Abstract
A number of studies have documented the relationship between smoking and depression in adolescent and adult populations. The purpose of this study was to examine the level of depressive symptoms among college-age smokers and to determine whether or not increases in cigarette use were associated with increases in the number of depressive symptoms. A nonexperimental cross-sectional design was used. Of the random sample (N = 895), 28% were current smokers, and more than 26% of the total sample reported high levels of depressive symptoms. Current smokers were more likely to report an increased level of depressive symptoms than non-smokers were. Correlates of depressive symptoms included grade point average, marijuana use, and increased work hours. When controlling for these variables, the number of days smoked was predictive of depressive symptoms. School nurses have an important role in preventing smoking and treating smokers, as well as in preparing adolescents for the transition to college where many begin smoking.
INTRODUCTION
School nurses play an integral role in the health care of students. Under their watchful eyes, an untold number of adolescents (and their parents) make a successful transition from high school to college. It is a time of new beginnings, and the possibilities seem endless. It is also a time when the developmental tasks of identity formation and an increase in autonomy collide and put students at risk for a number of behaviors that can have long-term health consequences. Unfortunately, smoking is one of the behaviors often encountered at this age. Cigarette smoking is a known contributor to cardiovascular disease, stroke, cancer, and obstructive lung disease (Centers for Disease Control and Prevention, [CDC], 2004a). Now, there is mounting evidence that cigarette smoking is associated with an increased incidence of depressive symptoms and major depression. The purpose of this study was to examine the level of depressive symptoms among college-age smokers and to determine whether or not increases in cigarette use were associated with increases in the number of depressive symptoms, after controlling for other factors associated with depressive symptoms.
REVIEW OF LITERATURE
Tobacco is the leading cause of death in the United States, accounting for more than 440,000 American lives lost every year (CDC, 2002). A nationwide study of 140 college campuses in 1999 found that nearly one third of college students 18–24 years of age were current smokers (smoked during the past 30 days; Rigotti, Lee, & Wechsler, 2000). Of the college students who smoke regularly (on a daily basis), more than one fourth transitioned to daily smoking while in college (Wechsler, Rigotti, Gledhill-Hoyt, & Lee, 1998). These findings suggest that during college, cigarette smoking becomes engrained, putting young people at risk for numerous tobacco-related diseases.
Of the college students who smoke regularly (on a daily basis), more than one fourth transitioned to daily smoking while in college.
Several large epidemiological studies of adults have found an association between smoking and major depression (Breslau, Kilbey, & Andreski, 1991; Glassman, Helzer, Convey, Cottler, Stetner, Tipp, & Johnson, 1990). A major depressive episode is a period longer than 2 weeks during which there is either a depressed mood or loss of interest in nearly all activities (American Psychiatric Association, 2000). The mood disturbance must be accompanied by at least four other symptoms (e.g., significant weight loss/gain, insomnia/hypersomnia, psychomotor agitation/retardation, fatigue, feelings of worthlessness/guilt, decreased concentration, and morbid thoughts). The symptoms must persist nearly every day for most of the day. The depressive symptoms must be accompanied by significant clinical distress, social or occupational distress, or a functional change in other important areas. Although a number of instruments exist to screen for the depressive symptoms, the diagnosis of depression must be made in the clinical setting to differentiate between major depressive episodes and other organic causes of mood disturbances.
Nicotine, the active drug in cigarettes, has a wide variety of psychopharmacologic effects. It is thought to increase serotonin levels and endogenous opioids and to stimulate dopamine receptors in the brain, thereby elevating mood (Pomerleau & Pomerleau, 1987; U.S. Department of Health and Human Services, 1988). Among current smokers, scores on measures of depressive symptoms have been correlated with the amount that both men and women smoked (Haukkala, Uutela, Vartiainen, McAlister, & Knekt, 2000). Prospective cohort studies with young adults (ages 21–30) showed that a history of major depression at baseline increased the risk of daily smoking, and daily smoking at baseline increased the risk of subsequent episodes of major depression (Breslau, Kilbey, & Andreski, 1993; Breslau, Peterson, Schultz, Chilcoat, & Andreski, 1998). Because most first-time smoking occurs in adolescence (U.S. Department of Health and Human Services, 1994), studies also have been conducted with this population to better understand the phenomena.
Among current smokers, scores on measures of depressive symptoms have been correlated with the amount that both men and women smoked.
In adolescents, studies also have shown links between smoking and depression (Escobedo, Reddy, & Giovino, 1998; Goodman & Capitman, 2000; Patton, Carlin, Coffey, Wolfe, Hibbert, & Bowes, 1998; Vogel, Hurford, Smith, & Cole, 2003; Wu & Anthony, 1999). Some researchers have found that higher levels of depressive symptoms at baseline increased the risk of smoking initiation in this age group (Escobedo, Reddy, & Giovino, 1998;Patton, Carlin, Coffey, Wolfe, Hibbert, & Bowes), whereas others found evidence for the opposite: Goodman and Capitman, as well as Wu and Anthony, noted that current smoking increased the risk of depressive symptoms at follow-up. The complexity of the causal relations has led some investigators to posit that smoking and depression have a shared etiology (Breslau, Peterson, Schultz, Chilcoat, & Andreski, 1998).
An increased level of depressive symptoms has been associated with many other factors besides smoking. Increased levels of depressive symptoms are found during late adolescence among college students, when compared with levels in the general population (Blumberrry, Oliver, & McClure, 1978; Nolan & Willson, 1994; Sherer, 1985; Wells, Klerman, & Deykin, 1987). In general, women are more likely than men to report depressive symptoms (Kessler, McGonagle, Swartz, Blazer, & Nelson, 1993; Weissman, Bland, Joyce, New-man, Wells, & Wittchen, 1993; Weissman & Klerman, 1977). Adolescents who use alcohol and other drugs or who have poor academic performance (Diego, Field, & Sanders, 2003), and college students who are heavy episodic drinkers are more likely to report an increased level of depressive symptoms (Weitzman, 2004). Lenz (2004) documented a sevenfold increase in smoking among young college students who reported a lifetime diagnosis of depression or treatment for depression. There are no known reports of the correlation between the frequency of cigarette smoking and the level of depressive symptoms in the college population. This study will examine smoking and level of depressive symptoms in a college population.
METHODS
Measures
The survey instrument was an 89-item questionnaire adapted from the CDC National College Health Risk Behavior Survey (NCHRBS) and included questions about safety, sleep, nutrition, exercise, alcohol, tobacco and other drugs, sexual health, and mental health (CDC, 1997). Current cigarette use was assessed using an ordinal measure of the frequency of smoking. Students reported how many days during the last 30 that they smoked. Responses ranged from 1 (none) to 7 (all 30 days). Binge drinking (five or more drinks at one sitting) was measured using the same 7-item ordinal scale. Marijuana use was also assessed using a 7-item ordinal scale. However, responses for the marijuana item ranged from 1 (none) to 7 (more than 100 times).
The Centers for Epidemiologic Studies Depression Scale (CES-D) was used to measure depressive symptoms (Radloff, 1977). The CES-D contains 20 questions that measure the presence of symptoms and complaints associated with depression. The items are scored from 0 (rarely or none of the time) to 3 (most or all of the time), which results in a total score ranging from 0–60. The standard cutoff score for an increased level of depressive symptoms is greater than or equal to 16. The Cronbach’s alpha is reported between .85 and .91 (McDowell & Newell, 1996). Demographic data also were collected.
Sample
This cross-sectional survey collected data from a random sample of 1,700 18-to 24-year-old undergraduate students who were registered full-time for classes during the Spring 2004 semester at a southeastern public university. The university registrar randomly selected a sample from the list of enrolled undergraduates (n = 18,000). Mailing labels were provided to the investigators.
Procedure
Participants received a mailed questionnaire with a cover letter explaining the purpose of the study and the importance of participating. The participants were encouraged to respond and to return the questionnaire in the enclosed, addressed, stamped envelope. Two dollars were included in the mailing as an incentive for completing the questionnaire. The study used a passive consent procedure. Completing and returning the questionnaire served as informed consent to participate. All research protocols and procedures were reviewed by the institutional review board of the university to ensure participant protection.
Stamped postcards with the subjects’ names also were included with the questionnaire. Respondents were asked to complete both the questionnaire and the postcard and to return the postcard separately to indicate completion of the questionnaire. This procedure preserved the anonymity for the participants while allowing the researchers to track who should receive a follow-up questionnaire. One week after the questionnaires were mailed, a postcard was sent reminding the participants of the importance of the study and requesting completion and return of the questionnaire. Approximately 3 weeks following the initial mailing, second follow-up letters and replacement questionnaires were mailed to all who had not returned a postcard. An attempt was made to find correct addresses for those whose questionnaires had been returned as undeliverable. After the second questionnaire mailing, those who did not return questionnaires were considered nonrespondents, and no further follow-up was made (Dillman, 2000).
RESULTS
A total of 895 completed questionnaires were received and 125 were undeliverable, yielding a response rate of 57%. More females (61%) responded than their proportion in the larger undergraduate sample from which they were selected. The participants were similar to the population in race (91% were White-non-Hispanic), and 27% held membership in a social fraternity or sorority. The participants were more likely to live in residence halls (34%) than was the student body as a whole (25%). The mean age was 21 years, and 98% were single. The majority (77%) described their health as good; 43% did not work for pay, and 19% worked 20 hours or more. The respondents were similar to the population in class standing: freshman, 26%, sophomores, 23%, juniors, 23%, and seniors 28%.
Analysis of Variables
Overall, 28% of the students reported smoking at least one time during the past 30 days. Of the current smokers, 50% smoked fewer than 10 days, 27% smoked between 10 and 29 days, and 23% smoked daily during the past month. Nearly 90% were light smokers, smoking 10 or fewer cigarettes per day. Only 1% of smokers reported smoking greater than one pack per day (more than 20 cigarettes). More than 50% of the students who ever smoked regularly (every day for 30 days) did so after age 17. Of the current smokers, 33% who ever smoked regularly reported doing so after age 17. See Table 1 for the demographic characteristics of the smoking and nonsmoking groups. The mean score on the CES-D was 11.9 (SD = 8.9), with scores ranging from 0–52. More than 26% of the sample reported an increased level of depressive symptoms with a CES-D score ≥16. More than 39% of the students reported that they did not binge-drink during the past 30 days. Of those who reported drinking five or more alcoholic beverages in one sitting, 61% reported between one and five episodes. The remainder (39%) reported binge drinking anywhere from 6 days to every day. Current marijuana use was reported by 21% of the students. Of those who used marijuana during the past 30 days, more than one third (35%) reported using the drug once or twice. The students were performing well academically. The majority (61%) reported a grade point average (GPA) of 3.0 or greater.
Nearly 90% were light smokers, smoking 10 or fewer cigarettes per day. Only 1% of smokers reported smoking greater than one pack per day.
Analysis of Smoking Groups. The smoking and non-smoking groups did not differ on age, race, place of residence, or GPA. The current smokers were more likely to be men (32%) than women (26%), χ 2(1, N = 895) = 4.1, p < .05. Current smokers reported binge drinking a greater number of times (t 892 = 12.8, p < .001) and smoking marijuana more frequently (t 762 = 4.7, p < .001) during the past month than the non-smoking group.
Analysis of Depressive Symptoms. A significantly higher mean score on the CES-D was reported by the current smokers (t 836 = 2.7, p = .01). When using the dichotomous form of the variable (score ≥16 on the CES-D), the current smokers continued to report higher levels of depressive symptoms than the nonsmokers, χ 2(1, N = 738) = 7.6, p < .01. Even when the more conservative cutoff point of ≥20 was used for the CES-D score, smokers continued to report more symptoms, χ2(1, N = 738) = 4.7, p < .05. The CES-D mean scores did not differ by sex or place of residence. Also, there were no differences when dichotomizing the variable by sex or place of residence. Minority participants did report a greater number of depressive symptoms (t 838 = 2.3, p = .02), and there were significant intercorrelations between the CES-D score and several study variables. Higher CES-D scores were associated with a lower GPA, a greater number of hours worked, increased marijuana use, and an increased number of days smoked (Table 2).
A significantly higher mean score on the CES-D was reported by the current smokers.
Multivariate Analysis. Hierarchical linear regression was conducted to explore the association between number of days smoked and level of depressive symptoms, while controlling for GPA, hours worked, marijuana use, and binge drinking. The covariates entered the model simultaneously in step one; next, the number of days smoked entered the model. When the CES-D score was regressed onto the predictor variables, the overall model was significant for predicting depressive symptoms (p < .01), and the number of days participants smoked was a significant predictor of depressive symptoms even after controlling for several known correlates of depression (Table 3).
DISCUSSION
National data show that the current smoking rates of college men and women are very similar, 22% and 23%, respectively (Johnston, O’Malley, & Bachman, 2004). However, the current study revealed that current smokers were more likely to be men than women. A greater number of women returned the survey compared with the reference population, and this may have skewed the results. Another possible explanation is that in a southeastern state known for tobacco production, a higher number of men in the 18-to 24-year-old age group actually smoke. Statewide data for the general population in the 18-to 24-year-old age group showed that 35% of men and 31% of women were smokers (CDC, 2004b). Therefore, the college students may be exhibiting the behaviors found in the larger population. For the daily smokers, there was no difference between men and women. This trend is similar to that found in the larger population of U.S. college students (Johnston, O’Malley, & Bachman, 2004).
Even though women were overrepresented in the sample, there was not a difference in the level of depressive symptoms that they reported, compared with men. This finding is consistent with earlier studies of depressive symptoms in college students (Nolan & Willson, 1994; Wells, Klerman, & Deykin, 1987). More recent studies of college students that measured psychological distress (including depressive symptoms) did find gender differences between the symptoms (Adlaf, Gliksman, Demers, & Newton-Taylor, 2001; Rosenthal & Schreiner, 2000). It is unclear why some samples of college women and men do not differ on the reporting of depressive symptoms and others do differ. In the overall sample, 26% of women and men reported an increased level of depressive symptoms. This rate is well above the adult rate (9.5%) for affective disorders in the United States (Regier, Narrow, Rae, Manderscheid, Locke, & Goodwin, 1993), but consistent with higher levels of depressive symptoms in other college samples (Wells, Klerman, & Deykin).
A key finding of this study was that college smokers report higher levels of depressive symptoms than non-smokers. Although the relationship is well documented in both adolescent and adult populations, this is one of the first reports in a college sample. Also, this study offers some support for the assumption that an increase in number of days smoked is associated with an increased level of depressive symptoms among college students.
A key finding of this study was that college smokers report higher levels of depressive symptoms than nonsmokers.
There are a number of limitations to the current study, the most important being the cross-sectional design. A prospective cohort study of college students that measures the change in behaviors over time could add greatly to the current literature. In addition to following a cohort of students, a study that uses multiple methods of data collection may produce more accurate results. This would help reduce the limitation of self-report data. The sample itself had some limitations. The study sample was mostly Caucasian, well educated, middle class, and from a state known for its tobacco production. Thus, the generalizability of the findings to other groups of college students may be limited. A follow-up study using a stratified sampling strategy would help reduce potential error.
IMPLICATIONS FOR SCHOOL NURSING PRACTICE
There are a number of recommendations that are important for school nurses. Prevention is paramount. Although for many, initiation and establishment of smoking begins in middle or high school, smoking continues to be a concern as students enter college. Continued dialogue between parents and their student about smoking is important for reducing the short-and long-term consequences of smoking. In preparation for the transition from high school to college, school nurses and others involved in the process of anticipatory guidance for college-bound students must promote health behaviors that may decrease or may prevent the adoption of smoking and other unhealthy behaviors. Of the current smokers in this study, 33% who ever smoked regularly reported doing so after age 17. School nurses have an important role in the prevention and treatment of the 67% of current smokers who started at earlier ages.
Although for many, initiation and establishment of smoking begins in middle or high school, smoking continues to be a concern as students enter college. Continued dialogue between parents and their student about smoking is important for reducing the short-and long-term consequences of smoking.
Other preventive strategies include policy changes to make all residence halls smoke-free. Weschler and colleagues (2001) found that smoke-free residence halls provided protection against smoking initiation after 19 years of age. Making all residence halls smoke-free may prevent the initiation of nonsmokers and the transition of current smokers to daily smokers, and may have an impact on mental health in the college-age population. Also, attention to the experience of depressive symptoms in college students, and particularly in those who are smoking as well as engaging in other risky behaviors, is important in reducing the immediate concerns related to depression. This is also an opportunity to prevent unhealthy behaviors and what could be lasting consequences of depression as it affects short-and long-term goals. Finally, current clinical guidelines recommend treating tobacco dependence in children and adolescents, as well as in adults (Fiore, Bailey, Cohen, Dorfman, Goldstein, Gritz, Heyman, Jaen, Kottke, Lando, Mecklenberg, Mullen, Nett, Robinson, Stitzer, Tommasello, Villejo, & Wewers, 2000). Appropriate interventions for children include counseling and behavioral therapy, whereas prescription bupropion or nicotine replacement therapy (NRT) may be considered when treating adolescents. There is no evidence that bupropion or NRT is harmful to adolescents (Fiore and colleagues); therefore, it is important to explore the use of pharmacotherapy in this population.
The 18-to 24-year-old age group is relatively healthy and free of chronic conditions, but their physical and mental well-being is compromised by smoking. School nurses are in a unique position to intervene and to affect the health of a large number of adolescents. Efforts to assess, to counsel, and to treat smoking among students will improve lives and will prevent disease sequelae for an entire generation.