Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Apr 15.
Published in final edited form as: Subst Use Misuse. 2016 Mar 22;51(5):553–564. doi: 10.3109/10826084.2015.1122063

Comparisons of Cocaine-Only, Opioid-Only and Users of Both Substances in the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC)

Robert F Leeman 1,✉, Ms Qisi Sun 2, Ms Devorah Bogart 3, Cheryl L Beseler 4, Mehmet Sofuoglu 5
PMCID: PMC5097614  NIHMSID: NIHMS778053  PMID: 27002858

Abstract

Background

Cocaine and opioid co-use is a notable public health concern, but little is known about correlates of this behavior. Most prior findings come from treatment samples and concern cocaine and heroin. Findings from a nationally-representative sample involving primarily prescription opioid misuse would expand knowledge.

Methods

Past-12-month cocaine and/or opioid users in Wave 1 of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) formed the sample (N=839). Cocaine-only, opioid-only and cocaine/opioid co-users were compared regarding sociodemographics, other substance involvement, psychiatric and medical conditions/events.

Results

Opioid-only users were the largest group (n=622), followed by cocaine-only (n=144) and co-users (n=73). The vast majority of opioid misuse was of prescription opioids (1.4% with past-12-month use of heroin). Notably, co-users did not differ from single drug users in frequency of use of either drug. Co-users did not have significantly greater incidence of any psychiatric conditions, medial conditions or events.

In preliminary analyses, co-users were more likely than either single use group to report several classes of other drug use. However, for most comparisons, opioid use did not add substantial risk beyond cocaine use. Differences on multiple sociodemographic variables suggested opioid-only users were at lowest risk of negative outcomes. These results may relate to a finding that opioid-only users were less likely to have sought treatment.

Conclusions/Importance

This sample of past-12-month cocaine and/or opioid users had greater involvement with other substances, more psychiatric and medical conditions compared to the general population. Co-users had greater involvement with other substances than opioid-only users in particular.

Keywords: cocaine, opioid, heroin, comorbidity, alcohol use disorder, poly-substance use, epidemiology, nicotine, cannabis, help-seeking

INTRODUCTION

Background on cocaine and opioid co-use

Concurrent use of cocaine and opioids is a notable public health concern. Across several studies, prevalence of cocaine use among heroin users, including treatment and non-treatment samples, has been between 30–80% (Leri, Bruneau, & Stewart, 2003). For instance, in a methadone program in Los Angeles designed to limit the spread of HIV/AIDS, just over half of patients had used cocaine both before and during a course of methadone treatment (Grella, Anglin, & Wugalter, 1997). In a study of New York City methadone clinics, among newly admitted patients, 73% of urines were cocaine-positive, with this rate declining later in treatment but remaining substantial at 43% in months 34–36 (Magura, Kang, Nwakeze, & Demsky, 1998). Studies in other countries (e.g., Australia) have also found substantial levels of cocaine use upon initiation of treatment for heroin (40% in Williamson, Darke, Ross, & Teesson, 2007). Studies in non-treatment samples, which have been less common than treatment studies, have also shown frequent cocaine use among heroin users. In a Canadian sample, about 70% of heroin users also injected cocaine (Leri, Stewart, Tremblay, & Bruneau, 2004). In a separate Canadian study, 57% of opioid-dependent individuals reported past week use of opioids (primarily heroin but with substantial prescription opioid use) and cocaine (Leri et al., 2005).

Concurrent cocaine/opioid use presents particular personal and public health costs. This includes simultaneous use of cocaine and opioids (often referred to as “speedballing”) and non-simultaneous use (Leri et al., 2003). For instance, in Australia, heroin users who also used cocaine exhibited a greater likelihood of homelessness, unemployment, crime and needle sharing, but were not more likely to suffer from severe physical disabilities (Williamson, Darke, Ross, & Teesson, 2007). Cocaine/opioid co-use has also been associated with increased likelihood of psychiatric diagnosis. For example, Malow and colleagues found that speedball users had greater depression and anxiety than cocaine-only users (Malow, West, Corrigan, Pena, & Lott, 1992) though the aforementioned Williamson et al. (2007) study did not find increased incidence of psychiatric conditions. Higher likelihood of relapse in treatment is another important negative consequence of cocaine and opioid co-use. Cocaine use has been related to negative outcomes in heroin treatment in multiple studies, including those in the United States (Magura et al., 1998), the Netherlands (Temorshuizen, Kroo, Prins, Geskus, van den Brink, & van Ameijden, 2005) and Australia (Williamson et al., 2007). Initial evidence shows that opioid use may undermine the efficacy of treatment for cocaine use as well (Greenwald, Lundahl, & Steinmiller, 2010).

Cocaine and prescription opioid co-use

Less is known about co-use of cocaine and prescription opioids. In the United States, since the 1980’s, there has been a 400% increase in prescriptions for opioids, along with a parallel 400% increase in opioid overdose deaths and 600% increase in treatment admissions for prescription opioid misuse (CDC, 2011). With this shift toward greater prescription opioid use and abuse, focus on co-use of cocaine and prescription opioids is of increasing importance. In the 2003 National Survey on Drug Use and Health, compared to those who abstained from cocaine, past-year cocaine users had a significantly greater likelihood of past-year prescription opioid use for non-medical purposes (18 times more likely among women and 14 times more likely among men) (Tetrault, Desai, Becker, Fiellin, Concato, & Sullivan, 2008).

Correlates of cocaine and opioid co-use

Due to the negative impact of cocaine and opioid use and, in particular, co-use, it is important to know more about correlates of cocaine and opioid use and specifically, characteristics that distinguish users of one substance from co-users. Correlates of co-use may contribute to enhanced risk associated with this behavior. Key correlates include use of other substances and meeting criteria for other substance use disorders. Prior reports have suggested that use of other substances is common among cocaine and opioid users (Frank & Galea, 1996; Monga et al., 2007). However, precise information regarding use of which substances and which types of substance use disorders are more common among cocaine and opioid co-users, compared to users of one or the other drug, are lacking.

In general, little is known about correlates of cocaine/opioid co-use, particularly among community samples, given that most recent findings come from treatment samples. In addition, much of our knowledge about cocaine/opioid co-use pertains to heroin, not prescription opioids. Therefore, findings from a representative community sample where the majority of opioid misuse was of prescription opioids rather than heroin would enhance completeness of knowledge in this area. The National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) is an outstanding source of data to learn more about correlates of cocaine use, opioid use and concurrent use of both substances.

The NESARC as a source of data on cocaine and opioid co-use

The large size of the NESARC offers an advantage, as does its representativeness of the United States population. While the NESARC is a two-wave prospective study, we opted to use only Wave 1 data. Wave 2 retention differed based on drug use status with rates significantly lower among cocaine-only compared to opioid-only users. The cocaine/opioid co-use group declined from 73 to only 54 in Wave 2, thus we had concerns about the small size of this group. These findings introduced concerns about validity of prospective analyses, thus we opted to utilize Wave 1 data only.

The present study is unique among prior published NESARC reports. We found no published analyses of the NESARC data that focused on cocaine and opioid co-use. Also, most prior Wave 1 NESARC analyses have focused on lifetime substance use and/or diagnoses (e.g., Florez-Salamanca et al., 2013; Grella, Karno, Warda, Niv, & Moore, 2009; Lev-Ran, Imtiaz, Rehm, & Le Foll, 2013; Saha et al., 2012; Vaughn, Fu, Perron, Bohnert, & Howard, 2010; Wu Woody, Yang, & Blazer, 2010; 2011; Yoon, Westermeyer, Warwick, & Kuskowski, 2012), however current status arguably has greater clinical relevance. Two investigations by Wu et al. (2010; 2011) are most relevant to the present study. Using Wave 1 data, they reported results involving data-driven (2010) and a priori (2011) groupings based on opioid misuse and reported on cocaine use among members of these groups, however no groupings centered on cocaine/opioid co-use and they focused on lifetime use. A report by Grella et al. (2009) from the Wave 1 data is also relevant. They examined gender differences and co-morbidity among those with lifetime diagnoses of opioid use disorder. They reported that almost half of these individuals met lifetime criteria for cocaine use disorder and that other psychiatric comorbidities were common as well with about 70% meeting lifetime criteria for mood or anxiety disorder. Prior NESARC studies have shown that lifetime prescription opioid misuse was associated with major depression and anxiety disorders (Martins, Fenton, Keyes, Blanco, Zhu, & Storr, 2012) and co-morbidity with other drug use disorders (Huang et al., 2006).

The present study

This study utilized data from Wave 1 of the NESARC. Specifically, we compared past-12-month users of cocaine only, opioids only and co-users with regard to sociodemographics, involvement with other substances, other psychiatric conditions and medical conditions/events. Our overall hypothesis was that co-users would be most likely to report a pattern of sociodemographics indicative of greater risk of negative outcomes (e.g., less likely to be married, less education). We also predicted that co-users would be more likely to use other substances, meet criteria for other substance use disorders and other psychiatric conditions and be more likely to suffer from various medical conditions.

METHODS

Sample

The NESARC is the largest, most comprehensive comorbidity survey to date (Grant et al., 2004). Wave 1 data collection (N = 43,093) occurred from 2001–2002. The main purposes of the NESARC included documenting prevalence rates for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Revision (DSM-IV) substance use, mood, anxiety and personality disorders, along with medical conditions and events among the non-institutionalized population 18 years and older in the United States. The survey gathered information from individuals living in households, group living conditions and off base as military personnel. The Census Supplementary Survey (C2SS) provided the sampling frame for the household component of the NESARC. Within primary sampling units, households were systematically selected, and one adult respondent age 18 or older was selected at random from each sample household. To ensure appropriate representation, young adults, Hispanics and non-Hispanic blacks were over-sampled.

Data were gathered through face-to-face interviews conducted by U.S. Census workers in participants’ homes, utilizing computer-assisted methods that protected confidentiality (Grant & Dawson, 2006). Individual, household and overall response rates were 93%, 89% and 82%, respectively. Informed consent and other research procedures received full review and approval from the U.S. Office of Management and Budget and the U.S. Censure Bureau. See Grant et al. (2004) for detailed methods.

The present study involves only those who self-reported use of cocaine and/or opioids at least once in the past 12 months. Opioid use was defined as use of heroin and/or opioid-based medications for non-medical purposes.

Measures

Sociodemographic

We examined several variables including sex; marital status (married or not); employment (employed full/part-time or unemployed) and age. Prescription opioid misuse is particularly common among both active military members (Jeffrey et al., 2013) and veterans (Barry et al., 2011), making military/veteran status an important sociodemographic variable to account for in the present study. We created a military affiliation variable by combining reports of current VA/military health care, indication of military job occupation or the armed forces as one’s employer. A recent report stated that about 60% of veterans of recent wars have utilized VA health care since 2002, thus this variable should capture a substantial proportion of current military and veterans (Dept of Veterans Affairs, 2015).

Substance use

Participants were asked about use of substances without a prescription; more often, in greater amounts, or for longer than prescribed; or for a purpose not prescribed by a physician. We created a 3-category grouping to examine use of cocaine only, opioids only and both substances at least once in the past 12 months. Reports of powder and crack cocaine use were combined. Use of heroin and prescription opioids were covered in separate questions but were combined for this study. Given added risk of exposure to HIV/AIDS and other diseases due to needle-based administration (Degenhardt & Hall, 2012), we also compared whether or not participants reported needle-based administration between the cocaine and co-use groups (parallel opioid questions were not included in the NESARC).

For alcohol, we focused on incidence of past-12-month heavy drinking days (≥5 drinks for men, ≥4 for women), typical drinks per drinking day and peak number of drinks in a 24-hour period. For other substances (e.g., cannabis, see Table 2), we examined whether or not participants reported using each at least once in the past 12 months.

Table 2.

Categorical substance use, other psychopathology and medical conditions/events for respondents (N = 839) reporting past 12-month use of cocaine only (n = 144), opioids only (n = 622) or both (n = 73): percentages within each grouping listed

Variables Cocaine Only Opioid Only Co-use Overall Chi-sq p-val
Percent with heavy drinking and use of cigarettes at least once in the past 12 months
≥ 1 heavy drinking day 80% a 59% b 85% a 65.4% <.001
Past-year cigarette use 72.6% b 58.2% c 84.8% a 63.2% <.001
Percent using categories of drugs at least once in the past 12 months
Prescription opioids 0% 99.4% 93.4% 83.1% .133
Heroin 0% 0.8% 8.1% 1.4% .088
Cocaine 100% 0% 100% 25.9% n/a
Sedatives 15% b 22.5% b 44.7% a 23.6% .002
Tranquilizers 12.6% b 21.1% b 51% a 22.8% <.001
Amphetamines 11.4% b 7.7% b 46.8% a 12.2% <.001
Cannabis 75.8% a 34.9% b 86.8% a 46.6% <.001
Hallucinogens 29.2% b 5.9% c 51.1% a 14.2% <.001
Inhalants 10.2% a 1.4% b 12.9% a 3.9% .001
Substance use disorders—current (past 12 months) unless noted (met criteria for abuse or dependence)
Opioid 1 0% 19% 31.1% 17.2% .091
Opioid: lifetime 5.1% b 31% a 46.4% a 28.4 <.001
Cocaine 2 47% 0% 48.6% 12.3% .809
Cocaine: lifetime 59.8% a 13.6% b 60.6% a 25.6% <.001
Alcohol 67.4% a 40.9% b 76.1% a 48.6% <.001
Nicotine dependence 45.5% b 43.2% b 65.1% a 45.8% .010
Other substance 44.2% a 20.4% b 55.4% a 27.7% <.001
Other drug-related variables
Needle-based route of cocaine administration 3.6% –– 17% 8.7% .015
Lifetime drug help-seeking 29.8% a 14% b 32.4% a 18.4% <.001
Other psychopathology
Current major depression 24.2% 16.9% 26.1% 19% .136
Current any anxiety disorder 22.1% 24.4% 29.1% 24.8% .706
Lifetime anxiety disorder due to stress 16.4% 11.5% 18.5% 13% .346
Current mania 10.8% 6.8% 6.3% 7.4% .526
Lifetime conduct or antisocial personality disorder 34.1% a 18.3% b 32.7% 22.2% .002
Lifetime any other personality disorder 34.5% 29.6% 28.6% 30.2% .629
Medical conditions/events occurring at least once within the past 12 months
Arthritis 14.7% 22.4% 12.3% 20.2% .070
Liver condition 5.7% 4.1% 7.9% 4.7% .615
Cardiac condition 29.8% 28.8% 26.2% 28.7% .900
Serious injury 26.1% b 32.8% 39.1% a 32.4% .250
Emergency room visit 34.7% 32.5% 41.7% 33.8% .470
Overnight hospital stay 28.7% a 14.4% b 4.8% c 15.7% <.001
Open in a new tab

Groups with differing superscript letters differed at p ≤.01

1

comparison between opioid-only and co-use group

2

comparison between cocaine and co-use group

Substance use disorder, family history and other psychiatric diagnoses

Substance use disorders, other psychiatric disorders and family history of drug problems were assessed using the Alcohol Use Disorders and Associated Disabilities Interview Schedule-DSM-IV (AUDADIS-IV), a reliable and valid structured diagnostic interview (Grant, Dawson, Stinson, Chou, Kay, & Pickering, 2003). Participants also reported whether or not they have ever sought help for drug use.

Participants were asked about familial history of problems with drugs among any biological family members. We defined positive family history as reporting at least one first-order relative had a drug problem.

We considered 4 classes of other, current psychiatric diagnoses as binary variables: current major depression; mania; anxiety disorder due to stress; and any anxiety disorder. The anxiety disorder due to stress variable was a combination of reports of single or multiple episodes of panic, generalized anxiety, social or specific phobia(s) “when thinking about an extremely stressful experience” (no PTSD diagnosis was available in Wave 1 of the NESARC). For parsimony, individuals were considered to have a current anxiety disorder if they met criteria for panic disorder, agoraphobia, social phobia, any specific phobia, or generalized anxiety disorder. We also considered two classes of lifetime personality disorder: conduct or antisocial and any other personality disorder assessed in Wave 1 of the NESARC. Individuals were considered to have met criteria for a lifetime other personality disorder if they met criteria for dependent, obsessive compulsive, schizoid, paranoid or histrionic personality disorder.

We selected for analysis a priori medical conditions/events that were particularly likely to be related to cocaine and/or opioid use. We considered past-12-month incidence of arthritis (as an example of a painful condition that may be related to opioid use), any liver condition, any cardiac condition, serious injury that “caused you to seek medical help or to cut down your usual activities for more than half a day,” overnight hospital stay and emergency room (ER) visit. The cardiac variable was a combination of items concerning hardening of arteries/atherosclerosis, high blood pressure/hypertension, chest pain/analgesia, rapid heartbeat/tachycardia, heart attack/myocardial infarction and other heart disease. The liver variable was a combination of items on cirrhosis and any other liver condition.

Statistical analyses

SAS, Version 9.3 was used. Distributions were examined for continuous variables. Where variables were skewed, winsorization down to mean + 3 SD, followed by transformation was attempted. Where this did not reduce skew substantially (typically due to a large number of zeros), binary versions were created (i.e., past-12-month heavy drinking, cigarette smoking, emergency room visit, serious injury). In preliminary analyses, we compared each variable across cocaine/opioid use groups, weighted to be representative of the US population. We used chi-square tests to evaluate differences across groups with the exception of t-tests when comparing two of the groups on continuous variables and ANOVA to compare across all groups. Comparisons were made using the SURVEYFREQ and SURVEYMEANS procedures, respectively, which enabled us to account for sample weights and strata based on certain sociodemographic criteria in the NESARC.

A subsequent series of logistic and linear regression models were conducted to determine which differences across cocaine/opioid use groups remained when holding constant any sociodemographic variables found to differ across cocaine/opioid use groups at p < .10 in the preliminary analyses. We again accounted for sample weights and strata, using SURVEYLOGISTIC and SURVEYREG procedures. In logistic regression models, dependent variables were past-12-month use of other substances; past-12-month diagnosis of other substance use disorders and other psychiatric disorders (e.g., major depression) with the exception of personality disorders, which were based on lifetime diagnosis; and various medical conditions/events (e.g., recent injuries). Independent variables were sociodemographic variables, family history and the cocaine/opioid use variable. In linear regressions, we modeled the continuous outcomes for number of drinks in a drinking day and peak number of drinks in a drinking day, including the same covariates. In all regression models, two dummy codes were utilized for the cocaine/opioid use variable: cocaine use only versus other and opioid use only versus other. Though sociodemographic variables were also included in these regression models, we report only on results involving cocaine and opioid use and family history for parsimony, however full regression results may be obtained from the authors. Due to multiple comparisons, we set alpha for statistical significance at .01.

To characterize race/ethnicity in the study sample, we reported on all classifications, however for preliminary comparison across cocaine/opioid use groups, we used a variable with 4 options: White, Black, Hispanic and “other.” For purposes of analysis in the main models, we utilized a binary White/non-White variable. Regarding education, for preliminary comparison across cocaine/opioid groups, we examined a 3-option variable of less than high school, high school/GRE and at least some college, however for parsimony, in the main models, we utilized a binary at least some college/not variable.

For descriptive purposes, we compared use of heroin and other opioids, frequency of combined opioid use and age of onset of any opioid use between the opioid-only and co-use groups. Similarly, we compared cocaine use frequency and age of cocaine use onset between the cocaine-only and co-use groups. Frequency of use items had 10 options ranging from every day to once a year, which were reversed for analysis.

The 3 cocaine/opioid use groups were compared with regard to their use of alcohol, cigarettes and other substances. In analyses involving substance use disorder diagnoses, we combined diagnosis of current DSM-IV abuse or dependence into a single classification, with the exception of nicotine for which there was no abuse diagnosis. For descriptive purposes, we compared whether or not participants met criteria for a current or lifetime cocaine or opioid use disorder. For alcohol, nicotine and other substances, we examined current diagnoses only. For parsimony, we also examined whether or not participants met diagnostic criteria for a substance use disorder for any of the remaining substances in the NESARC.

RESULTS

Sample description

Most of the sample was male and unmarried. About half had at least some college education. About one-third was unemployed. A similar proportion had a positive family history of drug problems (Table 1). Whites made up 73%, followed by Hispanics (10.4%), Blacks (7.9%), mixed race (4%), Asian (2.6%), American Indian (1.7%) and Native Hawaiian/Pacific Islander (.4%).

Table 1.

Categorical sociodemographic characteristics for respondents (N = 839) reporting past 12-month use of cocaine only (n = 144), opioids only (n = 622) or both (n = 73): percentages within each grouping listed

Variables Cocaine Only Opioid Only Co-use Overall Chi-sq p-val
Sociodemographic variables
Male 69.6% a 54.5% b 68.2% 58.2% .003
Married 22.1% b 35.3% a 11.2% b 30.8% <.001
Highest Educational
Attainment
1. <High School
2. High School/GED
3. At least some college		 1. 22.8%
2. 33.1%
3. 44.1%		 1. 14%
2. 30.8%
3. 55.2%		 1. 25.9%
2. 32.1%
3. 42%		 1. 16.6%
2. 31.3%
3. 52.1%		 .086
Unemployed not accounting for student status 37.4% 35.5% 42.1% 36.5% .665
Unemployed and not a current student 25.7% 28.5% 26% 27.8% .554
Military affiliation 21% 11.4% 19.2% 13.7% .081
Race
 1. White
 2. Black
 3. Hispanic
 4. Other		 1. 61.2% b
2. 17.3%
3. 14.2%
4. 7.3%		 1. 75.9% a
2. 6.4%
3. 9.2%
4. 8.5%		 1. 70% b
2. 3.9%
3. 13.6%
4. 12.5%		 1. 73%
2. 7.9%
3. 10.4%
4. 8.7%		 .005
Drug family history 36.2% 29.8% b 48.3% a 32.7% .018
Open in a new tab

Groups with differing superscript letters differed at p ≤.01

1

comparison between opioid-only and co-use group

2

comparison between cocaine and co-use group

In the full NESARC, 1.8% used opioids, while 0.5% used cocaine in the past 12 months (Table 2). Among NESARC Wave 1 respondents who reported past-12-month use of cocaine and/or opioid misuse (n=839), 84.2% used opioids while 25.9% used cocaine. Among past-12-month opioid users, 10.5% also used cocaine and among past-12-month cocaine users, 33.6% also used opioids (weighted estimates). The largest group was comprised of opioid users only (n=622, 74.1% of present study sample), followed by cocaine only (n=144, 17.2%) and co-users (n=73, 8.7%).

There were significant differences across cocaine/opioid use groups for age, sex, marital status and race, with a near significant trend for family history of drug problems. Cocaine/opioid co-users tended to be younger (Table 1) and were somewhat more likely than opioid-only users to have a family history of drug problems. The proportion of males was lowest in the opioid-only group, who were also the most likely to be married. Like co-users, opioid-only users had a high proportion of Whites and low proportion of Black respondents, compared to cocaine-only users.

Current substance use

Preliminary results

Prescription opioids made up the vast majority of opioid use with only 1.4% of the sample reporting past-12-month heroin use. Cocaine administration via needle was somewhat more common among cocaine/opioid co-users than among cocaine-only users (Table 2). There were no significant differences in cocaine use frequency between the cocaine-only and cocaine/opioid co-use groups. Similarly, there was no significant difference in opioid use frequency between the opioid-only and cocaine/opioid co-use groups. Co-users reported earlier ages of onset of use of each drug compared to single users (Table 3). The majority reported at least one heavy drinking day and cigarette smoking in the past 12 months. Heavy drinking rates were greater among co-users and cocaine-only users compared to opioid-only users. Number of drinks per drinking day and peak number of drinks were higher in co-users compared to opioid-only users with no significant differences from the cocaine-only group. Co-users were most likely to have smoked cigarettes, followed by the cocaine-only and the opioid-only group. Use of other substances was less common though nearly half reported past-12-month cannabis use. Co-users had the highest use rates of all other substances (Table 2).

Table 3.

Non-categorical sociodemographic and substance use characteristics for respondents (N = 839) reporting past 12-month use of cocaine only (n = 144), opioids only (n = 622) or both (n = 73): means and standard errors listed

Variables Cocaine Only Opioid Only Both Overall Anova p-val
Sociodemographic variable
Age 31.18 (0.95) 34.98 (0.62)a 27.88 (1.62)b 33.66 (0.50) <.001
Age of onset of drug use
Cocaine 22.18 (0.44) –– 19.51 (0.45) 21.14 (0.37) <.001
Opioids –– 25.72 (0.57) 19.31 (0.69) 24.94 (0.51) <.001
Past 12-month frequency of drug use
Cocaine1 4.28 (0.22) 0 3.81 (0.30) 4.09 (0.19) .312
Opioids1 0 2.62 (0.21) 3.51 (0.65) 2.73 (0.20) .218
Past 12-month alcohol use
Drinks per drinking day 5.75 (0.53) 3.96 (0.24)b 7.35 (0.91)a 4.59 (0.20) <.001
Peak drinks in a day 11.17 (1.03) 8.02 (0.43)b 16.37 (1.63)a 9.37 (0.38) <.001
Open in a new tab

Groups with differing superscript numbers differed at p ≤ .01

1

2 = 2 times per year, 3 = 3–6 times per year, 4 = 7–11 times per year

Regression results

Based on logistic regressions holding key sociodemographics constant, cocaine/opioid co-users were significantly more likely than users of either substance alone to have taken tranquilizers and amphetamines. Co-users were more likely than cocaine-only users to have taken sedatives in the past 12-months, while the difference with opioid-only users was borderline significant. Cocaine/opioid co-users were more likely than opioid-only users to smoke cigarettes and cannabis and take hallucinogens and inhalants, while there was no significant difference from cocaine-only users for these substances (Table 4). Cocaine/opioid use did not relate significantly to past-12-month heavy drinking. Based on linear regressions, co-users reported significantly greater drinks per drinking day and peak number of drinks than opioid-only users with no significant difference from cocaine-only users (Table 5).

Table 4.

Logistic regression analyses comparing respondents (N = 839) reporting past 12-month use of cocaine only (n = 144), opioids only (n = 622) or both (n = 73) with regard to past 12-month use of other substances and lifetime help-seeking for problems with drug use

Variable Heavy drinking
Cigarettes
Sedatives
OR 95% Cl p value OR 95% Cl p value OR 95% Cl p-value
Family history 2.01 1.32–3.07 .001 1.11 0.73–1.68 .620 1.75 1.15–2.66 .009
Cocaine use only 1.03 0.31–3.47 .961 0.54 0.22–1.32 .177 0.25 0.10–059 .002
Opioid use only 0.43 0.15–1.21 .110 0.33 0.15–0.72 .005 0.43 0.22–0.82 .011
Tranquilizers Cannabis Amphetamines
Family history 1.57 1.02–2.34 .040 1.40 0.91–2.17 .127 1.54 0.90–2.62 .114
Cocaine use only 0.15 0.07–034 <.001 0.60 0.27–1.32 .204 0.16 0.07–0.36 <.001
Opioid useonly 0.31 0.17–0.55 <.001 0.11 0.06–0.23 <.001 0.13 0.06–026 <.001
Hallucinogens Inhalants Life time drug help seeking
Family history 1.55 0.82–2.93 .177 0.90 0.33–2.44 .828 1.72 1.07–2.77 .021
Cocaine use only 0.53 0.22–1.26 .153 1.07 0.33–3.52 .910 0.83 0.39–1.76 .630
Opioid use only 0.07 0.03–0.15 <.001 0.16 0.05–0.48 .001 0.32 0.16–0.63 .001
Open in a new tab

Notes: For each model, dependent variable was at least one Incidence in the past 12 months compared to no incidences with the exception of lifetime drug help seeking where the dependent variable was at least one lifetime Incidence compared to no incidences. Cocaine use only and opioid use only were dummy codes with cocaine/oploid co-use as the reference group. Family history was coded 1 for at least one first-order relative with a history of a drug problem and 0 for no first-order relative with a history of a drug problem In each model, the following variables were held constant: sex, mantal status (currently married or unmarried), military status, educational attainment (at least some college or not), race/ethnicity (White or non-White), and age. Full results including findings for these covariates are available from the authors. OR: odds ratio, CI: confidence Interval.

Table 5.

Linear regression analyses comparing respondents (N = 839) reporting past 12-month use of cocaine only (n = 144), opioids only (n = 622) or both (n = 73) with regard to past-12-month alcohol use

Drinks per drinking day Peak number of drinks in a day
Variable beta S.E. p-val beta S.E. p-val
Family history 0.07 0.03 .01 0.10 0.03 .001
Cocaine use only −0.02 0.05 .730 −0.07 0.06 .241
Opioid use only −0.11 0.04 .009 −0.18 0.05 .001
Open in a new tab

Notes: Both alcohol variables winsorized, then log transformed to reduce skew. Cocaine use only and opioid use only were dummy codes with cocaine/opioid co-use as the reference group. Family history was coded 1 for at least 1 first order relative with a history of a drug problem and 0 for no first order relative with a history of a drug problem In each model, the following variables were held constant: sex, marital status (currently married or unmarried), military status, educational attainment (at least some college or not), race/ethnicity (White or Non-White) and age. Full results including findings for these covariates are available from the authors.

Substance use disorder diagnoses

Preliminary results

There were no significant differences between co-users and opioid-only users in likelihood of current or lifetime opioid use disorder or between co-users and cocaine-only users in likelihood of current or lifetime cocaine use disorder. Co-users had the highest rates of nicotine dependence. For alcohol and other substance use disorder and lifetime help-seeking for drug problems, rates were similar between co-users and cocaine only and noticeably higher than the opioid-only group (Table 2).

Regression results

Based on results from logistic regression models, co-users were significantly more likely than opioid-only users to meet criteria for current alcohol use disorder and other substance use disorder and also more likely to have sought help for drug problems. There were no significant differences with cocaine-only users. There were no significant differences for current nicotine dependence (Table 6).

Table 6.

Logistic regression analyses comparing respondents (N = 839) reporting past 12-month use of cocaine only (n = 144), opioids only (n = 622) or both (n = 73) with regard to diagnosis of current (past-12-month) DSM-IV alcohol use disorder, nicotine dependence and other substance use disorders.

Alcohol Use Disorder Nicotine Dependence Other Substance Use Disorder
Variable O.R. 95% C.I. for O.R. p-val O.R. 95% C.I. for O.R. p-val O.R. 95% C.I for O.R. p-val
Family history 1.44 0.93–2.23 .100 1.31 0.90–1.90 .156 2.02 1.32–3.09 .001
Cocaine use only 0.85 0.39–1.85 .679 0.50 0.26–0.96 .039 0.85 0.41–1.74 .648
Opioid use only 0.37 0.19–0.73 .004 0.58 0.32–1.03 .064 0.33 0.17–0.61 <.001
Open in a new tab

Notes: Other substance use disorder included diagnosis of current (past-12-month) abuse or dependence for sedatives, tranquilizers, cannabis, inhalants, hallucinogens, amphetamines or other drugs. Cocaine use only and opioid use only were dummy codes with cocaine/opioid co-use as the reference group. Family history was coded 1 for at least 1 first order relative with a history of a drug problem and 0 for no first order relative with a history of a drug problem In each model, the following variables were held constant: sex, marital status (currently married or unmarried), military status, educational attainment (at least some college or not), race/ethnicity (White or Non-White) and age. Full results including findings for these covariates are available from the authors. O.R.: odds ratio, C.I.: confidence interval

Other psychiatric diagnoses

In preliminary analyses, the groups differed significantly with respect to lifetime conduct or antisocial personality disorder only, which was more common among cocaine-only than among opioid-only users (Table 2). As shown by logistic regression models, cocaine and/or opioid use was not significantly related to any of these psychiatric diagnoses (Supplemental Table 1).

Medical conditions/events

In preliminary analyses, the groups differed only with regard to likelihood of a past-12 month overnight hospital stay, which was more common among cocaine-only than the other groups and also more common among opioid-only than among co-users (Table 2). The significant difference between cocaine-only and co-users on this variable held in a logistic regression (Supplemental Table 2).

DISCUSSION

The sample had elevated incidence of all substance use, substance use disorders and psychiatric conditions compared to the population as a whole (Grant et al., 2004). Notably, co-users did not differ significantly from users of either cocaine or opioids only in frequency of use of cocaine or opioids, respectively. Thus, any findings indicating higher incidence among co-users was not likely to be due primarily to elevated use of cocaine or opioids compared to the single use groups. Opioid misuse was almost exclusively of prescription opioids and was more common than cocaine use. Accordingly, opioid use among cocaine users was more common than cocaine use among opioid users. Compared to the other two groups, opioid-only users had lower incidence for most variables. With the exception of lifetime conduct or antisocial personality disorder in a preliminary analysis only, there were no significant differences across cocaine/opioid use groups for other psychopathology. The only medical condition/event with a significant difference across groups was overnight hospital stay, which was most common among users of cocaine only in both preliminary and regression analyses.

As predicted, cocaine/opioid co-use was more problematic than use of either substance alone for multiple variables, notably past-12-month use of tranquilizers, amphetamines and sedatives, based on regression results. However, it was somewhat surprising that cocaine/opioid co-use was not associated with greater risk than current cocaine use alone in more of the regression models we tested. Indeed, with regard to multiple variables, past-12-month opioid use did not add substantially to the risk associated with past-12-month cocaine use. Examples included past-12-month use of cannabis and cigarettes; peak number of alcoholic drinks in a day and number of drinks per drinking day; current alcohol use disorder and other substance use disorder. These findings differ from the treatment literature where the addition of opioid use has been associated with substantially worse outcomes (Lions et al., 2014; Sullivan et al., 2010). The 10.5% rate of past-12-month cocaine use among opioid users in the present study was also noticeably lower than in prior studies of treatment samples (Leri et al., 2003; Termorshuizen, Krol, Prins, Geskus, van den Brink, & van Ameijden, 2005; Williamson et al., 2007).

With the exception of lifetime conduct or antisocial personality disorder in a preliminary analysis, there were no significant differences across cocaine/opioid use groups for other psychopathology. Similarly, the only significant difference among medical conditions/events was for a greater likelihood of a past-12-month overnight hospital stay among cocaine-only users, which was found in a preliminary analysis and a logistic regression model.

One possible reason for a lack of findings indicating greater risk among cocaine/opioid co-users has to do with family history of drug problems having relationships both to substance use and other psychiatric disorders. Psychopathology such as conduct/antisocial disorder is highly genetically influenced (Kendler et al., 2006; 2008), thus, it is not surprising that family history would have a stronger relationship to this condition than current cocaine use did.

Again, this sample as a whole had higher incidence of other substance use, substance use disorders, psychopathology and several medical conditions/events than the population as a whole. For instance, participants in this sample were almost 3 times more likely than the population as a whole to report major depression; more than 4 times more likely to report mania; and more than 5 times more likely to report an alcohol use disorder (Grant et al., 2004). Those in the sample were about 2.5 times more likely to have had an overnight hospital stay and more than 2.5 times more likely to have a serious injury in the past 12 months, compared to the general population (Adams, Kirzinger, & Martinez, 2013). Thus, in most cases, the differences between members of this sample and the general population in the likelihood of these conditions/events was much greater than among the individual substance use groups in this study, making it less likely to find significant differences among the substance use groups. Further, there is precedent in the literature for a lack of evidence of greater likelihood of psychiatric and medical conditions among cocaine and opioid co-users compared to other substance use groups (Williamson et al., 2007).

There was an interesting pattern of findings with regard to opioid misuse in the present study. The sociodemographic profile of opioid-only users suggested somewhat lower risk than the other two substance use groups, including a lower proportion of males, and greater likelihood of being married. Thus, in a community sample, opioid misusers appear to be at elevated risk of myriad negative outcomes compared to the general population, but may face fewer negative outcomes compared to cocaine users and cocaine/opioid co-users. This pattern may also relate to the fact that opioid misuse was primarily of prescription opioids, which may be associated with different risk factors than heroin. For instance, a prior NESARC investigation found that lifetime prescription opioid users differed from heroin users with regard to which other drugs they endorsed using (Wu et al., 2011). The relatively lower risk profile among opioid-only users may have contributed to reduced perception of need for treatment, verified with their comparatively low rates of lifetime help seeking for drug problems in the present study. Reduced help-seeking may also relate to a perception that prescription opioid use is relatively safe (Manchikanti et al., 2012). The possibility that individuals who misuse prescription opioids are less likely to seek treatment is concerning and may contribute to the public health problem represented by these drugs. Clinical care providers should be made aware that opioid misusers may be less likely than users of other substances to present for treatment. In the present study, opioid users had elevated rates of overnight hospital stays and serious injuries compared to the general population. Thus, while these individuals may not be likely to present for treatment expressly for their substance use, they appear to be likely to present for other types of care, representing an opportunity for caregivers to intervene with these individuals.

The present investigation had limitations. These analyses were cross-sectional and thus do not speak to the extent to which cocaine/opioid co-use predicts subsequent outcomes. There are limitations in the assessment of treatment for pain and its relation to prescription opioid use in the NESARC. Data are not available regarding whether participants have received medical treatment for pain and further, whether their use of prescription opioids originated with treatment for a medical condition. We would have liked to have addressed these issues to learn more about the opioid-only group in comparison with the cocaine/opioid co-use group. We opted to combine prescription opioid misuse with heroin use into one category of opioid use in the present study, however given that there are likely to be differences between the group of prescription opioid and heroin users, this could also be considered a limitation. The military affiliation variable could be considered a limitation given that this was a variable we created based on reports of current VA/military health care, military job occupation or listing the armed forces as one’s employer and not an item posed directly in Wave 1 of the NESARC.

CONCLUSIONS

This study was unique in multiple respects. Cocaine and opioid co-use is a public health concern but under-addressed in research. In addition, most prior findings come from treatment samples and concern co-use of cocaine and heroin, whereas the present study utilized a community sample in which opioid misuse was primarily of prescription opioids. Cocaine/opioid co-users were more likely than single users of cocaine or opioids to have used multiple other substances in the past 12 months. However, in a number of other respects, use of opioids did not appear to add substantial risk above and beyond use of cocaine. It is important to note that in the sample as a whole, incidence of substance use, substance use disorders, other psychopathology and recent medical conditions/events was elevated compared to the general population. It was also notable that elevated risk was observable based only on use of cocaine and opioids rather than on meeting criteria for a substance use disorder for these drugs. Given prior findings that cocaine use can persist or even increase after opioid use declines due to treatment (e.g., Cunningham Giovanniello, Kunins, Roose, Fox, & Sohler, 2013; Sullivan et al. 2010), there is merit to addressing both addictive behaviors simultaneously in treatment.

Supplementary Material

Tables S1-S2

Acknowledgments

The National Epidemiologic Survey on Alcohol and Related Conditions was funded by the National Institute on Alcohol Abuse and Alcoholism with supplemental support from the National Institute on Drug Abuse. This study was supported in part by grants from the National Institutes of Health, the Veterans Administration Mental Illness Research, Education and Clinical Center (MIRECC), the Connecticut Department of Mental Health and Addiction Services and the Connecticut Mental Health Center. The sponsors had no further role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication. The authors alone are responsible for the content and writing of the paper.

Footnotes

Declaration of Interest

The authors report no conflicts of interest.

Contributor Information

Dr Robert F Leeman, Email: robert.leeman@yale.edu, Yale School of Medicine, Psychiatry, CMHC, 34 Park Street, New Haven, 06519 United States.

Ms Qisi Sun, Email: qisi.sun@yale.edu, Yale University, New Haven, United States.

Ms Devorah Bogart, Email: divibogart@gmail.com, Yale University, New Haven, United States.

Dr Cheryl L Beseler, Email: cbeseler@lamar.colostate.edu, Colorado State University, Fort Collins, United States.

Dr Mehmet Sofuoglu, Email: mehmet.sofuolglu@yale.edu, Yale School of Medicine, Psychiatry, New Haven, United States.

References

  1. Adams PF, Kirzinger WK, Martinez ME. Summary health statistics for the U.S. population: National Health Interview Survey, 2012. National Center for Health Statistics Vital Health Stat. 2013;10(259) [PubMed] [Google Scholar]
  2. Barry DT, Goulet JL, Kerns RK, Becker WC, Gordon AJ, Justice AC, Fiellin DA. Nonmedical use of prescription opioids and pain in veterans with and without HIV. Pain. 2011;152:1133–1138. doi: 10.1016/j.pain.2011.01.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Centers for Disease Control and Prevention. Vital signs: overdoses of prescription opioid pain relievers—United States, 1999–2008. MMWR: Morbidity and Mortality Weekly Report. 2011;60:1487–1492. [PubMed] [Google Scholar]
  4. Cunningham CO, Giovanniello A, Kunins HV, Roose RJ, Fox AD, Sohler NL. Buprenorphine treatment outcomes among opioid-dependent cocaine users and non-users. American Journal on Addictions. 2013;22:352–357. doi: 10.1111/j.1521-0391.2013.12032.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Degenhardt L, Hall W. Extent of illicit drug use and dependence, and their contribution to the global burden of disease. Lancet. 2012;379:55–70. doi: 10.1016/S0140-6736(11)61138-0. [DOI] [PubMed] [Google Scholar]
  6. Epidemiology Program, Post-Deployment Health Group, Office of Public Health, Veterans Health Administration, Department of Veterans Affairs. Analysis of VA Health Care Utilization among Operation Enduring Freedom, Operation Iraqi Freedom, and Operation New Dawn Veterans, from 1st Qtr FY 2002 through 4th Qtr FY 2014. Washington, DC: Author; 2014. Retrieved from http://www.publichealth.va.gov/docs/epidemiology/healthcare-utilization-report-fy2014-qtr1.pdf. [Google Scholar]
  7. Flórez-Salamanca L, Secades-Villa R, Hasin DS, Cottler LB, Wang S, Grant BF, Blanco C. Probability and predictors of transition from abuse to dependence on alcohol, cannabis, and cocaine: results from the National Epidemiologic Survey on Alcohol and Related Conditions. American Journal of Drug and Alcohol Abuse. 2013;39:168–79. doi: 10.3109/00952990.2013.772618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Frank B, Galea J. Cocaine trends and other drug trends in New York City, 1986–1994. Journal of Addictive Diseases. 1996;15:1–12. doi: 10.1300/J069v15n04_01. [DOI] [PubMed] [Google Scholar]
  9. Grant BF, Dawson DA, Stinson FS, Chou SP, Kay W, Pickering RP. The Alcohol Use Disorder and Associated Disabilities Interview Schedule-IV (AUDADIS-IV): reliability of alcohol consumption, tobacco use, family history of depression and psychiatric diagnostic modules in a general population sample. Drug & Alcohol Dependence. 2003;71:7–16. doi: 10.1016/s0376-8716(03)00070-x. [DOI] [PubMed] [Google Scholar]
  10. Grant BF, Stinson FS, Dawson DA, Chou SP, Dufour MC, Compton W, Pickering RP, Kaplan K. Prevalence and co-occurrence of substance use disorders and independent mood and anxiety disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Archives of General Psychiatry. 2004;61:807–816. doi: 10.1001/archpsyc.61.8.807. [DOI] [PubMed] [Google Scholar]
  11. Grant BF, Dawson DA. Introduction to the National Epidemiologic Survey on Alcohol and Related Conditions. Alcohol Research and Health. 2006;29:74–78. [Google Scholar]
  12. Greenwald MK, Lundahl LH, Steinmiller CL. Sustained release d-amphetamine reduces cocaine but not “speedball”-seeking in buprenorphine-maintained volunteers: A test of dual-agonist pharmacotherapy for cocaine/heroin polydrug abusers. Neuropsychopharmacology. 2010;35:2624–2637. doi: 10.1038/npp.2010.175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Grella CE, Angelin MD, Wugalter SE. Patterns and predictors of cocaine and crack use by clients in standard and enhanced methadone maintenance treatment. American Journal of Drug and Alcohol Abuse. 1997;23:15–42. doi: 10.3109/00952999709001685. [DOI] [PubMed] [Google Scholar]
  14. Grella CE, Karno MP, Warda US, Niv N, Moore AA. Gender and comorbidity among individuals with opioid use disorders in the NESARC Study. Addictive Behaviors. 2009;34:498–504. doi: 10.1016/j.addbeh.2009.01.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Huang B, Dawson DA, Stinson FS, Hasin DS, Ruan WJ, Saha TD, Grant BF. Prevalence, correlates, and comorbidity of nonmedical prescription drug use disorders in the United States: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Clinical Psychiatry. 2006;67:1062–1073. doi: 10.4088/jcp.v67n0708. [DOI] [PubMed] [Google Scholar]
  16. Jeffery DD, Babeau LA, Nelson LE, Kloc M, Klette K. Prescription drug misuse among U.S. Active duty military personnel: A secondary analysis of the 2008 DoD survey of health related behaviors. Military Medicine. 2013;178:180–195. doi: 10.7205/milmed-d-12-00192. [DOI] [PubMed] [Google Scholar]
  17. Kendler KS, Czajkowski N, Tambs K, Torgersen S, Aggen SH, Neale MC, Reichborn-Kjennerud T. Dimensional representations of DSM-IV Cluster A personality disorders in a population-based sample of Norwegian twins: a multivariate study. Psychological Medicine. 2006;36:1583–1591. doi: 10.1017/S0033291706008609. [DOI] [PubMed] [Google Scholar]
  18. Kendler KS, Aggen SH, Czajkowski N, Røysamb E, Tambs K, Torgersen S, Reichborn-Kjennerud T. The structure of genetic and environmental risk factors for DSM-IV personality disorders – a multivariate twin study. Archives of General Psychiatry. 2008;65:1438–1446. doi: 10.1001/archpsyc.65.12.1438. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Leri F, Bruneau J, Stewart J. Understanding polydrug use: review of heroin and cocaine co-use. Addiction. 2003;98:7–22. doi: 10.1046/j.1360-0443.2003.00236.x. [DOI] [PubMed] [Google Scholar]
  20. Leri F, Stewart J, Fischer B, Jürgen R, Marsh DC, Brissette S, Wild TC. Patterns of opioid and cocaine co-use: a descriptive study in a Canadian sample of untreated opioid dependent individuals. Experimental & Clinical Psychopharmacology. 2005;13:303–310. doi: 10.1037/1064-1297.13.4.303. [DOI] [PubMed] [Google Scholar]
  21. Leri F, Stewart J, Tremblay A, Bruneau J. Heroin and cocaine co-use in a group of injection drug users in Montreal. Journal of Psychiatry & Neuroscience. 2004;29:40–47. [PMC free article] [PubMed] [Google Scholar]
  22. Lev-Ran S, Imtiaz S, Rehm J, Le Foll B. Exploring the Association between Lifetime Prevalence of Mental Illness and Transition from Substance Use to Substance Use Disorders: Results from the National Epidemiologic Survey of Alcohol and Related Conditions. American Journal on Addictions. 2013;22:93–98. doi: 10.1111/j.1521-0391.2013.00304.x. [DOI] [PubMed] [Google Scholar]
  23. Lions C, Carrieri MP, Michel L, Mora M, Marcellin F, Morel A, Roux P. Predictors of non-prescribed opioid use after one year of methadone treatment: An attributable-risk appoach (ANRS-Methaville trial) Drug & Alcohol Dependence. 2014;135:1–8. doi: 10.1016/j.drugalcdep.2013.10.018. [DOI] [PubMed] [Google Scholar]
  24. Magura S, Kang SY, Nwakeze PC, Demsky S. Temporal patterns of heroin and cocaine use among methadone patients. Substance Use & Misuse. 1998;33:2441–2467. doi: 10.3109/10826089809059334. [DOI] [PubMed] [Google Scholar]
  25. Malow RM, West JA, Corrigan SA, Pena JM, Lott WC. Cocaine and speedball users: differences in psychopathology. Journal of Substance Abuse Treatment. 1992;9:287–291. doi: 10.1016/0740-5472(92)90021-f. [DOI] [PubMed] [Google Scholar]
  26. Manchikanti L, Helm S, II, Fellows B, Janata JW, Pampati V, Grider JS, Boswell MV. Opioid Epidemic in the United States. Pain Physician. 2012;15:ES9–ES38. [PubMed] [Google Scholar]
  27. Martins S, Fenton M, Keyes K, Blanco C, Zhu H, Storr C. Mood and anxiety disorders and their association with non-medical prescription opioid use and prescription opioid-use disorder: longitudinal evidence from the National Epidemiologic Study on Alcohol and Related Conditions. Psychological Medicine. 2012;42:1261–1272. doi: 10.1017/S0033291711002145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Monga N, Rehm J, Fischer B, Brissette S, Bruneau J, El-Guebaly N, Bahl S. Using latent class analysis (LCA) to analyze patterns of drug use in a population of illegal opioid users. Drug and Alcohol Dependence. 2007;88:1–8. doi: 10.1016/j.drugalcdep.2006.08.029. [DOI] [PubMed] [Google Scholar]
  29. Saha TD, Compton WM, Chou SP, Smith S, Ruan WJ, Huang B, Grant BF. Analyses related to the development of DSM-5 criteria for substance use related disorders: 1. Toward amphetamine, cocaine and prescription drug use disorder continua using Item Response Theory. Drug & Alcohol Dependence. 2012;122:38–46. doi: 10.1016/j.drugalcdep.2011.09.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Sullivan LE, Moore BA, O’Connor PG, Barry DT, Chawarski MC, Schottenfeld RS, Fiellin DA. The association between cocaine use and treatment outcomes in patients receiving office-based buprenorphine/naloxone for the treatment of opioid dependence. American Journal on Addictions. 2010;19:53–58. doi: 10.1111/j.1521-0391.2009.00003.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Termorshuizen F, Krol A, Prins M, Geskus R, van den Brink W, van Ameijden EJ. Prediction of relapse to frequent heroin use and the role of methadone prescription: An analysis of the Amsterdam Cohort Study among drug users. Drug & Alcohol Dependence. 2005;79:231–240. doi: 10.1016/j.drugalcdep.2005.01.013. [DOI] [PubMed] [Google Scholar]
  32. Tetrault JM, Desai RA, Becker WC, Fiellin DA, Concato J, Sullivan LE. Gender and non-medical use of prescription opioids: Results from a national U.S. survey. Addiction. 2008;103:258–268. doi: 10.1111/j.1360-0443.2007.02056.x. [DOI] [PubMed] [Google Scholar]
  33. Vaughn MG, Fu Q, Perron BE, Bohnert AS, Howard MO. Is crack cocaine use associated with greater violence than powdered cocaine use? Results from a national sample. The American Journal of Drug and Alcohol Abuse. 2010;36:181–186. doi: 10.3109/00952990.2010.491877. [DOI] [PubMed] [Google Scholar]
  34. Williamson A, Darke S, Ross J, Teesson M. The effect of baseline cocaine use ontreatment outcomes for heroin dependence over 24 months: Findings from the Australian Treatment Outcome Study. Journal of Substance Abuse Treatment. 2007;33:287–293. doi: 10.1016/j.jsat.2006.12.009. [DOI] [PubMed] [Google Scholar]
  35. Wu LT, Woody GE, Yang C, Blazer DG. Subtypes of nonmedical opioid users: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Drug & Alcohol Dependence. 2010;112:69–80. doi: 10.1016/j.drugalcdep.2010.05.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wu LT, Woody GE, Yang C, Blazer DG. How do prescription opioid users differ from users of heroin or other drugs in psychopathology: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Addiction Medicine. 2011;5:28–35. doi: 10.1097/ADM.0b013e3181e0364e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Yoon G, Westermeyer J, Warwick M, Kuskowski MA. Substance use disorders and adoption: findings from a national sample. PloS One. 2012;7:e49655. doi: 10.1371/journal.pone.0049655. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Tables S1-S2
NIHMS778053-supplement-Tables_S1-S2.docx (25KB, docx)

RESOURCES