- Email: [email protected]
Opioid abuse and depression in adolescents: Results from the National Survey on Drug Use and Health
Drug and Alcohol Dependence 152 (2015) 131–138
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Opioid abuse and depression in adolescents: Results from the National Survey on Drug Use and Health Mark J. Edlund a,∗ , Valerie L. Forman-Hoffman a , Cherie R. Winder a , David C. Heller a , Larry A. Kroutil a , Rachel N. Lipari b , Lisa J. Colpe c a
RTI International, Research Triangle Park, NC 27709-2194, United States Substance Abuse and Mental Health Services Administration, Rockville, MD 20857, United States c National Institute of Mental Health, Bethesda, MD 20892, United States b
a r t i c l e
i n f o
Article history: Received 8 December 2014 Received in revised form 7 April 2015 Accepted 10 April 2015 Available online 22 April 2015 Keywords: Opioid-related disorders Depression Drug abuse Prescription drug misuse Adolescent
a b s t r a c t Objective: To investigate the association of major depressive episode (MDE) with nonmedical prescription opioid use (NMPOU) and opioid abuse/dependence among adolescents aged 12 to 17. Methods: We analyzed 5 years of data from the National Survey on Drug Use and Health (NSDUH). We used logistic regressions to study the relationship between MDE and NMPOU among all adolescents, as well as the relationship of MDE with opioid abuse/dependence among adolescents with NMPOU. Other covariates included: sociodemographics, alcohol abuse/dependence, nonopioid drug abuse/dependence, delinquency, school performance, religious services attendance, and family support/supervision. Results: In the sample of all adolescents, 6% reported past year NMPOU, and 8% reported past year MDE. When NMPOU and MDE were comorbid, MDE usually preceded the NMPOU. In the sample of adolescents with NMPOU, 15% reported past year opioid abuse/dependence, and 20% reported past year MDE. In adjusted logistic regression, MDE was associated with both NMPOU (OR = 1.51, p < 0.001) among all adolescents and opioid abuse/dependence (OR = 2.18, p < 0.001) among adolescents with NMPOU. Conclusion: MDE occurs commonly in adolescents and is associated with NMPOU and opioid abuse/dependence. In terms of population attributable risk, which is a function of both the prevalence and the strength of the association, MDE is an important risk factor for NMPOU among adolescents and opioid abuse/dependence among adolescents with NMPOU. Preventive and clinical programs to decrease NMPOU and opioid abuse/dependence among adolescents should consider the prominent role of depression. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Prescription opioid misuse and abuse is a significant public health concern (Executive Office of the President of the United States, 2011). Further, opioids are the most common cause of accidental drug overdose in the United States (Compton and Volkow, 2006; Executive Office of the President of the United States, 2011; Paulozzi et al., 2006; Paulozzi and Xi, 2008), constituting an epidemic per the Centers for Disease Control and Prevention (2011). Adolescents are not immune from these problems (Banta-Green, 2012). Between 1999 and 2006, the annual death rate for fatal
∗ Corresponding author. Tel.: +1 919 597 5132; fax: +1 919 485 5555. E-mail address: [email protected] (M.J. Edlund). http://dx.doi.org/10.1016/j.drugalcdep.2015.04.010 0376-8716/© 2015 Elsevier Ireland Ltd. All rights reserved.
overdoses of opioids for individuals aged 15 to 24 rose 440%, from 0.7 per 100,000 to 3.8 per 100,000 (Warner et al., 2009), and adolescents are the age group in which the opioid overdose death rate increased most rapidly. According to data from the 2011 Monitoring the Future survey, 8.7% of 12th graders reported past year illicit use of prescription opioids, especially Vicodin, Percocet, and OxyContin (Johnston et al., 2012). Among adolescents, the National Survey on Drug Use and Health indicates that misuse of opioids is second only to misuse of marijuana in prevalence, and the number of past year initiates of opioid misuse is second only to the number who initiated marijuana use (Substance Abuse and Mental Health Services Administration, 2012c). Research from adults suggests the importance of depression as a risk factor for opioid abuse and dependence (Edlund et al., 2010, 2007; Schepis and Hakes, 2011). In terms of population-attributable
132
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
risk, which is a function of both the strength of the relationship and the frequency of the risk factor, depression may be the most important risk factor for opioid abuse/dependence in adults. The self-medication hypothesis is one prominent explanation for the empirically observed high rates of co-occurring depression and substance abuse, such as with opioids (Khantzian, 1997). This theory posits that individuals use psychoactive substances such as opioids to “self-medicate” painful or disturbing psychological symptoms. Much of the research on opioid abuse/dependence and depression among adolescents is from the National Survey on Drug Use and Health (NSDUH). NSDUH is conducted annually, with approximately 20,000 adolescents aged 12 to 17, is nationally representative, and contains measures of opioid abuse/dependence (American Psychiatric Association, 2000) and nonmedical prescription opioid use (NMPOU), a less severe form of opioid misuse. It also contains a measure of DSM-IV major depressive episode (MDE; American Psychiatric Association, 2000). According to 2012 NSDUH data, 13% of adolescents with MDE reported NMPOU in the past year versus 6% of those without MDE (SAMHSA, 2012d). In an analysis of misuse of any prescription medication (not just opioids) among adolescents using 2008 NSDUH data, major depression was associated with misuse of prescription drugs (OR = 2.60; Havens et al., 2011), but the relationship for NMPOU was not investigated. Wu and colleagues found that there was generally, but not always, an association between depression and opioid abuse/dependence in adolescents (Wu et al., 2008b). Schepis and Krishnan-Sarin (2008), using data on adolescents from the 2008 NSDUH, found MDE to be associated with NMPOU in unadjusted models but did not report data on the relationship for adjusted models. In Vaughn’s latent class analysis of adolescents with NMPOU, lifetime depression was not significantly associated with the latent class of nonmedical opioid use (Vaughn et al., 2012). This study builds upon past work by using NSDUH data to investigate three key issues not previously addressed in the literature. First, among adolescents with both NMPOU and MDE, we investigated the frequency with which the onset of NMPOU precedes the onset of MDE, and the frequency with which the onset of MDE precedes the onset of NMPOU. Although not definitive, temporal progression can give insight into causality. Consistent with the self-medication hypothesis (Khantzian, 1997), we hypothesized that MDE preceding NMPOU would be more common than NMPOU preceding MDE. Second, we used logistic regression to study the determinants of NMPOU among all adolescents and to study the determinants of opioid abuse/dependence among adolescents with NMPOU. These steps represent the development of abuse/dependence—going from no use to NMPOU (experimentation and recreational use) and, among those with NMPOU, from NMPOU to abuse/dependence (addiction). As an extension of the self-medication hypothesis, we posited that the effects of depression would be stronger for the second step than for the first. That is, some adolescents with depression and NMPOU find that opioids help their depression (at least temporarily) and increase their opioid use, progressing to addiction. Finally, we investigated possible moderators of the relationship of MDE with NMPOU and opioid abuse/dependence. We posited that among depressed adolescents, there are “buffering factors” that decrease the magnitude of the association of the relationship of MDE with NMPOU and opioid abuse/dependence. In particular, we hypothesized that the magnitude of the association of MDE with NMPOU and opioid abuse/dependence would be smaller among adolescents with greater parental support, compared with those with less parental support and smaller among those who attend church more frequently, compared with those who attend church less frequently.
2. Methods 2.1. Sample We used cross-sectional data from NSDUH, which the Substance Abuse and Mental Health Services Administration (SAMHSA) conducts annually and is the primary source of information on illicit drug, alcohol, and tobacco use in the United States. Consistent with SAMHSA policy, all sample sizes in the following text and tables have been rounded to the nearest hundred to protect confidentially of respondents. To increase the statistical power, we combined 2008–2012 NSDUH data. We investigated factors associated with NMPOU in the sample of all adolescents aged 12–17 (n = 112,600), and factors associated with opioid abuse/dependence among adolescents aged 12–17 with past year NMPOU (n = 7100). In the total sample, approximately 11,650 had missing data for one of the independent variables. Therefore, three independent variables (past year MDE, number of religious services attended in the past year, grades in last semester or grading period completed) were imputed using weighted sequential hot-deck imputation (Cox, 1980) with imputation classes based on age, race, and gender. After the study was described to participants, informed consent was obtained verbally from parents or guardians, and assent was obtained verbally from the adolescents. Written consent was not obtained because the names of respondents are not used in the screening and interview process. The study was approved by the RTI International IRB. Detailed information on the survey methodology is available in Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings (SAMHSA, 2014).
2.2. Measures Dependent variables: The two outcomes of interest were (1) past year NMPOU and (2) past year opioid abuse/dependence. These represent differing severity of misuse, with NMPOU including individuals with any misuse, and abuse/dependence representing those meeting formal criteria from DSM-IV (American Psychiatric Association, 2000). NMPOU was defined by answers to the following question: “Have you ever, even once, taken (names of prescription opioids) that was not prescribed for you or that you took only for the experience or feeling it caused?” Those answering affirmatively were coded as having NMPOU. Any past year opioid abuse/dependence was defined using standard DSM-IV criteria. Adolescents were also asked about age of first use and most recent use (past 30 days, past year, lifetime), from which the age of onset of NMPOU and past year NMPOU and opioid abuse/dependence were constructed. Independent variables: We identified a set of correlates of NMPOU and opioid abuse/dependence based on results from the literature (Havens et al., 2011; McCabe et al., 2012; Nargiso et al., 2015; Schepis and Krishnan-Sarin, 2008; SAMHSA, 2012b; Wu et al., 2008a; Young et al., 2012). Our independent variables comprised three domains: (1) diagnostic variables; (2) sociodemographic characteristics; (3) and other risk/protective factors. Diagnostic variables: In NSDUH, MDE for adolescents was assessed using a questionnaire based on the depression module in the National Comorbidity Survey—Adolescents (NCS-A, 2010; SAMHSA, 2012a). Alcohol use disorder was defined as meeting criteria for alcohol abuse or alcohol dependence in the past 12 months according to criteria specified in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV; American Psychiatric Association, 2000). We also constructed a variable describing past 12-month nonopioid drug abuse or dependence. Again, this was based on DSM-IV TR criteria, and included
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
133
Table 1 Sample characteristics among adolescents (aged 12–17), among adolescents with Past Year Nonmedical Prescription Opioid Use (NMPOU), and among Adolescents with Past Year Major Depressive Episode (MDE): Percentages, National Survey on Drug Use and Health (NSDUH) 2008 to 2012. Characteristic
Among adolescents aged 12–17
Among past year nonmedical users of opioids aged 12–17
Among adolescents aged 12–17 with past year MDE
Past year NMPOU Yes No
6.1 93.9
N/A N/A
14.2 85.8
Past year opioid abuse or dependence Yes No
0.9 99.1
15.1 84.9
4.1 95.9
Past year MDE Yes No
8.3 91.7
19.5 80.5
N/A N/A
Age group in years 12–13 14–15 16–17
31.6 33.9 34.5
14.9 32.2 52.9
17.4 36.6 46.0
Gender Male Female
51.1 48.9
46.4 53.6
27.7 72.3
Race/ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Non-Hispanic Other
57.2 14.6 20.5 7.7
60.6 13.0 20.0 6.4
59.6 12.7 20.5 7.2
Family income <$20,000 $20,000 to $49,999 $50,000 to $74,999 $75,000+
16.9 31.0 17.4 34.8
19.8 34.5 17.1 28.6
16.7 32.9 18.3 32.2
Rural/urban Rural Urban
17.9 82.1
18.5 81.5
17.6 82.4
4.3 95.7
23.3 76.7
11.5 88.5
4.0 96.0
27.2 72.8
11.1 88.9
Delinquency 0 1 2+
67.4 19.1 13.5
37.3 24.4 38.3
52.1 24.1 23.8
Grade for last semester or grading period completed A+, A, or A− Average B+, B, or B− Average C+, C, or C− Average D or Less Than D Average School Does Not Give These Grades Not in School
30.8 40.6 19.5 5.1 2.9 1.2
17.4 38.6 27.7 10.3 1.9 4.1
25.6 40.4 23.5 6.9 2.1 1.5
How many religious services attended in past year None 1–5 6–24 25–52 52+
33.7 22.4 12.9 13.3 17.7
39.4 26.3 12.2 9.6 12.5
33.3 22.8 14.2 12.3 17.4
Family support/supervision 0–3 4–5 6–7
17.3 33.2 49.5
31.1 37.3 31.5
31.3 37.2 31.5
Past year dependence or abuse status Alcohol abuse or dependence Yes No Abuse or dependence of an illicit drug other than opioids Yes No
N/A = not applicable.
marijuana, cocaine, stimulants, hallucinogens, inhalants, tranquilizers, and sedatives. Sociodemographic variables: Sociodemographic variables included gender, race/ethnicity, family income, and rural/urban
location. Education was not included because it is highly collinear with age in an adolescent population. Other risk/protective factors: Other risk and protective factors included the number of delinquent behaviors, grades in the last
134
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
Table 2 Past year NMPOU among adolescents aged 12–17, by characteristics: National Survey on Drug Use and Health (NSDUH) 2008 to 2012. Characteristic
Past year MDE MDE No MDE
Percentage (SE)
2.9 (0.1) 5.8 (0.2) 9.4 (0.2)
Gender Male Female
5.6 (0.1) 6.7 (0.1)
Race/ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Non-Hispanic Other
6.5 (0.1) 5.4 (0.2) 6.0 (0.2) 5.1 (0.3)
Family income <$20,000 $20,000–$49,999 $50,000–$74,999 $75,000+
7.2 (0.2) 6.8 (0.2) 6.0 (0.2) 5.0 (0.2)
Rural/urban Rural Urban
6.3 (0.2) 6.1 (0.1)
95% Confidence interval of adjusted odds ratio
1.51 1.00
(1.37–1.67) (1.00–1.00)
1.00 1.53 2.08
(1.00–1.00) (1.38–1.69) (1.88–2.30)
1.00 1.36
(1.00–1.00) (1.26–1.47)
1.00 0.73 0.75 0.81
(1.00–1.00) (0.65–0.82) (0.67–0.83) (0.71–0.94)
1.00 1.01 0.95 0.86
(1.00–1.00) (0.91–1.12) (0.84–1.07) (0.77–0.96)
1.00 0.94
(1.00–1.00) (0.86–1.03)
2.54 1.00
(2.26–2.85) (1.00–1.00)
4.63 1.00
(4.14–5.17) (1.00–1.00)
1.00 1.87 3.00
(1.00–1.00) (1.71–2.04) (2.74–3.29)
1.00 1.35 1.52 1.56 1.01 2.08
(1.00–1.00) (1.23–1.49) (1.36–1.68) (1.33–1.81) (0.79–1.29) (1.64–2.63)
1.00 0.99 0.93 0.83 0.83
(1.00–1.00) (0.91–1.08) (0.83–1.04) (0.73–0.94) (0.74–0.93)
1.00 0.84 0.62
(1.00–1.00) (0.77–0.92) (0.57–0.68)
366.0 (2), <.0001
35.2 (1), <.0001
9.2 (3), <.0001
28.5 (3), <.0001
1.0 (1), 0.32
Past year dependence or abuse status Alcohol abuse or dependence Yes No
33.5 (0.9) 4.9 (0.1)
Abuse or dependence of an illicit drug other than opiods Yes No
41.2 (1.0) 4.6 (0.1)
Delinquency 0 1 2+
3.4 (0.1) 7.8 (0.2) 17.4 (0.4)
Grade for last semester or grading period completed A+, A, or A− average B+, B, or B− average C+, C, or C− average D or less than D average School does not give these grades Not in school
3.5 (0.1) 5.8 (0.1) 8.7 (0.3) 12.4 (0.6) 4.0 (0.4) 20.7 (1.5)
Family support/supervision 0–3 4–5 6–7
Adjusted odds ratio
624.0 (1), <.0001 14.1 (0.5) 5.4 (0.1)
Age group in years 12–13 14–15 16–17
How many religious services attended in past year None 1–5 6–24 25–52 52+
Chi-square statistic (df), p value
2585.5 (1), <.0001
3625.2 (1), <.0001
1153.8 (2), <.0001
180.6 (5), <.0001
41.5 (4), <.0001 7.2 (0.2) 7.2 (0.2) 5.8 (0.2) 4.4 (0.2) 4.3 (0.2) 355.0 (2), <.0001 11.0 (0.3) 6.9 (0.2) 3.9 (0.1)
SE = standard error; df = degrees of freedom; MDE = major depressive episode; N/A = not applicable.
semester, encouragement received from family, and number of religious services attended in the past year. Delinquency was based on a count of seven delinquent behaviors in the past year: (1) took part in a serious fight at school or work, (2) took part in a fight in which groups fought groups, (3) carried a handgun, (4) sold illegal drugs, (5) stole or attempted to steal something worth more than $50, (6) attacked someone with serious intent to harm that person, or (7) was ever arrested and booked. These were then summed and classified as 0, 1, or 2+ delinquent behaviors, as in a previous study (Edlund et al., 2015). Grades in the last semester was
measured by a series of mutually exclusive binary variables representing A+, A, or A− average; B+, B, or B− average; C+, C, or C− average; D or less than D average; and dropout/other. In NSDUH, religious attendance in the past year was categorized as 0, 1–2, 3–5, 6–24, 25–52, or 52+ times. Family support and supervision was based on a count of seven items: (1) parents check if homework is done; (2) parents help with homework; (3) parents make adolescents do chores; (4) parents limit amount of television; (5) parents limit time out on school nights; (6) parents tell youths they have done a good job in the past year; and (7) parents tell youths that they
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
135
Table 3 Past year nonmedical use of prescription opioids among adolescents aged 12–17: moderator effects. Odds ratio
95% Confidence interval of odds ratio
Model 1 Age group (in years) × past year MDE MDE vs. No MDE Among persons aged 12–13 Among persons aged 14–15 Among persons aged 16–17
1.53 1.85 1.32
(1.16–2.01) (1.58–2.15) (1.15–1.52)
Model 2 Family support/supervision × past year MDE MDE vs. No MDE Among persons with family support 0–3 Among persons with family support 4–5 Among persons with family support 6–7
1.15 1.71 1.83
(0.97–1.36) (1.47–1.98) (1.52–2.20)
MDE = major depressive episode.
are proud of things done in the past year. As items were endorsed affirmatively by large majorities, the variable is coded as 0 to 3, 4 to 5, or 6 to 7. Using factor analysis, we confirmed that these variables loaded on a single factor rather than on two or more distinct factors. 2.3. Data analysis For regression analyses, we used two logistic regressions. In the first, we investigated the association, among all adolescents, of MDE and other covariates with NMPOU. In the second, we investigated the association, among adolescents with NMPOU, of MDE and other covariates with opioid abuse/dependence. For our analyses of moderators, a separate model was implemented for each moderator through the inclusion of the appropriate interaction term(s) of that moderator and the MDE variable. For interactions that were significant, we investigated contrasts between levels of interaction to further explain and describe the relationship between dependent variables and these moderators. All analyses combined NSDUH data from the 2008–2012 analytic files and used SUDAAN® (Research Triangle Institute, 2012) to account for the complex sample design. 3. Results 3.1. Any NMPOU Among all adolescents, 6.1% reported NMPOU in the past year, and 8.3% had past year MDE (Table 1). Among those with both lifetime MDE and lifetime NMPOU, 47% had the onset of MDE before the onset of NMPOU, 27.2% had the onset of NMPOU before the onset of MDE, and 25.8% had the onset in the same year (results not shown). In an adjusted logistic regression, MDE was significantly associated with NMPOU (OR = 1.51, 95% CI = 1.37–1.67; Table 2). Adolescents with a greater likelihood of NMPOU included those who were older, white, female, and those who had a nonopioid drug use disorder, an alcohol use disorder, lower family income, lower family support, lower grades, more delinquent behaviors, and less religious attendance (Table 2). The effects of major depression on NMPOU were moderated by family support (Wald F = 8.6, 2 df, p < 0.01), and age (Wald F = 5.2, 2 df, p < 0.01). The association between major depression and NMPOU was only significant for adolescents with moderate or high levels of family supervision; the association was not significant for adolescents with low levels of family supervision (OR = 1.15, 95% CI = 0.97–1.36) (Table 3, which reports results for statistically significant moderators). The association of MDE with NMPOU was significant for all adolescent age groups, and was largest for adolescents aged 14–15 (OR = 1.85, 95% CI = 1.58–2.15), followed by
adolescents aged 12–13 (OR = 1.53, 95% CI = 1.16–2.01) and adolescents aged 16–17 (OR = 1.32, 95% CI = 1.15–1.52). 3.2. Opioid abuse/dependence Among the sample with NMPOU, 15.1% had past year opioid abuse/dependence and 19.5% had past year MDE (Table 1). In an adjusted logistic regression, MDE was associated with opioid abuse/dependence (OR = 2.18, 95% CI = 1.77–2.70) (Table 4). Adolescents who had a greater likelihood of opioid abuse or dependence included those who were female, and those who had a nonopioid drug use disorder, an alcohol use disorder, less family support, more delinquent behaviors, and greater frequency of religious service attendance (fifty-two or more times per year). There were no variables that were statistically significant moderators of the depression and opioid abuse/dependence relationship. 4. Discussion Past year MDE was common among adolescents (8%), and among adolescents with NMPOU (20%). MDE was significantly associated with both NMPOU among the total sample of adolescents, and opioid abuse/dependence among those with NMPOU. As hypothesized, among individuals with comorbid depression and NMPOU, depression usually preceded the NMPOU. This suggests the importance of MDE in terms of population attributable risk for NMPOU and opioid abuse/dependence, which is a function both of prevalence and the magnitude of the association. The potentially sizable excess risk of opioid use and abuse/dependence in the total adolescent population potentially attributable to MDE may be underappreciated in the literature; two recent reviews of factors associated with non-medical use of prescriptions drugs among youth made no mention of depression (Nargiso et al., 2015; Young et al., 2012). As hypothesized, the OR measuring the magnitude of the association between MDE and opioid abuse/dependence among adolescents with NMPOU (OR = 2.30) was greater than the OR measuring the association between MDE and NMPOU among all adolescents (OR = 1.53). We believe this may provide indirect evidence for the self-medication hypothesis. We hypothesized that the magnitude of the association of MDE with NMPOU and opioid abuse/dependence would be smaller among adolescents with greater parental support, compared with those with less parental support, and smaller among those who attend church more frequently, compared with those who attend church less frequently, because these factors would buffer the effects of MDE. However, we found no evidence to support these hypotheses. The religious attendance–MDE interactions were nonsignificant. The parental support–MDE interaction was not
136
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
Table 4 Past year abuse opioid abuse/dependence among adolescents aged 12–17 with past year nonmedical prescription opioid use (NMPOU), by characteristics: National Survey on Drug Use and Health (NSDUH) 2008 to 2012. Characteristic
Percentage (SE)
Past year MDE MDE No MDE
28.4 (1.7) 11.9 (0.5)
Age group in years 12–13 14–15 16–17
10.7 (1.2) 16.2 (1.0) 15.6 (0.8)
Gender Male Female
12.3 (0.8) 17.5 (0.8)
Race/ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Non-Hispanic Other
15.9 (0.7) 10.1 (1.3) 15.5 (1.4) 16.6 (2.1)
Family income <$20,000 $20,000–$49,999 $50,000–$74,999 $75,000+
17.4 (1.3) 13.4 (0.9) 14.7 (1.3) 15.8 (1.1)
Rural/urban Rural Urban
15.5 (1.2) 15.0 (0.6)
Past year dependence or abuse status Alcohol abuse or dependence Yes No Abuse or dependence of an illicit drug other than pain relievers Yes No
Chi-square statistic (df), p value
Adjusted odds ratio
95% Confidence interval of adjusted odds ratio
2.18 1.00
(1.77–2.70) (1.00–1.00)
1.00 1.03 0.91
(1.00–1.00) (0.76–1.40) (0.68–1.22)
1.00 1.39
(1.00–1.00) (1.13–1.70)
1.00 0.85 1.03 1.03
(1.00–1.00) (0.60–1.20) (0.79–1.34) (0.70–1.50)
1.00 0.72 0.77 0.88
(1.00–1.00) (0.56–0.93) (0.57–1.04) (0.67–1.17)
1.00 0.94
(1.00–1.00) (0.74–1.19)
1.89 1.00
(1.53–2.32) (1.00–1.00)
2.73 1.00
(2.23–3.34) (1.00–1.00)
1.00 1.16 1.71
(1.00–1.00) (0.88–1.54) (1.34–2.19)
1.00 0.98 1.14 1.31 1.45 1.30
(1.00–1.00) (0.74–1.30) (0.84–1.53) (0.94–1.84) (0.78–2.70) (0.80–2.12)
1.00 1.11 1.09 0.99 1.60
(1.00–1.00) (0.88–1.39) (0.77–1.52) (0.68–1.45) (1.18–2.19)
1.00 0.73 0.62
(1.00–1.00) (0.59–0.90) (0.49–0.79)
120.2 (1), <.0001
6.1 (2), 0.002
19.4 (1), <.0001
3.9 (3), 0.009
2.5 (3), 0.06
0.1 (1), 0.74
183.1 (1), <.0001 29.4 (1.4) 10.7 (0.6) 269.4 (1), <.0001 30.6 (1.3) 9.3 (0.5)
Delinquency 0 1 2+
8.9 (0.7) 12.9 (1.1) 22.5 (1.0)
53.6 (2), <.0001
Grade For Last Semester Or Grading Period Completed A+, A, or A− average B+, B, or B− average C+, C, or C− average D or less than D average School does not give these grades Not in school
12.3 (1.2) 12.7 (0.8) 16.1 (1.1) 22.6 (2.0) 17.4 (3.9) 21.8 (3.2)
How many religious services attended in past year None 1–5 6–24 25–52 52+
14.7 (0.8) 15.9 (1.1) 14.3 (1.7) 12.7 (1.8) 17.0 (1.7)
Family support/supervision 0–3 4–5 6–7
20.8 (1.1) 13.9 (0.9) 10.8 (0.8)
7.6 (5), <.0001
1.0 (4), 0.43
30.3 (2), <.0001
SE = standard error; df = degrees of freedom; MDE = major depressive episode; N/A = not applicable.
significant in the regression predicting opioid abuse/dependence, and although the parental support–MDE interaction was significant in the regression predicting NMPOU, the effects of MDE were larger among those with greater parental support—the opposite of what we hypothesized. We know of no explanation in the literature for this finding. Further exploration of the effect of parental support on opioid use and dependence in the context of other risk and protective factors is warranted. Although the primary purpose of our study was to investigate the relationship of MDE with NMPOU and opioid abuse/
dependence, we believe that our other findings are important for the epidemiology of adolescent NMPOU and opioid abuse/ dependence. In a secondary analysis of Monitoring the Future data, McCabe et al. (2012) found no statistically significant differences in NMPOU between males and females. Some secondary analyses of NSDUH data have found that adolescent females are more likely to have NMPOU (Schepis and Krishnan-Sarin, 2008; Sung et al., 2005; Wu et al., 2008a, 2008b), whereas other analyses have found no statistical differences (Ford, 2008a, 2008b, 2009; Ford and Lacerenza, 2011; Simoni-Wastila et al., 2008). Our results provide strong
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
evidence that in adjusted models, adolescent females are more likely to have NMPOU, and among those with NMPOU, adolescent females are more likely to have opioid abuse/dependence than adolescent males. This is an important finding for clinicians, who may view adolescent boys as higher risk for opioid abuse/dependence, based on the large body of epidemiological research that suggests that males are in general at a higher risk for substance use disorders than females. Our results suggest that clinicians should maintain an equal, and perhaps even higher, index of suspicion for opioid abuse/dependence among girls than boys. Further, when there is an OUD, clinicians need to be aware that it is often comorbid with MDE. Our results illustrate that a two-step analytical approach, which first focuses on NMPOU, and then opioid abuse/dependence among those with NMPOU, may give a much more detailed understanding of the development of adolescent opioid abuse/dependence. For example, in the case of age, older adolescents were much more likely than younger adolescents to have NMPOU, which may indicate greater exposure to opioids and more experimentation. However, among adolescents with NMPOU, in adjusted models, older adolescents were not more likely than younger adolescents to have addiction (abuse/dependence). Thus, younger adolescents are significantly less likely than older adolescents to experiment, but those who do are as likely to progress to abuse/dependence as older adolescents. Because of this, opioid abuse/dependence is not necessarily rare among young adolescents. Among adolescents with opioid abuse/dependence, 1 out of 9 is aged 12 to 13, and about 1 out of 330 12 to 13 year olds have opioid abuse/dependence. This is particularly worrisome because earlier onset of drug abuse is associated with greater severity of disorder, and is usually more refractory to treatment. Although we know of no long-term studies of outcomes in individuals who develop opioid abuse/dependence in early adolescence, outcomes are likely extremely poor in all domains (i.e., physical and behavioral health, education, employment, and criminal justice system involvement). Our findings illustrate that efforts to prevent and treat opioid experimentation and addiction need to focus on even the youngest adolescents. Adolescents with less religious service attendance were more likely to have NMPOU, which is consistent with past studies that have found that those with lower levels of religiosity or religious attendance are less likely to abstain from alcohol and drugs (Michalak et al., 2007). On the other hand, among adolescents with NMPOU, we found that those who attended 52+ religious services per year were more likely to have opioid abuse or dependence than adolescents with NMPOU who attended religious services less frequently. The NSDUH does not include sufficient religiosity measures to allow us to determine why this is, and there are many possible causes. For example, prescription opioid abuse/dependence might be relatively less stigmatized than alcohol or illegal drug/abuse dependence among those with regular church attendance. Additional research focused on religiosity is needed to better understand this relationship. Several limitations of our study deserve mention. First, although NSDUH asks respondents about age of onset, ultimately the data is cross-sectional and not longitudinal; thus, our results should be viewed as associations, not causal. Second, NSDUH does not allow us to determine whether MDE was caused by major depressive disorder or bipolar disorder, although we note that in adolescents, major depressive disorder is about 3 times more common than bipolar disorder (Kessler et al., 2012). Third, NSDUH does not survey those younger than 12, and we should not assume that opioid misuse does not extend into preadolescents. In summary, MDE is associated with NMPOU and opioid abuse/dependence. When NMPOU and MDE are comorbid, MDE usually precedes the MDE. In terms of population-attributable risk,
137
which is a function of both the prevalence and the strength of the association, MDE is a key risk factor for NMPOU and opioid abuse/dependence. Preventive and clinical programs to decrease NMPOU and opioid abuse/dependence among adolescents should consider the prominent role of depression. Author disclosures Role of funding source The National Survey on Drug Use and Health (NSDUH) is funded by the Substance Abuse and Mental Health Services Administration (SAMHSA), Center for Behavioral Health Statistics and Quality (CBHSQ). This study was funded under Contract No. 283-201000003C, Project No. 0212800, which was supported by funding from the National Institute of Mental Health (NIMH). The study utilized data previously collected by SAMHSA. Co-authors on the paper were from SAMHSA (Dr. Lipari) and NIMH (Dr. Colpe), and they assisted with study design, analysis and interpretation of data, writing of the manuscript, and the decision to submit the manuscript for publication. Contributors Mark J. Edlund. Study design, supervision of statistical analyses, analysis and interpretation of data, writing of initial draft and revisions of manuscript. Valerie L. Forman-Hoffman. Study design, supervision of statistical analyses, analysis and interpretation of data, writing of manuscript. Cherie R. Winder. Statistical programming, analysis and interpretation of data, writing of manuscript. David C. Heller Statistical programming, analysis and interpretation of data, writing of manuscript. Larry A. Kroutil Study design, analysis and interpretation of data, writing of manuscript. Rachel N. Lipari Study design, analysis and interpretation of data, writing of manuscript. Lisa J. Colpe Study design, analysis and interpretation of data, writing of manuscript. All authors have reviewed and approved the final version of the manuscript. Conflict of interest All authors declare that they have no conflicts of interest. Acknowledgments We thank SAMHSA and NIMH, who provided funding for this study. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Substance Abuse and Mental Health Services Administration. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute of Mental Health. References American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental Disorders, Text Revision. American Psychiatric Association, Washington, DC. Banta-Green, C.J., 2012. Adolescent abuse of pharmaceutical opioids raises questions about prescribing and prevention. Arch. Pediatr. Adolesc. Med. 166, 865–866. Centers for Disease Control and Prevention, 2011. Vital signs: overdoses of prescription opioid pain relievers—United States, 1999–2008. MMWR 60, 1487–1492. Compton, W.M., Volkow, N.D., 2006. Major increases in opioid analgesic abuse in the United States: concerns and strategies. Drug Alcohol Depend. 81, 103–107.
138
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
Cox, B.G., 1980. The weighted sequential hot deck imputation procedure. In: Proceedings of the 1980 American Statistical Association, Survey Research Methods Section. American Statistical Association, Houston, TX, pp. 721–726. Edlund, M.J., Martin, B.C., Fan, M.Y., DeVries, A., Braden, J.B., Sullivan, M.D., 2010. Risks for opioid abuse and dependence among recipients of chronic opioid therapy: results from the TROUP Study. Drug Alcohol Depend. 112, 90–98. Edlund, M.J., Pettiford, A.G., Hampton, J., Hoffman, V., Ault, K.L., Colpe, L., Hedden, S.L., 2015. Adolescents’ assessments of the helpfulness of treatment for major depression: results from a national survey. Psych. Serv. (in press). Edlund, M.J., Steffick, D., Hudson, T., Harris, K.M., Sullivan, M., 2007. Risk factors for clinically recognized opioid abuse and dependence among veterans using opioids for chronic non-cancer pain. Pain 129, 355–362. Executive Office of the President of the United States, 2011. Epidemic: Responding to America’s Prescription Drug Abuse Crisis, Washington, DC. Ford, J.A., 2008a. Nonmedical prescription drug use and delinquency: an analysis with a national sample. J. Drug Issues 38, 493–516. Ford, J.A., 2008b. Social learning theory and nonmedical prescription drug use among adolescents. Sociol. Spectr. 28, 299–316. Ford, J.A., 2009. Nonmedical prescription drug use among adolescents: the influence of bonds to family and school. Youth Soc. 40, 336–352. Ford, J.A., Lacerenza, C., 2011. The relationship between source of diversion and prescription drug misuse, abuse, and dependence. Subst. Use Misuse 46, 819–827. Havens, J.R., Young, A.M., Havens, C.E., 2011. Nonmedical prescription drug use in a nationally representative sample of adolescents: evidence of greater use among rural adolescents. Arch. Pediatr. Adolesc. Med. 165, 250–255. Johnston, L.D., O’Malley, P.M., Bachman, J.G., Schulenberg, J.E., 2012. Monitoring the Future National Survey Results on Drug Use, 1975–2011. Volume I Secondary School Students, http://www.monitoringthefuture.org/pubs/ monographs/mtf-vol1 2011.pdf (accessed on December 8, 2014). Kessler, R.C., Avenevoli, S., Costello, E.J., Georgiades, K., Green, J.G., Gruber, M.J., He, J.P., Koretz, D., McLaughlin, K.A., Petukhova, M., Sampson, N.A., Zaslavsky, A.M., Merikangas, K.R., 2012. Prevalence, persistence, and sociodemographic correlates of DSM-IV disorders in the National Comorbidity Survey Replication Adolescent Supplement. Arch. Gen. Psychiatry 69, 372–380. Khantzian, E.J., 1997. The self-medication hypothesis of substance use disorders: a reconsideration and recent applications. Harv. Rev. Psychiatry 4, 231–244. McCabe, S.E., West, B.T., Teter, C.J., Boyd, C.J., 2012. Medical and nonmedical use of prescription opioids among high school seniors in the United States. Arch. Pediatr. Adolesc. Med. 166, 797–802. Michalak, L., Trocki, K., Bond, J., 2007. Religion and alcohol in the U.S. National Alcohol Survey: how important is religion for abstention and drinking. Drug Alcohol Depend. 87, 268–280. Nargiso, J.E., Ballard, E.L., Skeer, M.R., 2015. A systematic review of risk and protective factors associated with nonmedical use of prescription drugs among youth in the United States: a social ecological perspective. J. Stud. Alcohol Drugs 76, 5–20. NCS-A, 2010. National Comorbidity Survey—Adolescents, http://www.hcp.med. harvard.edu/ncs/ (accessed on November 25, 2012). Paulozzi, L.J., Budnitz, D.S., Xi, Y., 2006. Increasing deaths from opioid analgesics in the United States. Pharmacoepidemiol. Drug Saf. 15, 618–627. Paulozzi, L.J., Xi, Y., 2008. Recent changes in drug poisoning mortality in the United States by urban–rural status and by drug type. Pharmacoepidemiol. Drug Saf. 10, 997–1005, http://dx.doi.org/10.1002/pds.1626
Research Triangle Institute, 2012. SUDAAN Language Manual, Release 11.0. Research Triangle Institute, Research Triangle Park, NC. Schepis, T.S., Hakes, J.K., 2011. Non-medical prescription use increases the risk for the onset and recurrence of psychopathology: results from the National Epidemiological Survey on Alcohol and Related Conditions. Addiction 106, 2146–2155. Schepis, T.S., Krishnan-Sarin, S., 2008. Characterizing adolescent prescription misusers: a population-based study. J. Am. Acad. Child Adolesc. Psychiatry 47, 745–754. Simoni-Wastila, L., Yang, H.W., Lawler, J., 2008. Correlates of prescription drug nonmedical use and problem use by adolescents. J. Addict. Med. 2, 31–39. SAMHSA, 2012a. Results from the 2010 National Survey on Drug Use and Health: Mental Health Findings. HHS Publication (NSDUH Series H-42, HHS Publication No. SMA 11-4667). http://www.samhsa.gov/data/nsduh/ 2k10MH Findings/2k10MHResults.htm (accessed on January 6 2014). SAMHSA, 2012b. Results from the 2011 National Survey on Drug Use and Health: Mental Health Findings. HHS Publication (NSDUH Series H-45, HHS Publication No. SMA 12-4725). http://www.samhsa.gov/data/NSDUH/ 2k11MH FindingsandDetTables/2K11MHFR/NSDUHmhfr2011.htm (accessed on January 6 2014). SAMHSA, 2012c. Results from the 2011 National Survey on Drug Use and Health: Summary of National Findings. HHS Publication (NSDUH Series H44, HHS Publication No. SMA 12-4713). http://www.samhsa.gov/data/NSDUH/ 2k11Results/NSDUHresults2011.htm (accessed on January 6 2013). SAMHSA, 2012. Results from the 2011 National Survey on Drug Use and Health: Detailed Tables. Prevalence Estimates, Standard Errors, P Values, and Sample Sizes, http://www.samhsa.gov/data/NSDUH/ 2011SummNatFindDetTables/NSDUH-DetTabsPDFWHTML2011/2k11Detailed Tabs/Web/PDFW/NSDUH-DetTabsCover2011.pdf (accessed on January 6 2013). SAMHSA, 2014. Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings. HHS Publication (NSDUH Series H-48, HHS Publication No. (SMA) 14-4863). http://www.samhsa.gov/data/sites/default/files/ NSDUHresultsPDFWHTML2013/Web/NSDUHresults2013.pdf (accessed on April 7 2015). Sung, H.E., Richter, L., Vaughan, R., Johnson, P.B., Thom, B., 2005. Nonmedical use of prescription opioids among teenagers in the United States: trends and correlates. J. Adolesc. Health 37, 44–51. Vaughn, M.G., Fu, Q., Perron, B.E., Wu, L.T., 2012. Risk profiles among adolescent nonmedical opioid users in the United States. Addict. Behav. 37, 974–977. Warner, M., Chen, L.H., Makuc, D.M., 2009. Increase in fatal poisonings involving opioid analgesics in the United States, 1999–2006. In: NCHS Data Brief. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics. Wu, L.T., Pilowsky, D.J., Patkar, A.A., 2008a. Non-prescribed use of pain relievers among adolescents in the United States. Drug Alcohol Depend. 94, 1–11. Wu, L.T., Ringwalt, C.L., Mannelli, P., Patkar, A.A., 2008b. Prescription pain reliever abuse and dependence among adolescents: a nationally representative study. J. Am. Acad. Child Adolesc. Psychiatry 47, 1020–1029. Young, A.M., Glover, N., Havens, J.R., 2012. Nonmedical use of prescription medications among adolescents in the United States: a systematic review. J. Adolesc. Health 51, 6–17.
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Opioid abuse and depression in adolescents: Results from the National Survey on Drug Use and Health Mark J. Edlund a,∗ , Valerie L. Forman-Hoffman a , Cherie R. Winder a , David C. Heller a , Larry A. Kroutil a , Rachel N. Lipari b , Lisa J. Colpe c a
RTI International, Research Triangle Park, NC 27709-2194, United States Substance Abuse and Mental Health Services Administration, Rockville, MD 20857, United States c National Institute of Mental Health, Bethesda, MD 20892, United States b
a r t i c l e
i n f o
Article history: Received 8 December 2014 Received in revised form 7 April 2015 Accepted 10 April 2015 Available online 22 April 2015 Keywords: Opioid-related disorders Depression Drug abuse Prescription drug misuse Adolescent
a b s t r a c t Objective: To investigate the association of major depressive episode (MDE) with nonmedical prescription opioid use (NMPOU) and opioid abuse/dependence among adolescents aged 12 to 17. Methods: We analyzed 5 years of data from the National Survey on Drug Use and Health (NSDUH). We used logistic regressions to study the relationship between MDE and NMPOU among all adolescents, as well as the relationship of MDE with opioid abuse/dependence among adolescents with NMPOU. Other covariates included: sociodemographics, alcohol abuse/dependence, nonopioid drug abuse/dependence, delinquency, school performance, religious services attendance, and family support/supervision. Results: In the sample of all adolescents, 6% reported past year NMPOU, and 8% reported past year MDE. When NMPOU and MDE were comorbid, MDE usually preceded the NMPOU. In the sample of adolescents with NMPOU, 15% reported past year opioid abuse/dependence, and 20% reported past year MDE. In adjusted logistic regression, MDE was associated with both NMPOU (OR = 1.51, p < 0.001) among all adolescents and opioid abuse/dependence (OR = 2.18, p < 0.001) among adolescents with NMPOU. Conclusion: MDE occurs commonly in adolescents and is associated with NMPOU and opioid abuse/dependence. In terms of population attributable risk, which is a function of both the prevalence and the strength of the association, MDE is an important risk factor for NMPOU among adolescents and opioid abuse/dependence among adolescents with NMPOU. Preventive and clinical programs to decrease NMPOU and opioid abuse/dependence among adolescents should consider the prominent role of depression. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Prescription opioid misuse and abuse is a significant public health concern (Executive Office of the President of the United States, 2011). Further, opioids are the most common cause of accidental drug overdose in the United States (Compton and Volkow, 2006; Executive Office of the President of the United States, 2011; Paulozzi et al., 2006; Paulozzi and Xi, 2008), constituting an epidemic per the Centers for Disease Control and Prevention (2011). Adolescents are not immune from these problems (Banta-Green, 2012). Between 1999 and 2006, the annual death rate for fatal
∗ Corresponding author. Tel.: +1 919 597 5132; fax: +1 919 485 5555. E-mail address: [email protected] (M.J. Edlund). http://dx.doi.org/10.1016/j.drugalcdep.2015.04.010 0376-8716/© 2015 Elsevier Ireland Ltd. All rights reserved.
overdoses of opioids for individuals aged 15 to 24 rose 440%, from 0.7 per 100,000 to 3.8 per 100,000 (Warner et al., 2009), and adolescents are the age group in which the opioid overdose death rate increased most rapidly. According to data from the 2011 Monitoring the Future survey, 8.7% of 12th graders reported past year illicit use of prescription opioids, especially Vicodin, Percocet, and OxyContin (Johnston et al., 2012). Among adolescents, the National Survey on Drug Use and Health indicates that misuse of opioids is second only to misuse of marijuana in prevalence, and the number of past year initiates of opioid misuse is second only to the number who initiated marijuana use (Substance Abuse and Mental Health Services Administration, 2012c). Research from adults suggests the importance of depression as a risk factor for opioid abuse and dependence (Edlund et al., 2010, 2007; Schepis and Hakes, 2011). In terms of population-attributable
132
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
risk, which is a function of both the strength of the relationship and the frequency of the risk factor, depression may be the most important risk factor for opioid abuse/dependence in adults. The self-medication hypothesis is one prominent explanation for the empirically observed high rates of co-occurring depression and substance abuse, such as with opioids (Khantzian, 1997). This theory posits that individuals use psychoactive substances such as opioids to “self-medicate” painful or disturbing psychological symptoms. Much of the research on opioid abuse/dependence and depression among adolescents is from the National Survey on Drug Use and Health (NSDUH). NSDUH is conducted annually, with approximately 20,000 adolescents aged 12 to 17, is nationally representative, and contains measures of opioid abuse/dependence (American Psychiatric Association, 2000) and nonmedical prescription opioid use (NMPOU), a less severe form of opioid misuse. It also contains a measure of DSM-IV major depressive episode (MDE; American Psychiatric Association, 2000). According to 2012 NSDUH data, 13% of adolescents with MDE reported NMPOU in the past year versus 6% of those without MDE (SAMHSA, 2012d). In an analysis of misuse of any prescription medication (not just opioids) among adolescents using 2008 NSDUH data, major depression was associated with misuse of prescription drugs (OR = 2.60; Havens et al., 2011), but the relationship for NMPOU was not investigated. Wu and colleagues found that there was generally, but not always, an association between depression and opioid abuse/dependence in adolescents (Wu et al., 2008b). Schepis and Krishnan-Sarin (2008), using data on adolescents from the 2008 NSDUH, found MDE to be associated with NMPOU in unadjusted models but did not report data on the relationship for adjusted models. In Vaughn’s latent class analysis of adolescents with NMPOU, lifetime depression was not significantly associated with the latent class of nonmedical opioid use (Vaughn et al., 2012). This study builds upon past work by using NSDUH data to investigate three key issues not previously addressed in the literature. First, among adolescents with both NMPOU and MDE, we investigated the frequency with which the onset of NMPOU precedes the onset of MDE, and the frequency with which the onset of MDE precedes the onset of NMPOU. Although not definitive, temporal progression can give insight into causality. Consistent with the self-medication hypothesis (Khantzian, 1997), we hypothesized that MDE preceding NMPOU would be more common than NMPOU preceding MDE. Second, we used logistic regression to study the determinants of NMPOU among all adolescents and to study the determinants of opioid abuse/dependence among adolescents with NMPOU. These steps represent the development of abuse/dependence—going from no use to NMPOU (experimentation and recreational use) and, among those with NMPOU, from NMPOU to abuse/dependence (addiction). As an extension of the self-medication hypothesis, we posited that the effects of depression would be stronger for the second step than for the first. That is, some adolescents with depression and NMPOU find that opioids help their depression (at least temporarily) and increase their opioid use, progressing to addiction. Finally, we investigated possible moderators of the relationship of MDE with NMPOU and opioid abuse/dependence. We posited that among depressed adolescents, there are “buffering factors” that decrease the magnitude of the association of the relationship of MDE with NMPOU and opioid abuse/dependence. In particular, we hypothesized that the magnitude of the association of MDE with NMPOU and opioid abuse/dependence would be smaller among adolescents with greater parental support, compared with those with less parental support and smaller among those who attend church more frequently, compared with those who attend church less frequently.
2. Methods 2.1. Sample We used cross-sectional data from NSDUH, which the Substance Abuse and Mental Health Services Administration (SAMHSA) conducts annually and is the primary source of information on illicit drug, alcohol, and tobacco use in the United States. Consistent with SAMHSA policy, all sample sizes in the following text and tables have been rounded to the nearest hundred to protect confidentially of respondents. To increase the statistical power, we combined 2008–2012 NSDUH data. We investigated factors associated with NMPOU in the sample of all adolescents aged 12–17 (n = 112,600), and factors associated with opioid abuse/dependence among adolescents aged 12–17 with past year NMPOU (n = 7100). In the total sample, approximately 11,650 had missing data for one of the independent variables. Therefore, three independent variables (past year MDE, number of religious services attended in the past year, grades in last semester or grading period completed) were imputed using weighted sequential hot-deck imputation (Cox, 1980) with imputation classes based on age, race, and gender. After the study was described to participants, informed consent was obtained verbally from parents or guardians, and assent was obtained verbally from the adolescents. Written consent was not obtained because the names of respondents are not used in the screening and interview process. The study was approved by the RTI International IRB. Detailed information on the survey methodology is available in Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings (SAMHSA, 2014).
2.2. Measures Dependent variables: The two outcomes of interest were (1) past year NMPOU and (2) past year opioid abuse/dependence. These represent differing severity of misuse, with NMPOU including individuals with any misuse, and abuse/dependence representing those meeting formal criteria from DSM-IV (American Psychiatric Association, 2000). NMPOU was defined by answers to the following question: “Have you ever, even once, taken (names of prescription opioids) that was not prescribed for you or that you took only for the experience or feeling it caused?” Those answering affirmatively were coded as having NMPOU. Any past year opioid abuse/dependence was defined using standard DSM-IV criteria. Adolescents were also asked about age of first use and most recent use (past 30 days, past year, lifetime), from which the age of onset of NMPOU and past year NMPOU and opioid abuse/dependence were constructed. Independent variables: We identified a set of correlates of NMPOU and opioid abuse/dependence based on results from the literature (Havens et al., 2011; McCabe et al., 2012; Nargiso et al., 2015; Schepis and Krishnan-Sarin, 2008; SAMHSA, 2012b; Wu et al., 2008a; Young et al., 2012). Our independent variables comprised three domains: (1) diagnostic variables; (2) sociodemographic characteristics; (3) and other risk/protective factors. Diagnostic variables: In NSDUH, MDE for adolescents was assessed using a questionnaire based on the depression module in the National Comorbidity Survey—Adolescents (NCS-A, 2010; SAMHSA, 2012a). Alcohol use disorder was defined as meeting criteria for alcohol abuse or alcohol dependence in the past 12 months according to criteria specified in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV; American Psychiatric Association, 2000). We also constructed a variable describing past 12-month nonopioid drug abuse or dependence. Again, this was based on DSM-IV TR criteria, and included
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
133
Table 1 Sample characteristics among adolescents (aged 12–17), among adolescents with Past Year Nonmedical Prescription Opioid Use (NMPOU), and among Adolescents with Past Year Major Depressive Episode (MDE): Percentages, National Survey on Drug Use and Health (NSDUH) 2008 to 2012. Characteristic
Among adolescents aged 12–17
Among past year nonmedical users of opioids aged 12–17
Among adolescents aged 12–17 with past year MDE
Past year NMPOU Yes No
6.1 93.9
N/A N/A
14.2 85.8
Past year opioid abuse or dependence Yes No
0.9 99.1
15.1 84.9
4.1 95.9
Past year MDE Yes No
8.3 91.7
19.5 80.5
N/A N/A
Age group in years 12–13 14–15 16–17
31.6 33.9 34.5
14.9 32.2 52.9
17.4 36.6 46.0
Gender Male Female
51.1 48.9
46.4 53.6
27.7 72.3
Race/ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Non-Hispanic Other
57.2 14.6 20.5 7.7
60.6 13.0 20.0 6.4
59.6 12.7 20.5 7.2
Family income <$20,000 $20,000 to $49,999 $50,000 to $74,999 $75,000+
16.9 31.0 17.4 34.8
19.8 34.5 17.1 28.6
16.7 32.9 18.3 32.2
Rural/urban Rural Urban
17.9 82.1
18.5 81.5
17.6 82.4
4.3 95.7
23.3 76.7
11.5 88.5
4.0 96.0
27.2 72.8
11.1 88.9
Delinquency 0 1 2+
67.4 19.1 13.5
37.3 24.4 38.3
52.1 24.1 23.8
Grade for last semester or grading period completed A+, A, or A− Average B+, B, or B− Average C+, C, or C− Average D or Less Than D Average School Does Not Give These Grades Not in School
30.8 40.6 19.5 5.1 2.9 1.2
17.4 38.6 27.7 10.3 1.9 4.1
25.6 40.4 23.5 6.9 2.1 1.5
How many religious services attended in past year None 1–5 6–24 25–52 52+
33.7 22.4 12.9 13.3 17.7
39.4 26.3 12.2 9.6 12.5
33.3 22.8 14.2 12.3 17.4
Family support/supervision 0–3 4–5 6–7
17.3 33.2 49.5
31.1 37.3 31.5
31.3 37.2 31.5
Past year dependence or abuse status Alcohol abuse or dependence Yes No Abuse or dependence of an illicit drug other than opioids Yes No
N/A = not applicable.
marijuana, cocaine, stimulants, hallucinogens, inhalants, tranquilizers, and sedatives. Sociodemographic variables: Sociodemographic variables included gender, race/ethnicity, family income, and rural/urban
location. Education was not included because it is highly collinear with age in an adolescent population. Other risk/protective factors: Other risk and protective factors included the number of delinquent behaviors, grades in the last
134
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
Table 2 Past year NMPOU among adolescents aged 12–17, by characteristics: National Survey on Drug Use and Health (NSDUH) 2008 to 2012. Characteristic
Past year MDE MDE No MDE
Percentage (SE)
2.9 (0.1) 5.8 (0.2) 9.4 (0.2)
Gender Male Female
5.6 (0.1) 6.7 (0.1)
Race/ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Non-Hispanic Other
6.5 (0.1) 5.4 (0.2) 6.0 (0.2) 5.1 (0.3)
Family income <$20,000 $20,000–$49,999 $50,000–$74,999 $75,000+
7.2 (0.2) 6.8 (0.2) 6.0 (0.2) 5.0 (0.2)
Rural/urban Rural Urban
6.3 (0.2) 6.1 (0.1)
95% Confidence interval of adjusted odds ratio
1.51 1.00
(1.37–1.67) (1.00–1.00)
1.00 1.53 2.08
(1.00–1.00) (1.38–1.69) (1.88–2.30)
1.00 1.36
(1.00–1.00) (1.26–1.47)
1.00 0.73 0.75 0.81
(1.00–1.00) (0.65–0.82) (0.67–0.83) (0.71–0.94)
1.00 1.01 0.95 0.86
(1.00–1.00) (0.91–1.12) (0.84–1.07) (0.77–0.96)
1.00 0.94
(1.00–1.00) (0.86–1.03)
2.54 1.00
(2.26–2.85) (1.00–1.00)
4.63 1.00
(4.14–5.17) (1.00–1.00)
1.00 1.87 3.00
(1.00–1.00) (1.71–2.04) (2.74–3.29)
1.00 1.35 1.52 1.56 1.01 2.08
(1.00–1.00) (1.23–1.49) (1.36–1.68) (1.33–1.81) (0.79–1.29) (1.64–2.63)
1.00 0.99 0.93 0.83 0.83
(1.00–1.00) (0.91–1.08) (0.83–1.04) (0.73–0.94) (0.74–0.93)
1.00 0.84 0.62
(1.00–1.00) (0.77–0.92) (0.57–0.68)
366.0 (2), <.0001
35.2 (1), <.0001
9.2 (3), <.0001
28.5 (3), <.0001
1.0 (1), 0.32
Past year dependence or abuse status Alcohol abuse or dependence Yes No
33.5 (0.9) 4.9 (0.1)
Abuse or dependence of an illicit drug other than opiods Yes No
41.2 (1.0) 4.6 (0.1)
Delinquency 0 1 2+
3.4 (0.1) 7.8 (0.2) 17.4 (0.4)
Grade for last semester or grading period completed A+, A, or A− average B+, B, or B− average C+, C, or C− average D or less than D average School does not give these grades Not in school
3.5 (0.1) 5.8 (0.1) 8.7 (0.3) 12.4 (0.6) 4.0 (0.4) 20.7 (1.5)
Family support/supervision 0–3 4–5 6–7
Adjusted odds ratio
624.0 (1), <.0001 14.1 (0.5) 5.4 (0.1)
Age group in years 12–13 14–15 16–17
How many religious services attended in past year None 1–5 6–24 25–52 52+
Chi-square statistic (df), p value
2585.5 (1), <.0001
3625.2 (1), <.0001
1153.8 (2), <.0001
180.6 (5), <.0001
41.5 (4), <.0001 7.2 (0.2) 7.2 (0.2) 5.8 (0.2) 4.4 (0.2) 4.3 (0.2) 355.0 (2), <.0001 11.0 (0.3) 6.9 (0.2) 3.9 (0.1)
SE = standard error; df = degrees of freedom; MDE = major depressive episode; N/A = not applicable.
semester, encouragement received from family, and number of religious services attended in the past year. Delinquency was based on a count of seven delinquent behaviors in the past year: (1) took part in a serious fight at school or work, (2) took part in a fight in which groups fought groups, (3) carried a handgun, (4) sold illegal drugs, (5) stole or attempted to steal something worth more than $50, (6) attacked someone with serious intent to harm that person, or (7) was ever arrested and booked. These were then summed and classified as 0, 1, or 2+ delinquent behaviors, as in a previous study (Edlund et al., 2015). Grades in the last semester was
measured by a series of mutually exclusive binary variables representing A+, A, or A− average; B+, B, or B− average; C+, C, or C− average; D or less than D average; and dropout/other. In NSDUH, religious attendance in the past year was categorized as 0, 1–2, 3–5, 6–24, 25–52, or 52+ times. Family support and supervision was based on a count of seven items: (1) parents check if homework is done; (2) parents help with homework; (3) parents make adolescents do chores; (4) parents limit amount of television; (5) parents limit time out on school nights; (6) parents tell youths they have done a good job in the past year; and (7) parents tell youths that they
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
135
Table 3 Past year nonmedical use of prescription opioids among adolescents aged 12–17: moderator effects. Odds ratio
95% Confidence interval of odds ratio
Model 1 Age group (in years) × past year MDE MDE vs. No MDE Among persons aged 12–13 Among persons aged 14–15 Among persons aged 16–17
1.53 1.85 1.32
(1.16–2.01) (1.58–2.15) (1.15–1.52)
Model 2 Family support/supervision × past year MDE MDE vs. No MDE Among persons with family support 0–3 Among persons with family support 4–5 Among persons with family support 6–7
1.15 1.71 1.83
(0.97–1.36) (1.47–1.98) (1.52–2.20)
MDE = major depressive episode.
are proud of things done in the past year. As items were endorsed affirmatively by large majorities, the variable is coded as 0 to 3, 4 to 5, or 6 to 7. Using factor analysis, we confirmed that these variables loaded on a single factor rather than on two or more distinct factors. 2.3. Data analysis For regression analyses, we used two logistic regressions. In the first, we investigated the association, among all adolescents, of MDE and other covariates with NMPOU. In the second, we investigated the association, among adolescents with NMPOU, of MDE and other covariates with opioid abuse/dependence. For our analyses of moderators, a separate model was implemented for each moderator through the inclusion of the appropriate interaction term(s) of that moderator and the MDE variable. For interactions that were significant, we investigated contrasts between levels of interaction to further explain and describe the relationship between dependent variables and these moderators. All analyses combined NSDUH data from the 2008–2012 analytic files and used SUDAAN® (Research Triangle Institute, 2012) to account for the complex sample design. 3. Results 3.1. Any NMPOU Among all adolescents, 6.1% reported NMPOU in the past year, and 8.3% had past year MDE (Table 1). Among those with both lifetime MDE and lifetime NMPOU, 47% had the onset of MDE before the onset of NMPOU, 27.2% had the onset of NMPOU before the onset of MDE, and 25.8% had the onset in the same year (results not shown). In an adjusted logistic regression, MDE was significantly associated with NMPOU (OR = 1.51, 95% CI = 1.37–1.67; Table 2). Adolescents with a greater likelihood of NMPOU included those who were older, white, female, and those who had a nonopioid drug use disorder, an alcohol use disorder, lower family income, lower family support, lower grades, more delinquent behaviors, and less religious attendance (Table 2). The effects of major depression on NMPOU were moderated by family support (Wald F = 8.6, 2 df, p < 0.01), and age (Wald F = 5.2, 2 df, p < 0.01). The association between major depression and NMPOU was only significant for adolescents with moderate or high levels of family supervision; the association was not significant for adolescents with low levels of family supervision (OR = 1.15, 95% CI = 0.97–1.36) (Table 3, which reports results for statistically significant moderators). The association of MDE with NMPOU was significant for all adolescent age groups, and was largest for adolescents aged 14–15 (OR = 1.85, 95% CI = 1.58–2.15), followed by
adolescents aged 12–13 (OR = 1.53, 95% CI = 1.16–2.01) and adolescents aged 16–17 (OR = 1.32, 95% CI = 1.15–1.52). 3.2. Opioid abuse/dependence Among the sample with NMPOU, 15.1% had past year opioid abuse/dependence and 19.5% had past year MDE (Table 1). In an adjusted logistic regression, MDE was associated with opioid abuse/dependence (OR = 2.18, 95% CI = 1.77–2.70) (Table 4). Adolescents who had a greater likelihood of opioid abuse or dependence included those who were female, and those who had a nonopioid drug use disorder, an alcohol use disorder, less family support, more delinquent behaviors, and greater frequency of religious service attendance (fifty-two or more times per year). There were no variables that were statistically significant moderators of the depression and opioid abuse/dependence relationship. 4. Discussion Past year MDE was common among adolescents (8%), and among adolescents with NMPOU (20%). MDE was significantly associated with both NMPOU among the total sample of adolescents, and opioid abuse/dependence among those with NMPOU. As hypothesized, among individuals with comorbid depression and NMPOU, depression usually preceded the NMPOU. This suggests the importance of MDE in terms of population attributable risk for NMPOU and opioid abuse/dependence, which is a function both of prevalence and the magnitude of the association. The potentially sizable excess risk of opioid use and abuse/dependence in the total adolescent population potentially attributable to MDE may be underappreciated in the literature; two recent reviews of factors associated with non-medical use of prescriptions drugs among youth made no mention of depression (Nargiso et al., 2015; Young et al., 2012). As hypothesized, the OR measuring the magnitude of the association between MDE and opioid abuse/dependence among adolescents with NMPOU (OR = 2.30) was greater than the OR measuring the association between MDE and NMPOU among all adolescents (OR = 1.53). We believe this may provide indirect evidence for the self-medication hypothesis. We hypothesized that the magnitude of the association of MDE with NMPOU and opioid abuse/dependence would be smaller among adolescents with greater parental support, compared with those with less parental support, and smaller among those who attend church more frequently, compared with those who attend church less frequently, because these factors would buffer the effects of MDE. However, we found no evidence to support these hypotheses. The religious attendance–MDE interactions were nonsignificant. The parental support–MDE interaction was not
136
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
Table 4 Past year abuse opioid abuse/dependence among adolescents aged 12–17 with past year nonmedical prescription opioid use (NMPOU), by characteristics: National Survey on Drug Use and Health (NSDUH) 2008 to 2012. Characteristic
Percentage (SE)
Past year MDE MDE No MDE
28.4 (1.7) 11.9 (0.5)
Age group in years 12–13 14–15 16–17
10.7 (1.2) 16.2 (1.0) 15.6 (0.8)
Gender Male Female
12.3 (0.8) 17.5 (0.8)
Race/ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Non-Hispanic Other
15.9 (0.7) 10.1 (1.3) 15.5 (1.4) 16.6 (2.1)
Family income <$20,000 $20,000–$49,999 $50,000–$74,999 $75,000+
17.4 (1.3) 13.4 (0.9) 14.7 (1.3) 15.8 (1.1)
Rural/urban Rural Urban
15.5 (1.2) 15.0 (0.6)
Past year dependence or abuse status Alcohol abuse or dependence Yes No Abuse or dependence of an illicit drug other than pain relievers Yes No
Chi-square statistic (df), p value
Adjusted odds ratio
95% Confidence interval of adjusted odds ratio
2.18 1.00
(1.77–2.70) (1.00–1.00)
1.00 1.03 0.91
(1.00–1.00) (0.76–1.40) (0.68–1.22)
1.00 1.39
(1.00–1.00) (1.13–1.70)
1.00 0.85 1.03 1.03
(1.00–1.00) (0.60–1.20) (0.79–1.34) (0.70–1.50)
1.00 0.72 0.77 0.88
(1.00–1.00) (0.56–0.93) (0.57–1.04) (0.67–1.17)
1.00 0.94
(1.00–1.00) (0.74–1.19)
1.89 1.00
(1.53–2.32) (1.00–1.00)
2.73 1.00
(2.23–3.34) (1.00–1.00)
1.00 1.16 1.71
(1.00–1.00) (0.88–1.54) (1.34–2.19)
1.00 0.98 1.14 1.31 1.45 1.30
(1.00–1.00) (0.74–1.30) (0.84–1.53) (0.94–1.84) (0.78–2.70) (0.80–2.12)
1.00 1.11 1.09 0.99 1.60
(1.00–1.00) (0.88–1.39) (0.77–1.52) (0.68–1.45) (1.18–2.19)
1.00 0.73 0.62
(1.00–1.00) (0.59–0.90) (0.49–0.79)
120.2 (1), <.0001
6.1 (2), 0.002
19.4 (1), <.0001
3.9 (3), 0.009
2.5 (3), 0.06
0.1 (1), 0.74
183.1 (1), <.0001 29.4 (1.4) 10.7 (0.6) 269.4 (1), <.0001 30.6 (1.3) 9.3 (0.5)
Delinquency 0 1 2+
8.9 (0.7) 12.9 (1.1) 22.5 (1.0)
53.6 (2), <.0001
Grade For Last Semester Or Grading Period Completed A+, A, or A− average B+, B, or B− average C+, C, or C− average D or less than D average School does not give these grades Not in school
12.3 (1.2) 12.7 (0.8) 16.1 (1.1) 22.6 (2.0) 17.4 (3.9) 21.8 (3.2)
How many religious services attended in past year None 1–5 6–24 25–52 52+
14.7 (0.8) 15.9 (1.1) 14.3 (1.7) 12.7 (1.8) 17.0 (1.7)
Family support/supervision 0–3 4–5 6–7
20.8 (1.1) 13.9 (0.9) 10.8 (0.8)
7.6 (5), <.0001
1.0 (4), 0.43
30.3 (2), <.0001
SE = standard error; df = degrees of freedom; MDE = major depressive episode; N/A = not applicable.
significant in the regression predicting opioid abuse/dependence, and although the parental support–MDE interaction was significant in the regression predicting NMPOU, the effects of MDE were larger among those with greater parental support—the opposite of what we hypothesized. We know of no explanation in the literature for this finding. Further exploration of the effect of parental support on opioid use and dependence in the context of other risk and protective factors is warranted. Although the primary purpose of our study was to investigate the relationship of MDE with NMPOU and opioid abuse/
dependence, we believe that our other findings are important for the epidemiology of adolescent NMPOU and opioid abuse/ dependence. In a secondary analysis of Monitoring the Future data, McCabe et al. (2012) found no statistically significant differences in NMPOU between males and females. Some secondary analyses of NSDUH data have found that adolescent females are more likely to have NMPOU (Schepis and Krishnan-Sarin, 2008; Sung et al., 2005; Wu et al., 2008a, 2008b), whereas other analyses have found no statistical differences (Ford, 2008a, 2008b, 2009; Ford and Lacerenza, 2011; Simoni-Wastila et al., 2008). Our results provide strong
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
evidence that in adjusted models, adolescent females are more likely to have NMPOU, and among those with NMPOU, adolescent females are more likely to have opioid abuse/dependence than adolescent males. This is an important finding for clinicians, who may view adolescent boys as higher risk for opioid abuse/dependence, based on the large body of epidemiological research that suggests that males are in general at a higher risk for substance use disorders than females. Our results suggest that clinicians should maintain an equal, and perhaps even higher, index of suspicion for opioid abuse/dependence among girls than boys. Further, when there is an OUD, clinicians need to be aware that it is often comorbid with MDE. Our results illustrate that a two-step analytical approach, which first focuses on NMPOU, and then opioid abuse/dependence among those with NMPOU, may give a much more detailed understanding of the development of adolescent opioid abuse/dependence. For example, in the case of age, older adolescents were much more likely than younger adolescents to have NMPOU, which may indicate greater exposure to opioids and more experimentation. However, among adolescents with NMPOU, in adjusted models, older adolescents were not more likely than younger adolescents to have addiction (abuse/dependence). Thus, younger adolescents are significantly less likely than older adolescents to experiment, but those who do are as likely to progress to abuse/dependence as older adolescents. Because of this, opioid abuse/dependence is not necessarily rare among young adolescents. Among adolescents with opioid abuse/dependence, 1 out of 9 is aged 12 to 13, and about 1 out of 330 12 to 13 year olds have opioid abuse/dependence. This is particularly worrisome because earlier onset of drug abuse is associated with greater severity of disorder, and is usually more refractory to treatment. Although we know of no long-term studies of outcomes in individuals who develop opioid abuse/dependence in early adolescence, outcomes are likely extremely poor in all domains (i.e., physical and behavioral health, education, employment, and criminal justice system involvement). Our findings illustrate that efforts to prevent and treat opioid experimentation and addiction need to focus on even the youngest adolescents. Adolescents with less religious service attendance were more likely to have NMPOU, which is consistent with past studies that have found that those with lower levels of religiosity or religious attendance are less likely to abstain from alcohol and drugs (Michalak et al., 2007). On the other hand, among adolescents with NMPOU, we found that those who attended 52+ religious services per year were more likely to have opioid abuse or dependence than adolescents with NMPOU who attended religious services less frequently. The NSDUH does not include sufficient religiosity measures to allow us to determine why this is, and there are many possible causes. For example, prescription opioid abuse/dependence might be relatively less stigmatized than alcohol or illegal drug/abuse dependence among those with regular church attendance. Additional research focused on religiosity is needed to better understand this relationship. Several limitations of our study deserve mention. First, although NSDUH asks respondents about age of onset, ultimately the data is cross-sectional and not longitudinal; thus, our results should be viewed as associations, not causal. Second, NSDUH does not allow us to determine whether MDE was caused by major depressive disorder or bipolar disorder, although we note that in adolescents, major depressive disorder is about 3 times more common than bipolar disorder (Kessler et al., 2012). Third, NSDUH does not survey those younger than 12, and we should not assume that opioid misuse does not extend into preadolescents. In summary, MDE is associated with NMPOU and opioid abuse/dependence. When NMPOU and MDE are comorbid, MDE usually precedes the MDE. In terms of population-attributable risk,
137
which is a function of both the prevalence and the strength of the association, MDE is a key risk factor for NMPOU and opioid abuse/dependence. Preventive and clinical programs to decrease NMPOU and opioid abuse/dependence among adolescents should consider the prominent role of depression. Author disclosures Role of funding source The National Survey on Drug Use and Health (NSDUH) is funded by the Substance Abuse and Mental Health Services Administration (SAMHSA), Center for Behavioral Health Statistics and Quality (CBHSQ). This study was funded under Contract No. 283-201000003C, Project No. 0212800, which was supported by funding from the National Institute of Mental Health (NIMH). The study utilized data previously collected by SAMHSA. Co-authors on the paper were from SAMHSA (Dr. Lipari) and NIMH (Dr. Colpe), and they assisted with study design, analysis and interpretation of data, writing of the manuscript, and the decision to submit the manuscript for publication. Contributors Mark J. Edlund. Study design, supervision of statistical analyses, analysis and interpretation of data, writing of initial draft and revisions of manuscript. Valerie L. Forman-Hoffman. Study design, supervision of statistical analyses, analysis and interpretation of data, writing of manuscript. Cherie R. Winder. Statistical programming, analysis and interpretation of data, writing of manuscript. David C. Heller Statistical programming, analysis and interpretation of data, writing of manuscript. Larry A. Kroutil Study design, analysis and interpretation of data, writing of manuscript. Rachel N. Lipari Study design, analysis and interpretation of data, writing of manuscript. Lisa J. Colpe Study design, analysis and interpretation of data, writing of manuscript. All authors have reviewed and approved the final version of the manuscript. Conflict of interest All authors declare that they have no conflicts of interest. Acknowledgments We thank SAMHSA and NIMH, who provided funding for this study. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Substance Abuse and Mental Health Services Administration. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute of Mental Health. References American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental Disorders, Text Revision. American Psychiatric Association, Washington, DC. Banta-Green, C.J., 2012. Adolescent abuse of pharmaceutical opioids raises questions about prescribing and prevention. Arch. Pediatr. Adolesc. Med. 166, 865–866. Centers for Disease Control and Prevention, 2011. Vital signs: overdoses of prescription opioid pain relievers—United States, 1999–2008. MMWR 60, 1487–1492. Compton, W.M., Volkow, N.D., 2006. Major increases in opioid analgesic abuse in the United States: concerns and strategies. Drug Alcohol Depend. 81, 103–107.
138
M.J. Edlund et al. / Drug and Alcohol Dependence 152 (2015) 131–138
Cox, B.G., 1980. The weighted sequential hot deck imputation procedure. In: Proceedings of the 1980 American Statistical Association, Survey Research Methods Section. American Statistical Association, Houston, TX, pp. 721–726. Edlund, M.J., Martin, B.C., Fan, M.Y., DeVries, A., Braden, J.B., Sullivan, M.D., 2010. Risks for opioid abuse and dependence among recipients of chronic opioid therapy: results from the TROUP Study. Drug Alcohol Depend. 112, 90–98. Edlund, M.J., Pettiford, A.G., Hampton, J., Hoffman, V., Ault, K.L., Colpe, L., Hedden, S.L., 2015. Adolescents’ assessments of the helpfulness of treatment for major depression: results from a national survey. Psych. Serv. (in press). Edlund, M.J., Steffick, D., Hudson, T., Harris, K.M., Sullivan, M., 2007. Risk factors for clinically recognized opioid abuse and dependence among veterans using opioids for chronic non-cancer pain. Pain 129, 355–362. Executive Office of the President of the United States, 2011. Epidemic: Responding to America’s Prescription Drug Abuse Crisis, Washington, DC. Ford, J.A., 2008a. Nonmedical prescription drug use and delinquency: an analysis with a national sample. J. Drug Issues 38, 493–516. Ford, J.A., 2008b. Social learning theory and nonmedical prescription drug use among adolescents. Sociol. Spectr. 28, 299–316. Ford, J.A., 2009. Nonmedical prescription drug use among adolescents: the influence of bonds to family and school. Youth Soc. 40, 336–352. Ford, J.A., Lacerenza, C., 2011. The relationship between source of diversion and prescription drug misuse, abuse, and dependence. Subst. Use Misuse 46, 819–827. Havens, J.R., Young, A.M., Havens, C.E., 2011. Nonmedical prescription drug use in a nationally representative sample of adolescents: evidence of greater use among rural adolescents. Arch. Pediatr. Adolesc. Med. 165, 250–255. Johnston, L.D., O’Malley, P.M., Bachman, J.G., Schulenberg, J.E., 2012. Monitoring the Future National Survey Results on Drug Use, 1975–2011. Volume I Secondary School Students, http://www.monitoringthefuture.org/pubs/ monographs/mtf-vol1 2011.pdf (accessed on December 8, 2014). Kessler, R.C., Avenevoli, S., Costello, E.J., Georgiades, K., Green, J.G., Gruber, M.J., He, J.P., Koretz, D., McLaughlin, K.A., Petukhova, M., Sampson, N.A., Zaslavsky, A.M., Merikangas, K.R., 2012. Prevalence, persistence, and sociodemographic correlates of DSM-IV disorders in the National Comorbidity Survey Replication Adolescent Supplement. Arch. Gen. Psychiatry 69, 372–380. Khantzian, E.J., 1997. The self-medication hypothesis of substance use disorders: a reconsideration and recent applications. Harv. Rev. Psychiatry 4, 231–244. McCabe, S.E., West, B.T., Teter, C.J., Boyd, C.J., 2012. Medical and nonmedical use of prescription opioids among high school seniors in the United States. Arch. Pediatr. Adolesc. Med. 166, 797–802. Michalak, L., Trocki, K., Bond, J., 2007. Religion and alcohol in the U.S. National Alcohol Survey: how important is religion for abstention and drinking. Drug Alcohol Depend. 87, 268–280. Nargiso, J.E., Ballard, E.L., Skeer, M.R., 2015. A systematic review of risk and protective factors associated with nonmedical use of prescription drugs among youth in the United States: a social ecological perspective. J. Stud. Alcohol Drugs 76, 5–20. NCS-A, 2010. National Comorbidity Survey—Adolescents, http://www.hcp.med. harvard.edu/ncs/ (accessed on November 25, 2012). Paulozzi, L.J., Budnitz, D.S., Xi, Y., 2006. Increasing deaths from opioid analgesics in the United States. Pharmacoepidemiol. Drug Saf. 15, 618–627. Paulozzi, L.J., Xi, Y., 2008. Recent changes in drug poisoning mortality in the United States by urban–rural status and by drug type. Pharmacoepidemiol. Drug Saf. 10, 997–1005, http://dx.doi.org/10.1002/pds.1626
Research Triangle Institute, 2012. SUDAAN Language Manual, Release 11.0. Research Triangle Institute, Research Triangle Park, NC. Schepis, T.S., Hakes, J.K., 2011. Non-medical prescription use increases the risk for the onset and recurrence of psychopathology: results from the National Epidemiological Survey on Alcohol and Related Conditions. Addiction 106, 2146–2155. Schepis, T.S., Krishnan-Sarin, S., 2008. Characterizing adolescent prescription misusers: a population-based study. J. Am. Acad. Child Adolesc. Psychiatry 47, 745–754. Simoni-Wastila, L., Yang, H.W., Lawler, J., 2008. Correlates of prescription drug nonmedical use and problem use by adolescents. J. Addict. Med. 2, 31–39. SAMHSA, 2012a. Results from the 2010 National Survey on Drug Use and Health: Mental Health Findings. HHS Publication (NSDUH Series H-42, HHS Publication No. SMA 11-4667). http://www.samhsa.gov/data/nsduh/ 2k10MH Findings/2k10MHResults.htm (accessed on January 6 2014). SAMHSA, 2012b. Results from the 2011 National Survey on Drug Use and Health: Mental Health Findings. HHS Publication (NSDUH Series H-45, HHS Publication No. SMA 12-4725). http://www.samhsa.gov/data/NSDUH/ 2k11MH FindingsandDetTables/2K11MHFR/NSDUHmhfr2011.htm (accessed on January 6 2014). SAMHSA, 2012c. Results from the 2011 National Survey on Drug Use and Health: Summary of National Findings. HHS Publication (NSDUH Series H44, HHS Publication No. SMA 12-4713). http://www.samhsa.gov/data/NSDUH/ 2k11Results/NSDUHresults2011.htm (accessed on January 6 2013). SAMHSA, 2012. Results from the 2011 National Survey on Drug Use and Health: Detailed Tables. Prevalence Estimates, Standard Errors, P Values, and Sample Sizes, http://www.samhsa.gov/data/NSDUH/ 2011SummNatFindDetTables/NSDUH-DetTabsPDFWHTML2011/2k11Detailed Tabs/Web/PDFW/NSDUH-DetTabsCover2011.pdf (accessed on January 6 2013). SAMHSA, 2014. Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings. HHS Publication (NSDUH Series H-48, HHS Publication No. (SMA) 14-4863). http://www.samhsa.gov/data/sites/default/files/ NSDUHresultsPDFWHTML2013/Web/NSDUHresults2013.pdf (accessed on April 7 2015). Sung, H.E., Richter, L., Vaughan, R., Johnson, P.B., Thom, B., 2005. Nonmedical use of prescription opioids among teenagers in the United States: trends and correlates. J. Adolesc. Health 37, 44–51. Vaughn, M.G., Fu, Q., Perron, B.E., Wu, L.T., 2012. Risk profiles among adolescent nonmedical opioid users in the United States. Addict. Behav. 37, 974–977. Warner, M., Chen, L.H., Makuc, D.M., 2009. Increase in fatal poisonings involving opioid analgesics in the United States, 1999–2006. In: NCHS Data Brief. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics. Wu, L.T., Pilowsky, D.J., Patkar, A.A., 2008a. Non-prescribed use of pain relievers among adolescents in the United States. Drug Alcohol Depend. 94, 1–11. Wu, L.T., Ringwalt, C.L., Mannelli, P., Patkar, A.A., 2008b. Prescription pain reliever abuse and dependence among adolescents: a nationally representative study. J. Am. Acad. Child Adolesc. Psychiatry 47, 1020–1029. Young, A.M., Glover, N., Havens, J.R., 2012. Nonmedical use of prescription medications among adolescents in the United States: a systematic review. J. Adolesc. Health 51, 6–17.