- Journal List
- HHS Author Manuscripts
- PMC3813459
As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsem*nt of, or agreement with, the contents by NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice
J Am Acad Child Adolesc Psychiatry. Author manuscript; available in PMC 2014 May 1.
Published in final edited form as:
J Am Acad Child Adolesc Psychiatry. 2013 May; 52(5): 10.1016/j.jaac.2013.02.014.
Published online 2013 Apr 4. doi:10.1016/j.jaac.2013.02.014
PMCID: PMC3813459
NIHMSID: NIHMS455004
PMID: 23622852
Dr. Christian Hopfer, M.D., Dr. Stacy Salomonsen-Sautel, Ph.D., Dr. Susan Mikulich-Gilbertson, Ph.D., Dr. Sung-Joon Min, Ph.D., Dr. Matt McQueen, Ph.D., Dr. Thomas Crowley, M.D., Dr. Susan Young, Ph.D., Dr. Robin Corley, Ph.D., Dr. Joseph Sakai, M.D., Dr. Christian Thurstone, M.D., Dr. Analice Hoffenberg, M.D., M.S.P.H., Dr. Christie Hartman, Ph.D., and Dr. John Hewitt, Ph.D.
Author information Copyright and License information PMC Disclaimer
The publisher's final edited version of this article is available at J Am Acad Child Adolesc Psychiatry
Associated Data
- Supplementary Materials
Abstract
Objective
To examine the influence of conduct disorder (CD) on substance initiation.
Method
Community adolescents without CD (n= 1165, mean baseline age= 14.6), with CD (n= 194, mean baseline age= 15.3), and youth with CD recruited from treatment (n=268, mean baseline age= 15.7) were prospectively followed and re-interviewed during young adulthood (mean ages at follow-up respectively: 20, 20.8, and 24). Young adult retrospective reports of age of substance initiation for 10 substance classes were analyzed using Cox regression analyses. Hazard ratios of initiation for the CD cohorts (community without CD as the reference) at ages 15, 18, and 21 were calculated, adjusting for baseline age, gender, and race.
Results
Among community subjects, CD was associated with elevated adjusted hazards for initiation of all substances, with comparatively greater hazard ratios of initiating illicit substances at age 15. By age 18, the adjusted hazard ratios remained significant except for alcohol. At age 21, the adjusted hazard ratios were significant only for cocaine, amphetamines, inhalants and club drugs. A substantial portion of community subjects without CD never initiated illicit substances. Clinical youth with CD demonstrated similar patterns, with comparatively larger adjusted hazard ratios.
Conclusions
CD confers increase risk for substance initiation across all substance classes at age 15 with greater relative risk for illicit substances compared to licit substances. This effect continues until age 18, with the weakest effect for alcohol. It further diminishes for other substances by 21; although, the likelihood of initiating cocaine, amphetamines, inhalants and club drugs among those who have not initiated yet continues to be highly elevated by 21.
Keywords: conduct disorder, substance use disorders
Initiation of substance use is the first step along a multi-step pathway that may eventually lead to the development of substance use disorders (SUD). Early age of initiation of substance use has been associated with a greater risk of eventually developing a SUD for alcohol, tobacco, and illicit substances1–9. The association between early age of onset of substance initiation and the later development of SUD has important theoretical as well as public health implications. There has been substantial debate over whether early age of substance initiation is a causal risk factor for developing later SUD or whether it acts as a marker for correlated factors that confer the risk for developing SUD3. Some researchers, primarily studying the influence of alcohol use on adolescents have posited that early alcohol use alters adolescent brain development and developmental trajectories2,10–12. An alternative hypothesis has emphasized that early age of substance initiation is a noncausal risk factor for the later development of SUD that acts as a marker for a range of correlated adolescent problem behaviors which themselves confer the risk for later SUD4,13. These researchers have focused on the observation that adolescent antisocial behaviors, novelty seeking, early initiation of substance use are highly correlated and have emphasized that early substance use is a marker for correlated externalizing behaviors which confer risk for the development of SUD14–16.
From a public health perspective, if early age of initiation is a causal risk factor for later development of SUD, an implication would be that delaying the onset of substance use could result in a reduction in the number of persons who eventually develop a SUD. If the alternative hypothesis is correct, i.e., that age of initiation is a noncausal risk factor, this would imply that prevention programs may need to more broadly target a range of problematic adolescent behaviors, including antisocial behaviors, to reduce the development of SUD.
One class of problem behaviors consistently associated with early onset of substance use is conduct disorder (CD). CD is a syndrome characterized by aggressive behaviors, truancy, violating social norms, and lying (APA)17. Epidemiological studies have shown that youth with CD initiate substances early, have elevated rates of substance use, and elevated rates of SUD18,19. National surveys of adult populations have consistently demonstrated that persons with adult antisocial personality disorder, which requires adolescent CD as a precursor, have high rates of SUD as well as more severe SUD20,21. Most prospective studies examining the association between CD and substance use have focused on the initiation of commonly used substances, such as alcohol, tobacco, or marijuana22–24. Population-based epidemiologic surveys of adolescents have demonstrated that by 12th grade, 71%, 42%, 43%, and 25% have tried alcohol, cigarettes, marijuana, or another illicit substances during their lifetime25. Although CD has been associated with an earlier rate of initiation of substances, to our knowledge, whether the effect of CD on initiation varies by substance class has not been examined. In particular, since CD is associated with a pattern of violating social norms, we hypothesized that youth with CD may be comparatively at greater risk for initiating substances that are illicit. Alcohol and tobacco, although illicit substances for underage youth, are frequently used by adolescents and are legal to use once persons become 21 and 18, respectively. Marijuana, though once considered illicit for all ages, is also commonly used; in 18 states marijuana is now available through a medical recommendation26 and was recently legalized in two states for adult recreational consumption. Other illicit substances are experimented with by proportionally fewer adolescents25. The goals of this study were to: 1) examine the influence of CD on substance initiation, across substance classes; 2) to explore whether CD is associated with an earlier age of initiation across all substances; and 3) to examine whether CD conferred relatively greater hazards of initiating substances. We hypothesized that youth with CD are more likely to have ever used a substance, have a younger age of substance experimentation, and have greater hazards for substance initiation at earlier ages compared with youth without CD, and that these hazard ratios would be comparatively greater for illicit as compared to licit substances
Method
Sample
All subjects in this study were assessed at two time points, adolescence and young adulthood. Subjects were originally recruited from three sources: 1) the Colorado Twin Registry (CTR; n = 1,246); 2) a clinical sample of adolescents recruited from a treatment program for youth with serious conduct and substance problems (n = 280); and 3) an additional community sample of adolescents (n = 156) who were, at baseline, matched to the clinical sample on age, race/ethnicity, gender, and ZIP code of residence. Follow-up rates for the CTR are 86.1%, the clinical sample 68.7%, and the matched community sample 70.6%. The CTR includes a Community Twin Sample, recruited through the Colorado Department of Public Health and Environment and Colorado school districts as well as a Longitudinal Twin Sample, recruited through the Colorado Department of Vital Statistics. For this study, one twin was randomly selected from each twin pair. The second wave of data collection for both clinical and community subjects was completed between 2002–2008. Details for the subjects recruited from treatment for substance use disorders and the community sample have been described in detail elsewhere27.
These samples were divided into three groups: Community subjects (from either the twin registry or the additional community sample) who did not meet criteria for CD at either wave 1 or 2 (n =1165), community subjects who met the criteria for CD at either wave 1 or 2 (n = 194), and clinical subjects who met the criteria for CD at either wave 1 or 2 (n = 268).
Measures
For this study, self-report CD was used. A participant was considered to have a CD diagnosis if they endorsed 3 or more adolescent symptoms of CD reported concurrently (i.e., queried in adolescence) or retrospectively (i.e., queried in young adulthood). We utilized this approach because some subjects may have developed CD after initial assessment. During adolescence, CD diagnosis was obtained through a structured interview by a trained interviewer, using either the paper-pencil DSM-III-R or DSM-IV Diagnostic Interview Schedule for Children (DISC)28, or the DSM-IV Diagnostic Interview Schedule (DIS)29 for participants ages 18 and over. During young adulthood, the CD diagnosis was similarly obtained from self-report by a trained interviewer utilizing the computerized DSM-IV DIS.
Ages of initiation for 10 classes of substances as classified by the Composite International Diagnostic Interview–Substance Abuse Module (CIDI-SAM)30, which includes tobacco, alcohol, marijuana, cocaine, amphetamines, sedative, inhalants, hallucinogens, opiates, and club drugs, were extracted from a supplemental questionnaire. This questionnaire asks about ever using any substance, age of initial and regular use, typical pattern of use, and recency of use31. For these analyses, the outcome variables were based upon the answers to the question, “How old were you the first time you used [drug],” which was separately asked for each of the 10 substances the subject endorsed ever using in his/her lifetime. To assist subjects in accurately recalling past events a timeline32 approach was used as a guide. Subjects were asked to recall memorable events as well as their living arrangements and this was used to anchor developmental timepoints.
Analyses
One-way analysis of variances (ANOVAs; or a non-parametric alternative) and Pearson chi-square analyses were completed to compare CD group differences on demographics and ever using the 10 different substances. Analyses of covariances (ANCOVAs) were completed to compare CD group differences on initiation ages in the subset who had ever used each substance with covariates to control for baseline age, gender, and race (white versus nonwhite). Scheffé’s post hoc multiple comparison procedures were completed for both the ANOVAs and ANCOVAs to control for the type 1 experimentwise error rate. To determine the influence of CD on age of substance initiation among community and clinical youth, Kaplan-Meier plots were produced and Cox regression models were analyzed. Kaplan-Meier survival curves were plotted to visually compare the unadjusted effects of the three CD groups on time to onset of first substance use. Cox regression models, adjusting for covariates (baseline age, sex, and race) and, when necessary, incorporating the CD group by time interaction (non-proportional hazards), assessed the influence of CD group on initiation of 10 substances. The hazard ratios and confidence limits are calculated at ages 15, 18, and 21. The community without CD group is the reference group in the Cox regression models. All analyses were completed in SAS 9.3. To graphically display the adjusted hazard ratios, two line graphs were created to display the adjusted hazard ratios at each age between 12–22 years old for each of the 10 substances.
Results
Ever Using a Substance
Table 1 describes the baseline demographic characteristics of each group, their mean CD symptoms at wave 1, their mean age at initial assessment and wave 2 follow-up, as well as the number and percent of subjects who initiated a particular substance. Compared to those without, community subjects with CD were more likely to be male, nonwhite, and had more baseline CD symptoms. They were also older at baseline and at wave 2. Clinical subjects with CD showed a greater contrast to community subjects without CD.
Table 1
Demographic Description and Percent Ever Used Each Substance by Group
| Community without CD n = 1,165 | Community with CD n = 194 | Patients with CD n = 268 | Statistic | |
|---|---|---|---|---|
| Baseline Age, m (SD) | 14.59 (2.12) | 15.32 (2.07)a | 15.73 (1.21)b | F2,1624=41.19* |
| Gender (Male), n (%) | 558 (47.9) | 143 (73.7) | 246 (91.8) | χ22=194.33* |
| Race\Ethnicity (White), n (%) | 998 (85.7) n = 1164 | 148 (76.3) | 155 (57.8) | χ22=107.96* |
| Baseline CD Symptoms, m (SD) | 0.53 (0.68) | 3.02 (2.07) | 6.23 (2.52) n = 267 | Kruskal-Wallis Test, χ22=912.41* |
| Age at Wave 2, m (SD) | 19.99 (2.77) | 20.78 (2.62)a | 24.04 (2.82)b | F2,1624=234.17* |
| Ever used alcohol, n (%) | χ22=88.49* | |||
| Yes | 933 (80.1) | 186 (95.9) | 268 (100.0) | |
| No | 232 (19.9) | 8 (4.1) | 0 (0.0) | |
| Ever used tobacco, n (%) | χ22=230.30* | |||
| Yes | 632 (54.2) | 167 (86.1) | 264 (98.5) | |
| No | 533 (45.8) | 27 (13.9) | 4 (1.5) | |
| Ever used marijuana, n (%) | χ22=281.67* | |||
| Yes | 559 (48.0) | 160 (82.9) | 265 (98.9) | |
| No | 605 (52.0) n = 1164 | 33 (17.1) n = 193 | 3 (1.1) | |
| Ever used cocaine, n (%) | χ22=654.96* | |||
| Yes | 107 (9.2) | 72 (37.3) | 222 (82.8) | |
| No | 1,058 (90.8) | 121 (62.7) n = 193 | 46 (17.2) | |
| Ever used amphetamines, n (%) | χ22=477.87* | |||
| Yes | 105 (9.0) | 65 (33.7) | 185 (69.0) | |
| No | 1,060 (91.0) | 128 (66.3) n = 193 | 83 (31.0) | |
| Ever used club drugs, n (%) | χ22=392.52* | |||
| Yes | 75 (6.4) | 56 (29.0) | 149 (55.8) | |
| No | 1,090 (93.6) | 137 (71.0) n = 193 | 118 (44.2) | |
| Ever used sedatives, n (%) | χ22=163.62* | |||
| Yes | 54 (4.6) | 39 (20.2) | 79 (29.5) | |
| No | 1,111 (95.4) | 154 (79.8) n = 193 | 189 (70.5) | |
| Ever used opiates, n (%) | χ22=161.14* | |||
| Yes | 143 (12.3) | 64 (33.2) | 118 (44.0) | |
| No | 1,022 (87.7) | 129 (66.8) n = 193 | 150 (56.0) | |
| Ever used inhalants, n (%) | χ22=186.30* | |||
| Yes | 14 (1.2) | 12 (6.2) | 58 (21.6) | |
| No | 1,151 (98.8) | 181 (93.8) n = 193 | 210 (78.4) | |
| Ever used hallucinogens, n (%) | χ22=504.15* | |||
| Yes | 125 (10.7) | 81 (42.0) | 199 (74.3) | |
| No | 1,040 (89.3) | 112 (58.0) n = 193 | 69 (25.7) | |
Open in a separate window
Note:
asignificant (p ≤ 0.05) post hoc comparisons between community with conduct disorder (CD) versus community without CD.
bsignificant (p ≤ 0.05) post hoc comparisons between patients with CD versus community without CD.
*p < 0.001
A greater proportion of the community with CD group had ever used each of the 10 substances compared with the community without CD group (p < 0.001). In turn, a greater proportion of the clinical with CD group had ever used each of the 10 substances compared with the community without CD group (p < 0.001). Kaplan-Meier curves (see Figures S1, S2, and S3, available online) show the pattern of initiation for each substance by CD group, for the unadjusted effects of CD on time to onset of first substance use.
Age of Initiating Substances
Table 2 describes the age of initiation of substances by group. ANCOVAs were not completed with tobacco and sedative onset ages due to heterogeneity of slopes/variances. The ANCOVAs on the subset who reported ever using each substance revealed significant CD group differences in onset age for all substances, except amphetamines and inhalants. The post hoc comparisons revealed that the community with CD group had younger initiation ages for alcohol, marijuana, cocaine, and hallucinogens compared with the community without CD group. In turn, the clinical with CD group had younger initiation ages for alcohol, marijuana, cocaine, hallucinogens, opiates, and club drugs compared with the community without CD group.
Table 2
Onset Age (Years) of Substance Initiation by Group
| Community without CD Adjusted Means | Community with CD Adjusted Means | Patients with CD Adjusted Means | Statistic | |
|---|---|---|---|---|
| Onset Age 1st Alcohol Use | 16.13 n = 933 | 14.59a n = 186 | 12.24b n = 268 | F2,1380=229.03* |
| Onset Age 1st Marijuana Use | 16.23 n = 559 | 14.68a n = 160 | 12.08b n = 265 | F2,977=261.67* |
| Onset Age 1st Cocaine Use | 18.46 n = 107 | 16.97a n = 72 | 15.61b n = 222 | F2,395=47.94* |
| Onset Age 1st Amphetamine Use | 16.45 n = 105 | 15.84 n = 65 | 16.02 n = 185 | F2,349=0.56 |
| Onset Age 1st Inhalant Use | 15.30 n = 15 | 15.20 n = 12 | 14.37 n = 59 | F2,80=1.37 |
| Onset Age 1st Hallucinogen Use | 17.79 n = 125 | 16.63a n = 81 | 14.79b n = 199 | F2,399=75.59* |
| Onset Age 1st Opiate Use | 17.77 n = 144 | 16.97 n = 64 | 16.78b n = 118 | F2,320=4.27** |
| Onset Age 1st Club Drug Use | 18.39 n = 76 | 17.46 n = 56 | 17.30b n = 149 | F2,275=5.01** |
Open in a separate window
Note:
asignificant (p ≤ 0.05) post hoc comparisons between community with conduct disorder (CD) versus community without CD, adjusting for covariates.
bsignificant (p ≤ 0.05) post hoc comparisons between patients with CD versus community without CD, adjusting for covariates.
*p < 0.0001;
**p ≤ 0.01
Ever Using a Substance
A greater proportion of the community sample with CD group had ever used each of the 10 substances compared with the community sample without CD group (p = 0.0005). In turn, a greater proportion of the clinical sample with CD group had ever used each of the 10 substances compared with the community sample without CD group (p = 0.0005). Kaplan-Meier curves (see Figures S1, S2, and S3, available online) show the pattern of initiation for each substance by group, for unadjusted effects of CD on time to onset of first substance use.
Relative Hazard of Initiating Substances
Table 3 reports adjusted hazard ratios of initiating for community subjects with CD versus those without for 10 substances. Across all substances, youth with CD had an elevated hazard at age 15 for initiating a substance. Hazard is a measure of the probability of initiating a substance at a specific time conditioned on the fact that an individual has not initiated yet. At age 15 community youth with CD had hazard ratios ranging from 3–11 for initiating illicit substances compared to 2–3 for initiating licit substances.
Table 3
Comparing Community Subjects With Conduct Disorder (CD) to Those Without on First Use of Different Substances.
| Substance | Adjusted Hazard Ratio at Age 15 95% Confidence Limits | Adjusted Hazard Ratio at Age 18 95% Confidence Limits | Adjusted Hazard Ratio at Age 21 95% Confidence Limits |
|---|---|---|---|
| Alcohol | 1.99 1.69, 2.36 | 1.27 0.97, 1.65 | 0.80 0.52, 1.24 |
| Tobacco | 2.10 1.75, 2.53 | 1.47 1.09, 1.97 | 1.02 0.64, 1.63 |
| Marijuana | 2.76 2.29, 3.34 | 1.52 1.13, 2.06 | 0.84 0.50, 1.41 |
| Cocaine | 9.38 5.91, 14.87 | 4.06 2.93, 5.63 | 1.76 1.0007, 3.10 |
| Amphetaminesa | 4.07 2.96, 5.61 | 4.07 2.96, 5.61 | 4.07 2.96, 5.61 |
| Club Drugs | 8.24 4.89, 13.89 | 4.42 3.07, 6.36 | 2.37 1.29, 4.34 |
| Sedatives | 10.60 5.65, 19.91 | 3.99 2.51, 6.33 | 1.50 0.70, 3.21 |
| Opiates | 3.73 2.53, 5.51 | 2.51 1.82, 3.46 | 1.69 0.99, 2.87 |
| Inhalantsa | 4.73 2.17, 10.32 | 4.73 2.17, 10.32 | 4.73 2.17, 10.32 |
| Hallucinogens | 6.26 4.20, 9.31 | 3.22 2.32, 4.45 | 1.65 0.89, 3.07 |
Open in a separate window
Note: All Cox regression models adjusted for baseline age, gender, and race. Bold-type numbers indicate significant results.
aCD group by time interactions were not significant so the hazard ratios were time-independent.
Similarly, at age 18, except for alcohol, the hazard ratios are significantly greater than one. In most models, except amphetamine onset age and inhalant onset age, the CD group by time interaction was significant, indicating that the hazards change differentially in proportion, over time for each CD group. After age 18, there is a leveling off effect with fewer youth initiating substances and there are fewer significant hazard ratios. By adulthood, these differences in initiation have equalized, in part, because youth with CD have already initiated substances. A similar pattern occurs when comparing clinical subjects with CD to community subjects without CD as seen in Table 4. The adjusted hazard ratios comparing each substance use onset are significant at ages 15 and 18; although, the hazard ratios are smaller and some are non-significant at older ages. Figures 1 and and22 display these hazard ratios for both community and clinical CD groups as a function of age.
Figure 1
Adjusted Hazard Ratios for First Use in Community with conduct disorder (CD) versus Community without CD. Note: Y axis label = Adjusted Hazard Ratio; X axis label = Age (in years).
Figure 2
Adjusted Hazard Ratios for First Use in Clinical with conduct disorder (CD) versus Community without CD. Note: Y axis label = Adjusted Hazard Ratio; X axis label = Age (in years).
Table 4
Comparing Clinical Subjects With Conduct Disorder (CD) to Community Subjects Without CD on First Use of Different Substances
| Substance | Adjusted Hazard Ratio at Age 15 95% Confidence Limits | Adjusted Hazard Ratio at Age 18 95% Confidence Limits | Adjusted Hazard Ratio at Age 21 95% Confidence Limits |
|---|---|---|---|
| Alcohol | 3.77 3.07, 4.63 | 1.64 1.14, 2.36 | 0.72 0.41, 1.25 |
| Tobacco | 3.74 2.96, 4.74 | 1.67 1.13, 2.45 | 0.74 0.42, 1.32 |
| Marijuana | 6.04 4.69, 7.78 | 1.67 1.02, 2.75 | 0.46 0.21, 1.02 |
| Cocaine | 43.43 28.71, 65.68 | 12.62 9.45, 16.85 | 3.67 2.27, 5.92 |
| Amphetaminesa | 11.16 8.41, 14.79 | 11.16 8.41, 14.79 | 11.16 8.41, 14.79 |
| Club Drugs | 21.98 13.51, 35.77 | 9.97 7.20, 13.81 | 4.52 2.73, 7.49 |
| Sedatives | 16.05 8.79, 29.32 | 5.55 3.70, 8.33 | 1.92 1.06, 3.47 |
| Opiates | 5.85 4.05, 8.43 | 3.01 2.27, 3.99 | 1.55 1.01, 2.37 |
| Inhalantsa | 18.71 9.77, 35.85 | 18.71 9.77, 35.85 | 18.71 9.77, 35.85 |
| Hallucinogens | 23.29 16.32, 33.24 | 4.02 2.80,5.78 | 0.69 0.34, 1.40 |
Open in a separate window
Note: All Cox regression models adjusted for baseline age, gender, and race. Bold-type numbers indicate significant results.
aCD group by time interactions were not significant so the hazard ratios were time-independent.
Discussion
The primary purpose of this study was to examine whether the influence of CD on substance initiation differed among 10 substance classes. As hypothesized, CD was associated with earlier substance initiation across all substances. In addition, greater hazard ratios were present for illicit substances compared to licit substances, particularly at age 15, in groups with CD compared with the community without CD group. Illicit substance hazard ratios at age 15 for community youth with CD were between 2.8 (for marijuana) up to 10.6 (for sedatives). Even greater hazard ratios were seen in the clinical sample, which exhibited CD with a greater mean severity than the community sample without CD (mean CD symptoms = 6.23 vs. 0.53, respectively). Although the hazard ratios for initiation diminish with age, hazards continue to remain elevated for a number of illicit substances until age 21. As seen in Figures 1 and and2,2, the hazard ratios for initiation of licit substance approach 1 by young adulthood, reflecting the fact that the majority of youth eventually initiate the substance. However, for many illicit substances, the hazards remain elevated, even into young adulthood. In conclusion, youth with CD are more likely to ever experiment with a substance, particularly an illicit one, have an earlier age of initiating, as well as have an increased hazard of using a substance at earlier ages compared with youth without CD.
To our knowledge, this is the first comparative analysis examining the influence of CD across 10 substance classes. One important result, beyond demonstrating that CD is associated with earlier age of initiation and ever initiating, is that youth with CD compared with youth without CD have larger hazard ratios for initiating illicit substances which extends into young adulthood. These findings are important because individuals who never initiate a substance are protected from developing a substance use disorder for that substance. Adolescents who do not exhibit CD are unlikely to experiment with illicit substances (except for marijuana), and thus are protected. By contrast, youth with CD not only initiate earlier, they are more likely to ever initiate illicit substances, and thus, have a greater possibility for developing a SUD upon substances that they have initiated.
Limitations of the study include reliance on only self-report for a CD diagnosis, use of retrospective reports of substance initiation, and that the clinical sample was recruited from substance abuse treatment, potentially biasing these results. While a timeline approach was used to aid in accuracy of recall33, young adult age-of-onset reports typically suffer from a “telescoping” effect 32, however, any such effect is likely to be similar across groups. The clinical sample was included primarily to examine the influence of more severe CD on initiation of less commonly used, illicit substances, however, since this sample was recruited from substance abuse treatment this creates a confound and results should be interpreted cautiously.
Although this study does not specifically examine whether early initiation of a substance is a causal risk factor for the later development of substance use disorder or whether early initiation acts as a marker for correlated behaviors, one implication of these results is that intervention or prevention approaches targeting conduct disorder may influence substance experimentation and the eventual development of substance use disorders, particularly for illicit substances.
Clinical Guidance
Youth without conduct disorder are unlikely to experiment with illicit substances with the exception of Marijuana.
Youth with conduct disorder are at elevated risk for experimenting with all substances, particularly illicit ones, and this elevated risk persists into young adulthood.
Supplementary Material
01
Figure S1. Kaplan-Meier survival plots comparing patients and community conduct disorder groups on initiation of (A) alcohol, (B) tobacco, and (C) marijuana use.
Click here to view.(862K, tif)
02
Figure S2. Kaplan-Meier survival plots comparing patients and community conduct disorder groups on initiation of (A) cocaine, (B) amphetamine, and (C) club drug use.
Click here to view.(778K, tif)
03
Figure S3. Kaplan-Meier survival plots comparing patients and community conduct disorder groups on initiation of (A) sedative, (B) optiate, (C) inhalant, and (D) hallucinogen use.
Click here to view.(867K, tif)
Acknowledgments
This research was funded by the National Institute on Drug Abuse (DA-011015, DA-021913, DA-01284). This research was also funded by the National Institutes of Health grants T32 MH015442 (A.H.), 5T32AA007464 (S.S.-S.), and R01DA031761 and DA011015 (J.S.); and the Kane Family Foundation (J.S.).
The authors thank the many subjects who participated in this research.
Footnotes
Supplemental material cited in this article is available online.
Clinical guidance is available at the end of this article.
Disclosure: Dr. Sakai has received reimbursem*nt for completing a policy review for the WellPoint Office of Medical Policy and Technology Assessment, WellPoint, Inc., Thousand Oaks, CA. He has served as a board member of the Addiction Research and Treatment Services Foundation. Drs. Hopfer, Salomonsen-Sautel, Mikulich-Gilbertson, Min, McQueen, Crowley, Young, Corley, Thurstone, Hoffenberg, Hartman, and Hewitt report no biomedical financial interests or potential conflicts of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Contributor Information
Dr. Christian Hopfer, University of Colorado Anschutz Medical Campus.
Dr. Stacy Salomonsen-Sautel, University of Colorado Anschutz Medical Campus.
Dr. Susan Mikulich-Gilbertson, University of Colorado Anschutz Medical Campus.
Dr. Sung-Joon Min, University of Colorado Anschutz Medical Campus.
Dr. Matt McQueen, Institute for Behavioral Genetics at the University of Colorado at Boulder.
Dr. Thomas Crowley, University of Colorado Anschutz Medical Campus.
Dr. Susan Young, University of Colorado Anschutz Medical Campus.
Dr. Robin Corley, Institute for Behavioral Genetics at the University of Colorado at Boulder.
Dr. Joseph Sakai, University of Colorado Anschutz Medical Campus.
Dr. Christian Thurstone, Denver Health and Hospital Authority.
Dr. Analice Hoffenberg, University of Colorado Anschutz Medical Campus.
Dr. Christie Hartman, University of Colorado Anschutz Medical Campus.
Dr. John Hewitt, Institute for Behavioral Genetics at the University of Colorado at Boulder.
References
1. Grant BF, Dawson DA. Age at onset of alcohol use and its association with DSM-IV alcohol abuse and dependence: results from the National Longitudinal Alcohol Epidemiologic Survey. J Subst Abuse. 1997;9:103–110. [PubMed] [Google Scholar]
2. DeWit DJ, Adlaf EM, Offord DR, Ogborne AC. Age at first alcohol use: a risk factor for the development of alcohol disorders. Am J Psychiatry. 2000 May;157(5):745–750. [PubMed] [Google Scholar]
3. Prescott CA, Kendler KS. Age at first drink and risk for alcoholism: a noncausal association. Alcohol Clin Exp Res. 1999 Jan;23(1):101–107. [PubMed] [Google Scholar]
4. McGue M, Iacono WG, Legrand LN, Malone S, Elkins I. Origins and consequences of age at first drink. I. Associations with substance-use disorders, disinhibitory behavior and psychopathology, and P3 amplitude. Alcohol Clin Exp Res. 2001 Aug;25(8):1156–1165. [PubMed] [Google Scholar]
5. Grant BF, Dawson DA. Age of onset of drug use and its association with DSM-IV drug abuse and dependence: results from the National Longitudinal Alcohol Epidemiologic Survey. J Subst Abuse. 1998;10(2):163–173. [PubMed] [Google Scholar]
6. Chen CY, O’Brien MS, Anthony JC. Who becomes cannabis dependent soon after onset of use? Epidemiological evidence from the United States: 2000–2001. Drug Alcohol Depend. 2005 Jul;79(1):11–22. [PubMed] [Google Scholar]
7. Chen CY, Storr CL, Anthony JC. Early-onset drug use and risk for drug dependence problems. Addict Behav. 2009 Mar;34(3):319–322. [PMC free article] [PubMed] [Google Scholar]
8. Anthony JC, Petronis KR. Early-onset drug use and risk of later drug problems. Drug Alcohol Depend. 1995 Nov;40(1):9–15. [PubMed] [Google Scholar]
9. Janson H. Longitudinal patterns of tobacco smoking from childhood to middle age. Addict Behav. 1999 Mar-Apr;24(2):239–249. [PubMed] [Google Scholar]
10. Brown SA, Tapert SF. Adolescence and the trajectory of alcohol use: basic to clinical studies. Ann N Y Acad Sci. 2004 Jun;1021:234–244. [PubMed] [Google Scholar]
11. Tapert SF, Granholm E, Leedy NG, Brown SA. Substance use and withdrawal: neuropsychological functioning over 8 years in youth. J Int Neuropsychol Soc. 2002 Nov;8(7):873–883. [PubMed] [Google Scholar]
12. Jacobus J, Bava S, Cohen-Zion M, Mahmood O, Tapert SF. Functional consequences of marijuana use in adolescents. Pharmacol Biochem Behav. 2009 Jun;92(4):559–565. [PMC free article] [PubMed] [Google Scholar]
13. King SM, Iacono WG, McGue M. Childhood externalizing and internalizing psychopathology in the prediction of early substance use. Addiction. 2004 Dec;99(12):1548–1559. [PubMed] [Google Scholar]
14. Cosgrove VE, Rhee SH, Gelhorn HL, et al. Structure and etiology of co-occurring internalizing and externalizing disorders in adolescents. J Abnorm Child Psychol. 2011 Jan;39(1):109–123. [PMC free article] [PubMed] [Google Scholar]
15. Hicks BM, Iacono WG, McGue M. Consequences of an adolescent onset and persistent course of alcohol dependence in men: adolescent risk factors and adult outcomes. Alcohol Clin Exp Res. 2010 May;34(5):819–833. [PMC free article] [PubMed] [Google Scholar]
16. Young SE, Stallings MC, Corley RP, Krauter KS, Hewitt JK. Genetic and environmental influences on behavioral disinhibition. Am J Med Genet. 2000 Oct 9;96(5):684–695. [PubMed] [Google Scholar]
17. American-Psychiatric-Association. Diagnostic and Statistical Manual of Mental Disorders. 4. Washington, D.C: 2000. text rev. [Google Scholar]
18. Fergusson DM, Horwood LJ, Ridder EM. Conduct and attentional problems in childhood and adolescence and later substance use, abuse and dependence: results of a 25-year longitudinal study. Drug Alcohol Depend. 2007 Apr;88(Suppl 1):S14–26. [PubMed] [Google Scholar]
19. Costello EJ. Psychiatric predictors of adolescent and young adult drug use and abuse: what have we learned? Drug Alcohol Depend. Apr. 2007;88(Suppl 1):S97–99. [PubMed] [Google Scholar]
20. Hasin D, Kilcoyne B. Comorbidity of psychiatric and substance use disorders in the United States: current issues and findings from the NESARC. Current opinion in psychiatry. 2012 May;25(3):165–171. [PMC free article] [PubMed] [Google Scholar]
21. Compton WM, Conway KP, Stinson FS, Colliver JD, Grant BF. Prevalence, correlates, and comorbidity of DSM-IV antisocial personality syndromes and alcohol and specific drug use disorders in the United States: results from the national epidemiologic survey on alcohol and related conditions. J Clin Psychiatry. 2005 Jun;66(6):677–685. [PubMed] [Google Scholar]
22. Copeland WE, Angold A, Shanahan L, Dreyfuss J, Dlamini I, Costello EJ. Predicting persistent alcohol problems: a prospective analysis from the Great Smoky Mountain Study. Psychol Med. 2011 Dec 13;:1–11. [PMC free article] [PubMed] [Google Scholar]
23. Sung M, Erkanli A, Angold A, Costello EJ. Effects of age at first substance use and psychiatric comorbidity on the development of substance use disorders. Drug Alcohol Depend. 2004 Sep 6;75(3):287–299. [PubMed] [Google Scholar]
24. Elkins IJ, McGue M, Iacono WG. Prospective effects of attention-deficit/hyperactivity disorder, conduct disorder, and sex on adolescent substance use and abuse. Arch Gen Psychiatry. 2007 Oct;64(10):1145–1152. [PubMed] [Google Scholar]
25. Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE. Monitoring the Future national results on adolescent drug use: Overview of key findings, 2011. 2012. [Google Scholar]
26. Salomonsen-Sautel S, Sakai JT, Thurstone C, Corley R, Hopfer C. Medical marijuana use among adolescents in substance abuse treatment. J Am Acad Child Adolesc Psychiatry. 2012 Jul;51(7):694–702. [PMC free article] [PubMed] [Google Scholar]
27. Rhea SA, Gross AA, Haberstick BC, Corley RP. Colorado Twin Registry: An Update. Twin Res Hum Genet. 2012 Oct 23;:1–7. [PMC free article] [PubMed] [Google Scholar]
28. Sakai JT, Hopfer CJ, Hartman C, et al. Test of association between TaqIA A1 allele and alcohol use disorder phenotypes in a sample of adolescent patients with serious substance and behavioral problems. Drug Alcohol Depend. 2007 May 11;88(2–3):130–137. [PubMed] [Google Scholar]
29. Fergusson DM, Horwood LJ, Boden JM, Jenkin G. Childhood social disadvantage and smoking in adulthood: results of a 25-year longitudinal study. Addiction. 2007 Mar;102(3):475–482. [PubMed] [Google Scholar]
30. Cottler LB. The CIDI and CIDI-substance abuse module (SAM): cross-cultural instruments for assessing DSM-III, DSM-III-R and ICD-10 criteria. NIDA Res Monogr. 1990;105:220–226. [PubMed] [Google Scholar]
31. Crowley TJ, Mikulich SK, Ehlers KM, Whitmore EA, MacDonald MJ. Validity of structured clinical evaluations in adolescents with conduct and substance problems. J Am Acad Child Adolesc Psychiatry. 2001 Mar;40(3):265–273. [PubMed] [Google Scholar]
32. Parra GR, O’Neill SE, Sher KJ. Reliability of self-reported age of substance involvement onset. Psychol Addict Behav. 2003 Sep;17(3):211–218. [PubMed] [Google Scholar]
33. Hjorthoj CR, Hjorthoj AR, Nordentoft M. Validity of Timeline Follow-Back for self-reported use of cannabis and other illicit substances--systematic review and meta-analysis. Addict Behav. 2012 Mar;37(3):225–233. [PubMed] [Google Scholar]
