Alcohol and tobacco use is a growing concern all throughout the world. Twenty-two percent of Canadians and 14%–27% of the British population exceed the recommended guidelines (1, 2). Current statistics reveal that 21% of Canadians, 20% of Americans, and 26% of the British population ages 16 and over currently smoke (2–4).

The literature is clear regarding the harmful effects of smoking and excessive alcohol consumption. Excessive alcohol consumption has many physiological, social, and mental consequences, including impaired vision and motor coordination, elevated blood pressure and heart rate, risk of stroke and heart failure, introversion, and antisocial behavior (5). The current Canadian guidelines for “low-risk drinking” suggest that women and men drink no more than 9 to 14 standard drinks per week, respectively, and no more than two standard drinks per day (1). Smoking also has many physiological consequences, including cancer, emphysema, organ damage, and heart disease (6).

Despite the recent publicity regarding the harmful effects of alcohol and smoking, the largest proportion of smokers and drinkers is represented by adults between 20 and 24 years old (3, 4). The primary reasons cited for drinking in this age group are pleasure and social pressure. Medical students have been reported to consume alcohol and tobacco at levels exceeding the recommended guidelines (7–12). These results are surprising because medical students are supposedly more educated on the harmful effects of smoking and excessive alcohol consumption.

The purpose of this study was to quantify the extent of alcohol consumption and cigarette smoking in Canadian medical students and to assess their knowledge and attitudes toward these behaviors. Specifically, this questionnaire of first-, second-, and third-year medical students at the University of Calgary explored the prevalence of problem drinking (according to the CAGE questionnaire), the prevalence of tobacco consumption, and the relationship between perceptions of alcohol and tobacco use and the risk of problem drinking.

A questionnaire was handed out at lectures to both the first- and second-year medical students. Because the third-year students were in their clinical training year, questionnaires were placed in their mailboxes. An e-mail was also sent to all medical students informing them about the study and where they could obtain copies of the questionnaire in case they had not attended the lecture or to inform them that the questionnaires had been placed in their mailboxes. The questionnaire was anonymous, confidential, and approved by the University of Calgary Conjoint Health Research Ethics Board.

The questionnaire used in this study was similar to that used in previous studies of medical students in Britain, Turkey, and the United States (9–11).

With regards to alcohol consumption, participants were asked about the prevalence of drinking (yes or no), the age at which they started drinking, the frequency with which they consumed any amount of alcohol, the frequency with which they consumed more than five drinks in a single sitting, the amount of alcohol consumed over 1 week, the average number of drinks consumed in a single sitting, their reasons for drinking, and whether they experienced any negative consequences of drinking. Students who were at an increased risk of problem drinking were identified by a score of 2 or greater on the CAGE questionnaire (13).

With regards to cigarette smoking, participants were asked about their current and lifetime use of cigarettes (yes or no), the number of cigarettes smoked per day, the number of years they have smoked, and their reasons for smoking.

In addition to quantifying the extent of alcohol use and cigarette consumption, the students’ perceptions of these behaviors were assessed using the agreement or disagreement with the following statements on a 5-point Likert scale (1=strongly agree, 3=neutral, 5=strongly disagree):

Demographic characteristics were compared between smokers and nonsmokers, and between those students who were and were not at risk for problem drinking (score ≥2 on the CAGE questionnaire) using chi-square tests or independent samples t tests, as appropriate. Similarly, smoking status was also compared between the two alcohol consumption groups using chi-square tests.

Given the significant negative impact of problem drinking both within the general community and among college students, the predictors of problem drinking were explored. Independent samples t tests were first used to compare mean scores on the perceptions questionnaire items among those students who were and were not at risk for problem drinking according to the CAGE questionnaire. Gender, smoking status, and those perceptions significantly related to the risk for problem drinking were then entered into a backward logistic regression model in order to determine the independent correlates of an increased risk for problem drinking. All variables were entered in a single step, and the variable that contributed least significantly to the model was removed until only significant variables remained.

Recruitment

Questionnaires were sent to a total of 327 students. Responses were received from 72% of first-year students, 66% of second-year students, and 22% of third-year students. The lower response rate observed among third-year students was likely because the University of Calgary is a 3-year medical school and third-year students were on their clerkship training and therefore had less time to participate in this study.

Tobacco

Six percent of the medical students currently smoked, while 24% of medical students reported smoking at some point in their lives. Of the 11 students who smoked, only two smoked more than five cigarettes per day. Current smoking status was not significantly related to age, class, or marital status (p>0.05 in all cases, Table 1). While a significantly greater proportion of males were current smokers (χ²=7.52, p=0.006), gender did not have a significant impact on a history of cigarette use (χ²=1.84, p=0.18). There was a trend toward an increased proportion of students at risk for problem drinking among the group who currently smoked (χ²=2.81, p=0.09). The two most common reasons given for smoking were “social” and “stress relief.”

Alcohol

Eighty-six percent of students currently drank, with a majority (81%) drinking fewer than 11 drinks per week. As expected, a significantly smaller proportion of students were at risk for problem drinking, with only 15% scoring 2 or greater on the CAGE questionnaire (χ²=743.45, p<0.0005). The two most common reasons given for drinking were “social” and “taste” while the top three consequences of drinking were “engaging in risky behavior, “missing class,” and “memory loss.”

Age, class, and marital status were not significantly related to an increased risk for problem drinking (Table 2). While gender was also not significantly related to an increased risk for problem drinking, males did consume a greater amount of alcohol on a weekly basis, with 15% consuming 11 or more drinks per week, compared to only 1% of females consuming the same amount (χ²=10.60, p=0.001). As shown in Table 2, current and past smoking status also differed among students who were and were not at risk for problem drinking (current smokers: χ²=3.94, p=0.047; past smokers: χ²=4.51, p=0.031). Lastly, as expected, students at risk for problem drinking experienced a greater number of negative consequences associated with alcohol use (4.4 versus 2.3 for nonproblem drinkers; t₈₆=4.2, p<0.0005).

Perceptions

Surprisingly, while 50% of current smokers reported feeling in control or strongly in control of their cigarette consumption, 96% of students who were at risk for problem drinking according to the CAGE questionnaire reported feeling the same way regarding their alcohol consumption (χ²=21.16, p<0.0005).

Students scoring 2 or greater on the CAGE questionnaire believed more strongly that getting drunk is acceptable on occasion (t₁₇₃=3.13, p=0.003) and that as physicians they would foresee themselves advocating against excessive alcohol and cigarette consumption despite engaging in these behaviors themselves (t₁₇₂=2.20, p=0.03). They also believed less strongly that increased alcohol consumption has many negative health consequences (t₁₇₃=3.12, p=0.002), that they felt in control of their alcohol consumption (t₁₇₃=3.79, p<0.0005), and that they would look negatively upon their peers who they believed consumed unreasonable amounts of alcohol (t₁₇₂=4.37, p<0.0005). Trends were also observed for students who were at risk for problem drinking to feel less in control of their cigarette consumption (t₁₅₆=1.70, p=0.09) and feel less strongly that they would look negatively upon their peers who smoked cigarettes (t₁₇₃=1.92, p=0.06).

Backwards logistic regression analysis showed that a lack of control over alcohol consumption (p=0.02), a belief that getting drunk is acceptable on occasion (p=0.02), and perceptions regarding the negative consequences of alcohol consumption (p=0.05) were significant independent predictors of an increased risk for problem drinking. The remaining perceptions (that differed among those who were and were not at risk for problem drinking in previous t-tests), gender, and smoking status were removed from the final regression model (χ²=24.95, p<0.0005, Nagelkerke R²=0.23).

Eighty-six percent of students reported that they consume alcohol. This is comparable to the prevalence rates in medical schools in other countries (83%–97%) and that of the Canadian population ages 20–34 (85%–90%) (7, 8, 10, 11, 14–16). Similarly, the 15% of students at an increased risk for problem drinking is also consistent with international studies of problem drinking among medical students, which report prevalences of 4%–27% using various screening instruments (10, 15, 17). We elected to use the CAGE questionnaire as an indicator of problem drinking in our study sample, which has a reported sensitivity of 53%–69% and specificity of 81%–95% when used as a screening test for alcohol abuse (18). This suggests that the prevalence of problem drinking among Canadian medical students may be even higher than that reported in this survey. Nonetheless, our findings raise the concern that a substantial proportion of medical students are at risk for problem drinking.

Although we observed a greater proportion of current and past smokers among the students who screened positively on the CAGE questionnaire and found that men were more likely than women to consume large quantities of alcohol (≥11 drinks/week), gender and smoking status did not emerge as independent predictors of problem drinking in this sample of medical students. However, perceptions regarding the negative health consequences of alcohol, the acceptability of getting drunk, and control over alcohol consumption significantly predicted an increased risk for problem drinking in our logistic regression model. Previous investigations into the attitudinal correlates of problem drinking among medical students have suggested that a lack of control over temptation (19, 20), use of alcohol as a coping strategy (20), psychotic personality traits (21), and stronger beliefs regarding the benefits of drinking (19) are also related to excess alcohol consumption or problem drinking. Further assessment of such personal and attitudinal factors related to problem drinking may help in the identification of suitable targets for its treatment.

The high percentage of students in this study who believe that excessive alcohol consumption has negative health consequences (92%) might suggest a high awareness of the hazards of excessive alcohol consumption. “Social” and “taste” were the most common reasons given for drinking, and feeling less strongly regarding the negative consequences of increased alcohol consumption was significantly related to the presence of problem drinking. This suggests that although health concerns may not play a large role in a student’s decision to drink, a lack of concern regarding the negative consequences of alcohol consumption may be related to an increased risk of problem drinking among students who do drink.

Potential limitations of this study include the lack of a validated measure of nicotine dependence, such as the Fagerstrom Scale (22), and the possibilities of selection bias and reduced generalizability. Selection bias is one of the inherent limitations of using a questionnaire. We attempted to minimize this bias by e-mailing copies of the questionnaire to all students and providing hard copies to all students who attended classes. However, it is conceivable that the students who were less likely to respond to the survey consumed a greater amount of alcohol or displayed more problem drinking patterns. Administration of this questionnaire to students at medical colleges across Canada may help to minimize this bias by reaching a greater number of students at an increased risk for problem drinking. Cross-national administration of this questionnaire would also help to increase the generalizability of our findings and allow us to assess regional variations in the patterns and perceptions of alcohol and tobacco use.

Medical students at the University of Calgary consume comparable amounts of alcohol but fewer cigarettes than similar groups of medical students described previously in the literature. We observed a relatively high proportion of students who were at risk for alcohol abuse according to the CAGE questionnaire, which was also similar to previous studies of substance abuse among medical students. Attitudes related to an increased risk for problem drinking revolved around the negative health consequences of alcohol, the acceptability of getting drunk, and feelings of control over alcohol consumption. Assessment of the use and perceptions of alcohol and tobacco in medical students across Canada may help to improve the generalizability of our findings and select appropriate targets for intervention.

At the time of submission, the authors disclosed no competing interests.