Skip to main content
PMC home page
Nicotine & Tobacco Research logoLink to Nicotine & Tobacco Research
. 2010 Mar 15;12(5):474–482. doi: 10.1093/ntr/ntq027

Course of comorbidity of tobacco and marijuana use: Psychosocial risk factors

Judith S Brook 1,✉, Jung Yeon Lee 1, Stephen J Finch 2, Elaine N Brown 1
PMCID: PMC2861886  PMID: 20231241

Abstract

Introduction:

This longitudinal study examined the psychosocial factors associated with the comorbidity of pairs of tobacco and marijuana use trajectories from adolescence extending into adulthood in two ethnic groups, Blacks and Puerto Ricans.

Methods:

Data on psychosocial functioning and tobacco and marijuana use at four points in time were obtained.

Results:

The association between the trajectories of tobacco and marijuana use was quite high. Pairs of comorbid trajectories of tobacco and marijuana use may share at least three kinds of influence: (a) a constellation of externalizing personality risk factors, (b) Depressive Mood and low Ego Integration, and (c) identification with certain group values.

Discussion:

Knowledge of the risk and protective factors for pairs of comorbid trajectories of use may strengthen the foundation for individual and group targets for prevention and treatment programs.

Introduction

There is a wealth of data supporting the relationship between tobacco and marijuana use (Lai, Lai, Page, & McCoy, 2000; Richter et al., 2004). An estimated 9.5 million Americans (Office of Applied Studies, 2008) currently smoke both substances nationwide. Comorbidity of the two substances is significant in that, in addition to the separate effects of tobacco and marijuana use on psychosocial functioning (Hall, Degenhardt, & Lynskey, 2002; Mathers, Toumbourou, Catalano, Williams, & Patton, 2006), concurrent use of these substances can have a cumulative effect on physical functioning (e.g., chronic pulmonary problems, Taylor et al., 2002). Thus, the adverse consequences of comorbid substance use present significant public health concerns. Accordingly, it is important to identify the psychosocial indicators, which persist over time, of comorbid tobacco and marijuana use.

This is one of the first studies designed to address the important issue of predictors of specific comorbidities in patterns of use, or trajectories, of these substances. The research is unique in that it covers a wide developmental span from early adolescence to adulthood and studies the comorbidities in the understudied population of Puerto Ricans and Blacks. Furthermore, the psychosocial variables associated with pairs of comorbid trajectories of tobacco and marijuana use are related to five important domains in an individual’s life, including the individual’s personality attributes and social network. Understanding the relationship of these domains to pairs of comorbid trajectories of use is essential to improve treatment programs.

A small number of studies have focused on the relationship between trajectories of use of different substances (e.g., Flory, Lynam, Milich, Leukefeld, & Clayton, 2004; Orlando, Tucker, Ellickson, & Klein, 2005). Jackson, Sher, and Schulenberg (2008), using a large national sample, followed individuals from late adolescence to adulthood and identified five trajectories of tobacco use and four trajectories of marijuana use. Among 20 possible pairs of trajectories of comorbid tobacco and marijuana use, 7 occurred more frequently than expected. In order to isolate predictors of pairs of such comorbid trajectories of tobacco and marijuana use, Jackson et al. first identified factors that were common to trajectories of use of both substances.

A number of important influences on trajectories of substance use have been described in Family Interactional Theory (FIT). FIT (Brook, Brook, Gordon, Whiteman, & Cohen, 1990) is a multidimensional theory of the developmental pathways to substance use and other problem behaviors. The model incorporates interrelated domains, which function as proximal and distal influences on the individual’s behavior, namely components of the individual’s personality (e.g., Ego Integration, Depressive Mood, Risk Taking, Rebellion, Delinquency), social influences (e.g., Religious Attendance, Peer Deviance, Peer Substance Use), parent personality and parenting, and ecology. These domains are linked to substance use and other problem behaviors via three primary mechanisms: social modeling, parent–child attachment, and identification with values and behaviors as a result of the attachment relationship between parents and child.

Studies have found that psychosocial variables such as Sensation Seeking, Depressive Symptoms, Delinquency, Religious Attendance, and Peer Substance Use are significantly related to trajectories of use of a single substance (Brown, Flory, Lynam, Leukefeld, & Clayton, 2004; Windle & Wiesner, 2004). We add to this line of research by examining the associations of many such factors from the personality and social influence domains of FIT with pairs of comorbid trajectories of tobacco and marijuana use. Though little research has been done on the role of peer factors in predicting comorbid trajectories of use of these substances, the link between peer behavior and the individual’s substance use (Hoffman, Monge, Chou, & Valente, 2007) and trajectories of use of a single substance (Windle & Wiesner) is likely to extend to comorbid trajectories of tobacco and marijuana use.

Our research builds on the work of Jackson et al. (2008) in two important ways. First, in addition to examining late adolescent through adult stages of development, we are also examining early adolescence, a critical developmental period. Second, we examine two understudied minority populations, namely Puerto Ricans and Blacks, living in an urban area.

We propose the following hypotheses: (a) There will be comorbidity between the trajectories of tobacco and of marijuana use. (b) The comorbidity of pairs of trajectories of tobacco and marijuana use will be accounted for, in part, by internalizing problems, such as Depressive Mood and low Ego Integration. (c) The comorbidity of pairs of trajectories of tobacco and marijuana use will be accounted for, in part, by externalizing problems, such as Delinquency, Risk Taking, and Rebellion. (d) The comorbidity of pairs of trajectories of tobacco and marijuana use will be accounted for, in part, by peer influences, such as Peer Deviance, Peer Tobacco, and Peer Marijuana Use.

Methods

Sample and procedure

Data are from a four-wave longitudinal study of Black and Puerto Rican adolescents and adults. The time 1 (T1) data were collected in 1990 and the majority of the time 2 (T2), time 3 (T3), and time 4 (T4) data were collected in 1994, 2000, and 2002, respectively. The study’s procedures for data collection were Institutional Review Board approved. Participants at T1 (N = 1,331) came from Grades 7–10 in 11 schools serving the East Harlem area of New York City. The T1 data collection took place in classrooms, while the T2, T3, and T4 data were collected primarily via in-person home interviews. The T2 response rate was 89% of those who participated at T1. Because of budget limitations, the T3 data collection was a subsample of the T2 sample (T3 N = 660). To ensure sufficient Ns on our dependent variables, we oversampled respondents who reported using marijuana and/or having a child at T2. At T4, again due to budget restrictions, we took a subsample of the T3 participants (T4 N = 475). As the T4 data collection emphasized tobacco use, smokers were oversampled. The 475 participants present at T4 were used in the trajectory analyses in this paper.

We ran three sets of comparisons: those interviewed at T1 only compared with those participating at both T1 and T4, those interviewed at T2 but not T4 compared with those participating at T2 and T4, and those interviewed at T3 but not T4 compared with those participating at T3 and T4. Participants who took part at T1 and T4, compared with those interviewed at T1 only, reported less Ego Integration and greater Peer Deviance, Cigarette Use, and Marijuana Use (p ≤ .05). Participants who took part at T2 and T4, compared with those who participated at T2 but not T4, reported less Ego Integration, greater Peer Deviance, and greater Peer Cigarette Use (p ≤ .05). Participants who participated at T3 and T4, compared with those who participated at T3 but not T4, demonstrated no statistically significant differences on the demographic or psychosocial variables (p ≤ .05).

Of the 475 participants, 51% (n = 243) were Black and 49% (n = 232) were Puerto Rican. Females comprised 50.7% (n = 241) of the sample. Mean ages were 13.9 (SD = 1.3) at T1, 19.3 (SD = 1.5) at T2, 24.4 (SD = 1.3) at T3, and 26.1 (SD = 1.4) at T4. The median educational level at T4 was having completed at least 1 year of business or technical school. With regard to the occupational level at T4, 17.3% were employed in semiskilled jobs (e.g., factory worker), 11.0% in skilled jobs (e.g., mechanic), 33.4% in clerical positions, 13.9% had professional level jobs, and 24.4% were unemployed. Of those who were unemployed, 17.2% were attending school. At T4, 20.6% of the participants were cohabiting, 16.0% were married and living together, 2.8% were married but separated, and 60.6% were single.

Measures

The respondents were asked about the number of cigarettes currently smoked at each wave (T1–T4). Response options included “none” (coded 1), “a few cigarettes or less a week” (2), “one to five cigarettes a day” (3), “about half a pack a day” (4), “about one pack a day” (5), and “more than one pack a day” (6).

The respondents were asked about the frequency of their marijuana use. The response options included “never” (1), “a few times a year or less” (2), “about once a month” (3), “several times a month” (4), and “once a week or more” (5).

Table 1 presents the demographic variables and the psychosocial variables with their Cronbach’s alphas and source (see Table 1). Each psychosocial variable is the sum of all items from T1 to T4. The Cronbach’s alphas were adequate. The psychosocial variables have been found in previous research to predict substance use and psychopathology (Brook, Whiteman, Czeisler, Shapiro, & Cohen, 1997; Crawford, Cohen, & Brook, 2001). Externalizing and internalizing personality sets were patterned after the work of Achenbach (1999).

Table 1.

Psychosocial variables: Sources and Cronbach’s alphas

T1–T4 psychosocial variables Source No. of items Sample item Response range (minimum, maximum) Cronbach’s alphaa
Demographics
    Genderb – 1 Are you (female or male)? Female (1), male (2) n/a
    Race/Ethnicityb – 1 Which (race/ethnicity) best describes you? Black (1), Puerto Rican (2) n/a
Religion
    Religious Attendancec – 4 How often do you attend religious services? Never (1), once a week or more (5) n/a
Internalizing personality attributes
    Ego Integration Brook and Pahl (2005) 12 (How well does this describe you?) You feel like swearing Completely true (1), completely false (4) .68
    Depressive Mood Derogatis (1994) 20 (How well does this describe you?) You sometimes feel unhappy, sad or depressed Completely false (1), completely true (4) .87
Externalizing personality attributes
    Delinquency Gold (1966); Huizinga, Menard and Elliott (1989) 46 How often have you gotten in trouble with the police for something you did? Never (1), 5 or more times (5) .88
    Risk Taking Jackson (1997) 12 (How well does this describe you?) You would do almost anything on a dare Completely false (1), completely true (4) .84
    Rebellion Smith and Fogg (1979) 12 (How well does this describe you?) Sometimes you enjoy seeing how much you can get away with Completely false (1), completely true (4) .76
Peer factors
    Peer Deviance Gold (1966); Huizinga, Menard and Elliott (1989) 12 How many of your friends have gotten into a serious fight at school or work? None (1), most (4) .85
    Peer Tobacco Usec Siqueira and Brook (2003) 4 How many of your friends smoke cigarettes on a regular basis? None (1), most (4) n/a
    Peer Marijuana Usec Siqueira and Brook (2003) 4 How many of your friends have ever used marijuana or hashish(pot, grass)? None (1), most (4) n/a
Open in a new tab

Note. n/a = not applicable. aCronbach’s alphas were computed based on responses to each of the items that comprised the psychosocial measures four points in time.

b

Items only asked at T1.

c

Variable used is the sum of a single item asked at four different timepoints.

Analytic plan

We applied the SAS Traj procedure (Jones & Nagin, 2007; Jones, Nagin, & Roeder, 2001) to explore the trajectories of the participants’ tobacco use and marijuana use using the censored normal distribution (White, Pandina, & Chen, 2002).

Since Brook, Ning, and Brook (2006) and Brook, Balka, Ning, and Brook (2007) reported a four-trajectory group model for tobacco use using this sample, we used four tobacco use trajectory groups. For marijuana use, the model having the maximum value of the Bayesian information criterion (BIC) and Akaike’s information criterion (AIC) was selected. We assigned trajectory group membership using modal posterior probabilities.

In line with Jackson et al. (2008), we evaluated comorbidity between tobacco and marijuana use from a cross-tabulation of trajectory group memberships (see Table 2).

Table 2.

Cross-tabulations of frequency (expected frequency, Eij) of group membership in tobacco use trajectories and marijuana use trajectories

Marijuana use
Tobacco use Non/low user Maturing out Late onset Chronic Marginals
Non/low use 257 (202.9)↑ 14 (27.6) 23 (41.7) 11 (32.8) 305 (64.21%)
Maturing out 3 (4.0) 0 (0.0) 2 (0.8) 1 (0.6) 6 (1.26%)
Late onset 44 (73.9) 14 (10.5) 30 (15.2)↑ 23 (11.9) 111 (23.37%)
Chronic 12 (35.3) 15 (4.8)↑ 10 (7.3) 16 (5.7)↑ 53 (11.16%)
Marginals 316 (66.53%) 43 (9.05%) 65 (13.68%) 51 (10.74%) 475
Open in a new tab

Note. χ2(9, n = 475) = 141.78, p < .0001; Φ = 0.55; Cramer’s V = 0.32. Numbers with up arrows (↑) indicate values that are significantly greater than would be expected under independence using the normal approximation to the binomial (p < .0001), that is, Inline graphic where Oij is the observed value of ith row and jth column and Eij is the expected value of ith row and jth column.

We tested observed versus expected cell frequencies in the trajectories of tobacco and marijuana use contingency table to determine those trajectory pairs that occur more frequently than expected under independence (Lienert & Krauth, 1975; von Eye, 2002). A pair of trajectories (i, j) was selected when Inline graphic, where Oij is the observed value of ith row and jth column and Eij is the expected value of ith row and jth column with 3.72 chosen to set p < .0001.

For each selected pair of tobacco and marijuana use trajectories, first, we predicted Yk, the indicator of the selected marijuana use trajectory group for participant k from the indicator of the selected tobacco use trajectory group for participant k, Xk. The logistic regression model is Yi = β0 + β1Xi + ϵi, i = 1, … ,475, where ϵk is the residual error for kth participant under the model. The overall odds ratio (OR) for the selected pair of trajectories is Inline graphic (Agresti, 1996). These ORs are reported in the first line of Table 3.

Table 3.

Adjusted odds ratios of selected pairs of tobacco use trajectories and marijuana use trajectories from logistic regression analyses after controlling each specified risk factor

Risk factors controlled for Non/low tobacco, non/low marijuana Chronic tobacco, maturing-out marijuana Late onset tobacco, late onset marijuana Chronic tobacco, chronic marijuana
None 10.1 (6.5, 15.7)a 5.6 (2.7, 11.3)a 3.5 (2.0, 6.0)a 4.8 (2.4, 9.4)a
Gender 9.9 (6.3, 15.4) 5.5 (2.7, 11.2) 3.2 (1.8, 5.5) 5.1 (2.6, 10.2)
Ethnicity 10.2 (6.6, 15.9) 5.5 (2.7, 11.2) 3.5 (2.0, 6.0) 5.2 (2.6, 10.3)
Religious Attendance 10.0 (6.3, 15.7) 5.5 (2.6, 11.4) 3.3 (1.9, 5.8) 3.9 (1.9, 7.8)**
Ego Integration 10.5 (6.6, 16.7) 5.2 (2.5, 10.8) 3.4 (1.9, 5.9) 3.4 (1.7, 7.0)**
Depressive Mood 10.3 (6.5, 12.3) 4.7 (2.2, 9.9) 3.4 (1.9, 5.9) 3.4 (1.7, 7.0)**
Delinquency 7.2 (4.4, 11.7)** 4.4 (2.0, 9.2) 3.1 (1.7, 5.4) 2.4 (1.1, 5.2)***
Risk Taking 9.5 (5.9, 15.2) 4.7 (2.3, 9.8) 3.0 (1.7, 5.3) 3.7 (1.8, 7.6)*
Rebellion 8.2 (5.1, 13.2)* 3.8 (1.8, 8.1)** 3.0 (1.7, 5.3) 2.7 (1.3, 5.7)***
Peer Deviance 9.1 (5.6, 14.6) 4.0 (1.9, 8.4)** 3.2 (1.8, 5.7) 2.9 (1.4, 6.0)***
Peer Tobacco Use 7.8 (4.8, 12.6)* 3.0 (1.4, 6.8)** 3.5 (1.9, 6.2) 1.9 (0.9, 4.1)***
Peer Marijuana Use 7.4 (4.5, 12.2)** 2.9 (1.4, 6.3)*** 2.9 (1.6, 5.1)* 2.1 (1.0, 4.4)***
Open in a new tab

Note. The tobacco and marijuana use trajectory groups are defined in the results section. Stars are used to indicate that the odds ratio (OR) after controlling for risk factor(s) was significantly less than the OR with no risk factors controlled (*p < .05, **p < .01, ***p < .001). Each column describes a logistic regression on N = 475 subjects. The dependent variable is the indicator of membership in the specified marijuana use trajectory, and the independent trajectory variable is membership in the specified tobacco use trajectory. The second independent variable is specified in the row. For example, the second column indicates the OR of membership in the non/low marijuana use trajectory group as predicted by membership in the non/low tobacco use trajectory alone (first row) and with control on each of the specified risk factors (Rows 2–12).

a

OR significant at the level of .0001.

We then conducted further logistic regression analyses to see whether a risk variable, R, reduced the OR in the pair of trajectories. That is, we fit the logistic regression model Yk = γ0 + γ1Xk + γ2Rk + νk, k = 1, … ,475, where Rk is the value of the risk variable R, Yk and Xk are defined as above, and νk is the residual error for kth participant under this model. The value Inline graphic is the OR between trajectory groups controlling for risk variable R.

We compare the OR controlling for R with the overall OR using the test statistic Inline graphic (Clogg, Petkova, & Haritou, 1995). Under the null hypothesis of no change in OR, T is approximately standard normal.

We then examined the five sets of variables specified in Table 4. The logistic regression of the trajectory of marijuana use was estimated with control on the tobacco use trajectory and all variables in each of the sets (see Table 4). We also estimated the logistic regression of the trajectory of marijuana use with control on the tobacco use trajectory and with control on all the psychosocial variables simultaneously. We then tested whether the OR with a given tobacco use trajectory variable was significantly reduced when a set of predictors was added as a control.

Table 4.

Adjusted odds ratios and 95% CI for selected pairs of tobacco use trajectories and marijuana use trajectories from logistic regression analyses with control on each of 5 sets of risk factors, and on a set of all the risk factors

Risk factors controlled for Non/low tobacco, non/low marijuana Chronic tobacco, maturing-out marijuana Late onset tobacco, late onset marijuana Chronic tobacco, chronic marijuana
None 10.1 (6.5, 15.7) 5.6 (2.7, 11.3) 3.5 (2.0, 6.0) 4.8 (2.4, 9.4)
Demographics 10.0 (6.4, 15.6) 5.4 (2.6, 11.1) 3.2 (1.8, 5.5) 5.5 (2.7, 11.2)
    Gender
    Ethnicity
Religion 10.0 (6.3, 15.7) 5.5 (2.6, 11.4) 3.3 (1.9, 5.8) 3.9 (2.0, 7.8)
    Religious Attendance
Internalizing personality attributes 10.2 (6.4, 16.3) 4.7 (2.2, 9.9) 3.4 (1.9, 5.9) 3.1 (1.5, 6.3)***
    Ego Integration
    Depressive Mood
Externalizing personality attributes 6.6 (4.0, 11.0)*** 3.8 (1.7, 8.3)* 2.9 (1.7, 5.2)* 2.1 (1.0, 4.7)***
    Delinquency
    Risk taking
    Rebellion
Peer factor 10.7 (6.0, 18.9) 2.9 (1.3, 6.5)*** 3.3 (1.8, 5.9) 1.9 (0.9, 4.2)***
    Peer Deviance
    Peer Tobacco Use
    Peer Marijuana Use
Gender 7.0 (3.8, 12.9)*** 3.7 (1.6, 8.5)* 2.4 (1.3, 4.4)** 1.3 (0.5, 3.3)***
Ethnicity
Religious Attendance
Ego Integration
Depressive Mood
Delinquency
Risk Taking
Rebellion
Peer Deviance
Peer Tobacco Use
Peer Marijuana Use
Open in a new tab

Note. The tobacco and marijuana use trajectory groups are defined in the results section. Stars are used to indicate that the odds ratio (OR) after controlling for risk factor(s) was significantly less than the OR with no risk factors controlled (*p < .05, **p < .01, ***p < .001). Each column describes a logistic regression on N = 475 subjects. The dependent variable is the indicator of membership in the specified marijuana use trajectory, and the independent trajectory variable is membership in the specified tobacco use trajectory. The additional independent variable(s) is(are) specified in the row. For example, the second column indicates the OR of membership in the non/low marijuana use trajectory group as predicted by membership in the non/low tobacco use trajectory alone (first row) and with control on each of the sets of risk factors (Rows 2–7).

Results

Mixture modeling: Extracting trajectories

There were four tobacco use trajectory groups. The average trajectory for “non/low tobacco users” had means corresponding to not smoking at all or smoking a few cigarettes or less a week at all four times (i.e., 1.29, 1.07, 1.10, and 1.14). The average trajectory for “maturing-out tobacco users” had means corresponding to smoking one pack a day at T1 (5.02), not smoking at all or smoking a few cigarettes or less a week at T2 and T3 (1.05 and 1.17), and smoking more than a few cigarettes per week at T4 (2.23). The average trajectory for “late onset tobacco users” had means corresponding to not smoking at all or smoking a few cigarettes or less a week at T1 and T2 (1.34 and 1.74), smoking one to five cigarettes a day at T3 (2.96), and smoking more than one to five cigarettes a day at T4 (3.24). The average trajectory for “chronic tobacco users” had means corresponding to not smoking at all or smoking a few cigarettes or less a week at T1 (1.62) and smoking a half a pack a day or more at T2, T3, and T4 (3.97, 4.26, and 4.08). Estimated prevalences for the four trajectory groups were 64.2% non/low tobacco users, 1.3% maturing-out tobacco users, 23.4% late onset tobacco users, and 11.1% chronic tobacco users.

For marijuana use, we computed solutions for two through five components. Since both the BIC (−2,409) and the AIC (−2,359) had the largest value for the four-trajectory group model, we chose a four-component model for marijuana use.

The average trajectory for “non/low marijuana users” had means corresponding to not using marijuana at all or using marijuana a few times a year or less at all four waves of data collection (i.e., 1.08, 1.25, 1.19, and 1.24). The average trajectory for “maturing-out marijuana users” had means corresponding to not using marijuana at all or using marijuana a few times a year or less at T1 (1.48), using more than several times a month at T2 (4.50), using from a few times a year or less to once a month at T3 (2.28), and not using marijuana at all or using marijuana a few times a year or less at T4 (1.73). The average trajectory for “late onset marijuana users” had means corresponding to not using marijuana at all or using marijuana a few times a year or less at T1 and T2 (1.15 and 1.78), using more than once a month at T3 (3.44), and more than several times a month at T4 (4.28). The average trajectory for “chronic marijuana users” had means corresponding to not using marijuana at all or using marijuana a few times a year or less at T1 (1.72) and using marijuana more than several times a month T2, T3, and T4 (4.45, 4.65, and 4.65). Estimated prevalences of the four trajectory groups were 66.5% non/low marijuana users, 10.7% maturing-out marijuana users, 13.7% late onset marijuana users, and 9.1% chronic marijuana users.

Comorbidity

Table 2 presents the cross-tabulation of trajectory group memberships. The trajectories of tobacco and marijuana use were significantly associated, χ2(9, n = 475) = 141.78, p < .0001; Φ = 0.55; Cramer’s V = 0.32. There were four trajectory pairs (indicated by up arrows [↑]) that had counts significantly greater than would be expected under independence. As noted in Table 3, the ns and the ORs for the groups were as follows: non/low tobacco users and non/low marijuana users (n = 257, OR = 10.1; p < .001), chronic tobacco users and maturing-out marijuana users (n = 15, OR = 5.6; p < .001), late onset tobacco users and late onset marijuana users (n = 30, OR = 3.5; p < .001), and chronic tobacco users and chronic marijuana users (n = 16, OR = 4.8; p < .01). More than 66% of participants were in a comorbid trajectory pair with most in the non/low tobacco use and non/low marijuana use trajectory pair.

Correlates of comorbidity

Table 3 shows ORs when a single risk variable was controlled. The ORs of the comorbidity of the pairs of trajectories of tobacco and marijuana use were reduced with control on many of the psychosocial risk variables. For the non/low tobacco use and non/low marijuana use trajectory pair, no single risk factor reduced the OR below 7.2 (OR reduced from 10.1 to 7.2, t = 3.0, p < .01). For the non/low tobacco use and non/low marijuana use trajectory pair, the greatest reduction was obtained by controlling for Delinquency. For each of the four comorbid trajectory pairs, controlling for Peer Marijuana Use generated the largest or next largest reduction in OR. The chronic tobacco use and chronic marijuana use trajectory pair controlling for Peer Tobacco use had the largest reduction in OR. For the chronic tobacco use and chronic marijuana use trajectory group, all of the psychosocial risk variables except the demographic variables (i.e., Gender and Ethnicity) significantly reduced the ORs (see Table 3).

Table 4 presents the results of six sets of multivariate logistic regression analyses (one for each row of the table) for the four selected pairs of tobacco and marijuana use trajectory groups (see Table 4). The leftmost column specifies the variables controlled for. As shown in Table 4, for the non/low tobacco use and non/low marijuana use trajectory pair, which included more than 50% of the sample, no set of risk factors reduced the OR below 6.6. Controlling for the set of externalizing personality attributes produced the greatest reduction in OR for the non/low tobacco use and non/low marijuana use trajectory pair and the late onset tobacco use and late onset marijuana use trajectory pair. It also reduced the OR significantly for the other pairs. Controlling for the set of peer variables had the greatest reduction in OR for the chronic tobacco use and maturing-out marijuana use trajectory pair and the chronic tobacco use and chronic marijuana use trajectory pair. Controlling for the internalizing personality attributes significantly reduced the OR only for the chronic tobacco use and chronic marijuana use trajectory pair. Controlling for all the variables at once significantly reduced the OR for each of the four comorbid trajectory pairs.

Discussion

We used the growth mixture model to identify multiple trajectories of tobacco and marijuana use. We explored the developmental course of comorbid tobacco and marijuana use beginning in adolescence and extending into adulthood, and we identified the associated factors that reduced the comorbidity of pairs of tobacco and marijuana use trajectories. This is the first study to examine the psychosocial factors that are common to pairs of comorbid trajectories of tobacco and marijuana use in Blacks and Puerto Ricans.

Comorbidities in substance use

The trajectories of tobacco and marijuana use are strongly related (see Table 2). The following four pairs are more common than expected under independence: non/low tobacco use and non/low marijuana use, chronic tobacco use and maturing-out marijuana use, late onset tobacco use and late onset marijuana use, and chronic tobacco use and chronic marijuana use (see Table 2). All of the pairs of comorbid trajectories were consistent with those reported by Jackson et al. (2008).

The comorbid tobacco and marijuana use trajectories have implications for similar developmental timing in the use of tobacco and marijuana. This may be due to the interaction of these two substances or to common developmental transitions (e.g., living situation, traditional roles associated with a new career and family relations) and personality factors.

Prediction of comorbidity by risk factors

Some of the individual risk factors explained, in part, the comorbidity of the pairs of trajectories of tobacco use and marijuana use (see Table 3). The pattern of risk factors suggests three kinds of influence on comorbidity. The first is identification with certain group values, hence the importance of deviant peer groups and participation in religious groups. The second pattern refers to a personality disposition manifested in impulsivity and ignoring the consequences of one’s behavior. The third draws on the significance of Depressive Mood and possible relief from internal distress.

Table 4 indicates that comorbidity of pairs of trajectories of tobacco and marijuana use may be explained in part by a constellation of externalizing personality risk factors (e.g., Delinquency, Risk Taking, and Rebellion). Thus, those who are more extreme in unconventional personality attributes are more likely to be polydrug users. Our findings are consistent with Problem Behavior Theory (Donovan & Jessor, 1985; Turbin, Jessor, & Costa, 2000), which posits a subset of adolescent behaviors including Delinquency, tobacco use, and illicit drug use that are linked and often co-occur.

Internalizing behavior only reduced the OR for the comorbid pair of trajectories of chronic tobacco and chronic marijuana use. However, our findings that internalizing behavior does not lead to a reduction in the ORs of the other comorbid pairs of trajectories of tobacco and marijuana use may be due to a smaller effect size. The self-medication hypothesis (Khantzian, 1997) suggests that the abuse of substances may function as a means of relieving or making a “subjective state of distress” (such as Depressive Mood) more tolerable. Our finding that those with internalizing personality problems are more likely to be chronic users of both tobacco and marijuana (data not shown) supports the relationship between “distress” and chronic polysubstance use.

The externalizing factor appears to be more influential than the internalizing factor (Table 4). This is in accord with previous research on the effect of such personality factors on trajectories of use of only one drug (Windle & Wiesner, 2004). The externalizing personality attributes partially reduced the comorbidity of pairs of trajectories of tobacco and marijuana use, despite our small sample size. In future research using larger samples, the relationship of a variety of biopsychosocial factors to pairs of comorbid trajectories of tobacco and marijuana use may be even more apparent.

Peer factors partly reduced the ORs of the comorbidity of pairs of trajectories of tobacco and marijuana use (i.e., chronic tobacco use and maturing-out marijuana use and chronic tobacco use and chronic marijuana use). One possibility for this effect is that deviant peers have an adverse effect on Peer Substance Use through socialization processes (Hoffman et al., 2007). Moreover, substance-using individuals seek out peers with similar behaviors (Ennett & Bauman, 2006). In general, these findings are in accord with FIT, which emphasizes the influence of peers through such processes.

There were no major differences in Ethnicity or Gender in the risk and protective factors related to pairs of comorbid trajectories of use, in accord with Jackson et al. (2008). The results have considerable generalizability for male and female and for Black and Puerto Rican individuals. Clinical programs designed to deal with the comorbidity of tobacco and marijuana use might be similar for both Blacks and Puerto Ricans. Nevertheless, as Compton, Grant, Colliver, Glantz, and Stinson (2004) have noted, interventions need to be linguistically appropriate and culturally relevant.

Limitations

First, it remains possible that the associations between the predictors and pairs of comorbid trajectories may arise from genetic risk factors and other environmental variables (e.g., school influences) that were not examined in this study. Second, our data are based on self-reports rather than on external measurements from official records, such as police records, though studies have shown that use of this type of self-report data yields reliable results (Harrison, Martin, Enev, & Harrington, 2007). Third, the sample sizes for some of the comorbid trajectory pairs are limited in size. Given the relatively small N, and the restriction of our sample to New York City, the study of additional samples is warranted.

Despite these limitations, the study supports and extends the literature in a number of important ways. We assess psychosocial variables over a span of 12 years. The current study demonstrates the feasibility and substantive importance of this approach to modeling comorbidity. Since comorbidity of tobacco and marijuana use trajectories occurs in many individuals, the presence of additional substance use problems needs to be assessed in treatment programs for the use of one of these substances. Once identified, treatment for all substance use problems should be coordinated. Without comprehensive treatment, chronic tobacco and chronic marijuana users are at risk for the adverse psychosocial and health consequences associated with concurrent heavy use of tobacco and marijuana over a number of years. Additionally, the study highlights the predictors of the comorbid pair of trajectories of chronic tobacco use and chronic marijuana use: namely dimensions involving identification with group values and behaviors (e.g., Religious Attendance, Peer Substance Use), personality dispositions (e.g., Risk Taking), and Depressive Mood. This knowledge may strengthen the foundation for both prevention and treatment programs that address the development of comorbid use of tobacco and marijuana.

Funding

This work was supported by a research scientist award (DA00244) and a research grant (DA005702) from the National Institutes of Drug Abuse and by a research grant from the National Cancer Institute (CA084063).

Declaration of Interests

None declared.

Acknowledgments

The authors thank Dr. Martin Whiteman and David Brook, M.D., for critical review of the manuscript and two anonymous reviewers for their helpful comments.

References

  1. Achenbach TM. The child behavior checklist and related instruments. In: Maruish ME, editor. The use of psychological testing for treatment planning and outcomes assessment. Mahwah, NJ: Erlbaum; 1999. pp. 429–466. [Google Scholar]
  2. Agresti A. An introduction to categorical data analysis. New York: Wiley; 1996. [Google Scholar]
  3. Brook JS, Balka EB, Ning Y, Brook DW. Trajectories of cigarette smoking among African Americans and Puerto Ricans from adolescence to young adulthood: Associations with dependence on alcohol and illegal drugs. American Journal on Addictions. 2007;16:195–201. doi: 10.1080/10550490701375244. [DOI] [PubMed] [Google Scholar]
  4. Brook JS, Brook DW, Gordon AS, Whiteman M, Cohen P. The psychosocial etiology of adolescent drug use: A family interactional approach. Genetic, Social, and General Psychology Monographs. 1990;116:111–267. [PubMed] [Google Scholar]
  5. Brook JS, Ning Y, Brook DW. Personality risk factors associated with trajectories of tobacco use. American Journal on Addictions. 2006;15:426–433. doi: 10.1080/10550490600996363. [DOI] [PubMed] [Google Scholar]
  6. Brook JS, Pahl K. The protective role of ethnic and racial identity and aspects of an Africentric orientation against drug use among African-American young adults. Journal of Genetic Psychology. 2005;166:329–345. doi: 10.3200/GNTP.166.3.329-345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brook JS, Whiteman M, Czeisler LJ, Shapiro J, Cohen P. Cigarette smoking in young adults: Childhood and adolescent personality, familial, and peer antecedents. Journal of Genetic Psychology. 1997;158:172–188. doi: 10.1080/00221329709596660. [DOI] [PubMed] [Google Scholar]
  8. Brown TL, Flory K, Lynam DR, Leukefeld C, Clayton RR. Comparing the developmental trajectories of marijuana use of African American and Caucasian adolescents: Patterns, antecedents, and consequences. Experimental and Clinical Psychopharmacology. 2004;12:47–56. doi: 10.1037/1064-1297.12.1.47. [DOI] [PubMed] [Google Scholar]
  9. Clogg CC, Petkova E, Haritou A. Symposium on applied regression: Statistical methods for comparing regression coefficients between models. American Journal of Sociology. 1995;100:1261–1293. [Google Scholar]
  10. Compton WM, Grant BF, Colliver JD, Glantz MD, Stinson FS. Prevalence of marijuana use disorders in the United States: 1991–1992 and 2001–2002. Journal of the American Medical Association. 2004;291:2114–2121. doi: 10.1001/jama.291.17.2114. [DOI] [PubMed] [Google Scholar]
  11. Crawford TN, Cohen P, Brook JS. Dramatic-erratic personality disorder symptoms: II. Developmental pathways from early adolescence to adulthood. Journal of Personality Disorders. 2001;15:336–350. doi: 10.1521/pedi.15.4.336.19185. [DOI] [PubMed] [Google Scholar]
  12. Derogatis LR. Symptoms checklist 90-R administration, scoring procedures manual. 3rd ed. Minneapolis, MN: National Computer Systems; 1994. [Google Scholar]
  13. Donovan JE, Jessor R. Structure of problem behavior in adolescence and young adulthood. Journal of Consulting and Clinical Psychology. 1985;53:890–904. doi: 10.1037//0022-006x.53.6.890. [DOI] [PubMed] [Google Scholar]
  14. Ennett ST, Bauman KE. The contribution of influence and selection to adolescent peer group homogeneity: The case of adolescent cigarette smoking. In: Levine JM, Moreland RL, editors. Small groups: Key readings in social psychology. New York: Psychology Press; 2006. pp. 21–36. [DOI] [PubMed] [Google Scholar]
  15. Flory K, Lynam D, Milich R, Leukefeld C, Clayton R. Early adolescent through young adult alcohol and marijuana use trajectories: Early predictors, young adult outcomes, and predictive utility. Development and Psychopathology. 2004;16:193–213. doi: 10.1017/s0954579404044475. [DOI] [PubMed] [Google Scholar]
  16. Gold M. Undetected delinquent behavior. Journal of Research in Crime and Delinquency. 1966;3:27–46. [Google Scholar]
  17. Hall W, Degenhardt L, Lynskey M. The health and psychological effects of cannabis use (Monograph Series No. 44). Canberra: Commonwealth of Australia; 2002. [Google Scholar]
  18. Harrison LD, Martin SS, Enev T, Harrington D. Comparing drug testing and self-report of drug use among youths and young adults in the general population (DHHS Publication No. SMA 07-4249, Methodology Series M-7) Rockville, MD: Substance Abuse and Mental Health Services Administration, Office of Applied Studies; 2007. [Google Scholar]
  19. Hoffman BR, Monge PR, Chou C.-P., Valente TW. Perceived peer influence and peer selection on adolescent smoking. Addictive Behaviors. 2007;32:1546–1554. doi: 10.1016/j.addbeh.2006.11.016. [DOI] [PubMed] [Google Scholar]
  20. Huizinga DH, Menard S, Elliott DS. Delinquency and drug use: Temporal and developmental patterns. Justice Quarterly. 1989;6:419–455. [Google Scholar]
  21. Jackson DN. Jackson Personality Inventory-Revised manual. Port Heron, MI: Sigma Assessment Systems; 1997. [Google Scholar]
  22. Jackson KM, Sher KJ, Schulenberg JE. Conjoint developmental trajectories of young adult substance use. Alcoholism: Clinical and Experimental Research. 2008;32:723–737. doi: 10.1111/j.1530-0277.2008.00643.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jones BL, Nagin DS, Roeder K. A SAS procedure based on mixture models for estimating developmental trajectories. Sociological Methods & Research. 2001;29:374–393. [Google Scholar]
  24. Jones BL, Nagin DS. Advances in group-based trajectory modeling and a SAS procedure for estimating them. Sociological Methods & Research. 2007;35:542–571. [Google Scholar]
  25. Khantzian EJ. The self-medication hypothesis of substance use disorders: A reconsideration and recent applications. Harvard Review of Psychiatry. 1997;4:231–244. doi: 10.3109/10673229709030550. [DOI] [PubMed] [Google Scholar]
  26. Lai S, Lai H, Page JB, McCoy CB. The association between cigarette smoking and drug abuse in the United States. Journal of Addictive Diseases. 2000;19:11–24. doi: 10.1300/J069v19n04_02. [DOI] [PubMed] [Google Scholar]
  27. Lienert GA, Krauth J. Configural frequency analysis as a statistical tool for defining types. Educational and Psychological Measurement. 1975;35:231–238. [Google Scholar]
  28. Mathers M, Toumbourou JW, Catalano RF, Williams J, Patton GC. Consequences of youth tobacco use: A review of prospective behavioural studies. Addiction. 2006;101:948–958. doi: 10.1111/j.1360-0443.2006.01438.x. [DOI] [PubMed] [Google Scholar]
  29. Office of Applied Studies. Results from the 2007 National Survey on Drug Use and Health: Detailed tables. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2008. [Google Scholar]
  30. Orlando M, Tucker JS, Ellickson PL, Klein DJ. Concurrent use of alcohol and cigarettes from adolescence to young adulthood: An examination of developmental trajectories and outcomes. Substance Use & Misuse. 2005;40:1051–1069. doi: 10.1081/JA-200030789. [DOI] [PubMed] [Google Scholar]
  31. Richter KP, Kaur H, Resnicow K, Nazir N, Mosier MC, Ahluwalia JS. Cigarette smoking among marijuana users in the United States. Substance Abuse. 2004;25:35–43. doi: 10.1300/j465v25n02_06. [DOI] [PubMed] [Google Scholar]
  32. Siqueira LM, Brook JS. Tobacco use as a predictor of illicit drug use and drug-related problems in Colombian youth. Journal of Adolescent Health. 2003;32:50–57. doi: 10.1016/s1054-139x(02)00534-7. [DOI] [PubMed] [Google Scholar]
  33. Smith GM, Fogg CP. Psychological antecedents of teenage drug use. In: Simmons R, editor. Research in community and mental health: An annual compilation of research. Vol. 1. Greenwich, CT: JAI; 1979. pp. 87–102. [Google Scholar]
  34. Taylor DR, Fergusson DM, Milne BJ, Horwood LJ, Moffitt TE, Sears MR, et al. A longitudinal study of the effects of tobacco and cannabis exposure on lung function in young adults. Addiction. 2002;97:1055–1061. doi: 10.1046/j.1360-0443.2002.00169.x. [DOI] [PubMed] [Google Scholar]
  35. Turbin MS, Jessor R, Costa FM. Adolescent cigarette smoking: Health-related behavior or normative transgression? Prevention Science. 2000;1:115–124. doi: 10.1023/a:1010094221568. [DOI] [PubMed] [Google Scholar]
  36. von Eye A. The odds favor antitypes: A comparison of tests for the identification of configural types and antitypes. Methods of Psychological Research Online. 2002;7:1–29. [Google Scholar]
  37. White HR, Pandina RJ, Chen P.-H. Developmental trajectories of cigarette use from early adolescence into young adulthood. Drug and Alcohol Dependence. 2002;65:167–178. doi: 10.1016/s0376-8716(01)00159-4. [DOI] [PubMed] [Google Scholar]
  38. Windle M, Wiesner M. Trajectories of marijuana use from adolescence to young adulthood: Predictors and outcomes. Development and Psychopathology. 2004;16:1007–1027. doi: 10.1017/s0954579404040118. [DOI] [PubMed] [Google Scholar]

Articles from Nicotine & Tobacco Research are provided here courtesy of Oxford University Press

RESOURCES