Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2010 Sep 4.
Published in final edited form as: Prev Med. 2008 Sep 26;47(6):629–634. doi: 10.1016/j.ypmed.2008.09.009

Improving Targets for the Prevention of Drug Use Disorders: Sociodemographic Predictors of Transitions Across Drug Use Stages in the National Comorbidity Survey Replication

Joel Swendsen 1, James C Anthony 2, Kevin P Conway 3, Louisa Degenhardt 4, Lisa Dierker 5, Meyer Glantz 3, Jianping He 6, Amanda Kalaydjian 6, Ronald C Kessler 7, Nancy Sampson 7, Kathleen R Merikangas 6
PMCID: PMC2933380  NIHMSID: NIHMS84815  PMID: 18926848

Abstract

Objectives

Models of drug use etiology and prevention require precise information concerning the expression of population-based risk factors across the continuum of drug use. However, the majority of previous epidemiologic research on this topic has not addressed transitions between specific drug stages. The present investigation examined the sociodemographic predictors of progression across six stages of drug use in the National Comorbidity Survey Replication (NCS-R), a nationally representative household survey of the U.S. population conducted between February, 2001 and April, 2003.

Methods

Lifetime history of opportunity to use illicit substances, initial drug use, and DSM-IV drug use disorders were collected using in-person structured diagnostic interviews.

Results

The median age of first opportunity to use drugs as well as drug use, abuse and dependence each occurred prior to age 20, while the median remission from abuse and dependence occurred at 26 and 30 years, respectively. Most sociodemographic variables, in particular sex and ethnicity, demonstrated highly differential associations with transitions depending on the stage examined.

Conclusions

The findings may partially explain the effectiveness of strategies designed to reduce drug use, abuse and dependence, and indicate that increased correspondence is needed between available epidemiologic data and existing models of etiology or prevention.

Keywords: Prevention, Substance-Related Disorders, Epidemiology, Demographic Factors

INTRODUCTION

Numerous drug prevention programs in the United States incorporate strategies to address psychosocial or demographic risk factors (Winters et al., 2007) based on findings from community samples of adolescents (Degenhardt et al., 2000; Konings et al., 1995; Lynskey et al., 1999; Melchior et al., 2007; Perkonigg, Pfister et al., 2006; von Sydow et al., 2002; Wallace et al., 2003) and adults (Anthony et al., 1994; Compton et al., 2007; Johnson & Gerstein, 1998; Kerr et al., 2007; Kessler et al., 2005; Medina-Mora et al., 2005; Oakley Browne et al., 2006; Warner et al., 1995). The age at which initial stages are reached is also strongly associated with abuse and dependence (Anthony, & Petronis, 1995; Chen et al., 2005; Grant & Dawson, 1998; McCabe et al., 2007) and has provided additional information for both selective and indicated prevention programs.

Although progress has been made particularly in reducing incident cases of disorder onset among drug users (see Cuijpers, 2003; Hawkins et al., 1992; Toumbourou et al., 2007), greater advances may be achieved by decomposing previously documented aggregate associations in an effort to pinpoint the precise transitions across the drug use trajectory that are most associated with specific risk factors. The few published epidemiologic studies on this topic confirm the significant role of certain sociodemographic characteristics (Anthony et al., 1994; Chen et al., 2005; Grant & Dawson, 1998; Wagner & Anthony, 2007; Stone et al., 2007), but also that other risk factors, including gender and ethnicity, may have markedly different effects depending on stage of use (Grant & Dawson, 1998; O’Brien & Anthony, 2005; Van Etten et al., 1999).

Detailed descriptive information concerning how sociodemographic risk factors vary across the full trajectory of drug use is therefore needed to adjust risk formulas and potentially improve the precision of prevention strategies. Using data from the National Comorbidity Survey Replication, the present investigation examines this question across six stages of drug involvement.

METHODS

NCS-R Sampling and Field Procedures

The NCS-R is a nationally representative face-to-face household survey of the prevalence and correlates of a wide range of DSM-IV mental disorders carried out between February, 2001 and April, 2003 (for details see Kessler, Berglund et al., 2004; Kessler & Merikangas, 2004). The sampling frame was English speaking adults selected using a multi-stage clustered area probability design from the civilian household population of the continental United States. The NCS-R procedures were approved by the human subjects committees of Harvard Medical School and the University of Michigan.

The NCS-R interview was carried out in two parts. The Part I interview was administered to a nationally representative household sample of 9282 respondents ages 18 and older. The response rate was 70.9% and informed consent was obtained from all participants. Part I included the core mental disorders assessed in the survey along with a battery of sociodemographic variables. Part II was administered to 5692 of the 9282 Part I respondents, including all Part I respondents with a lifetime core disorder plus a probability subsample of other respondents. The Part II sample was weighted to adjust for differential probabilities of within-household selection, differential probabilities of selection from the Part I sample into the Part II sample based on Part I responses, and for discrepancies between the sample and the US population on sociodemographic and geographic factors assessed in the 2000 Census.

Diagnostic Assessment and Drug Use Stages

The opportunity to use drugs, drug use, and DSM-IV diagnoses of abuse and dependence were assessed using version 3.0 of the World Health Organization Composite International Diagnostic Interview (CIDI; Kessler & Ustun, 2004). Good concordance was found between CIDI and SCID diagnoses for substance use disorders and for a wide range of anxiety, mood, and externalizing disorders assessed in the survey (Haro et al., 2006). The drug module of the CIDI was administered to the entire Part II sample. An initial screening question asked respondents if they ever used marijuana or hashish, cocaine in any form, prescription drugs either without the recommendation of a health professional or for any reason other than what a health professional said they should be used for, or other illicit drug. If the respondent endorsed use of any of these drugs, the interviewer then asked questions concerning age of first use, frequency of use, and presence of DSM-IV criteria for drug abuse (a repetitive pattern of drug use causing legal or interpersonal problems, increasing physical danger, or preventing fulfillment of major role obligations). Respondents were also asked about the frequency of such problems, the age of the respondent when he or she had the problem for the first and last time, and the number of years they had the problem. If respondents endorsed any problems related to drugs, they were also asked questions about DSM-IV drug dependence (symptoms of physical tolerance, withdrawal, uncontrolled drug use or inability to stop use, excessive time invested in order to acquire drugs, reduction in regular activities due to drugs, and continued drug use despite harmful or undesirable consequences), as well as age of first onset of symptoms and timing of most recent symptoms. All respondents were asked their age the first time they had the opportunity to use any illicit drug (regardless of type), defined as either being offered a drug or being present while others were using a drug and being able to use it if desired. Past users who no longer used drugs at the time of interview were asked their age of most recent use. Age of remission from drug abuse or dependence was defined as the age two years after the respondent most recently used drugs. An age-of-onset variable was examined for each of the six stages of drug use, and variables were created to represent the duration of transitions between each drug use stage.

Sociodemographic and drug classification variables

Sociodemographic variables include birth year cohort: 1972–1983, 1957–1971, 1942–1956, and 1941 or before; sex; race/ethnicity (Non-Hispanic White, Non-Hispanic Black, Hispanic, and Other); completed years of education [less than high school (0–11), high school (12), some college (13–15), and college graduate (16 or more)], student status (student vs. non-student); and marital status (never married, previously married, and currently married/cohabitating). Information collected for education, student status, and marital status was used to treat these variables as time-varying covariates in survival analyses. Drug classification variables included number of drugs and drug type.

Statistical Analyses

Estimated projections of the cumulative probability of drug use stages as of age 60 were obtained by the actuarial method implemented in PROC LIFETEST in SAS (version 9.1.3, SAS Institute, Cary, N.C.). The actuarial method is similar to the more familiar Kaplan-Meier method but handles ties in a more intuitive fashion. The associations between timing and speed of transitions between drug use stages were estimated in discrete-time multivariable logistic regression analyses with person-year as the unit of analysis (Efron, 1988). Standard errors and significance tests were estimated using the Taylor series linearization method (Wolter, 1985) implemented in SUDAAN to adjust for design effects (Research Triangle Institute, 2004). The person-year variable was defined as the number of years since onset of the earlier drug use category up to either age of onset of a later drug use stage or age at interview (for censored cases). Each data array began with year 0 (i.e., the same year as the earlier transition) to allow for the fact that respondents could make multiple transitions in the same year (e.g., start use and become an abuser in the same year as they first had an opportunity to use). Individuals who progress from one stage to the next within the same year were treated as having made two transitions in the same year. When predicting remission from drug abuse or dependence, the person-year variable was defined as the number of years since the onset of drug abuse or dependence, respectively, up to either age of remission of abuse or dependence or age at interview (for censored cases). Discrete-time survival models included sociodemographic variables, number of drugs and drug type as predictors of a given drug-related outcome, and assessed the influence of age of onset of each previous stage as well as speed of transition between stage categories. Due to existence of linear combinations among these predictors, it was not possible to simultaneously test for their effects within the same model. We therefore ran multiple preliminary models with different combinations of age of onset and speed of transition variables, and then selected the variables that were most consistently associated with the outcome for use in the final models.

RESULTS

Prevalence and Age of Onset of Drug Use Stages

Table 1 presents the lifetime prevalence of drug use stages in the Part II sample (n=5692). The prevalence of drug use, abuse and dependence, as well as remission from abuse and dependence, was generally highest among individuals born between 1957 and 1971. Concerning conditional prevalence rates, this cohort was also the most likely to have used drugs if given the opportunity. Individuals in the oldest cohort were the least likely to use drugs if they had the opportunity, or to abuse drugs if they were drug users, but a greater portion of these individuals developed dependence if they abused drugs.

Table 1.

Lifetime prevalence of classes of drug use, disorders and remission for Part II sample (n=5692). U.S. National Comorbidity Survey Replication (NCS-R) data collected between February 2001 and April 2003

N Birth Cohort Total
1972–1983 1957–1971 1942–1956 1941 or before
n % (SE) n % (SE) n % (SE) n % (SE) n % (SE)
Unconditional
 Opportunity to use drugs 4428 1371 88.5 (1.1) 1826 87.9 (1.2) 1521 76.5 (1.3) 974 28.5 (1.6) 5692 72.4 (0.8)
 Use of any illicit drug 2940 1371 54.2 (1.9) 1826 60.6 (1.9) 1521 47.1 (1.9) 974 7.4 (1.0) 5692 44.2 (1.1)
 Abuse of any illicit drug 651 1371 11.1 (0.9) 1826 12.1 (1.0) 1521 6.8 (0.7) 974 0.3 (0.1) 5692 8.0 (0.4)
 Dependence on any illicit drug 248 1371 3.9 (0.5) 1826 5.1 (0.7) 1521 2.3 (0.4) 974 0.2 (0.1) 5692 3.1 (0.2)
 Remission from drug abuse 344 1371 4.7 (0.5) 1826 6.3 (0.6) 1521 4.3 (0.5) 974 0.1 (0.1) 5692 4.1 (0.3)
 Resmission from drug dependence 212 1371 2.9 (0.4) 1826 4.9 (0.6) 1521 2.2 (0.4) 974 0.2 (0.1) 5692 2.6 (0.2)
Conditional
 Use given opportunity 2940 1232 61.2 (1.8) 1670 68.8 (1.8) 1220 61.5 (2.3) 306 25.8 (3.2) 4428 61.0 (1.3)
 Abuse given use 651 824 20.5 (1.6) 1218 20.0 (1.6) 806 14.5 (1.4) 92 3.9 (2.0) 2940 18.0 (0.9)
 Dependence given abuse 248 208 35.2 (3.6) 291 42.6 (3.5) 146 33.9 (3.7) 6 65.4 (21.0) 651 38.4 (2.2)
 Remission given abuse 344 134 65.9 (4.4) 166 91.4 (2.2) 100 95.7 (2.4) 3 100.0 (0.0) 403 83.7 (1.9)
 Remission given dependence 212 74 73.4 (5.3) 125 90.9 (3.4) 46 94.1 (3.0) 3 100. 0 (0.0) 248 86.4 (2.8)
Open in a new tab

Age-of-onset distributions of Transitions among Drug Use Stages

Figure 1 illustrates that the median age of first opportunity was 16 years and the median onset for the subsequent stages of use, abuse and dependence each occurred prior to age 20. The median age of remission was 26 for respondents with abuse, and 30 for those with dependence. As illustrated in Figure 2, the median delay between first opportunity and any illicit drug use was one year. Half of drug users continued to use illicit substances for three years before abuse occurred, but the speed of transition from abuse to dependence took less than one year. Drug abuse persisted for five years until remission occurred, while dependence persisted for seven years.

Figure 1.

Figure 1

Open in a new tab

Age of onset of drug use stages. U.S. National Comorbidity Survey Replication (NCS-R) data collected between February 2001 and April 2003

Figure 2.

Figure 2

Open in a new tab

Speed of transition between drug use stages. U.S. National Comorbidity Survey Replication (NCS-R) data collected between February 2001 and April 2003

Sociodemographic predictors of Transitions among Drug Use Stages

The results of the discrete-time logistic regression analyses for opportunity to use and transitions to drug use, abuse or dependence are presented in Table 2. Despite the important association of birth cohort with the early stages of opportunity and use of drugs, no significant associations were observed by cohort in the transition from use to abuse, or from abuse to dependence. Women were less likely to have the opportunity to use drugs or to make the transition to using or abusing drugs, but women meeting criteria for drug abuse were more likely to become dependent. ‘Other’ minority races were less likely than Whites to have the opportunity to use drugs, but these same individuals were more than twice as likely to make the transition from drug use to abuse. Hispanics and Blacks were less likely than Whites to use drugs when given the opportunity. Compared to individuals with the highest education, all other education levels had a probability three times greater of having the opportunity to use drugs. Although significant odds ratios were also observed for lower education levels in the transition from drug use to abuse, only current student status was associated with greater probability of using drugs when given the opportunity. Concerning marital status, never-married respondents were more likely to have the opportunity to use drugs, as well as to continue on to drug use and abuse stages. Those who were formerly married were more likely to use drugs when given the opportunity, and more likely to transition from use to abuse. An earlier age of first drug use and a shorter time between opportunity to use and first drug use were both associated with an increased probability of developing abuse. In comparison to cannabis, the likelihood of transitioning to drug abuse was increased for cocaine, extra-medical drugs or other illicit drugs. In turn, cocaine and other illicit drugs also increased the probability of transition from abuse to dependence.

Table 2.

Sociodemographic predictors of first drug use opportunity and transitions to use, abuse and dependence. U.S. National Comorbidity Survey Replication (NCS-R) data collected between February 2001 and April 2003

1st opportunity to use in Part II of sample 1st drug use among those having opportunity Drug abuse among drug users Drug dependence among drug abusers
Predictors Category OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
Person year INT† 1.07* 1.06 – 1.07 0.76* 0.73 – 0.79 0.89* 0.87 – 0.91 0.78* 0.70 – 0.87
X21 [p] 213.2 [.000] 250.3 [.000] 131.0 [.000] 20.4 [.000]
Birth Cohort 1972–1983 27.46* 20.74 – 36.35 3.00* 2.15 – 4.18 2.48 0.79 – 7.83 0.41 0.11 – 1.49
1957–1971 24.11* 18.21 – 31.92 3.26* 2.35 – 4.53 1.94 0.65 – 5.81 0.43 0.13 – 1.45
1942–1956 12.36* 9.32 – 16.38 2.95* 2.06 – 4.24 1.81 0.59 – 5.57 0.38 0.12 – 1.18
1941 or before 1 --- 1 --- 1 --- 1 ---
X23 [p] 714.0 [.000] 56.4 [.000] 4.5 [.214] 3.3 [.350]
Sex Female 0.68* 0.62 – 0.74 0.81* 0.69 – 0.95 0.62* 0.52 – 0.73 1.51* 1.11 – 2.07
Male 1 --- 1 --- 1 --- 1 ---
X21 [p] 76.9 [.000] 7.4 [.007] 32.6 [.000] 7.2 [.007]
Race-ethnicity Hispanic 0.86 0.74 – 1.00 0.81* 0.66 – 1.00 0.91 0.62 – 1.34 0.78 0.35 – 1.73
Black 0.98 0.87 – 1.10 0.73* 0.60 – 0.89 0.96 0.71 – 1.30 1.31 0.81 – 2.10
Other 0.62* 0.48 – 0.79 0.98 0.68 – 1.40 2.09* 1.32 – 3.31 1.84 0.94 – 3.61
White 1 --- 1 --- 1 --- 1 ---
X23 [p] 17.4 [.001] 13.3 [.004] 13.3 [.004] 5.1 [.163]
Education Student 3.01* 2.16 – 4.20 1.54* 1.16 – 2.06 2.03* 1.24 – 3.35 1.21 0.56 – 2.61
Low 3.14* 1.95 – 5.06 1.31 0.93 – 1.84 2.79* 1.69 – 4.63 1.52 0.70 – 3.31
Low/Med 3.06* 1.98 – 4.74 1.14 0.86 – 1.52 1.91* 1.11 – 3.30 0.99 0.51 – 1.90
Medium 3.31* 2.25 – 4.89 0.92 0.69 – 1.22 2.06* 1.12 – 3.79 0.82 0.41 – 1.63
High 1 --- 1 --- 1 --- 1 ---
X24 [p] 54.3 [.000] 18.6 [.001] 20.2 [.000] 7.4 [.117]
Marital status Never 7.46* 5.42 – 10.25 1.51* 1.28 – 1.80 1.89* 1.38 – 2.58 1.51 0.80 – 2.83
Formerly 1.02 0.74 – 1.41 2.64* 1.92 – 3.63 2.26* 1.36 – 3.77 1.83 0.83 – 4.01
Currently 1 --- 1 --- 1 --- 1 ---
X22 [p] 167.4 [.000] 46.0 [.000] 24.2 [.000] 2.5 [.291]
Additional Predictors Age of 1st opportunity --- --- 1.00 0.98 – 1.01 --- --- --- ---
X21 [p] --- --- 0.4 [.532] --- --- --- ---
Age of 1st drug used --- --- --- --- 0.94* 0.92 – 0.97 --- ---
X21 [p] --- --- --- --- 23.6 [.000] --- ---
Time: opp. to use --- --- --- --- 0.92* 0.87 – 0.98 --- ---
X21 [p] --- --- --- --- 8.4 [.004] --- ---
Age of drug abuse --- --- --- --- --- --- 1.00 0.93 – 1.07
X21 [p] --- --- --- --- --- --- .01 [.903]
Time: opp. to abuse --- --- --- --- --- --- 0.99 0.88 – 1.12
X21 [p] --- --- --- --- --- --- .04 [.843]
Time: use to abuse --- --- --- --- --- --- 1.05 0.94 – 1.17
X21 [p] --- --- --- --- --- --- .8 [.384]
Number drugs used: 1 --- --- --- --- 1 --- 1 ---
Number drugs >=2 --- --- --- --- 0.87 0.49 – 1.55 1.39 0.61 – 3.16
X21 [p] --- --- --- --- 0.2 [.628] 0.6 [.421]
Drug: Cannabis --- --- --- --- 1 --- 1 ---
Drug: Cocaine --- --- --- --- 5.88* 4.08 – 8.49 2.20* 1.17 – 4.15
Drug: Extra-medical --- --- --- --- 3.61* 2.38 – 5.48 1.16 0.69 – 1.96
Drug: other --- --- --- --- 5.37* 3.73 – 7.74 1.97* 1.12 – 3.47
X23 [p] --- --- --- --- 140.4 [.000] 13.3 [.004]
Open in a new tab
†

person-year indicator (INT) is based on chronological age for first opportunity to use drugs, and years since last stage for drug use, abuse and dependence.

*

OR significant at .05 level, two-tailed test

Whereas the three younger cohorts were generally more likely to remit from drug abuse or dependence (Table 3), this effect was particularly salient for individuals born between 1972 and 1983. Women were also more likely than men to achieve remission from both disorders. However, the likelihood of remission from either drug abuse or dependence was unrelated to race-ethnicity, marital status, or education. An older age of drug abuse was associated with increased probability of remission from this disorder. Finally, in comparison with cannabis, the use of cocaine and extra-medical drugs were associated with a decreased probability of remission from drug abuse. No specific drug type was associated with remission from drug dependence.

Table 3.

Sociodemographic predictors of remission from drug abuse and dependence. U.S. National Comorbidity Survey Replication (NCS-R) data collected between February 2001 and April 2003

Remission from Drug Abuse w/o Dependence Remission from Drug Dependence
Predictors Category OR (95% CI) OR (95% CI)
Person year INT† 1.03* 1.01 – 1.06 1.11* 1.08 – 1.15
X21 [p] 8.1 [.005] 47.3 [.000]
Birth Cohort 1972–1983 7.41* 4.33 – 12.68 13.31* 3.77 – 46.95
1957–1971 3.01* 2.15 – 4.22 3.77* 1.36 – 10.40
1942–1956 1.93* 1.43 – 2.59 1.32 0.52 – 3.36
1941 or before 1 --- 1 ---
X23 [p] 57.8 [.000] 48.0 [.000]
Sex Female 1.53* 1.10 – 2.11 1.58* 1.14 – 2.20
Male 1 --- 1 ---
X21 [p] 6.9 [.009] 8.0 [.005]
Race-ethnicity Hispanic 0.84 0.54 – 1.31 0.66 0.27 – 1.65
Black 0.64 0.40 – 1.02 0.67 0.40 – 1.12
Other 1.24 0.72 – 2.13 1.43 0.80 – 2.55
White 1 --- 1 ---
X23 [p] 5.9 [.116] 4.1 [.249]
Education Student 0.66 0.38 – 1.15 0.48* 0.23 – 0.99
Low 0.83 0.50 – 1.39 0.74 0.38 – 1.43
Low/Med 0.86 0.53 – 1.38 0.90 0.50 – 1.61
Medium 1.28 0.79 – 2.09 0.89 0.51 – 1.54
High 1 --- 1 ---
X24 [p] 9.8 [.045] 5.2 [.269]
Marital status Never 1 0.73 – 1.36 1.38 0.87 – 2.19
Formerly 1.28 0.86 – 1.91 1.17 0.72 – 1.89
Currently 1 --- 1 ---
X22 [p] 1.7 [.434] 2.0 [.367]
Additional Predictors Age of drug abuse 1.08* 1.04 – 1.13 --- ---
X21 [p] 6.3 [.012] --- ---
Time: drug use to abuse 0.99 0.95 – 1.03 --- ---
X21 [p] 0.2 [.624] --- ---
Age of drug dependence --- --- 0.97 0.93 – 1.01
X21 [p] --- --- 2.1 [.153]
Time: drug abuse to dependence --- --- 0.99 0.94 – 1.04
X21 [p] --- --- 0.3 [.581]
Number drugs used: 1 1 --- --- ---
Number drugs >=2 2.10 0.39 – 11.22 --- ---
X21 [p] 0.8 [.371] --- ---
Drug: Cannabis 1 --- 1 ---
Drug: Cocaine 0.14* 0.04 – 0.45 0.33 0.07 – 1.54
Drug: Extra-medical 0.13* 0.04 – 0.50 0.33 0.07 – 1.65
Drug: other 0.24* 0.07 – 0.84
X23 [p] | X22 [p] 15.4 [.002] 6.1 [.047]
Open in a new tab
†

person-year indicator (INT) is based on chronological age for first opportunity to use drugs, and years since last drug use stage

*

OR significant at .05 level, two-tailed test

DISCUSSION

The most common sociodemographic risk factors identified through aggregate analyses in epidemiologic investigations were used to predict transitions across specific stages of drug use. Birth cohort was significantly associated with transitions to having the opportunity to use drugs, to initial drug use, and to remission from drug use disorders. However, it was not associated with the risk of abuse among drug users, or with dependence among drug abusers. The effects of education and marital status were linked primarily to transitions among the first three stages. By contrast, transitions to abuse, dependence or remission were significantly associated with drug type or history of drug use.

Additional important findings concern the manner in which some variables exerted qualitatively different effects depending on stage of use. In particular, female drug abusers were significantly more likely than males to progress from drug abuse to dependence. This finding is opposite to investigations reporting a reduced risk of drug dependence among women when estimating such risk relative to the overall population (Compton et al., 2007; Oakley Browne et al., 2006; Warner et al., 1995; Chamberlain et al., 2004; Farrell et al., 2001). Although it is similar to other studies of drug stage transitions (Anthony et al., 1994, O’Brien, & Anthony, 2005), it suggests that the increased risk for women may involve only the transition from abuse to dependence. Women were also more likely to make the transition to remission, a finding consistent recent evidence that drug-dependent men in treatment oscillate more frequently between drug use and abstinence (Gallop et al., 2007). Whereas ‘other’ minority groups were less likely than Whites to have the opportunity to use drugs, they were more likely to transition to abuse. Most epidemiologic investigations have demonstrated lower drug use disorders among ethnic minorities, with the notable exception of Native Americans (Wallace et al., 2003; Compton et al., 2007). The present findings argue for the need to increase resources to reduce the risk of progression after the initial stages of drug involvement in these populations.

Study Limitations and Strengths

The present findings should be interpreted in light of the small subgroups for specific variables and relative to the statistical methods or software applied. Alternative analytic strategies might include longitudinal latent growth mixture models or multivariate latent transition models, and researchers may wish to explore if more optimal results can be achieved by instantaneous risk estimates. The retrospective assessment of drug use history in the NCS-R may also increase the risk of “forward telescoping” (e.g. Johnson & Schultz, 2005), while the prevalence of drug use in older cohorts may be biased downward due to greater stigmatization of drug use, greater time since events, changes in perception of one’s drug use, or because of mortality risks that have had more time to accumulate in older cohorts. Despite its minimal impact on the role of sociodemographic factors (Degenhardt et al., in press; Degenhardt et al., 2007), an additional limitation is that drug dependence was assessed only among respondents with a history of drug abuse. The principal strengths of the study, in comparison, include its use of a nationally-representative sample and assessment of multiple stages of the drug use trajectory. The findings therefore provide novel descriptive information for selective prevention programs and for advancing hypothesis-driven research concerning underlying mechanisms of risk.

CONCLUSION

Commonly-cited sociodemographic risk factors may have markedly different associations depending on stage of the drug use trajectory. The present findings underscore the extent to which knowledge of transitions between stages is necessary for etiologic models and may be a critical determinant of the precision at which existing prevention programs apply findings from epidemiologic research.

Acknowledgments

The National Comorbidity Survey Replication (NCS-R) was supported by grant U01-MH60220 from the National Institute of Mental Health (NIMH) with supplemental support from the National Institute of Drug Abuse (NIDA), The Substance Abuse and Mental Health Services Administration; grant 044708 from The Robert Wood Johnson Foundation and the John W. Alden Trust. Manuscript preparation was also supported by an ATIP award from the French National Center for Scientific Research (Dr. Swendsen), the Intramural Research Program of the National Institutes of Health, National Institute of Mental Health (Drs. Kalaydjian and Merikangas), grant K01 DA15454 from the National Institute of Drug Abuse and an Investigator Award from the Patrick & Catherine Weldon Donaghue Medical Research Foundation (Dr. Dierker).

Footnotes

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.

References

  1. Anthony JC, Petronis KR. Early-onset drug use and risk of later drug problems. Drug Alcohol Depend. 1995;40:9–15. doi: 10.1016/0376-8716(95)01194-3. [DOI] [PubMed] [Google Scholar]
  2. Anthony JC, Warner LA, Kessler RC. Comparative epidemiology of dependence on tobacco, alcohol, controlled substances and inhalants: Basic findings from the National Comorbidity Survey. Exp Clin Psychopharmacol. 1994;2:244–268. [Google Scholar]
  3. Chamberlain CD, Mutaner C, Walrath C. Gender, occupational, and socioeconomic correlates of alcohol and drug abuse among U.S. rural, metropolitan, and urban residents. Am J Drug Alcohol Abuse. 2004;30:409–428. doi: 10.1081/ada-120037385. [DOI] [PubMed] [Google Scholar]
  4. 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;79:11–22. doi: 10.1016/j.drugalcdep.2004.11.014. [DOI] [PubMed] [Google Scholar]
  5. Compton WM, Thomas YF, Stinson FS, Grant BF. Prevalence, correlates, disability, and comorbidity of DSM-IV Drug Abuse and Dependence in the United States. Arch Gen Psychiatry. 2007;64:566–576. doi: 10.1001/archpsyc.64.5.566. [DOI] [PubMed] [Google Scholar]
  6. Cuijpers P. Three decades of drug prevention research. Drugs: Educ Prev Policy. 2003;10:7–20. [Google Scholar]
  7. Degenhardt, Bohnert, Anthony Assessment of cocaine and other drug dependence in the general population: “gated versus “ungated” approaches. Drug Alcohol Depend. 2008;93:227–32. doi: 10.1016/j.drugalcdep.2007.09.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Degenhardt L, Cheng H, Anthony J. Assessing cannabis dependence in community surveys: Methodological issues. Int J Methods Psychiatr Res. 2007;16:43–51. doi: 10.1002/mpr.206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Degenhardt L, Lynskey M, Hall W. Cohort trends in the age of initiation of drug use in Australia. Aust N Z J Public Health. 2000;24:421–426. doi: 10.1111/j.1467-842x.2000.tb01605.x. [DOI] [PubMed] [Google Scholar]
  10. Efron B. Logistic regression, survival analysis, and the Kaplan-Meier curve. J Am Stat Assoc. 1988;83:414–425. [Google Scholar]
  11. Farrell M, Howes S, Bebbington P, Brugha T, Jenkins R, Lewis G, Marsden J, Taylor C, Meltzer H. Nicotine, alcohol and drug dependence and psychiatric comorbidity. Br J Psychiatry. 2001;179:432–437. doi: 10.1192/bjp.179.5.432. [DOI] [PubMed] [Google Scholar]
  12. Gallop RJ, Crits-Christoph P, Ten Have TR, Barber JP, Frank A, Griffin ML, Thase ME. Differential transitions between cocaine use and abstinence for men and women. J Consult Clin Psychol. 2007;75:95–103. doi: 10.1037/0022-006X.75.1.95. [DOI] [PubMed] [Google Scholar]
  13. 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;10:163–173. doi: 10.1016/s0899-3289(99)80131-x. [DOI] [PubMed] [Google Scholar]
  14. Haro JM, Arbabzadeh-Bouchez S, Brugha TS, de Girolamo G, Guyer ME, Jin R, et al. Concordance of the Composite International Diagnostic Interview Version 3.0 (CIDI 3.0) with standardized clinical assessments in the WHO World Mental Health surveys. Int J Methods Psychiatr Res. 2006;15:167–180. doi: 10.1002/mpr.196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hawkins D, Catalano R, Miller J. Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychol Bull. 1992;112:64–105. doi: 10.1037/0033-2909.112.1.64. [DOI] [PubMed] [Google Scholar]
  16. Johnson EO, Schultz L. Forward telescoping bias in reported age of onset: an example from cigarette smoking. Int J Methods Psychiatr Res. 2005;14 (3):119–129. doi: 10.1002/mpr.2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Johnson RA, Gerstein DR. Initiation of use of alcohol, cigarettes, marijuana, cocaine, and other substances in US birth cohorts since 1919. A J Public Health. 1998;88:27–33. doi: 10.2105/ajph.88.1.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kerr WC, Greenfield TK, Bond J, Ye Y, Rehm J. Age-period-cohort influences on trends in past year marijuana use in the U.S. from the 1984, 1990, 1995 and 2000 National Alcohol Surveys. Drug Alcohol Depend. 2007;86:132–138. doi: 10.1016/j.drugalcdep.2006.05.022. [DOI] [PubMed] [Google Scholar]
  19. Kessler RC, Berglund P, Chiu WT, Demler O, Heeringa S, Hiripi E, et al. The US National Comorbidity Survey Replication (NCS-R): design and field procedures. Int J Methods Psychiatr Res. 2004;13:69–92. doi: 10.1002/mpr.167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62:593–602. doi: 10.1001/archpsyc.62.6.593. [DOI] [PubMed] [Google Scholar]
  21. Kessler RC, Merikangas KR. The National Comorbidity Survey Replication (NCS-R): background and aims. Int J Methods Psychiatr Res. 2004;13:60–68. doi: 10.1002/mpr.166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kessler RC, Ustun TB. The World Mental Health (WMH) Survey Initiative Version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI) Int J Methods Psychiatr Res. 2004;13:93–121. doi: 10.1002/mpr.168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Konings E, Dubois-Arber F, Narring F, Michaud PA. Identifying adolescent drug users: results of a national survey on adolescent health in Switzerland. J Adolesc Health. 1995;16:240–247. doi: 10.1016/1054-139X(94)00098-Y. [DOI] [PubMed] [Google Scholar]
  24. Kraemer H, Kazdin A, Offord D, Kessler RC, Jensen P, Kupfer D. Coming to terms with the terms of risk. Arch Gen Psychiatry. 1997;54:337–43. doi: 10.1001/archpsyc.1997.01830160065009. [DOI] [PubMed] [Google Scholar]
  25. Lynskey M, White V, Hill D, Letcher T, Hall W. Prevalence of illicit drug use among youth: results from the Australian School Students’ Alcohol and Drugs Survey. Aust N Z J Public Health. 1999;23:519–524. doi: 10.1111/j.1467-842x.1999.tb01309.x. [DOI] [PubMed] [Google Scholar]
  26. McCabe SE, West BT, Morales M, Cranford JA, Boyd CJ. Does early onset of non-medical use of prescription drugs predict subsequent prescription drug abuse and dependence? Results from a national study. Addiction. 2007 doi: 10.1111/j.1360-0443.2007.02015.x. Epub ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Medina-Mora ME, Borges G, Lara C, Benjet C, Blanco J, Fleiz C, Villatoro J, Rojas E, Zambrano J. Prevalence, service use, and demographic correlates of 12-month DSM-IV psychiatric disorders in Mexico: results from the Mexican National Comorbidity Survey. Psychol Med. 2005;35:1773–1783. doi: 10.1017/S0033291705005672. [DOI] [PubMed] [Google Scholar]
  28. Melchior M, Chastang JF, Goldberg P, Fombonne E. High prevalence rates of tobacco, alcohol and drug use in adolescents and young adults in France: Results from the GAZEL Youth study. Addict Behav. 2007 doi: 10.1016/j.addbeh.2007.09.009. Epub ahead of print. [DOI] [PubMed] [Google Scholar]
  29. Oakley Browne MA, Wells JE, Scott KM, McGee MA. New Zeal and Mental Health Survey Research Team, 2006. Lifetime prevalence and projected lifetime risk of DSM-IV disorders in Te Rau Hinengaro: the New Zealand Mental Health Survey. Aust N Z J Psychiatry. 40:865–874. doi: 10.1080/j.1440-1614.2006.01905.x. [DOI] [PubMed] [Google Scholar]
  30. O’Brien MS, Anthony JC. Risk of becoming cocaine dependent: Epidemiological estimates for the Unites States, 2000–2001. Neuropsychopharmacology. 2005;30:1006–1018. doi: 10.1038/sj.npp.1300681. [DOI] [PubMed] [Google Scholar]
  31. Perkonigg A, Pfister H, Hofler M, Frohlich C, Zimmermann P, Lieb R, Wittchen HU. Substance use and substance use disorders in a community sample of adolescents and young adults: incidence, age effects and patterns of use. Addict Res. 2006;12:187–196. doi: 10.1159/000094421. [DOI] [PubMed] [Google Scholar]
  32. Research Triangle Institute. SUDAAN: Professional Software for Survey Data Analysis, 9.0. Research Triangle Park, NC: Research Triangle Institute; 2004. [Google Scholar]
  33. Stone AL, O’Brien MS, De La Torre A, Anthony JC. Who is becoming hallucinogen dependent soon after hallucinogen use starts? Drug Alcohol Depend. 2007;87:153–163. doi: 10.1016/j.drugalcdep.2006.08.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Toumbourou JW, Stockwell T, Neighbors C, Marlatt GA, Sturge J, Rehm J. Interventions to reduce harm associated with adolescent substance use. Lancet. 2007;369:1391–1401. doi: 10.1016/S0140-6736(07)60369-9. [DOI] [PubMed] [Google Scholar]
  35. Van Etten ML, Neumark YD, Anthony JC. Male-female differences in the earliest stages of drug involvement. Addiction. 1999;94:1413–1419. doi: 10.1046/j.1360-0443.1999.949141312.x. [DOI] [PubMed] [Google Scholar]
  36. von Sydow K, Lieb R, Pfister H, Höfler M, Wittchen HU. Use, abuse and dependence of ecstasy and related drugs in adolescents and young adults-a transient phenomenon? Results from a longitudinal community study. Drug Alcohol Depend. 2002;66:147–159. doi: 10.1016/s0376-8716(01)00195-8. [DOI] [PubMed] [Google Scholar]
  37. Wagner FA, Anthony JC. Male-female differences in the risk of progression from first use to dependence upon cannabis, cocaine, and alcohol. Drug Alcohol Depend. 2007;86:191–198. doi: 10.1016/j.drugalcdep.2006.06.003. [DOI] [PubMed] [Google Scholar]
  38. Wallace JM, Bachman JG, O’Malley PM, Schulenberg JE, Cooper SM, Johnston LD. Gender and ethnic differences in smoking, drinking and illicit drug use among American 8th, 10th, and 12th grade students 1976–2000. Addiction. 2003;98:225–234. doi: 10.1046/j.1360-0443.2003.00282.x. [DOI] [PubMed] [Google Scholar]
  39. Warner LA, Kessler RC, Hughes M. Prevalence and correlates of drug use and dependence in the United States: Results from the National Comorbidity Survey. Arch Gen Psychiatry. 1995;52:219–229. doi: 10.1001/archpsyc.1995.03950150051010. [DOI] [PubMed] [Google Scholar]
  40. Winters KC, Fawkes T, Fahnhorst T, Botzet A, August G. A synthesis of exemplary drug abuse prevention programs in the United States. J Subst Abuse Treat. 2007;32:371–380. doi: 10.1016/j.jsat.2006.10.002. [DOI] [PubMed] [Google Scholar]
  41. Wolter KM. Introduction to Variance Estimation. New York: Springer-Verlag; 1985. [Google Scholar]

RESOURCES