The relationship between substance use and scores on the Mini Markers Five Factors Personality scale in college and high school students.

Dear Editor:

Eysenck (1967) and more recently Cloninger and Svrakic (1997) have proposed models of personality that purportedly predict reactions to various psychoactive substances. The majority of empirical studies in this area have examined characterological correlates of alcohol use. Baer (2002) reviewed college-drinking studies conducted during the last thirty years within United States colleges and universities, to assess the influence of various personality factors, including impulsivity, extraversion, and neuroticism, and reported that various studies have found a link between impulsivity/sensation seeking and alcohol consumption.

A small number of studies have investigated the personality correlates of tobacco use, and still fewer have assessed whether personality characteristics are differentially associated with marijuana use. Kopstein, Crum, Celentano, & Martin (2001) investigated the association between sensation seeking and disinhibition and adolescent cigarette and marijuana use. High disinhibition 8th graders were found to be significantly more likely to smoke tobacco and marijuana. Thrill and adventure seeking did not predict cigarette or marijuana use; peer use was the most important explanatory variable for both current marijuana and cigarette smoking.

Other researchers have simultaneously explored the relationship between personality and multiple forms of alcohol and drug use. Both Zuckerman (1994) and Webb, Baer, Francis, and Card (1993) offer findings consistent with the notion that sensation seeking has both direct and indirect effects on adolescent alcohol and drug use. Some individuals may be more sensitive to direct appetitive stimulation from drug use due to their high level of sensation seeking. As a result, they may have higher motivation to use substances. Other high sensation seekers users may be motivated more by the illegality of the drug and the opportunity to reject authority.

Comeau, Stewart, & Loba (2001) examined the relationships among trait anxiety, anxiety sensitivity, and sensation seeking and adolescents' motivation for substance use (alcohol, cigarettes, and marijuana). They found that all of the personality factors they examined were associated with substance use, especially alcohol use.

Some maintain that common factors underlie problems with substance use, and that treatment need not be structured on the basis of the particular substance a client abuses (Carroll & Chambliss, 1990). Empirical findings among both males and females consistently demonstrate greater similarities than differences between alcohol-dependent and drug-dependent patients, when adequate statistical controls are exercised for critical demographics variables such as age, race, and sex.

Alternatively, others argue that discrete etiologic and personality factors determine choice of abused substance, and that different treatments are needed to address these varying causes (Cloninger & Svrakic, 1997). Empirical findings have been mixed, but many studies have found substantial rates of multiple substance use, which challenge the assumption that personality dictates specific selection of substance. These results suggest that once an inhibitory threshold is crossed, and an individual begins to use any psychoactive substance, the probability of their experimenting with other substances in the future increases. Some direct brain effects of substances are specific to particular psychoactive agents, while others are nonspecific effects associated with any type of altered state of consciousness. If individuals are simply seeking any deviation from their ordinary subjective experience, users might be quite undiscerning in their choice of substance.

We conducted a study examining this issue in a nonclinical population, in order to assess rates of multiple substance use among adolescents as a function of personality attributes. If self medication motivation governs adolescent substance use, personality differences should predict type of substance used, because personality variability is associated with different symptom tendencies (e.g., anxiety versus boredom). Alternatively, if sensation seeking underlies most adolescent substance use, experimentation with multiple substances with contradictory psychoactive effects might be most prevalent. Personality variables other than those related to risk taking and rebellion should have little relationship to substance use.

Our study attempted to evaluate the apparent role of self-medication factors versus situational factors in determining adolescents' selection of substances, by assessing the relationship between personality and substance use at different developmental stages. We investigated the relationship between the five basic personality dimensions and three types of substance use by surveying students. Both high school and college undergraduate students were assessed in order to permit the evaluation of developmental period differences in the relationship between personality and substance use. Information gleaned from this investigation may assist those developing interventions to help reduce adolescents' experimentation with various substances, by allowing the specification of especially high-risk groups of students.

In our study, participants were 115 college students (30 males, 85 females) from a small liberal arts college from a suburban area in the Northeast United States and 115 high school students (46 males, 64 females) attending a public school in the same area. One hundred forty-nine female students and 80 male students, with a combined mean age of 17.78 years, completed the survey. The survey was administered to college students enrolled in an introductory psychology course, and high school students enrolled in health education classes.

The survey consisted of the Mini Markers Five Factors Personality Scale (MMFFP, Saucier, 1992) and author devised items assessing the subject's reported substance abuse behavior and demographic variables. The MMFFP consists of 40 alphabetized self-descriptive personality characteristics that respondents endorse using a 10-point Likert scale; it yields summary scores on five basic personality traits (openness, conscientiousness, extreme agreeableness, and neuroticism). The frequency of the participant's cigarette, alcohol, and marijuana use were also assessed using self-report items (Wechsler, Rigotii, Gledhill-Hoyt & Lee, 1998).

It was found that reported use of alcohol, cigarettes, and marijuana were all significantly intercorrelated (r range: .36 to .59; p < .001). One quarter of the students reported smoking cigarettes within the past month; one half reported never using tobacco. Roughly one half (53%) of the participants reported using alcohol within the past month; one fifth reported never using alcohol. One fifth of the students reported using marijuana within the past month; roughly one half (57%) reported never using marijuana.

Directionally adjusted items were totaled to create summary scores for extraversion, neuroticism, agreeableness, conscientiousness, and openness. A median split was performed on the summary scores for each of the five personality factors to create high and low level groups for each of the five traits.

Multivariate analyses of variance were used to evaluate sex, developmental period, and personality variable effects for each of the five factors assessed. A 2 x 2 x 2 MANOVA (sex: male and female; developmental period: high school and college; extraversion: low and high) on each of the three types of self-reported substance use (cigarette, alcohol, and marijuana) revealed a significant sex main effect for cigarette use (males: x = 2.54, SD = 1.63, n = 70 versus females: x = 2.11, SD = 1.43, n = 149; F = 6.60, df 1/211, p < .01). Significant developmental period main effects were also obtained for cigarette (F = 10.87, df = 1/211, p < .01) and alcohol (F = 27.33, df = 1/211, p < .001) use, and a trend was noted for marijuana use (p = .06) (see Table 1). This analysis revealed no significant main effect for extraversion or significant interaction effects.

A 2 x 2 x 2 MANOVA (sex: male and female; developmental period: high school and college; neuroticism: low and high) on each of the three types of self-reported substance use (cigarette, alcohol, and marijuana) was performed in order to evaluate main and interaction effects for neuroticism. There was no significant main effect for neuroticism. A significant neuroticism by sex interaction effect was found for alcohol use (F = 4.35, df = 1/211, p < .04) (see Table 2).

A 2 x 2 x 2 MANOVA (sex: male and female; developmental period: high school and college; agreeableness: low and high) on each of the three types of self-reported substance use (cigarette, alcohol, and marijuana) was performed in order to evaluate main and interaction effects for agreeableness. There were no significant main effects for agreeableness and there were also no significant interaction effects among agreeableness, sex, and developmental period.

A 2 x 2 x 2 MANOVA (sex: male and female; developmental period: high school and college; conscientiousness: low and high) on each of the three types of self-reported substance use (cigarette, alcohol, and marijuana) was performed in order to evaluate main and interaction effects for conscientiousness. There was no significant main effect for conscientiousness; however, a significant 3-way interaction effect was found among sex, developmental period, and conscientiousness for marijuana use (F = 4.03, df = 1/211, p < .05) (see Table 3).

A 2 x 2 x 2 MANOVA (sex: male and female; developmental period: high school and college; openness: low and high) on each of the three types of self-reported substance use (cigarette, alcohol, and marijuana) was performed in order to evaluate main and interaction effects for openness. There were no significant main effects for openness. However, a significant openness by developmental period interaction was found for both alcohol and marijuana use (alcohol: F = 10.31, df= 1/211, p < .01; marijuana: F = 7.77, df = 1/211, p < .01) (See Table 4 and 5).

The finding that all forms of substance use were associated supports the notion that a common factor such as thrill seeking, rebellion, or peer support may underlie the decision to experiment in this way during adolescence. However, the magnitude of these associations was modest, emphasizing the importance of other, possibly more substance-specific factors as well.

The results indicated that males were more likely to engage in cigarette smoking behavior than females. Because the traditional male stereotype emphasizes disregard for authority, males may be more likely to smoke cigarettes as an expression of their autonomy, given that adult norms increasingly stigmatize this behavior. Developmental period predicted all three forms of substance use. College students may be more likely to use alcohol, cigarettes, and marijuana because of their increased independence and access to these substances.

The failure to find any significant personality main effects suggests the need to consider these factors in combination with other determinants of substance use. Contrary to some previous research, this study failed to find a link between extraversion and reported use of any of the three substances evaluated. This was surprising, given the extraverted subjects' typical preference for social situations, which are standard settings for alcohol use during the high school and college years. However, when personality factors were evaluated in conjunction with gender and developmental period, some interesting links to substance use emerged.

Neuroticism was differentially related to alcohol use for males and females. Among males, neuroticism seemed to inhibit drinking, while highly neurotic females drank more than their less neurotic counterparts. This may suggest that the motives underlying alcohol use during the adolescent period vary across the sexes. Perhaps females are more likely to use alcohol to attenuate anxiety, while males use alcohol as a form of thrill seeking. This would account for why high neurotic females, who presumably experience greater anxiety, make more frequent use of alcohol than their calmer female peers. It would also explain why high neurotic anxiety males would use alcohol less, since presumably they prefer less thrill seeking than their less anxious counterparts. Another possibility is that neuroticism differentially affects popularity in males and females. Sex role stereotypes may make anxiety more acceptable in females than in males. As a result, high neurotic females may be more likely than high neurotic males to participate in social contexts that facilitate and support alcohol use.

When developmental period, sex, and personality were considered in combination, an interesting interaction emerged. College males with low levels of conscientiousness were the most likely subgroup to smoke marijuana. In contrast, high school females reporting high levels of conscientiousness were the least likely to engage in marijuana use. Differential gender socialization may promote obedience and conformity to parental expectations for females that is especially pronounced among highly conscientious young women while they reside at home. This may strikingly inhibit marijuana use among more conscientious high school females. The impact of living away from parents and having greater access to illegal substances may be particularly disinhibiting for less conscientious young men.

The way in which openness interacted with developmental period in predicting substance use was different for alcohol and marijuana use. Among college students, low openness was especially associated with greater alcohol use. Students low in openness to intellectual experience may find the college academic environment, which places an obvious premium on intellectual curiosity and exploration, alienating. This mismatch between personality style and context may set the stage for compensatory or even escapist alcohol consumption.

When marijuana use was evaluated, while low openness again functioned as a risk factor among high school students, it actually seemed protective for college students. The highest reported use of marijuana was found among high openness college students. On college campuses, experimentation with marijuana may be seen as a means of exploring altered states of mind, which might be especially appealing to students reporting a high preference for novel mental experiences. This is consistent with earlier studies of college students indicating that marijuana use is associated with high levels of openness, due to its hallucinogenic qualities.

In contrast, marijuana use among high school students may represent an extension of illegal, thrill seeking behavior that typically begins with alcohol use. Since low openness high school students use alcohol more, they might be more likely to cross into experimentation with marijuana. Low openness high school students are less likely to be engaged and successful in school, which increases their likelihood of being grouped academically with peers who reinforce antiestablishment norms and endorse substance abuse.

One limitation of the correlational design used in this study involves the problem of possible circularity. Although the MMFFP was designed to tap relatively stable features of personality, it is possible that substance use colored participants' endorsement of the personality items, thereby accounting for the observed relationships. Future studies employing a cross-lagged longitudinal design might clarify the actual causal direction of the results obtained in this study.

Table 1
Self-reported substance use for high school
versus college respondents

                   High School            College
                     n = 107              n = 112

Cigarette Use   x = 1.93 SD = 1.39   x = 2.56 SD = 1.55
Alcohol Use     x = 2.56 SD = 1.24   x = 3.44 SD = 1.12
Marijuana Use   x = 1.87 SD = 1.29   x = 2.16 SD = 1.33

Table 2
Self-reported alcohol use for high neurotic and
low neurotic male and female respondents

                     Males      Females

Low Neuroticism     x = 3.18    x = 2.89
                   SD = 1.27   SD = 1.31
                    n = 34      n = 65

High Neuroticism    x = 2.86    x = 3.10
                   SD = 1.25   SD = 1.21
                    n = 36      n = 84

Table 3
Self-reported marijuana use for high conscientious
and low conscientious male and female
high school and college respondents

                                    Males      Females

       Low                         x = 1.94    x = 1.97
Conscientiousness   High School   SD = 1.29   SD = 1.40
                                   n = 32      n = 30

                                   x = 2.86    x = 2.03
                      College     SD = 1.51   SD = 1.30
                                   n = 14      n = 30

      High                         x = 2.27    x = 1.59
Conscientiousness   High School   SD = 1.49   SD = 1.10
                                   n = 11      n = 34

                                   x = 2.00    x = 2.10
                      College     SD = 1.35   SD = 1.28
                                   n = 13      n = 55

Table 4
Self-reported alcohol use for high openness and
low openness for high school and college respondents

                High School    College

High Openness     x = 2.29     x = 2.72
                 SD = 1.18    SD = 1.28
                  n = 62       n = 72

Low Openness      x = 2.93     x = 3.27
                 SD = 1.23    SD = 1.18
                  n = 45       n = 77

Table 5
Self-reported marijuana use for high openness and
low openness for high school and college respondents

                High School    College

High Openness     x = 1.68     x = 2.40
                 SD = 1.14    SD = 1.47
                  n = 62       n = 52

Low Openness      x = 2.13     x = 1.95
                 SD = 1.44    SD = 1.17
                  n = 45       n = 60

References

Baer, J. S. (2002). Student Factors: Understanding individual variation in college drinking. Special Issue: College drinking, what it is, and what to do about it: Review of the state of the science. Journal of Studies on Alcohol, 14, 40-53.

Carroll, J. F. & Chambliss, C. A. (1990). A comparison of the self-reported problem of alcoholics, drug dependent persons, and multiple drug dependent persons at admission. Alcoholism Treatment Quarterly, 7(4) 3-39.

Cloninger, C. R. & Svrakic, D. M. (1997). Integrative psychobiological approach to psychiatric assessment and treatment. Psychiatry: Interpersonal and Biological Processes, 60, 120-141.

Comeau, N., Stewart, S. H., & Loba, P. (2001). The relations of trait anxiety, anxiety sensitivity, and sensation seeking to adolescents' motivations for alcohol, cigarette, and marijuana use. Addictive Behaviors, 26, 803-825.

Eysenck, H. J. (1967). The Biological Basis of Personality. Springfield, IL: Charles C. Thomas.

Kopstein, A. N., Crum, R. M., Celentano, D. D., & Martin, S. S. (2001). Sensation seeking needs among 8th and 11th graders: characteristics associated with cigarette and marijuana use. Drug and Alcohol Dependence, 62, 195-203.

Saucier, G. (1992). Mini-Markers: A brief version of Goldberg's unipolar Big-Five markers. Journal of Personality, 63, 506-512.

Wechsler, H., Rigotii, N. A., Gledhill-Hoyt, J., & Lee, H. (1998). Increased levels of cigarette use among college students: A cause for national concern. Journal of the American Medical Association, 280 (19), 1673-1678.

Webb, J. A., Baer, P. E., Francis, D. J., & Card, C. D. (1993). Relationship among social and intrapersonal risk, alcohol expectancies, and alcohol usage among early adolescents. Addictive Behaviors, 18, 127-134.

Zuckerman, M. (1994). Behavioral Expressions and Biosocial Bases of Sensation Seeking. New York: Cambridge University Press.

Catherine Chambliss, Megan Austin, Joanne Brosh, Gina Iannella, Rebecca Outten, Margaret Rowles Ursinus College [e-mail]