes (i.e., 5 or more drinks in one sitting; Delva et al., 2004). Because engaging in many of these strategies tends to limit consumption, whether PBS buffer experienced negative consequences over and above both quantity and frequency of alcohol use is unknown. Second, no study to date has examined this link longitudinally. Concurrent relations alone provide only a starting point for understanding the relation of PBS to negative consequences of drinking.
PBS may be especially important for the well-being of women. They are more likely than men to use PBS, perhaps partly explaining why women are also less likely to experience negative consequences (e.g., Benton et al., 2004; Delva et al., 2004; Park & Grant, 2005). Further, Benton et al. (2004) found that, whereas only heavy-drinking men experienced buffering effects of PBS, benefits of using PBS were seen for women across the drinking spectrum. Further, Delva and colleagues (2004) found that, compared to women exercising the greatest number of PBS, women who used the fewest were 6.5 times more likely to report having alcohol-related problems. This ratio was only 1.74 for men. Thus, a growing body of research suggests benefits for using PBS for both men and women, but perhaps especially so for women.
What Might Amplify Harmful Drinking Consequences for Women: Risk-Amplifying Behaviors (RAB)
Whereas some behaviors may buffer the effects of drinking on experiencing negative consequences, other behaviors while drinking may increase the risk for experiencing negative consequences. Paralleling the PBS literature, risk-amplifying behaviors (RAB) are specific behaviors that individuals engage in once they are intoxicated that increase their risk for experiencing consequences; they are not simply an outcome of drinking. As we define them, RAB are distinct from the compounding effect of a negative consequence leading to additional consequences (e.g., not doing school work leads to a failing grade). Examples of RAB might include use of illicit drugs in combination with alcohol which may increase the likelihood of sexual assault (Mohler-Kuo, Dowdall, Koss, & Wechsler, 2004). Likewise, driving while intoxicated can result in problems with authorities, bodily injury, or other social problems (e.g., fights with friends). Importantly, identification of students' RAB may help health professionals focus on specific drinking behaviors rather than globally addressing negative consequences.
Indeed, the RAB examples described above have been found to be positively associated with high-risk drinking (Naimi et al., 2003; NIAAA, 2004; Marlatt et al., 1998; Wechsler et al., 1995). Yet, the extent to which RAB put individuals at risk for negative consequences is less clear. This may be partially due to behaviors (e.g., drug use while drinking, drinking and driving) being defined as consequences, not risk-amplifying factors (e.g., Benton et al., 2004). Similarly, when risk has been measured directly, it has been typically conceptualized as consumption, personal risk factors (e.g., family history), or as a general disposition (e.g., risk-taking style, sensation seeking; Benton et al., 2006; McCabe, 2002; Nolen-Hoeksema, 2004; Wechsler et al., 1995). This area is further limited by reliance on cross-sectional studies which may or may not control for consumption levels. Understanding RAB's relation to negative alcohol-related outcomes is crucial given that college students tend to underestimate the amount of alcohol that leads to both risk-taking behaviors and negative consequences (Mallett et al., 2006).
Although the prevailing sentiment seems to suggest men are more likely to engage in risky behaviors or RAB, risk-taking behaviors in women may be underestimated in alcohol research (Wilke, Siebert, Delva, Smith, & Howell, 2005). In evaluating drinking problem domains, more college men have alcohol-related encounters with authorities, but women and men have been equally identified in the risky/reckless behavior domain (e.g., having unprotected or unplanned sex; Vik, Carrello, Tate, & Field, 2000). Further, meta-analytic work indicates that in 40 percent of studies evaluated, women engaged in as much or more risk taking than men, with the gap appearing to decrease in recent years (Byrnes, Miller, & Schafer, 1999). Despite compelling evidence that women's risky behaviors are increasingly similar to their male counterparts, prior work has focused primarily on only men's drinking experiences.
Do Negative Drinking Consequences Spur Behavior Change?
Authors of previous studies examining PBS and risky behaviors (e.g., Benton et al., 2004) have suggested that experiencing negative consequences may actually drive changes in either PBS or RAB. But without longitudinal studies to evaluate this link, the direction of the relationship between these variables and negative consequences is unknown. Knowing this causal relationship has potential value for designing and implementing prevention and intervention programs more accurately. Having women focus on the negative results of their drinking could alter future alcohol-related behaviors. For example, getting in trouble or experiencing negative consequences might increase an individual's use of PBS and decrease RAB as a means of learning from consequences. Yet, Mallett and colleagues (2006) found that individuals who engage in risky behaviors and experience negative consequences are still likely to repeat engagement in the same risky behavior and receive the same consequences again in the future. This suggests that individuals are not necessarily changing their behaviors because of negative experiences, but longitudinal studies are necessary to assess this relationship.
The Current Study
The current study seeks to identify behaviors done within the context of drinking that either buffer or amplify harmful drinking outcomes, over and above consumption, for college women. In line with Benton et al. (2006), we contend that PBS and RAB are related, yet independent, constructs such that an individual may engage in both types of behaviors simultaneously. As stated, a major limitation of past research is the lack of longitudinal studies. Although useful in establishing relations between alcohol-related constructs, cross-sectional studies potentially lead to false assumptions about directionality of effects and may leave health professionals uninformed about how best to focus interventions. Overall, we hypothesized that: (a) PBS will be inversely associated with negative consequences, both concurrently and longitudinally, controlling for alcohol consumption; and (b) RAB will be positively associated with negative consequences both concurrently and longitudinally; but (c) like Mallet et al. (2006) found, negative consequences will not be related to an increase in PBS or a decrease in RAB over time.
Participants included 189 female undergraduates enrolled at a Midwestern university. In order to obtain a sample with variation in drinking behavior (i.e., not only sampling moderate drinkers or only heavy drinkers), participants were recruited from three sources: (a) during a general wellness fair (labeled "Fair"; estimated consent rate = 72% based on estimated attendance figures of women [N=110] at fair); (b) prior to two brief workshops on alcohol, one specifically targeting sorority members, presented during the University's Alcohol Responsibility Month (labeled "Workshop"; consent rate = 75%); and (c) prior to an individual interview or group-administered workshop for students found in minor violation of the university's alcohol use policy (labeled "Violation"; consent rate = 77%).
One hundred sixty-three participants completed Time 2 packets (86% of Time 1). Although students found in the Violation group were less likely to complete the Time 2 survey ([chi square] (2) = 11.13, p<. 05), no other differences on study variables (i.e., drinking quantity and frequency, negative consequences, PBS, or RAB) were found between those who did and did not complete both time points. Of those participants completing packets at both time points, 138 reported drinking at least one alcoholic drink on one occasion in the past month at both data time points (four months apart) and these were included in the analyses (N = 55, 38, and 47 for the Fair, Workshop, and Violation groups, respectively).
Drinking quantity and frequency. Participants were asked to provide the number of drinks consumed and the duration of drinking, in hours, for a typical week in the past month. They provided a specific number for each day of the week. Drinking quantity was calculated as the mean blood alcohol content (BAC) per episode of drinking. Participants were provided with a guide for what constitutes one drink (e.g., 12 oz of beer, 4 oz of wine) and BAC was calculated using a formula obtained from the National Highway Traffic Safety Administration (1994) using participants' reported bodyweight. Drinking frequency was assessed with one item that asked, "How often did you drink alcohol during the past month?" Responses were on a seven-point scale from 0 (I did not drink at all) to 6 (Once a day or more). Although concerns have been raised about validity of self-reported alcohol use and bodyweight, past studies suggest that self-report of alcohol consumption is valid (Midanik, 1988), and differences between self-reported and observed weight, though significant, are relatively small (Jeffery, 1996; Stewart, 1992). Further, online administration of the second assessment period necessitated use of self-report methodology.
Negative consequences. The Rutgers Alcohol Problem Index (RAPI; White & Labouvie, 1989) was used to assess negative consequences resulting from drinking. The RAPI is a 23-item, self-report measure. Participants indicated on scale from 0 (never) to 4 (more than 10 times) how many times in the past four months each item occurred (e.g., "Not able to do your homework or study for a test"), and scores are summed to form a scale ranging from 0 to 92, with higher scores indicating more alcohol-related problems. Note, the timeframe was modified from six months to four months to avoid measurement overlap between assessment time points. Coefficient alpha for our sample was .88 and .90 at Time 1 and Time 2, respectively.
Protective Behavioral Strategies. PBS were measured with six items that assessed the extent to which participants planned before drinking and were observant while drinking. Items were written for this study based on the clinical experiences of the second author with women who abuse alcohol (see Appendix A items 2, 3, 4, and 6), in conjunction with those safe practices utilized in research by Delva and colleagues (2004; see appendix A items 1 and 5). Respondents indicated on a five-point scale from 0 (Not at all true) to 4 (Very true) how true each item was for them, and the mean item score provides the total PBS score. The scale had adequate internal consistency; coefficient alpha in our sample was .64 and .76 at Time 1 and Time 2, respectively.
Risk-amplifying Behaviors. RAB were assessed with four items that assessed poly-drug use, intoxicated driving, and unplanned and unprotected sexual behavior done in the context of drinking (see Appendix A for specific items). Participants indicated on a five-point scale ranging from 0 (Never) to 4 (More than 10 times) to what extent in the past four months they engaged in each behavior when drinking alcohol. The mean of the four item scores was the RAB score, with higher scores indicating engagement in more risky behavior. Internal consistency estimates for this scale were lower than for other measures in our study, .70 and .60 at Time 1 and Time 2, respectively. This was not unexpected as this construct includes variable types of risk-amplifying behaviors (e.g., sexual behaviors, poly-drug use). We included these behaviors in a single scale because risk taking across multiple domains may put individuals at heightened risk for negative consequences.
Time 1 surveys were administered in paper format during a one-month period in the fall semester. Exactly four months following completion of the Time 1 battery, participants were emailed an invitation to complete a second set of surveys online. Existing research supports similarity in reliability and validity for web-based and more traditional administrations of alcohol measures (e.g., Miller et al., 2002). The timing of the second assessment was designed to be prior to the university's spring recess in an effort to avoid reporting of unrepresentative patterns of alcohol use during the break. Responses from Time 1 and Time 2 were linked via a participant-created password comprised of letters and numbers derived from answers to personal questions (e.g., "first letter of your mother's name"; "month of birth"). With the exception of those in the violation group who completed the assessment as part of their intervention, participants received a small token of appreciation (i.e., a water bottle) for completing Time 1 surveys. For Time 2, all participants received a $20 bookstore gift certificate.
Missing Data Analyses
Approximately 13% of data were missing. Analyses suggested no significant differences on demographic or other variables of interest between those with and without missing data. Thus, we used NORM (Schafer, 1999), a multiple imputation program in which missing multivariate data are simulated m > 1 times, to impute estimates of missing values. First, the EM algorithm was applied to all cases to obtain maximum likelihood estimates of means, variances and covariances, and to provide initial estimates of starting values for all model parameters of interest to the current study (including covariates drinking quantity and frequency, but not sorority status). Then, based on the iterations needed for the EM algorithm to converge and using Markov chain Monte Carlo techniques for data augmentation with the NORM program, five imputations were generated (n = 5) to yield our final estimates.
Descriptive statistics of the study variables are presented for Time 1 and Time 2 separately in Table 1. Ninety-four percent of participants identified themselves as Caucasian. In this sample, 47% reported being first-year status, 29% sophomore, 15%junior, and 9% senior status. At Time 1, median age of participants was 19 years (range 18-23) with 14% of women age 21 or older. Sixty-six percent were members of a social sorority.
On average, the sample experienced a moderate number of negative consequences from drinking. The Time 1 mean was higher than nonclinical norms (M = 7.4 for college women; White & Labouvie, 1989), while the Time 2 mean was similar to nonclinical norms. Women averaged approximately .09 to. 10 BAC per time drinking, and reported drinking about one to two times per week. Paired samples t tests were used to test whether changes in values from Time 1 to Time 2 were significant. Of note, negative consequences (t = 5.24., p <.01, [[eta].sup.2] =. 17) and RAB (t = 4.87., p <.01, [[eta].sup.2] = .15) decreased significantly from the fall semester to the spring semester. Notably, there was no intervention effect. All three groups showed decreases in consequences and RAB. Bivariate Pearson product moment correlations of all study measures are also presented in Table 1. Measures were moderately stable across time points and did not differ based on recruitment group status. Importantly, PBS and RAB were negatively associated with one another. However, the moderate correlation suggests independence between these constructs (i.e., they are not simply two ends of a single continuum). Also, PBS were unrelated to drinking frequency at and across both time points.
For our primary analyses, we tested both concurrent and predictive relations of negative consequences, PBS, and RAB. Hierarchical multiple linear regressions were used to test for significant relationships while covarying age and drinking quantity and frequency. Although the Violations group varied significantly from the Fair and Workshop groups on drinking quantity and frequency at both time points as was expected, recruitment group status did not moderate the relations of interest described below. Coupled with the lack of specific intervention effects described above, we chose to combine groups for all analyses. Also, analyses were initially performed with the inclusion of interaction terms with Greek status due to the association with sorority membership and high-risk drinking (Wechsler et al., 1995). No significant moderations emerged; thus, the following analyses do not include Greek status.
Do PBS or RAB Predict Negative Consequences Concurrently?
Analyses first tested concurrent associations between PBS, RAB, and negative consequences. For each time point, a regression analysis was conducted with RAPI scores as the outcome variable predicted from age, drinking quantity, drinking frequency, and either PBS or RAB all measured at the same time point. Because two models were tested at each time point, Type 1 error was controlled via a Bonferroni correction ([alpha] = .05 / 2 = .025) (see Table 2). PBS were unrelated to negative consequences (i.e., RAPI scores) at Time 1 (PBS Model Stepl, upper panel Table 2) and Time 2 (PBS Model Step l, lower panel Table 2). In contrast, RAB were positively associated with RAPI scores at each time point (RAB Model Step 1, upper and lower panels Table 2) indicating that engagement in RAB may put women who drink at higher risk for negative consequences over and above their drinking amount and frequency.
Given their overlap, we were also interested in testing for unique effects of PBS over and above RAB and vice-versa. To test for these unique effects we added RAB as a second step in the regressions predicting RAPI scores from PBS, and vice versa, and then tested whether addition of the new variable significantly accounted for additional variance in the model (see Step 2, upper and lower panels Table 2). We also tested the interaction of PBS and RAB. This interaction term was not significant at either time point; therefore, only main effects models are show in Table 2. At both Time 1 and Time 2, RAB predicted additional variance over and above PBS, drinking quantity, drinking frequency, and age. In fact, at both time points, PBS was not significantly related to negative consequences in the models. This finding suggests that, concurrently, RAB are more influential predictors of negative consequences than are PBS.
Do PBS or RAB Predict Negative Consequences Longitudinally?
Next, we tested whether PBS and RAB measured at Time 1 predicted change in negative consequences across time. See Table 3 for a summary of these regressions. Again, Type 1 error was controlled by employing a Bonferroni correction ([alpha] = .05 / 2 = .025). For the first analysis (PBS Model Step 1, Table 3), Time 2 RAPI scores were predicted from Time 1 PBS controlling for Time 1 RAPI scores (i.e., to predict change), age, and Time 1 quantity and frequency. Initial PBS were found to negatively predict later RAPI scores, suggesting that engaging in safety behaviors reduces the likelihood of experiencing additional future negative consequences. We ran a similar analysis replacing Time 1 PBS with Time 1 RAB (RAB Model Step 1, Table 3). RAB were unrelated to Time 2 negative consequences.
Like before, we also were interested in the unique effects of PBS and RAB in predicting negative consequences longitudinally.
To this end, we added RAB as a second step in the regressions predicting Time 2 RAPI scores from Time 1 protective behavioral strategies, and vice versa, and then tested whether addition of the new variable significantly accounted for additional variance in the model (see Step 2 in Table 3). We also tested the interaction between PBS and RAB in the longitudinal model, and again this interaction term was not significant. Thus, only main effects models are shown in Table 3. In these models, PBS contributed additional variance over and above RAB, drinking quantity, drinking frequency, and age in predicting change in RAPI scores. RAB did not significantly relate to negative consequences in the models including both PBS and RAB. This finding suggests that the PBS, but not the RAB, in which a person engages influences experience of negative consequences over time. Women initially practicing more PBS are likely to experience reduction in future negative consequences.
Do Negative Consequences Predict Changes in PBS or RAB Longitudinally?
Given the possibility that experienced consequences may lead an individual to change future behavior, we also conducted a regression analysis to test whether initial (Time 1) negative consequences predicted later (Time 2) PBS, controlling for Time 1 values of PBS, age, drinking quantity and frequency. Time 1 RAPI scores did not significantly predict changes in PBS (b = .00, [beta] = -.01, t = -.20, ns). A similar analysis was conducted predicting Time 2 RAB from Time 1 negative consequences, controlling for Time 1 values of RAB, drinking quantity and frequency. Likewise, Time 1 RAPI scores did not predict changes in RAB significantly (b = .00, [beta] = -.01, t = -.01, ns). Taken together, results do not support the idea that experiencing negative drinking consequences either increases future PBS or decreases future RAB.
The current investigation examined transactional relations of using PBS and engaging in RAB to experiences of alcohol-related negative consequences using a four-month, prospective design in a sample of undergraduate women. To our knowledge, this study is the first to examine these potential protective and risk factors in a longitudinal context and to examine effects over and above both quantity and frequency of consumption. The sampled college women were engaging in high risk episodic drinking (approximately .09 to. 10 BAC) on a moderate basis and reported experiencing mild to moderate negative consequences, supporting past indications that college women's drinking is of concern (NIDA, 2003). Although RAB and RAPI scores decreased over time, potentially a result of a developmental maturing process (Marlatt et al., 1998) or of contextual factors (e.g., higher drinking during collegiate football season or warm weather, both at Time 1 for our sample; e.g., Neal & Fromme, 2007), important relations of RAB and PBS to negative consequences emerged. Both PBS and RAB affect alcohol-related outcomes for women, but differently. Our findings may have specific implications for both broad-based prevention efforts and for intervention with women who may already be engaging in problematic drinking patterns.
Concurrent Relations of PBS and RAB to Negative Consequences
The extent to which women engaged in PBS was not significantly related to experiencing fewer negative consequences at either Time 1 or Time 2 over and above alcohol consumption, indicating that women did not experience immediate benefits of using PBS. This finding is contrary to our hypotheses and findings of past studies (Benton et al. 2006; Benton et al., 2004; Delva et al., 2004; Martens et al., 2004; Martens et al., 2005). The lack of an association may be a result of controlling for both quantity and frequency of drinking, which was not done previously, and is more easily interpreted when also considering the longitudinal findings for PBS (discussed below).
On the other hand, engaging in RAB was related to experiencing more negative drinking consequences, even when controlling for consumption. This relation was similarly robust at both time points. Given the moderate correlation between PBS and RAB at both time points, it was important to examine the interplay of these two constructs in predicting alcohol-related outcomes. Yet, when PBS, RAB, and their interaction were entered in the model, only the main effect of RAB remained a significant predictor of negative consequences, at both time points, with similar strength. Overall, the findings suggest RAB is an important concurrent predictor of increased negative consequences in that individuals are likely to experience the immediate impact of RAB. For example, a woman who has unprotected sex when drinking is more likely to experience negative effects (e.g., STIs, pregnancy, guilt) immediately and intensely. Above consumption, RAB accounted for almost 10 percent of the variance in negative consequences, and for every one unit increase in RAB, RAPI scores increased by roughly 6.5 points. Both facts suggest that the construct of RAB warrants future examination. Overall, RAB findings align with our hypotheses and suggest that interventions aimed at decreasing women's RAB may be beneficial in reducing a greater number of negative drinking consequences.
Longitudinal Relations of PBS, RAB, and Negative Consequences
As noted, one major limitation in the literature is the lack of longitudinal research examining PBS, RAB, and drinking patterns. Our findings provide a clearer picture about the predictive relationship of PBS and RAB with regard to negative alcohol-related outcomes. Contrary to concurrent findings, but in line with our hypotheses, use of more PBS was related to change in drinking outcomes over time. That is, the more PBS endorsed at Time 1, the lower the number of alcohol-related problems at Time 2, when controlling for Time 1 RAPI scores and alcohol consumption. On the other hand, in contrast to concurrent relations and hypotheses, increased RAB at Time 1 did not predict increased alcohol-related problems at Time 2. Further, when RAB, PBS, and their interaction were included in the longitudinal model, only the main effect of PBS (and notably, neither quantity nor frequency of use) predicted change of negative consequences over time such that initial increased use of PBS predicted a reduction in later consequences.
Thus, PBS, but not RAB, appears to be an important predictor of alcohol-related consequences over time. Even when controlling for consumption, increased use of PBS by female college students may help to prevent alcohol-associated problems from developing. For example, a woman who counts her servings each time she drinks will likely have a more accurate and efficient estimate of her consumption amount which may, in turn, reduce the negative consequences she experiences later. This longitudinal finding is particularly important given that a significant concurrent association was not found in the present study. The long-term benefits of PBS may be especially potent and the cross-sectional studies conducted to date may actually underestimate the effect of PBS on negative consequences.
Negative Consequences Do Not Predict Changes in PBS or RAB
Finally, authors of previous studies examining PBS and risky behaviors (e.g., Benton et al., 2004) have suggested that, because their respective studies were cross-sectional, the possibility that the experience of negative consequences drives changes in either PBS or RAB could not be ruled out. The current study was able to directly test this hypothesis, and the findings suggest that students do not increase use of PBS or reduce RAB in response to experiencing negative consequences. These results corroborate Mailer and colleagues' (2006) finding that individuals who experienced a negative consequence from a risky behavior were likely to engage in the same risky behavior again and reexperience the same consequence in the future. Previous work has also suggested that change after a negative alcohol incident is more likely to occur for lighter drinkers who experience more severe consequences (Barnett, Goldstein, Murphy, Colby, & Monti, 2006). On average, negative consequences may not have been severe enough in our sample of moderate to heavy drinkers to elicit changes in behavior. Relying on natural consequences of drinking to change risky behaviors is not sufficient. Prevention and intervention efforts, especially those focusing on increasing PBS or reducing RAB, appear to be a necessary part of reducing alcohol-related negative consequences.
Limitations and Future Research
This study's results should be considered within its limitations. One purposeful limitation was our focus solely on college women. Consequentially, the results may not be as applicable to men or even similarly-aged women not enrolled in college. However, this study indicated that women drink at risky-levels, experience significant negative consequences, and engage in behaviors that amplify risk. Further, our sample was primarily Caucasian. White students may engage in high-risk drinking behaviors more than non-White students (e.g., Wilke et al., 2005). Thus, whether our findings regarding PBS and RAB generalize to non-White college women is unknown. Finally, this study focused on a narrow set of PBS and was an exploratory investigation of RAB. As such, internal consistency was somewhat low for PBS at time one and RAB at time two. However, it is not completely unexpected that individuals who engage in one type of PBS or RAB do not necessarily engage highly in others. And, there are likely additional PBS (e.g., putting extra ice in mixed drinks) and RAB (e.g., drinking in an unfamiliar setting) that college women engage in, and relations of these unstudied behaviors to negative outcomes may be even stronger.
Our goal was to conduct an initial examination of PBS and RAB and investigate the causal relation between these constructs and negative consequences of alcohol use. Yet, the environmental context in which drinking occurs may be an important moderator of the relations among drinking, PBS, RAB, and negative outcomes. Considering whether women (and men) are more likely to use PBS or engage in RAB in certain situations (e.g., being more likely to set limits before drinking if going out in public versus drinking in private dorms or sorority houses) generates many novel hypotheses and exciting ideas for future theory and research. Despite these limitations, our results inform a better understanding of women's drinking patterns and can contribute to better prevention and intervention programs.
Conclusions and Implications
Findings have implications for alcohol education. First, interventions that target only drinking quantity and frequency may be effective in reducing consumption, but are only addressing a portion of the overall high-risk drinking problem. Our results imply that PBS and RAB are two additional components that should be addressed in intervention programs. Second, the time points at which each construct was significant not only suggest that RAB and PBS are distinct constructs, but may provide two avenues for affecting alcohol related outcomes. Efforts that target RAB versus PBS will likely have differential results. Focusing on RAB is likely to result in immediate reduction of negative consequences, while teaching PBS will assist in building better drinking practices with benefits emerging over time. In this regard, decreasing RAB may be an effective intervention whereas increasing PBS may be an important target for prevention. Moreover, PBS and RAB are important constructs to target because they are specific actions in which individuals engage that impact alcohol-related outcomes, even if the individuals have other less controllable risk factors (e.g., family history of alcohol abuse). Overall, the most effective treatments would likely incorporate both efforts to increase PBS and to decrease RAB.
Correspondence concerning this article should be addressed to: Aaron M. Luebbe, 210 McAlester Hall, University of Missouri-Columbia, Columbia, MO 65211, (573) 771-9093, (Fax) 8821751 or email: AML5F5@mizzou.edu.
Funding for this study was provided by a Next Level Grant, #50700, awarded to the third author. We wish to thank Mary J. Heppner, Ph.D. for helpful comments on a previous version of this manuscript
Items measuring Protective Behavioral Strategies:
1. Before starting to drink, I set a limit on how much I want to drink.
2. When I go out drinking, I have a plan of how I am going to get home.
3. When at a party, bar, or club, I keep my drink in my possession.
4. When drinking, I know where my drink is at all times.
5. I eat within a few hours before I start drinking.
6. When my friends and I are drinking, we make sure to watch out for the physical safety of each other.
Items measuring Risk Amplifying Behaviors:
1. Used an illegal substance in combination with alcohol?
2. Drove shortly after having more than four drinks?
3. Engaged in unplanned sexual activity?
4. Not used protection during sexual activity?
Aaron M. Luebbe
Dept. of Psychological Sciences
University of Missouri-Columbia
Shiloh Varvel & Kim Dude
Wellness Resource Center
University of Missouri-Columbia
Barnett, N.P., Goldstein, A.L., Murphy, J. G., Colby, S. M., Monti, P. M. (2006) "I'll Never Drink Like That Again": Characteristics of alcohol-related incidents and predictors of motivation to change in college students. Journal of Studies on Alcohol, 67, 754-763.
Benton, S. L., Benton, S. A., & Downey, R. G. (2006). College student drinking, attitudes toward risks, and drinking consequences. Journal of Studies on Alcohol, 67, 543-551.
Benton S. L., Schmidt, J. L., Newton, E B., Shin, K., Benton, S. A., & Newton, D. W. (2004). College student protective strategies and drinking consequences. Journal of Studies on Alcohol, 65, 115-121.
Byrnes, J. P., Miller, D. C., & Schafer, W. D. (1999). Gender differences in risk taking: A meta-analysis. Psychological Bulletin, 125(3), 367-383.
Clapp, J. D., Shillington, A. M., & Segars, L. B. (2000). Deconstructing contexts of binge drinking among college students. American Journal of Drug and Alcohol Abuse, 26(1), 139-154.
Delva, J., Smith, M. P., Howell, R. L., Harrison, D. F., Wilke, D., & Jackson, L. (2004). A study of the relationship between protective behaviors and drinking consequences among undergraduate college students. Journal of American College Health, 53(1), 19-26.
Gruenwald, P.J., Johnson, F. W., Light, J. M., Lipton, R., & Saltz, R. F. (2003). Understanding college drinking: Assessing dose response from survey self-reports. Journal of Studies on Alcohol, 64, 500-514.
Jeffery, R.W. (1996) Bias in reported body weight as a function of education, occupation, health, and weight concern. Addictive Behavior, 21, 217-222.
Mallett, K. A., Lee, C. M., Neighbors, C., Larimer, M. E., & Turrisi, R. J. (2006). Do we learn from our mistakes? An
examination of the impact of negative alcohol-related consequences on college students' drinking patterns and perceptions. Journal of Studies on Alcohol, 67, 269-276.
Martens, M. P., Ferrier, A. G., Sheehy, M. J., Corbett, K., Anderson, D. A., & Simmons, A. (2005). Development of the protective behavioral strategies survey. Journal of Studies on Alcohol, 66, 698-705.
Martens, M. P., Taylor, K. K., Damann, K. M., Page, J. C., Mowry, E. S., & Cimini, M. D. (2004). Protective behavioral strategies when drinking alcohol and their relationship to negative alcohol-related consequences in college students. Psychology of Addictive Behaviors, 18(4), 390-393.
McCabe, S. E. (2002). Gender differences in collegiate risk factors for heavy episodic drinking. Journal of Studies on Alcohol, 63, 49-56.
Miller, E. T., Neal, D. J., Roberts, L. J., Baer, J. S., Cressler, S. O., Metrik, J., & Marlatt, G. A. (2002). Test-retest reliability of alcohol measures: Is there a difference between Internet-based assessment and traditional methods? Psychology of Addictive Behaviors, 16, 56-63.
Mohler-Kuo, M., Dowdall, G. W., Koss, M. P., & Wechsler, H. (2003). Correlates of rape while intoxicated in a national sample of college women. Journal of Studies on Alcohol, 65. 37-45.
Naimi, T. S., Brewer, R. D., Mokdad, A., Denny, C., Serdula, M. K., & Marks, J. S. (2003). Binge drinking among US adults. Journal of the American Medical Association, 289(1), 70-75.
National Highway Traffic Safety Administration (NHTSA) (1994). Computing a BAC estimate. U.S. Department of Transportation.
National Institute on Alcohol Abuse and Alcoholism (NIAAA). (2003). Overview of NIDA research on women's health and gender differences. Retrieved December 12, 2005, http:// www.drugabuse.gov/WHGD/WHGDIntro.html.
National Institute on Alcohol Abuse and Alcoholism (NIAAA). (2004). NIAAA council approves definition of binge drinking. NIAAA Newsletter, 2004(3). Retrieved November 18, 2005, http://www.niaaa.nih.gov.
Neal, D. J., & Fromme, K. (2007). Hook 'em horns and heavy drinking: Alcohol use and collegiate sports. Addictive Behaviors, 31, 2681-2693.
Nelson, C. B., Heath, A. C., & Kessler, R. C. (1998). Temporal progression of alcohol dependence symptoms in the U.S. household population: Results from the national comorbidity survey. Journal of Consulting and Clinical Psychology, 66, 474-483.
Nolen-Hoeksema, S. (2004). Gender differences in risk factors and consequences for alcohol use problems. Clinical Psychology Review, 2, 981-1010.
Park, C. L. & Grant, C. (2005). Determinants of positive and negative consequences of alcohol consumption in college students: alcohol use, gender, and psychological characteristics. Addictive Behaviors, 30, 755-765.
Presley, C. A. & Pimentel, E. R. (2006). The introduction of the heavy and frequent drinker: A proposed classification to increase accuracy of alcohol assessments in postsecondary educational settings. Journal of Studies on Alcohol, 67, 324-331.
Rohsenow, D. J. (1998). Alcoholism. In: Blechman, E. A., Brownell, K. D. (Eds.), Behavioral medicine and women: A comprehensive handbook. New York, NY: The Guilford Press, pp. 402-406.
Schaefer, J. L. (1999). NORM: Multiple imputation of incomplete multivariate data under a normal model, version 2. Software available from http://www.stat.psu.edu/-jls/misoftwa.html.
Stewart, A. L. (1982). The reliability of self-reported weight and height. Journal of Chronic Disease, 34, 295-305.
Vik, E W., Carrello, P., Tate, S. R., & Field, C. (2000). Progression of consequences among heavy-drinking college students. Psychology of Addictive Behaviors, 14(2), 91-101.
Wechsler, H., Dowdall, G., Davenport, A., & Castillo, S. (1995). Correlates of college student binge drinking. American Journal of Public Health, 85(7), 921-926.
White, H. R. & Labouvie, E. W. (1989). Towards the assessment of adolescent problem drinking. Journal of Studies on Alcohol, 50, 30-37.
Wilke, D. J., Siebert, D. C., Delva, J., Smith, M. P., & Howell, R. L. (2005). Gender differences in predicting high-risk drinking among undergraduate students. Journal of Drug Education, 35(1), 79-94.
TABLE 1. Means, Standard Deviations, and Intercorrelations of Initial (Timely and Later(Time 2) Negative Consequences, Protective Behavioral Strategies, Risk-amplifying behaviors, Drinking Quantity, and Drinking Frequency (N = 138) M (SD) RAPI1 PBS1 RAB1 QTY1 RAPI1 11.12 (9.97) -- PBS1 3.27 (.52) -.31 ** -- RAB1 .54 (.57) .54 ** -.52 ** -- QTY1 .10 (.07) .47 ** -.67 ** .49 ** -- FRQ1 2.65 (.96) .40 ** -.16 .36 ** .47 ** RAPI2 7.38 (7.33) .57 ** -.40 ** .43 ** .32 ** PBS2 3.30 (.65) -.15 ** .46 ** -.34 ** -.24 ** RAB2 .37 (.50) .42 ** -.54 ** .71 ** .42 ** QTY2 .09 (.06) .43 ** -.24 ** .38 ** .62 ** FRQ2 2.64 (.90) .37 ** -.13 .25 ** .40 ** FRQ1 RAPI2 PBS2 RAB2 QTY2 RAPI1 PBS1 RAB1 QTY1 FRQ1 -- RAPI2 .36 ** -- PBS2 .01 -.11 -- RAB2 .37 ** .51 ** -.31 ** -- QTY2 .48 ** .39 ** .01 .39 ** -- FRQ2 .63 ** .44 ** .08 .33 ** .49 ** Note. The number following the acronym denotes time of assessment. RAPI = Rutgers Alcohol Problem Index. PBS = Protective behavioral strategies. RAB = Risk-amplifying behaviors. QTY = Drinking quantity. FRQ = Drinking frequency. * p < .05. ** p < .01. TABLE 2. Concurrent Predictions at Time 1 and Time 2 ofNegative Consequences from Protective Behavioral Strategies and Risk-amplifying Behaviors Controlling for Age, Drinking Quantity, and Drinking Frequency Predicting RAPI scores at TIME1 T1 PBS Model Step 1 Pred. [beta] t value p AGE -.07 -.94 .35 QTY .27 3.07 .00 ** FRQ .23 2.76 .01 * PBS -.17 -2.20 .03 RAB -- -- -- [R.sup.2] .28 F for [DELTA] [R.sup.2] 13.24 Predicting RAPI scores at TIME1 T1 RAB Model Step 1 Pred. [beta] t value p AGE -.08 -1.14 .25 QTY .18 2.08 .04 FRQ .16 1.99 .05 PBS -- -- -- RAB .39 4.85 .00 ** [R.sup.2] .37 F for [DELTA] [R.sup.2] 19.55 Predicting RAPI scores at TIME1 T1 Step 2 Pred. [beta] t value p AGE -.08 -1.16 .25 QTY .18 1.98 .05 FRQ .16 2.01 .05 PBS -.03 -0.30 .77 RAB .38 4.24 .00 ** [R.sup.2] .37 F for [DELTA] [R.sup.2] PBS Model = 18.85, p < .001 RAB Model = 0.05. Ns Predicting RAPI scores at TIME2 T2 PBS Model Step 1 Pred. [beta] t value p AGE .06 .77 .44 QTY .24 2.76 .01 * FRQ .34 3.92 .00 ** PBS -.14 -1.92 .06 RAB -- -- -- [R.sup.2] .26 F for [DELTA] [R.sup.2] 11.43 Predicting RAPI scores at TIME2 T2 RAB Model Step 1 Pred. [beta] t value p AGE .04 .54 .58 QTY .13 1.53 .13 FRQ .25 3.12 .00 ** PBS -- -- -- RAB .39 4.88 .00 ** [R.sup.2] .35 F for [DELTA] [R.sup.2] 18.02 Predicting RAPI scores at TIME2 T2 Step 2 Pred. [beta] t value p AGE .04 .56 .58 QTY .13 1.54 .13 FRQ .26 3.12 .00 ** PBS -.02 -0.32 .75 RAB .37 4.42 .00** [R.sup.2] .35 F for [DELTA] [R.sup.2] PBS Model = 19.55, p < .01 RAB Model = 0.09. Ns * p < .025. ** p < .005. TABLE 3 Summary of Regression Analyses Examining Associations of Initial (Time 1) Protective Behavioral Strategies and Risk-amplifying Behaviors with Later, (Time 2) Negative Consequences ControllingPredicting Quantity, and Drinking Frequency. Predicting RAPI scores at TIME1 T1 PBS Model Step 1 Pred [beta] t value p AGE .08 1.13 .26 RAPI .46 5.75 .00 ** QTY -.05 -.53 .56 FRQ .17 2.14 .03 PBS -.24 -3.26 .00 ** [R.sup.2] -- -- -- .40 F for [DELTA] [R.sup.2] 17.57 Predicting RAPI scores at TIME1 T1 RAB Model Step 1 Pred [beta] t value p AGE .09 1.24 .22 RAPI .45 5.17 .00** QTY -.O1 -.10 .92 FRQ .14 1.74 .09 PBS -- -- -- [R.sup.2] .14 1.59 .11 .37 F for [DELTA] [R.sup.2] 15.07 Predicting RAPI scores at TIME1 T1 Step 2 Pred [beta] t value p AGE .08 1.10 .28 RAPI .45 5.23 .00 ** QTY -.OS -0.60 .55 FRQ .17 2.07 .04 PBS -.23 -2.84 .00 ** [R.sup.2] .04 .39 .70 .40 F for [DELTA] [R.sup.2] PBS Model = 18.85, p < .001 RAB Model = 0.05. Ns Note. RAPI = Rutgers Alcohol Problem Index. Pred. = Predictor. PBS = Protective behavioral strategies. RAB Risk-ampli- fying behaviors. QTY = Drinking quantity. FRQ = Drinking frequency. Bonferroni-corrected alpha of .025 used. * p < .025. ** p < .005.
|Gale Copyright:||Copyright 2009 Gale, Cengage Learning. All rights reserved.|