Perceived Racial/Ethnic Harassment and Tobacco Use Among African American Young Adults

Abstract

We examined the association between perceived racial/ethnic harassment and tobacco use in 2129 African American college students in North Carolina. Age-adjusted and multivariate analyses evaluated the effect of harassment on daily and less-than-daily tobacco use. Harassed participants were twice as likely to use tobacco daily (odds ratio = 2.01; 95% confidence interval=1.94, 2.08) compared with those with no reported harassment experiences. Experiences of racial/ethnic harassment may contribute to tobacco use behaviors among some African American young adults.

Tobacco use remains a leading preventable cause of death in the United States, particularly among African American individuals who experience disproportionately high rates of tobacco-related chronic disease incidence and mortality.1–3

For some, tobacco use may serve as a means of managing psychosocial stress.4 This type of use may be especially common among African American individuals,5–7 who, throughout the life course, may encounter heightened levels of psychosocial stress from perceived racial discrimination.8–10 Perceived racial discrimination has been associated with adverse health-related outcomes, including tobacco use.11–15

Few studies have examined the health consequences of experiencing racial/ethnic harassment, a behavioral manifestation of racial discrimination. We suspected that as an acute, episodic, and intense psychosocial exposure, racial/ethnic harassment might be associated with tobacco use behaviors among African American young adults.

METHODS

Sample

These data were drawn from a study of health behaviors at historically Black colleges and universities in North Carolina. The study sample included 729 males (36.9%) and 1246 females (63.1%); the participants’ ages ranged from 17 to 53 years.

All of North Carolina’s 11 historically Black colleges and universities were invited to participate in the study, and 8 were enrolled (73%). A convenience sample was recruited through seminars, academic counseling centers, and other university-based activities to complete a brief self-report questionnaire.16 The sample’s composition approximated the enrollment of the represented institutions.17 All subjects were volunteers, and each institution approved data collection methods. Data collection occurred during the 2000 to 2001 academic year.

Measures

Our harassment item was drawn from a standardized survey16: “indicate whether any of the following have happened to you within the last year while you were in and around campus: ethnic or racial harassment.” A yes or no response option was provided. Participants reported their tobacco use over the previous 30-day period; 7 response options were available, ranging from “zero” to “all 30 days.” Occupational status was categorized as part time, full time, or not working. Because of limited variability, age at tobacco use initiation was dichotomized as either adolescence or older. Grade point average ranged from A+ to F. Current residence was considered either on or off campus.

Analyses

Tobacco use occurring on all of the past 30 days was considered daily use. Use on any, but not all, of the previous 30 days constituted less-than-daily tobacco use. We first tested for differences in each potential covariate by harassment. To allow for the potential nonindependence of students within a university, we used generalized estimating equations with a binary distribution. Age-adjusted models were estimated, and, subsequently, additional covariates were included.

RESULTS

Descriptive data are presented in Table 1 ▶. Of the 219 participants, 10% reported experiencing racial/ethnic harassment. Harassed participants were more likely to be male (45.7%), work full time (13.7%), and reside on campus (71.4%), compared with those who did not report harassment (35.9%, 7.0%, and 78.5%, respectively).

TABLE 1—

Characteristics of Study Participants, by Report/Nonreport of Racial/Ethnic Harassment: African American College Students From 11 Historically Black Colleges, North Carolina, 2000–2001

	Harassed (n = 219), %a	Not Harassed (n = 1910), %a	Total (N = 2129), %a
Daily tobacco use*	7.8	4.0	4.4
< Daily tobacco use**	23.3	15.3	16.1
Male**	45.7	35.9	36.9
Working**
Full time	13.7	7.0	7.7
Part time	36.4	36.7	36.7
Living on campus*	71.4	78.5	77.8
Adolescent at smoking initiation	29.6	29.0	29.1
Mean age, y	20.3	20.0	20.0
Median grade point average	B+	B+	B+

^aPercentages are based on varying numbers of subjects because of missing values.

* P < .05; **P < .01.

In age-adjusted analyses, participants reporting harassment were twice as likely to use tobacco daily (odds ratio [OR] = 2.01; 95% confidence interval [CI] = 1.94, 2.08; see Table 2 ▶). This effect persisted after adjustment for other covariates (OR = 2.10; 95% CI = 1.23, 3.58), including gender, campus residence, grades, occupational status, and age at tobacco use initiation.

TABLE 2—

Odds Ratios (ORs) and 95% Confidence Intervals (CIs) of Tobacco Use, by Report of Racial/Ethnic Harassment: African American College Students From 11 Historically Black Colleges

	Daily Tobacco Use		Tobacco Use on Any of the Previous 30 Days
	Age Adjusted OR (95% CI)	Multivariatea OR (95% CI)	Age Adjusted OR (95% CI)	Multivariatea OR (95% CI)
Experienced harassment	2.01 (1.94, 2.08)	2.10 (1.23, 3.58)	1.70 (1.21, 2.41)	1.45 (0.89, 2.35)
No reported experience of harassment	1.0	1.0	1.0	1.0
Age	1.10 (1.06, 1.14)	1.11 (1.06, 1.17)	1.03 (1.00, 1.05)	1.04 (0.98, 1.09)
Gender
Male		1.12 (0.74, 1.69)		1.72 (1.47, 2.01)
Female		1.0		1.0
Working		0.67 (0.47, 0.95)		0.77 (0.61, 0.98)
Campus residence
On campus		0.49 (0.36, 0.66)		0.69 (0.52, 0.91)
Off campus		1.0		1.0
Grade point average		0.81 (0.73, 0.90)		0.92 (0.88, 0.96)
Age at tobacco use initiation
Adolescent		6.44 (4.17, 9.96)		8.46 (6.27, 11.43)
Older than adolescent		1.0		1.0

^aAdjusted for age, sex, living on campus, grades, adolescent smoking initiation, working.

Harassed participants were 70% more likely to have used tobacco on any of the previous 30 days, compared with the nonharassed participants (OR=1.70; 95% CI=1.21, 2.41); however, this effect diminished after adjustment for covariates (OR = 1.45; 95% CI = 0.89, 2.35).

DISCUSSION

We found that African American college students who reported experiencing racial/ethnic harassment were twice as likely to use tobacco products daily over a 30-day period. Our study supports and extends previous findings13,14 and suggests that tobacco may be used to manage the psychosocial stress associated with experiencing racial/ethnic harassment. Harassment was more common for participants who were employed full time and on-campus residents, contexts perhaps where interracial interaction was more common. Future work should contextualize the construct, evaluate potential modifiers (e.g., coping), and evaluate the influence of harassment, compared with other known determinants of young adult tobacco use.

Several limitations should be noted. Because of institutional concerns, the recruitment of a convenience sample was necessary. We used a nonvalidated harassment measure that did not assess the intensity, frequency, or setting of harassment episodes. However, we consider this measurement approach justified given the preliminary nature of the study and our speculation that a face-valid item might have minimized measurement error. Additionally, socioeconomic position data were unavailable; however, some evidence indicates that adjustment for socioeconomic position might have limited influence on tobacco use outcomes in comparable populations.18–20 Although we posited that harassment preceded tobacco use, evidence regarding the causal direction between psychosocial or affective exposures and tobacco use is mixed.21,22 Finally, we were unable to adjust for selected social contextual factors that have been associated with young adult tobacco use.23,24

Further work is necessary, but our findings appear to support the disaggregation of harassment in future studies of discrimination. Given African American individuals’ delayed tobacco use initiation,18,23–32 recent resurgence of tobacco use,25 and higher tobacco-related mortality rates,1,3 our findings highlight a potentially important determinant and setting (historically Black colleges and universities) for intervention.