Alcohol abuse and dependence in college and noncollege samples: a ten-year prospective follow-up in a national survey *.

CONSPICUOUS ATTENTION HAS BEEN GIVEN recently to the consequences of excessive drinking by college students (College Task Force, 2002; Wechsler et al., 2002). That attention is certainly warranted. In 2001, 600,000 college students were unintentionally injured from alcohol-related causes, 1,700 college students died from alcohol-related injuries, and 696,000 were assaulted or hit by another drinking college student (Hingson et al., 2005). Furthermore, 31% of undergraduate college students meet criteria for alcohol abuse according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV; American Psychiatric Association, 1994), and another 6% can be classified as DSM-IV alcohol-dependent (Knight et al., 2002). Recent national estimates for DSM-IV alcohol-use disorders (AUDs) among college students 18-29 years of age state that 8% are alcohol abusers and 10.9% are alcohol dependent (Dawson et al., 2004). These statistics lead to the view that the college environment is particularly problematic, acting like a facilitator or incubator of excessive drinking, alcohol abuse, and alcohol dependence.

However, college nonattenders are also subject to some of the influences that contribute to problematic drinking. All young people explore and develop their individual drinking repertoires within a general environment of weakening parental controls, the approaching legal drinking age, relative freedom from many adult roles, and engagement in the social rituals surrounding sexual experimentation and mate selection. Thus, it is reasonable to ask whether the college environment does more to facilitate heavy drinking, alcohol abuse, and alcohol dependence than environments experienced by young adults who do not go to college (e.g., workplace and military environments). We can address this question by comparing youth who have attended college with those who finish only high school and those who drop out of high school.

Bingham et al. (2005) report increases in at-risk alcohol use between 12th grade and young adulthood (age 24) with higher increases among college-educated men and women than those who did not attend college. Despite the higher increase in risk in the college samples, college men and women had the lowest negative drinking consequences when compared with men and women who did not attend college. Casswell et al. (2003) found increases in the frequency of drinking between ages 18 and 26 and that those who were less educated drank more at all ages. Other studies have found only slight increases in heavy drinking among college attenders compared with nonattenders (Dawson et al., 2004; O'Malley and Johnston, 2002; Paschall, 2003; Slutske et al., 2004).

A few studies have examined the longer-term consequences of alcohol use in college and noncollege samples. A 27-year longitudinal study of college students found that a large majority of college students mature out of their drinking problems (Fillmore et al., 1979). Sher et al. (2001) have shown that heavy drinking associated with membership in fraternities and sororities does not typically persist in the years after college. In a 25-year prospective study of inner-city black children, Crum et al. (1998) reported that dropping out of high school was associated with a twofold increase in lifetime alcohol abuse and dependence, according to the criteria of DSM, Third Edition, Revised (American Psychiatric Association, 1987). Other prospective studies have shown similar associations with educational status, but the findings in these studies were limited to a short observation period of 1 year (Crum et al., 1992, 1993).

Muthen and Muthen (2000), drawing on the National Longitudinal Survey of Labor Market Experience in Youth (NLSY), examined the severity of alcohol dependence (sum of 22 symptom items) from ages 25 to 37, adjusting for early onset of alcohol use, family history of alcoholism, gender, and race/ethnicity. Their results show that at age 25 there was little difference in the severity of alcohol dependence by educational status. At age 37, however, the severity of alcohol dependence was significantly higher among respondents who dropped out of high school than that of high school graduates. In contrast, going on to college was associated with significantly lower levels of alcohol dependence with increasing age. The authors conclude that a major finding in their study highlights the detrimental effect that dropping out of high school has on alcohol-related problems.

To our knowledge, no studies have compared the long-term risk, using DSM-IV AUD diagnoses, among college students and their noncollege peers. In view of the slight variations in alcohol-related risks found among young adults with different educational status, knowledge of their long-term risk has implications for intervention in the young-adult years. As noted by Crum et al. (1992, 1998), the presence of early problem behaviors influences both educational achievement and alcohol abuse and dependence. Prospective studies using the NLSY have reported associations between alcohol abuse and dependence with age at onset of alcohol use (Grant et al., 2001a; Muthen and Muthen, 2000), antisocial behavior (Harford and Muthen, 2000), and family history of alcoholism (Muthen and Muthen, 2000). The major objectives in the current study were to estimate the effect of educational status on the subsequent risk for DSM-IV AUDs 10 years later and determine whether other risk factors (e.g., early problem behaviors, family history of alcoholism, poverty status, early onset of drinking, heavy drinking, alcohol dependence symptoms) account for the association between educational status and subsequent risk for DSM-IV AUDs. This study conducts an analysis of data from the NLSY comparing baseline years of education when respondents were 19-25 years of age and most likely to have had exposure to college, with assessments of DSM-IV alcohol abuse and dependence in 1994. It extends the analysis of Muthen and Muthen (2000) by focusing on the educational status of young adults, the number of years of education beyond high school, and inclusion of additional risk factors (i.e., early illicit ...