JOURNAL OF ADOLESCENT HEALTH 1992;13:487-492
MARK J. WERNER,
Entering college freshmen 61 = 3013)completed a ques-
tionnaire which assessed drinking behaviors and identified students at risk for problem drinking as defined by the CAGE (focuses on Cutting down on drinki noyance by criticism by others about drinking feelings about drinking, and the use of an Eye opener) questionnaire and Perceived-Benefit-of-Drinking Scale (PBDS). Students were 50%male with a mean age of 17.9 years. In the past month, 17% had drunk on 10 or more occasions, and 18% had binged on 6 or more occasions. by 21% and CAGE scores of 2 or greater were obtai PBDS scores of 3 or greater by 29%, re ng high risk for problem drinking, High-risk CAGE and PBDS scores wexe associated with frequent drinking and binging. Student reports of parental problem drinking were not associated with high risk for problem drinking. Intent to join a fraternity or sorority (the Greek system) was associated with frequent drinking, binging, and high-risk CAGE and PBDS scores. Approaches to screening for problem drinking which emphasize attitudes and beliefs may be useful. The Greek system appears to be attractive to high-risk students and should be a focus of prevention programming. KEYWORDS: Alcohol Screening for problem drinking College students
among college students has drawn the attention of university administrators and
The use of alcohol
From the Division of Adolescent Medicine, Varlderbilt University, Nashville, Tennessee. Address reprint requests to: Mark 1. Werner, M.D., Division of Adolescent Medicine, 436 Medical Center South, Vanderbilt University, Nashville, TN 37232. Manscript accepted November 13, 1991.
faculty, government agencies, and a concerned public. Alcohol consumption has steadily risen during the last 20 years with annual use of alcohol by college students reaching 90-92%, and daily use near 7% (l-4). Approximately 25-50% of college students can be classified as heavy drinkers; such drinking is more common among males, whites, freshmen, and sophomores (3,5). Problems due to drinking that are frequently experienced by college students include personal injury, accidents, missing class, blackouts, legal difficulties, poorer academic performance, acquaintance rape, sexually transmitted disease, and unplanned pregnancy (6-12). Almost 30% of college students report loss of some hours of normal functioning while recovering from drinking during the preceding week (6). The influence of peers, the accessibility of alcohol, and pressure to be accepted by a group affect behavior and alcohol consumption by college students (13). Members of fraternal organizations are more likely to drink and to incur alcohol-related problems (4,5). Students who are frequent heavy drinkers appear disproportionately among those planning to join fraternities and sororities (5). The socialization process inherent in the (Greek) fraternity system places an unusual emphasis on drinking, with heavy alcohol consumption among fraternity pledges strongly correlated with a positive view of the socialization value of alcohol (14,15). The actual influence that the college environment has upon students’ drinking behaviors remains poorly de, particularly in comparison wit and drinking patterns established prior to entering college. Despite the perceived prevalence of problem drinking among college students, a reliable and efficient means of screening students has not been
0 Societyfor Adolescent Medicine, 1992 Published by Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, New York, NY 10010
467
1054139xJ9245.00
488
WERNER AND GREENE
developed. This is partly complicated by the difficulty that there is no standardized definition of problem drinking. The criteria usually applied to problem drinking focus on a pattern of excessive alcohol use and the development of physical, social, or psychologic problems related to drinking (16,17). The Michigan Alcoholism Screening Test (MAST) has been administered on several college campuses with &29% of students found to be possibly alcoholic or at risk for alcoholism (l&19). However, the MAST has been found to produce a significant number of false positive identifications in the college population (20). Four clinical interview questions, the CAGE questions, focus on Cutting down on drinking, Annoyance by criticism of others about drinking, Guilty feelings about drinking, and the use of an Eye opener. The CAGE questionnaire is a screening instrument commonly used in health care settings (21) that has been used with college students but may not be adequately sensitive (16,17). Both MAST and CAGE were developed for detection of alcoholism rather than problem drinking. Although approaches to screening for alcohol abuse have traditionally relied upon assessment of drinking frequency, quantity, and psychosocial consequences (22,23), an alternative approach utilizes the adolescent’s perceptions of why they drink. Petchers and Singer (24) and Petchers et al. (25) developed the Perceived-Benefit-of-Drinking Scale (PBDS) which probes, from the adolescent’s point of view, the specific reinforcements received from drinking. The more benefits an adolescent attributes to drinking, the more likely he or she is to drink. The PBDS scores are significantly higher for adolescents who report being drunk more often, who claim to have had a problem with drinking, and who have reported trouble due to drinking. The scale’s strong relationship with drinking behavior indicators makes it a potentially useful clinical screening device because it can serve as a proxy measure to assess drinking behaviors. The present study was undertaken to assess problem drinking among entering college freshmen, as identified by the PBDS and the CAGE questionnaire, and to assess the utility of these measures as screening tools that may be useful in a student health care setting. Students were studied at the start of their freshman year, prior to exposure to the college environment and the (Greek) fraternity/sorority SYSkm. Relationships between attitudes toward alcohol, problem drinking, and interest in the Greek system are evaluated. Implications for screening and intervention for problem drinking are discussed.
JOURNAL OF ADOLESCENT HEALTH Vol. 13, No. 6
Methods A sample of 308 freshmen students, repre,senting 20% of the 1500 students comprising the entering freshman class at a private southern university, were administered an anonymous questionnaire (during mandatory freshmen orientation programs conducted in each dormitory and directed at alcohol and drug issues. Students attended the orientation programs in groups of 30-50. Groups were randomly selected for participation such thit! there would be an approximately equal number of students from each coeducational dormitory. All students present at the selected orientation programs voluntarily participated. The programs occurred within the first 3 weeks of the fall semester. All freshmen students at the university are required to live on campus in either of two dormitories. Participants were 153 female and 155 male students from 16 to 19 years of age (mean, 17.9 years; SD = 0.5). Prior approval for the study was obtained through the Institutional Review Board. The questionnaire consisted of demographic information; self-report of frequency of alcohol, tobacco, and other drug use; self-report of intentions to join the Greek system; the PBDS; the CAGE questions; and questions about attitudes toward drinking. For the purposes of this study, frequent drinking was defined as drinking on 10 or more occasions in the past 30 days, and frequent hinging as drinking 5 or more drinks on 6 or more occasions in the past 30 days. A copy of the questionnaire is available upon request. The PBDS is a brief clinical screening instrument useful in identifying adolescents who may be alcohol abusers. It consists of endorsing as true or false each of the following five statements: drinking helps me forget, drinking helps me be friendly, drinking helps me feel good about myself, drinking helps me relax, and drinking helps me be friends with others who drink. Its reliability and validity are well established (24,25). This scale assesses the benefits an adolescent attributes to drinking and is not depend,ent upon direct information about drinking patterns or negative consequences resulting from drinking. A full evaluation of potential drinking problems is strongly recommended for those scoring 3 or greater (2425). The CAGE questions have proven use screening for alcoholism (21). The questions may be self-administered. Although a positive response to the CAGE questions is not diagnostic of alcoholism, a score of 2 or greater is highly suspicious and warrants further evaluation.
September 1992
PROBLEM DRINKING AMONG FRESHMEN
Students were grouped according to gender or their intention to join a fraternity or sorority. X2analyses were performed to test differences regarding frequent drinking, frequent binging, and the prevalence of high-risk CAGE and FBDS scores for these groups. Similarly, X2 analyses were used to test associations between high-risk CAGE or PBDS scores and frequent drinking, frequent binging, parental problem drinking, and drinking attitudes.
489
frequent drinking for female students only (p < 0.02). A history of frequent drinking or frequent binging may not necessarily indicate problem drinking as defined by CAGE or PBDS scores. High-risk CAGE scores were obtained by only 35% of students reporting frequent drinking and 42% of those reporting frequent binging. Similarly, high-risk PBDS scores were obtained by only 46% of students reporting frequent drinking and 50% of those reporting frequent bingeing. Of the students, 70% reported having a parent
Results
who drinks; 10% felt a parent had a drinking problem. Student report of parental problem drinking was not associated with frequent drinking, frequent Wnging, CAGE, or PBDS scores. Of the 308 students, 190 (62%) expressed an intent to join a fraternity or sorority, representing 50% of males and 73% of females. Those intending to join included 93% of those reporting frequent drinking, 89% of those reporting frequent binging, 75% of students with high-risk CAGE scores, and 76% of students with high-risk PBDS scores. Intent to join the Greek system was significantly associated with frequent binging (p < O.OOl), high-risk CAGE (p < 0.005), and high-&k PBDS scores (p < 0.02) for male students (Table 2). Analyses for female students were limited by the small number of female students not intending to join the Greek system with frequent drinking or high-risk scores. The same general trends, however, appear to hold for femalestudents (Table2). There was no associationbetweenparental problem drinking and intention to join the Greek system for male or female students. Students were also questioned about their attitudes toward drinking;17% felt drinkingmakes sexual experiences easier and more enjoyable, 55% feel alcoholic beverages make parties more fun, and 37%
Of the subjects 75% had drunk alcohol in the past 30 days, 51% had drunk on three or more occasions, and 17% on ten or more occasions. When asked about binge drinking, defined as five or more drinks on one occasion, 56% admitted to at least one binge episode in the last 30 days, 18% admitted to binging on 6 or more occasions in the last month. Drinking frequency did not differ between sexes; however, binge drinking was more frequent in males (p c 0.02). The distribution of PBDS scores and CAGE scores is shown in Table 1. PBDSscores of 3 or greater were obtained by 29% and CAGE scores of 2 or greater by 21%, reflecting high risk for problem drinking. Of the students, 122 (40%) had high-risk scores on either the PBDS or CAGE, and 33 students (11%) had high-risk scores on both the PBDS and CAGE. CAGE or PBDS scores were not significantly different between sexes. High-risk CAGE scores were associated with frequent drinking (p c 0.004) and frequent binging (p< 0.001) for male students only. On the other hand, high-risk PBDS scores were associated with frequent binging for both male and female students (p < 0.02) and p c 0.009, respectively), and with
Table 1. Perceived-Benefit-of-Drinking (PBDS)Scale and CAGE”Scores for College Freshmen CAGE ScoresC
PBDS Scoresh
(%)
Total
@)
40
(26)
85
(28)
86
(56)
28
(18)
55
(18)
32
(21)
39
(26)
78
(25)
27
(17)
(14)
27
(18)
48
(16)
8
15
(10)
10
(7)
25
(8)
2
8
(5)
8
(5)
16
(5)
Score
Male
0
4s
(29)
1
27
(17)
2
39
(25)
3
21
4 5
(%)
Female
Male
@)
(%)
Total
(%)
88
(58)
174
(56)
36
(24)
69
(22)
19
(12)
46
(15)
(5)
6
(4)
14
(5)
(1)
3
(2)
5
(2)
Female
“CAGE questionnaire focuses on Cutting down on drinking, Annoyance by criticism of others about drinking, Guilty feelings about drinking, and the use of an Eye opener. ‘Scores 2 3 indicate high risk, x2 male vs. female p < 0.8. ‘Scores 2 2 indicate high rusk, x2 male vs. female p < 0.7.
498
WERNERAND GREENE
JOURNAL OF ADOLESCENTHEALTH Vol. 13, No. 6
Table 2.
Problem Drinking and Intentions to Fraternity/Sorority
JOhI
a
Intention to join Yes n = 190
No II = 113
78 111
75 38
Frequent drinking” Male Female Total
31 19 50
2 2 4
Frequent hinging* Male Female Total
32 19 51
5 1 6
p < 0.001
High-risk CAGE’score (2 2) Male 26 Female 22 Total 48
10 5 15
p < 0.005 p < 0.02
High-risk PBDSdscore (2 3) Male 30 Female 38 Total 68
14 7 21
p c 0.02 p < 0.08 p < 0.002
Parent problem drinking Male 4 Female 14 Total 18
7 6 13
Gender Male Female
p < 0.001
“Frequent drinking, drinking on 10 or more occasions in the Past 30 days. bFrequentbinging, drinking 5 or more drinks on 6 or more occasions in the Past 30 days. ‘CAGE questionnaire focuses 07 Cutting down on drinking, Annoyance by criticism of others about drinking, Guilty feelings about drinking, and the use of an Eye opener. “PBDS, Perceived-Benefit-of-Drinking Scale.
agree that people feel powerful when they drink. In comparison, 91% of students reporting frequent drinking and 91% of those reporting frequent binging endorsed alcohol as making parties more fun (p < 0.001 and p < 0.001, respectively). Both highrisk CAGE and PBDS scores were associated with the same endorsement (p < 0.005 and p < 0.001, respectively).
Discussion Problem drinking on the college campus continues to be a significant public health concern. The fmdings of this study are generally in agreement with others conducted in a variety of university settings (3,5,22,23). Although drinking frequency did not differ between sexes, binge drinking was more com-
mon with male students. This study used two approaches to defining problem drinking which have not been used to our knowledge with college students; both are known to correlate highly with problem drinking or alcoholism. The PBDS defines problem drinking based on beliefs and attitudes, and is not dependent upon direct information about drinking patterns or negative consequences resulting from drinking. Beliefs about the effects that drinking will produce are known to influence both consumption and the subjective experience of intoxication (24,27). Previous studies have found that expectancies are associated with differing patterns of alcohol use and are capable of predicting transition to problem drinking in adolescents (28,29). Viewed in behavioral terms, the greater the benefit perceived from drinking, the more likely such behavior is to be reinforced, and thereby possibly increase (24,25). On this basis, 29% of the students studied received scores indicative of high risk for problem drinking. Because of ;is simple brief indirect approach, the PBDS may be a very useful clinical screening measure. The CAGE questionnaire, which relies on a more traditional approach to alcoholism screening, identified 21% of the students as high risk for problem drinking. While the sensitivity of the CAGE for problem drinking among college students remains unclear, its specificity is consistently above 85% (16,17). As such, these results may underestimate students at risk for problem drinking. Of the 122 students with high-risk scores on either the CAGE or PBDS, only 33 (27%) had high-risk scores on both measures. This suggests that the CAGE and PBDS may not be identifying the same at-risk population. The CAGE appears to be more applicable to identifying students with heavy drinking patterns and a high risk for alcoholism; particularly male students. In contrast, the PBDS identifies a larger number of male and female students who appear to have problem drinking and attitudes that reinforce their drinking patterns, but are not necessarily at risk for alcoholism. As such, the PBDS may be a more sensitive screening measure for students at risk for problem drinking and its associated health morbidities, that can be used in conjunction with the CAGE. Student report of parental problem drinking does not appear to help in identifying high-risk students; however, this study did not make in-depth efforts to assess parental drinking and possible alcoholism. Previous research indicates that most children and adolescents from families with alcoholism do function well (30,31).
September 1992
In the college setting, it can be very difficult to define what constitutes problem drinking. The interpretation of heavy drinking may depend on whether this form of drinking in college is a type of incipient problem drinking or simply a manifestation of usual adolescent development in an influential environment (32). Any evaluation of college students’ d&&ing patterns needs to consider the social environment and situational factors involved in college students’ decisions about drinking. Reliance upon self-report of drinking frequency and quantity may overestimate problem drinking in this population. Approaches that utilize an understanding of students’ beliefs, expectations, and motivations may be more useful. Similarly, given the influence of the college environment, there are likely to be significant numbers of students who suffer an adverse health or psychosocial outcome as a result of drinking but do not have alcoholism. Screening approaches should attempt to identify students who are at risk for alcohol-related problems as a result of their drinking behaviors and not rely solely on traditional alcoholism screening tools. This study is in agreement with others that college students intending to join the Greek system are more likely to be frequent, heavy, problem drinkers (5,23). Students appear to arrive at college with the expectation that the Greek system will support a heavy drinking pattern. Obviously, it is possible that a particular university or Greek system may be more attractive to students already involved in problem drinking. Many entering college students have prior knowledge of campus culture, traditions related to drinking, and Greek social systems. Nonetheless, attention to fraternities and sororities needs to be to be a significant part of any university alcohol prevention programming. Although the student population sampled in this study comes from a selective university, it is not different from students at a large number of other universities. It is important to note that this study was done at the start of a fall semester, a time when episodic heavy drinking may be more frequent. Longitudinal research focusing upon changes in drinking patterns throughout the college years and the association between drinking patterns and various psychosocial and medical problems is needed. Approaches to screening w drinking behaviors and associated problems are needed. Such studies are now underway. Problem drinking on college campuses is a complex issue requiring comprehensive campus-wide approaches to prevention intervention, and treat-
PROBLEM DRINKING A
491
ment. Screening for problem drinking should be a routine part of student health services. Many approaches are available; however, those which rely upon indirect measures of alcohol use through an understanding of students’ beliefs and attitudes may be most useful. Supported in part by a grant from the National Institute on Alcohol Abuse and Alcoholism and the National institute of Drug Abuse, RFA-AA-90-2.
References 1. Saltz R, Elandt D. College student drinking studies 19761985. Contemp Drug Probl1986;13:117-59. 2. Johnston LD, O’Malley PM, Bachman JG. Drug use among Americanhigh school students, college students and other young adults: National trends through 1985. Rockville, MD: National Intstute on Drug Abuse, 1986. 3. Pope HG, Ionescu-Pioggia M, Aizley HG, et al. Drug use and lie style among college undergraduates in 1989: A comparison with 1969 and 1978. Amer J Psychiatr 1990;147:998-1001, 4. Johnston LD, O’MaRey PM, Bachman JG. Illicit drug use, smoking, and drinking by American high school students, college students, and young adults, 1975-1987. Washington, DC: Government Printing Office, 1988; NIDA, ADAMHA, PHS, US Dept HHS. DHHS publication no. (ADM) 89-1602. 5. Canterbury RJ, Gressard CF, Vieweg WVR, et al. Psychosocial inventory among first-year college students by patterns of alcohol use. Adv Alcohol Subst Abuse 1990;9:1-11. 6. Meilman PW, Stone JE, Gaylor MS, et al. Alcohol consumption by college undergraduates: Current use and lo-year trends. J Stud Alcohol 1990;51:389-95. 7. Berkowitz AD, Perkins HW. Problem drinking among college students. J Amer Co11Health 1986;35:21-8. 8. Jessor R, Chase JA, Donovan JE. Psychosocial correlates of marijuana use and problem drinking in a national sample of adolescents. Amer J Public Health. 1980;70604-13. 9. Nicholi AM. The nontherapeutic use of psychoactive drugs: A modem epidemic. N Engl J Med 1983;308:925-33. 10. Nicholi AM. Characteristics of college students who use psychoactive drugs for nonmedical reasons. J Amer Co11Health 1985;33:189-92. 11. Gliksman L. Consequences of alcohol use: Behavior changes and problems during the first year of university. J Addict 1988;23:1281-95. 12. Engs RC, Hanson DJ. Gender differences in drinking patterns and problems among college students: A review of the literature. J Alcohol Drug Educ 1990;35:36-47. 13. Hanson DJ. Trends in drinking and behaviors among college students. J Alcohol Drug Educ 1974;19:6-14. 14. Leavy RL. An alternative source of information on a popular topic: A course in drinking. Teaching Psychology 1979;6:97100. 15. Faulkner RR, Alcorn JD, Garvin RB. Prediction of akohol consumption among fraternity pledges. J Alcohol Drug Educ 1989;34:12-20. 16. Smith DS, Collins M, Kreisberg JP, et al. Screening for problem drinking in college freshmen. J Amer Co8 Health 1987; 36:89-94.
492
WERNERAND GREENE
17. Heck EJ. Developing a screening questionnaire for problem drinking in college students. J Amer Coll Health 1991; 39227-31. 18. Favazza AR, Cannell B. Screening for alcoholism among college students. Amer J Psychiatr 1977;134:1414-16. Silber TJ, Capon M, Kuperschmit I. Administration of the 19. Michigan Alcoholism Screening Test (MASTJ at a student health service. J Amer Coll Health 1985;33229-33. 20. Martin CS, Liepman MR, Young CM. The Michigan Alcoholism §creening Test: False positives in a college sample. Alcohol Clin Exp Res 1990;14:853-5. 21. Ewing JE. Detecting alcoholism: The CAGE questionnaire. JAMA 1984;252:1905-7. 22. Temple M, Trends in collegiate drinking in California, 19791984. J Studies Alcohol. 1986;47:274-82. 23. Haworth-Hoeppner S, Globetti G, Stem J, et al. The quantity and frequency of drinkin;: among undergraduates at a southem university. Int J Addict 1989;24:829-57. 24. Petchers MK, Singer MI. perceived-Benefit-of-Drinking Scale: Approach to screening for adolescent alcohol abuse. J Pediatr 1987;110:977-81, 25. Petchers MK, Singer Ml, Angelotta JW, Chow J. Revalidation and expansion of an adolescent substance abuse screening measure. J Dev Behav Pediatr 1988,9:25-9.
JOURNAL OF ADOLESCENT HEALTH Vol. 13, No. 6
26. Marlatt GA, Rohsenow DR. Cognitive processes in alcohol use: Expectancy and the balanced p!acebo design. In: Meltow NK, ed. Advances in Substances Abuse. Greenwich, CT: JAI Fkss, 1980:155-99.
27. Sher KJ. Subjective effects of alcohol: The influence of setting and individual differences in alcohol expectancies. J Stud Alcohol 1985;46:137-46. 28. Brown SA, Christiansen BA, Goldman MS. The alcohol expectancy questionnaire: An instrument for the assessment of adolescent and adult alcohol expectancies. J Stud Alcohol 1987;483483-91. 29. Christiansen BA, Smith GT, RoehBng PV, et al. Using alcohol expectancies to predict adolescent drinking after one year. J Consult Clinic Psycho1 1989;57:93-9. 30. Tweed SH, Ryff CD. Adult children of alcoholics: Profiles of wellness amidst distress. J Stud Alcohol 1991;52:133-41. 31. Werner EE. Resilient offspring of alcoholics: A longitudinal study from birth to 18. J Stud Alcohol 1985;47:34-4O. 32. Scheller G, Gomberg ES, Clay M. College students and alcohol: An exploration of observations and opinions. J Ale Drug Educ 1979;24:30-41.
MARK J. WERNER,
Entering college freshmen 61 = 3013)completed a ques-
tionnaire which assessed drinking behaviors and identified students at risk for problem drinking as defined by the CAGE (focuses on Cutting down on drinki noyance by criticism by others about drinking feelings about drinking, and the use of an Eye opener) questionnaire and Perceived-Benefit-of-Drinking Scale (PBDS). Students were 50%male with a mean age of 17.9 years. In the past month, 17% had drunk on 10 or more occasions, and 18% had binged on 6 or more occasions. by 21% and CAGE scores of 2 or greater were obtai PBDS scores of 3 or greater by 29%, re ng high risk for problem drinking, High-risk CAGE and PBDS scores wexe associated with frequent drinking and binging. Student reports of parental problem drinking were not associated with high risk for problem drinking. Intent to join a fraternity or sorority (the Greek system) was associated with frequent drinking, binging, and high-risk CAGE and PBDS scores. Approaches to screening for problem drinking which emphasize attitudes and beliefs may be useful. The Greek system appears to be attractive to high-risk students and should be a focus of prevention programming. KEYWORDS: Alcohol Screening for problem drinking College students
among college students has drawn the attention of university administrators and
The use of alcohol
From the Division of Adolescent Medicine, Varlderbilt University, Nashville, Tennessee. Address reprint requests to: Mark 1. Werner, M.D., Division of Adolescent Medicine, 436 Medical Center South, Vanderbilt University, Nashville, TN 37232. Manscript accepted November 13, 1991.
faculty, government agencies, and a concerned public. Alcohol consumption has steadily risen during the last 20 years with annual use of alcohol by college students reaching 90-92%, and daily use near 7% (l-4). Approximately 25-50% of college students can be classified as heavy drinkers; such drinking is more common among males, whites, freshmen, and sophomores (3,5). Problems due to drinking that are frequently experienced by college students include personal injury, accidents, missing class, blackouts, legal difficulties, poorer academic performance, acquaintance rape, sexually transmitted disease, and unplanned pregnancy (6-12). Almost 30% of college students report loss of some hours of normal functioning while recovering from drinking during the preceding week (6). The influence of peers, the accessibility of alcohol, and pressure to be accepted by a group affect behavior and alcohol consumption by college students (13). Members of fraternal organizations are more likely to drink and to incur alcohol-related problems (4,5). Students who are frequent heavy drinkers appear disproportionately among those planning to join fraternities and sororities (5). The socialization process inherent in the (Greek) fraternity system places an unusual emphasis on drinking, with heavy alcohol consumption among fraternity pledges strongly correlated with a positive view of the socialization value of alcohol (14,15). The actual influence that the college environment has upon students’ drinking behaviors remains poorly de, particularly in comparison wit and drinking patterns established prior to entering college. Despite the perceived prevalence of problem drinking among college students, a reliable and efficient means of screening students has not been
0 Societyfor Adolescent Medicine, 1992 Published by Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, New York, NY 10010
467
1054139xJ9245.00
488
WERNER AND GREENE
developed. This is partly complicated by the difficulty that there is no standardized definition of problem drinking. The criteria usually applied to problem drinking focus on a pattern of excessive alcohol use and the development of physical, social, or psychologic problems related to drinking (16,17). The Michigan Alcoholism Screening Test (MAST) has been administered on several college campuses with &29% of students found to be possibly alcoholic or at risk for alcoholism (l&19). However, the MAST has been found to produce a significant number of false positive identifications in the college population (20). Four clinical interview questions, the CAGE questions, focus on Cutting down on drinking, Annoyance by criticism of others about drinking, Guilty feelings about drinking, and the use of an Eye opener. The CAGE questionnaire is a screening instrument commonly used in health care settings (21) that has been used with college students but may not be adequately sensitive (16,17). Both MAST and CAGE were developed for detection of alcoholism rather than problem drinking. Although approaches to screening for alcohol abuse have traditionally relied upon assessment of drinking frequency, quantity, and psychosocial consequences (22,23), an alternative approach utilizes the adolescent’s perceptions of why they drink. Petchers and Singer (24) and Petchers et al. (25) developed the Perceived-Benefit-of-Drinking Scale (PBDS) which probes, from the adolescent’s point of view, the specific reinforcements received from drinking. The more benefits an adolescent attributes to drinking, the more likely he or she is to drink. The PBDS scores are significantly higher for adolescents who report being drunk more often, who claim to have had a problem with drinking, and who have reported trouble due to drinking. The scale’s strong relationship with drinking behavior indicators makes it a potentially useful clinical screening device because it can serve as a proxy measure to assess drinking behaviors. The present study was undertaken to assess problem drinking among entering college freshmen, as identified by the PBDS and the CAGE questionnaire, and to assess the utility of these measures as screening tools that may be useful in a student health care setting. Students were studied at the start of their freshman year, prior to exposure to the college environment and the (Greek) fraternity/sorority SYSkm. Relationships between attitudes toward alcohol, problem drinking, and interest in the Greek system are evaluated. Implications for screening and intervention for problem drinking are discussed.
JOURNAL OF ADOLESCENT HEALTH Vol. 13, No. 6
Methods A sample of 308 freshmen students, repre,senting 20% of the 1500 students comprising the entering freshman class at a private southern university, were administered an anonymous questionnaire (during mandatory freshmen orientation programs conducted in each dormitory and directed at alcohol and drug issues. Students attended the orientation programs in groups of 30-50. Groups were randomly selected for participation such thit! there would be an approximately equal number of students from each coeducational dormitory. All students present at the selected orientation programs voluntarily participated. The programs occurred within the first 3 weeks of the fall semester. All freshmen students at the university are required to live on campus in either of two dormitories. Participants were 153 female and 155 male students from 16 to 19 years of age (mean, 17.9 years; SD = 0.5). Prior approval for the study was obtained through the Institutional Review Board. The questionnaire consisted of demographic information; self-report of frequency of alcohol, tobacco, and other drug use; self-report of intentions to join the Greek system; the PBDS; the CAGE questions; and questions about attitudes toward drinking. For the purposes of this study, frequent drinking was defined as drinking on 10 or more occasions in the past 30 days, and frequent hinging as drinking 5 or more drinks on 6 or more occasions in the past 30 days. A copy of the questionnaire is available upon request. The PBDS is a brief clinical screening instrument useful in identifying adolescents who may be alcohol abusers. It consists of endorsing as true or false each of the following five statements: drinking helps me forget, drinking helps me be friendly, drinking helps me feel good about myself, drinking helps me relax, and drinking helps me be friends with others who drink. Its reliability and validity are well established (24,25). This scale assesses the benefits an adolescent attributes to drinking and is not depend,ent upon direct information about drinking patterns or negative consequences resulting from drinking. A full evaluation of potential drinking problems is strongly recommended for those scoring 3 or greater (2425). The CAGE questions have proven use screening for alcoholism (21). The questions may be self-administered. Although a positive response to the CAGE questions is not diagnostic of alcoholism, a score of 2 or greater is highly suspicious and warrants further evaluation.
September 1992
PROBLEM DRINKING AMONG FRESHMEN
Students were grouped according to gender or their intention to join a fraternity or sorority. X2analyses were performed to test differences regarding frequent drinking, frequent binging, and the prevalence of high-risk CAGE and FBDS scores for these groups. Similarly, X2 analyses were used to test associations between high-risk CAGE or PBDS scores and frequent drinking, frequent binging, parental problem drinking, and drinking attitudes.
489
frequent drinking for female students only (p < 0.02). A history of frequent drinking or frequent binging may not necessarily indicate problem drinking as defined by CAGE or PBDS scores. High-risk CAGE scores were obtained by only 35% of students reporting frequent drinking and 42% of those reporting frequent binging. Similarly, high-risk PBDS scores were obtained by only 46% of students reporting frequent drinking and 50% of those reporting frequent bingeing. Of the students, 70% reported having a parent
Results
who drinks; 10% felt a parent had a drinking problem. Student report of parental problem drinking was not associated with frequent drinking, frequent Wnging, CAGE, or PBDS scores. Of the 308 students, 190 (62%) expressed an intent to join a fraternity or sorority, representing 50% of males and 73% of females. Those intending to join included 93% of those reporting frequent drinking, 89% of those reporting frequent binging, 75% of students with high-risk CAGE scores, and 76% of students with high-risk PBDS scores. Intent to join the Greek system was significantly associated with frequent binging (p < O.OOl), high-risk CAGE (p < 0.005), and high-&k PBDS scores (p < 0.02) for male students (Table 2). Analyses for female students were limited by the small number of female students not intending to join the Greek system with frequent drinking or high-risk scores. The same general trends, however, appear to hold for femalestudents (Table2). There was no associationbetweenparental problem drinking and intention to join the Greek system for male or female students. Students were also questioned about their attitudes toward drinking;17% felt drinkingmakes sexual experiences easier and more enjoyable, 55% feel alcoholic beverages make parties more fun, and 37%
Of the subjects 75% had drunk alcohol in the past 30 days, 51% had drunk on three or more occasions, and 17% on ten or more occasions. When asked about binge drinking, defined as five or more drinks on one occasion, 56% admitted to at least one binge episode in the last 30 days, 18% admitted to binging on 6 or more occasions in the last month. Drinking frequency did not differ between sexes; however, binge drinking was more frequent in males (p c 0.02). The distribution of PBDS scores and CAGE scores is shown in Table 1. PBDSscores of 3 or greater were obtained by 29% and CAGE scores of 2 or greater by 21%, reflecting high risk for problem drinking. Of the students, 122 (40%) had high-risk scores on either the PBDS or CAGE, and 33 students (11%) had high-risk scores on both the PBDS and CAGE. CAGE or PBDS scores were not significantly different between sexes. High-risk CAGE scores were associated with frequent drinking (p c 0.004) and frequent binging (p< 0.001) for male students only. On the other hand, high-risk PBDS scores were associated with frequent binging for both male and female students (p < 0.02) and p c 0.009, respectively), and with
Table 1. Perceived-Benefit-of-Drinking (PBDS)Scale and CAGE”Scores for College Freshmen CAGE ScoresC
PBDS Scoresh
(%)
Total
@)
40
(26)
85
(28)
86
(56)
28
(18)
55
(18)
32
(21)
39
(26)
78
(25)
27
(17)
(14)
27
(18)
48
(16)
8
15
(10)
10
(7)
25
(8)
2
8
(5)
8
(5)
16
(5)
Score
Male
0
4s
(29)
1
27
(17)
2
39
(25)
3
21
4 5
(%)
Female
Male
@)
(%)
Total
(%)
88
(58)
174
(56)
36
(24)
69
(22)
19
(12)
46
(15)
(5)
6
(4)
14
(5)
(1)
3
(2)
5
(2)
Female
“CAGE questionnaire focuses on Cutting down on drinking, Annoyance by criticism of others about drinking, Guilty feelings about drinking, and the use of an Eye opener. ‘Scores 2 3 indicate high risk, x2 male vs. female p < 0.8. ‘Scores 2 2 indicate high rusk, x2 male vs. female p < 0.7.
498
WERNERAND GREENE
JOURNAL OF ADOLESCENTHEALTH Vol. 13, No. 6
Table 2.
Problem Drinking and Intentions to Fraternity/Sorority
JOhI
a
Intention to join Yes n = 190
No II = 113
78 111
75 38
Frequent drinking” Male Female Total
31 19 50
2 2 4
Frequent hinging* Male Female Total
32 19 51
5 1 6
p < 0.001
High-risk CAGE’score (2 2) Male 26 Female 22 Total 48
10 5 15
p < 0.005 p < 0.02
High-risk PBDSdscore (2 3) Male 30 Female 38 Total 68
14 7 21
p c 0.02 p < 0.08 p < 0.002
Parent problem drinking Male 4 Female 14 Total 18
7 6 13
Gender Male Female
p < 0.001
“Frequent drinking, drinking on 10 or more occasions in the Past 30 days. bFrequentbinging, drinking 5 or more drinks on 6 or more occasions in the Past 30 days. ‘CAGE questionnaire focuses 07 Cutting down on drinking, Annoyance by criticism of others about drinking, Guilty feelings about drinking, and the use of an Eye opener. “PBDS, Perceived-Benefit-of-Drinking Scale.
agree that people feel powerful when they drink. In comparison, 91% of students reporting frequent drinking and 91% of those reporting frequent binging endorsed alcohol as making parties more fun (p < 0.001 and p < 0.001, respectively). Both highrisk CAGE and PBDS scores were associated with the same endorsement (p < 0.005 and p < 0.001, respectively).
Discussion Problem drinking on the college campus continues to be a significant public health concern. The fmdings of this study are generally in agreement with others conducted in a variety of university settings (3,5,22,23). Although drinking frequency did not differ between sexes, binge drinking was more com-
mon with male students. This study used two approaches to defining problem drinking which have not been used to our knowledge with college students; both are known to correlate highly with problem drinking or alcoholism. The PBDS defines problem drinking based on beliefs and attitudes, and is not dependent upon direct information about drinking patterns or negative consequences resulting from drinking. Beliefs about the effects that drinking will produce are known to influence both consumption and the subjective experience of intoxication (24,27). Previous studies have found that expectancies are associated with differing patterns of alcohol use and are capable of predicting transition to problem drinking in adolescents (28,29). Viewed in behavioral terms, the greater the benefit perceived from drinking, the more likely such behavior is to be reinforced, and thereby possibly increase (24,25). On this basis, 29% of the students studied received scores indicative of high risk for problem drinking. Because of ;is simple brief indirect approach, the PBDS may be a very useful clinical screening measure. The CAGE questionnaire, which relies on a more traditional approach to alcoholism screening, identified 21% of the students as high risk for problem drinking. While the sensitivity of the CAGE for problem drinking among college students remains unclear, its specificity is consistently above 85% (16,17). As such, these results may underestimate students at risk for problem drinking. Of the 122 students with high-risk scores on either the CAGE or PBDS, only 33 (27%) had high-risk scores on both measures. This suggests that the CAGE and PBDS may not be identifying the same at-risk population. The CAGE appears to be more applicable to identifying students with heavy drinking patterns and a high risk for alcoholism; particularly male students. In contrast, the PBDS identifies a larger number of male and female students who appear to have problem drinking and attitudes that reinforce their drinking patterns, but are not necessarily at risk for alcoholism. As such, the PBDS may be a more sensitive screening measure for students at risk for problem drinking and its associated health morbidities, that can be used in conjunction with the CAGE. Student report of parental problem drinking does not appear to help in identifying high-risk students; however, this study did not make in-depth efforts to assess parental drinking and possible alcoholism. Previous research indicates that most children and adolescents from families with alcoholism do function well (30,31).
September 1992
In the college setting, it can be very difficult to define what constitutes problem drinking. The interpretation of heavy drinking may depend on whether this form of drinking in college is a type of incipient problem drinking or simply a manifestation of usual adolescent development in an influential environment (32). Any evaluation of college students’ d&&ing patterns needs to consider the social environment and situational factors involved in college students’ decisions about drinking. Reliance upon self-report of drinking frequency and quantity may overestimate problem drinking in this population. Approaches that utilize an understanding of students’ beliefs, expectations, and motivations may be more useful. Similarly, given the influence of the college environment, there are likely to be significant numbers of students who suffer an adverse health or psychosocial outcome as a result of drinking but do not have alcoholism. Screening approaches should attempt to identify students who are at risk for alcohol-related problems as a result of their drinking behaviors and not rely solely on traditional alcoholism screening tools. This study is in agreement with others that college students intending to join the Greek system are more likely to be frequent, heavy, problem drinkers (5,23). Students appear to arrive at college with the expectation that the Greek system will support a heavy drinking pattern. Obviously, it is possible that a particular university or Greek system may be more attractive to students already involved in problem drinking. Many entering college students have prior knowledge of campus culture, traditions related to drinking, and Greek social systems. Nonetheless, attention to fraternities and sororities needs to be to be a significant part of any university alcohol prevention programming. Although the student population sampled in this study comes from a selective university, it is not different from students at a large number of other universities. It is important to note that this study was done at the start of a fall semester, a time when episodic heavy drinking may be more frequent. Longitudinal research focusing upon changes in drinking patterns throughout the college years and the association between drinking patterns and various psychosocial and medical problems is needed. Approaches to screening w drinking behaviors and associated problems are needed. Such studies are now underway. Problem drinking on college campuses is a complex issue requiring comprehensive campus-wide approaches to prevention intervention, and treat-
PROBLEM DRINKING A
491
ment. Screening for problem drinking should be a routine part of student health services. Many approaches are available; however, those which rely upon indirect measures of alcohol use through an understanding of students’ beliefs and attitudes may be most useful. Supported in part by a grant from the National Institute on Alcohol Abuse and Alcoholism and the National institute of Drug Abuse, RFA-AA-90-2.
References 1. Saltz R, Elandt D. College student drinking studies 19761985. Contemp Drug Probl1986;13:117-59. 2. Johnston LD, O’Malley PM, Bachman JG. Drug use among Americanhigh school students, college students and other young adults: National trends through 1985. Rockville, MD: National Intstute on Drug Abuse, 1986. 3. Pope HG, Ionescu-Pioggia M, Aizley HG, et al. Drug use and lie style among college undergraduates in 1989: A comparison with 1969 and 1978. Amer J Psychiatr 1990;147:998-1001, 4. Johnston LD, O’MaRey PM, Bachman JG. Illicit drug use, smoking, and drinking by American high school students, college students, and young adults, 1975-1987. Washington, DC: Government Printing Office, 1988; NIDA, ADAMHA, PHS, US Dept HHS. DHHS publication no. (ADM) 89-1602. 5. Canterbury RJ, Gressard CF, Vieweg WVR, et al. Psychosocial inventory among first-year college students by patterns of alcohol use. Adv Alcohol Subst Abuse 1990;9:1-11. 6. Meilman PW, Stone JE, Gaylor MS, et al. Alcohol consumption by college undergraduates: Current use and lo-year trends. J Stud Alcohol 1990;51:389-95. 7. Berkowitz AD, Perkins HW. Problem drinking among college students. J Amer Co11Health 1986;35:21-8. 8. Jessor R, Chase JA, Donovan JE. Psychosocial correlates of marijuana use and problem drinking in a national sample of adolescents. Amer J Public Health. 1980;70604-13. 9. Nicholi AM. The nontherapeutic use of psychoactive drugs: A modem epidemic. N Engl J Med 1983;308:925-33. 10. Nicholi AM. Characteristics of college students who use psychoactive drugs for nonmedical reasons. J Amer Co11Health 1985;33:189-92. 11. Gliksman L. Consequences of alcohol use: Behavior changes and problems during the first year of university. J Addict 1988;23:1281-95. 12. Engs RC, Hanson DJ. Gender differences in drinking patterns and problems among college students: A review of the literature. J Alcohol Drug Educ 1990;35:36-47. 13. Hanson DJ. Trends in drinking and behaviors among college students. J Alcohol Drug Educ 1974;19:6-14. 14. Leavy RL. An alternative source of information on a popular topic: A course in drinking. Teaching Psychology 1979;6:97100. 15. Faulkner RR, Alcorn JD, Garvin RB. Prediction of akohol consumption among fraternity pledges. J Alcohol Drug Educ 1989;34:12-20. 16. Smith DS, Collins M, Kreisberg JP, et al. Screening for problem drinking in college freshmen. J Amer Co8 Health 1987; 36:89-94.
492
WERNERAND GREENE
17. Heck EJ. Developing a screening questionnaire for problem drinking in college students. J Amer Coll Health 1991; 39227-31. 18. Favazza AR, Cannell B. Screening for alcoholism among college students. Amer J Psychiatr 1977;134:1414-16. Silber TJ, Capon M, Kuperschmit I. Administration of the 19. Michigan Alcoholism Screening Test (MASTJ at a student health service. J Amer Coll Health 1985;33229-33. 20. Martin CS, Liepman MR, Young CM. The Michigan Alcoholism §creening Test: False positives in a college sample. Alcohol Clin Exp Res 1990;14:853-5. 21. Ewing JE. Detecting alcoholism: The CAGE questionnaire. JAMA 1984;252:1905-7. 22. Temple M, Trends in collegiate drinking in California, 19791984. J Studies Alcohol. 1986;47:274-82. 23. Haworth-Hoeppner S, Globetti G, Stem J, et al. The quantity and frequency of drinkin;: among undergraduates at a southem university. Int J Addict 1989;24:829-57. 24. Petchers MK, Singer MI. perceived-Benefit-of-Drinking Scale: Approach to screening for adolescent alcohol abuse. J Pediatr 1987;110:977-81, 25. Petchers MK, Singer Ml, Angelotta JW, Chow J. Revalidation and expansion of an adolescent substance abuse screening measure. J Dev Behav Pediatr 1988,9:25-9.
JOURNAL OF ADOLESCENT HEALTH Vol. 13, No. 6
26. Marlatt GA, Rohsenow DR. Cognitive processes in alcohol use: Expectancy and the balanced p!acebo design. In: Meltow NK, ed. Advances in Substances Abuse. Greenwich, CT: JAI Fkss, 1980:155-99.
27. Sher KJ. Subjective effects of alcohol: The influence of setting and individual differences in alcohol expectancies. J Stud Alcohol 1985;46:137-46. 28. Brown SA, Christiansen BA, Goldman MS. The alcohol expectancy questionnaire: An instrument for the assessment of adolescent and adult alcohol expectancies. J Stud Alcohol 1987;483483-91. 29. Christiansen BA, Smith GT, RoehBng PV, et al. Using alcohol expectancies to predict adolescent drinking after one year. J Consult Clinic Psycho1 1989;57:93-9. 30. Tweed SH, Ryff CD. Adult children of alcoholics: Profiles of wellness amidst distress. J Stud Alcohol 1991;52:133-41. 31. Werner EE. Resilient offspring of alcoholics: A longitudinal study from birth to 18. J Stud Alcohol 1985;47:34-4O. 32. Scheller G, Gomberg ES, Clay M. College students and alcohol: An exploration of observations and opinions. J Ale Drug Educ 1979;24:30-41.
