Relationships between binge drinking and substance-free reinforcement in a sample of college students A preliminary investigation

Abstract

Heavy episodic drinking is a relatively common phenomenon among college students, and students who engage in binge drinking are at increased risk for a variety of adverse consequences. This paper investigates relationships between substance use and reinforcement derived from specific categories of substance-free activities among a sample of 256 college undergraduates. Data from a standardized behavioral inventory were used to compare the frequency, pleasure, and reinforcement potential of substance-free events and activities experienced by binge drinkers and a comparison group. Binge drinkers reported significantly lower scores across a variety of substance-free activity categories and, in the majority of the cases, the relationship between binge drinking and decreased reinforcement density remained significant after accounting for the effects of the use of other drugs and demographic variables. These results are consistent with a growing body of evidence linking substance use to deprivation of substance-free reinforcement.

1. Introduction

In a recent survey of college students, close to half of the men and 40% of the women reported at least one occasion of binge drinking during the past 2 weeks (Wechsler, Dowdall, Maenner, Gledhill-Hoyt, & Lee, 1998). Students who engage in binge drinking are 7–10 times more likely than nonbinge drinkers to engage in unsafe sexual practices, suffer an injury, and drive while drunk (Wechsler, Davenport, Dowdall, Moeykens, & Castillo, 1994). Behavioral theories on choice suggest that alcohol use is best understood within the context of competing sources of reinforcement (Vuchinich & Tucker, 1996). Correia, Simons, Carey, and Borsari (1998) applied behavioral choice theory to college student substance use. The results indicated that reinforcement from substance-free activities contributed to the prediction of substance use. Specifically, the frequency of substance use increased as reinforcement from substance-free activities decreased.

Van Etten, Higgins, Budney, and Badger (1998) explored the relationship between cocaine abuse and density of positive reinforcement derived from a variety of activities. These authors compared 100 cocaine abusers enrolled in outpatient substance abuse treatment to a sample of matched community controls. Cocaine abusers were less likely to engage in nonsocial, introverted, passive outdoor, and mood-related activities. Involvement in nonsocial, introverted, and mood-related activities was associated with greater cocaine abstinence during treatment. In light of the analysis and results reported in Van Etten et al. (1998), we conducted a new analysis of the data collected by Correia et al. (1998). We compared the frequency, pleasure, and reinforcement potential of specific categories of substance-free activities experienced by college binge drinkers and a comparison group. The current study sought to identify patterns of substance-free reinforcement deprivation associated with binge drinking.

2. Method

2.1. Participants

Participants were 256 undergraduates at a large private university who participated in research to fulfill a course requirement. The mean age of the sample was 18.98 (S.D. = 2.19); 57% were female and 16% were minorities. Participants were classified as binge drinkers (53%) if they reported consuming five or more standard drinks of alcohol on at least two occasions during the last 30 days. This definition of binge drinking is similar to other definitions reported in the literature (Meilman, Cashin, McKillip, & Presley, 1998). The remaining 47% of the sample served as the comparison group. Males were more likely to be classified as binge drinkers, χ²(1, N = 256) = 10.10, P < .001; minority participants were more likely to be classified as comparisons, χ²(1, N = 256) = 17.34, P < .001. A secondary set of analyses excluded the 96 participants who reported recent (last 30 days) use of illicit substance. The majority of those excluded reported marijuana use (n = 88). The restricted sample did not significantly differ from the full sample in terms of age and, as in the full sample, minorities were more likely to be classified as comparison participants, χ²(1, n = 160) = 9.93, P < .01. Females in the restricted sample were also more likely to be classified as comparison participants, χ²(1, n = 160) = 4.49, P < .05.

2.2. Procedures and measures

Participants provided written informed consent and completed a set of questionnaires addressing alcohol use and other topics in group sessions. Open-ended questions were used to assess the frequency and quantity of their drug and alcohol use for the last 30 days, including the number of occasions of binge drinking (five or more drinks on one occasion). Lifetime alcohol-related problems were assessed using the 23-item Rutgers Alcohol Problem Index (RAPI; White & Labouvie, 1989). This scale was designed for adolescents between the ages of 12 and 21, making it an appropriate tool for use with college students.

The Pleasant Events Schedule (PES) (MacPhillamy & Lewinsohn, 1982) is designed to measure the frequency and subjective pleasure of 320 potentially reinforcing events or activities. A cross-product score (Frequency × Subjective Pleasure) indicates the amount of reinforcement potential associated with engagement in each activity, which is considered a useful approximation of obtained positive reinforcement. Responses are also summarized in 10 rationally and empirically derived subscales; the subscales and representative items are listed in Table 1. Two notable modifications were made to the PES for the current study (see Correia et al., 1998 for a more complete description). First, participants were asked to provide two frequency and pleasure ratings for each activity: one set of ratings for times when the participants were substance-free, and the second set for times when participants engaged in the activity while using or under the influence of drugs or alcohol. Second, the original PES items that explicitly mention substance use were not used in calculations of summary scores or considered in the analyses. Thus, the substance-related reinforcement scores used in the current study refer to reinforcement from activities enhanced by substance use.

Table 1.

Subscales and sample items for the PES

Subscale	Sample items
Social	Having a frank and open conversation
	Going to a party
	Being with friends
Nonsocial	Watching TV
	Just sitting and thinking
	Taking a nap
Male	Going to a sports event
	Camping
	Gambling
Female	Taking a bath
	Giving gifts
	Complimenting or praising someone
Introverted	Doing art work
	Reading stories, novels, poems, or plays
	Going to the library
Extroverted	Driving fast
	Dancing
	Playing basketball
Solitary and tranquil	Having spare time
	Having daydreams
	Being alone
Passive outdoor	Looking at the moon or stars
	Seeing beautiful scenery
	Going on a nature walk or trip
Sexual	Getting a massage or backrub
	Being noticed as sexually attractive
	Having sexual relations with a partner of the opposite sex
Mood	Thinking about something good in the future
	Laughing
	Seeing good things happen to my family or friends

3. Results

Binge drinkers drank more frequently, consumed higher quantities of alcohol per occasion, and reported greater alcohol-related consequences than did members of the comparison group during the last 30 days (see Table 2). Unadjusted t test comparisons revealed that binge drinkers reported significantly higher scores on every PES substance-related frequency, pleasure, and cross-product subscale. Differences on the substance-free PES subscale scores are depicted in Table 3. Differences for unadjusted t test comparisons are considered significant at the P < .05 level. However, two different alpha levels (.01 and .05) are presented on the table to emphasize a more conservative evaluation of the significant effects while retaining useful information regarding patterns in the data.

Table 2.

Recent drinking practices of binge drinkers and comparison

	Binge drinkers (n = 136)		Comparison (n = 120)		t
	Mean	S.D.	Mean	S.D.
Days of alcohol use during the last 30 days	9.95	6.20	2.40	3.12	8.32
Days of binge drinking during the last 30 days	7.50	5.78	0.28	0.45	13.67
Number of drinks per occasion during the last 30 days	7.33	3.22	2.34	3.89	11.07
RAPI score	7.71	5.07	2.93	4.08	8.22

RAPI scores range from 0 to 23, P < .001 for all four t test comparisons variables.

Table 3.

Comparisons of binge drinker and comparison groups on PES substance-free subscales

	Full sample						Restricted sample
	Binge drinkers (n = 136)		Comparison (n = 120)		t Test P value	Regression P value	Binge drinkers (n = 66)		Comparison (n = 94)		t Test P value	Regression P value
	Mean	S.D.	Mean	S.D.			Mean	S.D.	Mean	S.D.
Frequency scores
Social	1.0	0.32	1.09	0.31	NS	NS	1.06	0.33	1.13	0.29	NS	NS
Nonsocial	0.92	0.22	0.99	0.21	< .01	< .05	0.92	0.22	1.01	0.21	< .05	< .05
Masculine	0.36	0.26	0.33	0.22	NS	NS	0.35	0.23	0.35	0.23	NS	NS
Feminine	0.85	0.26	0.93	0.25	< .01	NS	0.89	0.27	0.96	0.24	NS	NS
Introverted	0.59	0.25	0.68	0.25	< .01	< .01	0.58	0.23	0.69	0.26	< .01	< .01
Extroverted	0.76	0.26	0.72	0.24	NS	NS	0.76	0.24	0.75	0.24	NS	NS
Solitary	1.34	0.36	1.38	0.33	NS	NS	1.37	0.36	1.39	0.32	NS	NS
Passive outdoor	0.71	0.35	0.81	0.34	< .05	< .5	0.70	0.36	0.83	0.34	< .05	< .05
Sexual	0.95	0.42	0.98	0.45	NS	NS	0.92	0.44	0.98	0.47	NS	NS
Mood-related	1.16	0.29	1.21	0.25	NS	NS	1.16	0.30	1.23	0.25	NS	NS
Pleasure scores
Social	1.07	0.34	1.16	0.33	< .05	< .05	1.08	0.36	1.21	0.29	< .05	< .01
Nonsocial	0.82	0.25	0.90	0.26	< .01	< .05	0.80	0.28	0.92	0.26	< .01	< .05
Masculine	0.34	0.25	0.33	0.24	NS	NS	0.34	0.25	0.36	0.25	NS	NS
Feminine	0.84	0.29	0.95	0.28	< .01	< .05	0.58	0.31	0.69	0.30	< .05	< .05
Introverted	0.56	0.29	0.67	0.29	< .01	< .01	0.57	0.31	0.68	0.30	< .05	< .01
Extroverted	0.72	0.28	0.71	0.26	NS	NS	0.72	0.29	0.74	0.27	NS	NS
Solitary	1.37	0.42	1.49	0.37	< .05	< .05	1.38	0.45	1.51	0.37	NS	NS
Passive outdoor	0.76	0.38	0.91	0.39	< .001	< .01	0.72	0.41	0.94	0.39	< .01	< .001
Sexual	1.05	0.45	1.15	0.48	NS	NS	1.02	0.46	1.16	0.49	NS	NS
Mood-related	1.22	0.33	1.33	0.29	< .01	< .01	1.22	0.37	1.23	0.27	< .01	< .01
Cross-products
Social	1.69	0.66	1.81	0.65	NS	NS	1.73	0.70	1.90	0.63	NS	NS
Nonsocial	1.34	0.47	1.47	0.51	< .05	NS	1.34	0.51	1.51	0.52	NS	NS
Masculine	0.51	0.38	0.48	0.41	NS	NS	0.50	0.38	0.53	0.44	NS	NS
Feminine	1.38	0.53	1.57	0.54	< .01	NS	1.46	0.58	1.64	0.52	< .05	NS
Introverted	0.86	0.48	1.04	0.51	< .01	< .05	0.86	0.48	1.06	0.52	< .05	< .05
Extroverted	1.15	0.50	1.11	0.48	NS	NS	1.16	0.50	1.16	0.50	NS	NS
Solitary	2.22	0.83	2.32	0.75	NS	NS	2.26	0.87	2.36	0.74	NS	NS
Passive outdoor	1.17	0.67	1.39	0.68	< .01	< .05	1.15	0.71	1.43	0.68	< .05	< .01
Sexual	1.66	0.85	1.78	0.92	NS	NS	1.63	0.89	1.80	0.93	NS	NS
Mood-related	1.98	0.67	2.14	0.59	< .05	NS	2.00	0.72	2.19	0.57	< .05	NS

Carriers in the regression models are gender, age, race, and binge drinker classification (binge, comparison). Regression P value refers to the t test associated with binge drinker classification in the regression model. The restricted sample excludes participants who reported use of illicit drugs during the previous 30 days. NS: P > .05.

A series of multiple regressions analyses were conducted to determine if binge drinking continued to be associated with PES subscale scores after the effects of demographic variables were considered. In each regression, a PES substance-free subscale score was entered as the criterion variable, and age, gender, race, and binge drinker classification (binge drinker, comparison) were simultaneously entered as carries. Binge drinker classification was considered a significant predictor of the PES substance-free subscale score if it entered the model at P < .05. As Table 3 indicates, binge drinking displayed a continued relationship with the frequency of nonsocial, introverted, and passive outdoor activities. All relationships between binge drinking and pleasure subscales remained significant. Regarding the cross-products, the reinforcement potential derived from introverted and passive outdoor activities remained significantly related to binge drinking.

A secondary set of analyses that excluded recent users of other illicit drugs was conducted. Since all of the 160 participants selected for inclusion in the secondary analyses denied use of illicit drugs during the last 30 days, the substance-related PES scores effectively referred to occasions of alcohol use. These analyses are meant to eliminate the potential confound of other drug use and are reported along with the full sample analyses on Table 3. As with the full sample, when compared to the comparison group, binge drinkers reported more occasions of recent alcohol use (9.29 vs. 1.93, t = 10.01, P < .0001), recent occasions of binge drinking (6.55 vs. 0.27, t = 10.87, P < .0001), number of drinks per occasion (6.68 vs. 1.77, t = 11.09, P < .0001), and a greater number of alcohol-related consequences (6.29 vs. 2.46, t = 6.19, P < .0001). Binge drinkers from the restricted sample also reported higher scores on every substance-related frequency, pleasure, and cross-product subscale. Regarding the substance-free PES scales, analyses with the restricted sample revealed that the pattern of results was similar to the pattern found with the full sample, with a few notable exceptions. Binge drinkers and comparisons no longer differed on the frequency of engagement in feminine activities, the amount of pleasure derived from solitary activities, or the reinforcement potential derived from nonsocial activities after recent users of other drugs were excluded from the analyses.

4. Discussion

Students who engaged in binge drinking two or more times during the past month derived less reinforcement from certain types of substance-free experiences than students in the comparison group. Both Van Etten et al. (1998) and the current study suggest that heavy or high-risk substance use, broadly defined, is associated with reinforcement deprivation in specific types of substance-free activities. Within the current findings, the differences found on the nonsocial, passive outdoor, and introverted activities appear most robust, as binge drinkers reported lower frequency and pleasure scores, and the relationships between these scales and binge drinking persisted after accounting for demographic variables. Although the PES may not provide a comprehensive sample of activities available to our sample, the documented existence of substance-free reinforcement deficits may have implications for prevention efforts.

Individuals who engage in heavy substance use may select situations and environments that support activities compatible with substance use. As a greater percentage of total activity is associated with substance use, less time is available for substance-free activities. According to this perspective, the nonsocial, passive outdoor, and introverted activity categories identified in this study, and by Van Etten et al. (1998), appear to be the set least likely to be retained. It is also possible that heavier substance users may not experience reinforcement from substance-free activities that compete with substance use. Substance use may be used to alleviate feelings of boredom and constraint some adolescents associate with school and other daily activities (Larson, Csikszentmihalyi, & Freeman, 1984). Skill deficits may also limit the ability of some individuals to derive reinforcement from certain types of substance-free activities. Further work in this area, including longitudinal studies, will be needed to determine the nature of the relationship between high-risk substance use and substance-free activities.

The current study is limited by its post hoc design. Future studies specifically designed to investigate relationships between binge drinking and substance-free reinforcement would be improved by employing state-of-the-art definitions of binge drinking (e.g., frequent vs. occasional vs. nonbinge drinkers, gender-specific criteria). Although the current analyses focussed on the potential effects of binge drinking, the effects of recent use of other substances remain a confounding influence. The secondary analyses with the restricted sample addressed this problem by excluding the participants who reported recent use of other drugs. However, the results from the restricted sample are less likely to generalize to other college student samples. Another limiting factor is the number of comparisons. Given the exploratory nature of the paper, we elected to present the full range of finding to encourage continued exploration of relationships suggested by the behavioral choice perspective.