Chasing the Bean: Prescription Drug Smoking among Socially Active Youth

Abstract

Background

Alternative consumption practices of prescription drug misuse have been less well monitored than general prevalence. We describe prescription drug smoking among socially active youth and highlight correlates of this practice. We also examine its association with drug problems, drug dependence, and mental health.

Methods

We surveyed 404 young adults recruited from nightlife venues in New York via time-space sampling. We use linear and logistic regression models to examine the probability of smoking prescription drugs and its association with drug problems, dependence, and mental health. Qualitative findings supplement the survey data.

Results

Males have higher odds than females (OR=3.4) and heterosexuals have higher odds than sexual minority youth (OR=2.3) of smoking prescription drugs. Those involved in Electronic Dance Music nightlife have higher odds (OR=2.1) compared to those who do not participate in that scene, while those in college bar scenes have lower odds (OR=0.4) of having smoked prescription drugs. Prescription drug smokers report more drug problems (β=0.322) and greater symptoms of dependence (β=0.298) net of the frequency of misuse and other characteristics. Prescription drug smokers do not report greater mental health problems. Qualitative interview data support these survey findings.

Conclusions

Prescription drug smoking is a significant drug trend among socially active youth. It is associated with drug problems and symptoms of dependence net of frequency of misuse. Prevention and intervention efforts for youth who misuse prescription drugs should address the issue of prescription drug smoking, and this may be an area for clinicians to address with their adolescent patients.

INTRODUCTION

Epidemiological systems have monitored the 21^st century trend of prescription drug misuse.¹ The misuse of prescription drugs includes consuming a prescription drug in a way other than as prescribed by a doctor, consuming prescription drugs from a non-medical source, or consuming prescription drugs for non-medical or recreational purposes. The prevalence of prescription drug misuse is especially high among youth.² Studies have shown the misuse of prescription drugs to be common among college students^3–5 and street-based youth,^{6, 7} and almost half of nightlife involved youth report misusing prescription drugs.⁸ In this regard, the 21^st century trend of prescription drug misuse is a wide reaching phenomenon among diverse groups of young people. Yet, the emergence of alternative prescription drug misuse practices beyond oral consumption has been poorly monitored.

Prescription drug smoking includes the inhalation of prescription pills through smoking or vaporization. The transition to smoking prescription pills is a significant act for young people in that it represents an escalation of drug use. The normative manner of drug consumption for pills is oral consumption, an act reinforced by the standard pattern of use as pharmaceuticals. As such, pill smoking represents a more extreme form of prescription drug misuse, one that decreases the time to onset of drug effects and increases the intensity of the high.⁹

The escalation of route of administration is noteworthy in that it has been tied to drug dependence and social problems for users of cocaine and heroin^{10, 11} as well as infectious disease risk.^{12, 13} Additionally, the escalation of drug use practices is an indication of dependence on that substance.^14–16 Studies that have examined the escalation of route of administration have primarily focused only on transitions to injection drug use among illicit drug users.^{11, 17–19} Considerations of escalation to smoking prescription pills therefore remains important in working towards health promotion among youth.

In this paper, we present an overview of prescription drug smoking among youth. We have three primary goals. First, using survey data, we describe this trend among young prescription drug misusers and highlight correlates of this practice. Second, we examine the association of prescription drug smoking with drug problems, drug dependence, and mental health. Finally, we utilize qualitative data to describe this practice from the point of view of young prescription drug misusers. Collectively, these analyses provide a descriptive epidemiological profile of pill smoking among young prescription drug misusers.

METHODS

Sampling

To generate our sample, we utilized time-space sampling in nightlife venues in New York City (NYC), supplemented briefly by targeted online recruitment. Time-space sampling was first developed to capture hard-to-reach populations,^20–22 but is also useful to generate samples of venue-based populations.²³ As such, we used venues as our basic sampling unit to systematically generate a sample of socially active youth.

We randomized “time” and “space” – the days/times for sampling individuals and the venues attended – using a sampling frame of venues and times of operation. To construct the sampling frame, ethnographic fieldwork during the previous twelve months enabled the assessment of “socially viable” venues for each day of the week. A venue was viable if a preponderance of youth patron traffic existed on that day of the week. We generated lists of venues for each day of the week across several key nightlife scenes – e.g. electronic dance music (EDM) clubs and indie rock clubs. For each day of the week, socially viable venues were listed and assigned a number. Using a random digit generator, a random number was drawn corresponding to a particular venue on a particular day, yielding our schedule for recruitment.

Once at the venue, research staff attempted to survey as many individuals as possible, aiming to achieve saturation at the venue. The number of staff depended upon the size of the venue. Staff approached a patron, identified themselves, described the survey, and requested verbal consent for participation in the anonymous survey conducted on an iPod Touch^®. For those who provided consent, the surveys were initially administered by trained staff (consent, age, and NYC residency) and respondents self-reported more sensitive information (race, sexual identity, gender, and substance use). Staff were trained not to administer surveys to individuals who were impaired by intoxication to ensure the capacity to consent. Screening survey response rates (75.0%) were high considering the difficult conditions of nightclub settings and the lack of compensation. Survey software determined whether the screened individual was eligible for the study (9.4% of those screened met inclusion criteria). A majority of those deemed eligible (77.4%) provided contact information for study participation.

Near the end of the project, venue recruitment was supplemented by targeted recruitment via online groups associated with nightlife scenes of interest. We first identified web-based groups relevant to the youth cultures of interest using a form of web-based social mapping. Group members who were between the ages of 18–29 and resided in the NYC metropolitan area saw an advertisement for the study; if they clicked on it, they were directed to a Qualtrics® survey that screened for eligibility and, if eligible, collected their contact information. Less than 5% of the sample was recruited via this supplemental method.

Regardless of recruitment method, staff contacted participants by phone and e-mail to provide more information about the study, confirm eligibility, and schedule the initial assessment. Eligibility criteria were as follows: (1) Aged 18–29; (2) Reported the misuse of prescription drugs at least three times in the past six months; and (3) Reported the misuse of prescription drugs at least once in the past three months. The conformation of age and identity via ID diminished the possibility of duplicate assessments. During their initial assessment, participants completed the informed consent process and the survey. Half of the participants also completed a qualitative interview. Participants were compensated $50 in cash, check, or Amazon.com gift card (depending on their preference) upon completion of the assessment, whether both modules or survey only. All procedures were reviewed and approved by university Institutional Review Boards.

Measures

Prescription Drug Misuse and Mode of Administration

We used the following operational definition of prescription drug misuse, which was provided to subjects: “…using prescription drugs obtained from a non-medical source, using more than the prescribed dose, or using prescription drugs for a non-medical or recreational purpose. Non-medical use may occur whether you do or do not have a prescription for that drug” (based on Compton & Volkow⁹). Respondents reported their frequency of misuse of three prescription drug types (pain killers, sedatives, and stimulants) during the previous three months, with examples provided for each type. Respondents also self-reported whether they had misused prescription drugs through smoking. For this article, we use a dichotomous variable indicating whether the respondent had ever smoked prescription drugs.

Drug Problems, Dependence, and Mental Health

The Composite International Diagnostic Interview (CIDI) Substance Abuse Module was tailored to assess symptoms of drug dependence related to prescription drug misuse.²⁴ This 8-item measure is widely used to assess symptoms of drug dependence. The Short Inventory of Problems with Alcohol and Drugs (SIP-AD) was tailored to assess problems associated with prescription drug misuse. The SIP-AD is a 15-item inventory of problems associated with substance use.²⁵ The Brief Symptom Inventory (BSI-18) was used to capture symptoms of mental health problems over three domains (depression, anxiety, and somatization) using the domain specific subscales. Research demonstrates that the BSI-18 is a strong measure of mental health symptoms among drug users.²⁶

Demographics

Participants self-reported age, gender, sexual identity, race/ethnicity (recoded as White vs. non-White), highest education completed, and parental socio-economic status.

Nightlife Scenes

We also include the involvement in particular nightlife scenes associated with various youth cultures. Specifically, respondents self-reported involvement in the following 7 nightlife scenes during the previous 3 months: Indie Rock, EDM, Hip Hop, Jam Band, Lesbian Nightlife, Gay Nightlife, or College Bars. These categories are not mutually exclusive.

Statistical Modeling

When examining the probability of smoking prescription drugs, we use logistic regression models, with nested regressions that subsequently add demographics, prescription drug use, and nightlife participation. For the drug problems, dependence, and mental health scales, we use linear regression to examine the effect of prescription drug smoking while controlling for demographics and frequency of prescription drug misuse. The linear models use the log of each of these five scales, as indicated as necessary by regression diagnostics.

Qualitative Interviews

We qualitatively interviewed the first 214 of 404 participants, which ensured a sufficient yet manageable number of interviews. The qualitative data for this paper come from interviews with 41 participants who reported smoking or vaporizing prescription drugs. All interviews were conducted in private rooms to ensure confidentiality. The semi-structured interview contained critical incident measures to draw out specific narratives about participants’ prescription drug misuse practices.²⁷ Critical incident measures reduce recall bias and provide context for behaviors rooted within specific events.²⁸ Participants were asked to provide information about the first and last time they smoked a prescription drug type. Interviewers probed around the techniques used to smoke prescription drugs, the motivations behind this route of administration, and what participants liked and did not like about smoking prescription drugs. All interviews were recorded and transcribed.

A thematic analysis of the data generated descriptive and contextual understandings of the patterns present in young adults’ practices of smoking prescription drugs.²⁹ The analyses primarily focused on coding narratives of initiation into prescription drug smoking and the most recent prescription drug smoking event. A goal of thematic analysis is to identify general sentiments across the sample while leaving outlying individual assertions aside. The thematic analyses were performed utilizing a codebook constructed in NVIVO data analysis software. Subsequent waves of coding and analysis were conducted to facilitate more nuanced understandings of prescription drug smoking. We use these qualitative data to describe the practices and contexts of prescription drug smoking as well as to explicate our findings from the survey data.

RESULTS

Sample characteristics are presented in Table 1. The respondents had an average age of 24.6 years. The sample had slightly more males than females and two-thirds were White. Many were either currently enrolled in college or had completed a college degree. The smoking of prescription drugs was common among the youth in our study. Almost one out of five (18.1%) reported smoking prescription drugs.

Table 1.

Descriptive statistics of the sample

Demographic characteristics	Total (N = 404)
Age (mean, sd)	24.57 (2.69)
Sex
Male	220 (54 .73%)
Female	182 (45.27%)
Race
White	270 (66 .83%)
Nonwhite	134 (33.17%)
Educational achievement
High school or less	28 (6. 93%)
Some college or Associates degree	65 (16.09%)
Currently enrolled in college	83 (20.54%)
Bachelor's degree or higher	228 (56.44%)
Sexual orientation
Heterosexual	210 (51 .98%)
LGBTQ	194 (48.02%)
Parental class
Rich	20 (4. 99%)
Upper middle class	132 (32.92%)
Middle class	157 (39.15%)
Working class/poor	92 (22.94%)
Substance use characteristics
Past 90 days prescription drug misuse
Sedatives (mean, sd)	11.94 ( 19.99)
Pain killers (mean, sd)	10.03 (18.24)
Stimulants (mean, sd)	12.93 (21.28)
Ever smoked prescription drug	73 (18.07%)
CIDI score (mean, sd)	2.23 (2.23)
SIP-AD score (mean, sd)	5.01 (6.63)
BSI depression score (mean, sd)	6.25 (5.51)
BSI anxiety score (mean, sd)	6.18 (4.98)
BSI somatization score (mean, sd)	3.96 (3.92)
Participation in club scenes (n, %)
Indie scene	252 (62 .38%)
EDM scene	212 (52.48%)
Hip hop scene	108 (26.73%)
Hippie scene	109 (26.98%)
Lesbian scene	87 (21.53%)
Gay scene	170 (42.08%)
College scene	161 (39.85%)

Survey Results

Table 2 presents models, in which we consider three nested logistic regressions to understand the association of smoking prescription drugs with demographics, frequency of recent prescription drug misuse, and subcultural nightlife scenes, respectively. In Model 1, the effect for males on smoking is statistically significant, and maintains significance with the additional controls for recent misuse and nightlife scenes in Models 2 and 3, respectively. From Model 3, with the full complement of controls, males have 3.4 times greater odds than females to have smoked prescription drugs (p < .001). The effect of identifying as heterosexual relative to sexual minority identity is also significant in Model 1, with the former having 2.3 times greater odds to have smoked prescription drugs (p < .01). This effect, however, is reduced to non-significance when controlling for the nightlife scenes in Model 3, most likely due to its strong association with participation in gay and lesbian nightlife scenes.

Table 2.

Logistic regression predicting probability of having smoked prescription drugs.

	Model 1		Model 2		Model 3
	Coef. (SE)	O.R.	Coef. (SE)	O.R.	Coef. (SE)	O.R.
Age	−0.013 (0.061)	0.987	−0.048 (0.066)	0.953	−0.041 (0.07)	0.960
Parents' class
Upper middle class	−0.402 ( 0.646)	0.669	−0.357 (0.694)	0.700	−0.336 (0.721)	0.715
Middle class	−0.190 (0.636)	0.827	−0.047 (0.678)	0.954	−0.015 (0.707)	0.985
Working class/poor	0.500 (0.668)	0.606	−0.305 (0.711)	0.737	−0.365 (0.742)	0.694
Educational achievement
Some college or Associates degree	−0.120 ( 0.563)	0.887	−0.028 (0.598)	0.973	−0.106 (0.623)	0.899
Currently in college	−0.160 (0.542)	0.852	0.064 (0.572)	1.066	0.207 (0.611)	1.230
Bachelor's degree or higher	−0.863 (0.530)	0.422	−0.466 (0.574)	0.628	−0.537 (0.604)	0.585
Nonwhite	−0.365 (0.314)	0.694	−0.435 (0.327)	0.647	−0.355 (0.346)	0.701
Male	1.024***(0.302)	2.786	1.095***(0.320)	2.988	1.211***(0.353)	3.357
Heterosexual	0.806**(0.285)	2.238	0.881**(0.299)	2.413	0.597 (0.400)	1.817
Sedatives days of misuse (past 90 days)			0.014 (0.008)	1.015	0.012 (0.008)	1.012
Pain killers days of misuse (past 90 days)			0.013 (0.008)	1.013	0.017*(0.008)	1.017
Stimulants days of misuse (past 90 days)			0.009 (0.006)	1.009	0.009 (0.006)	1.009
Indie scene					0.456 (0.333)	1.578
EDM scene					0.754*(0.335)	2.126
Hip hop scene					−0.209 (0.356)	0.812
Hippie scene					−0.168 (0.354)	0.845
Lesbian scene					0.144 (0.492)	1.154
Gay scene					−0.518 (0.403)	0.596
College scene					−0.910**(0.336)	0.403
Constant	−1.37 (1.616)		−1.465 (1.693)		−1.701 (1.867)
Log likelihood	−173.609		−164.530		−156.317
Model chi-squared	29.50***		47.66***		64.08***

^*p < .05;

^**p < .01;

^***p < .001

In Model 3, we also observe a significant association between frequency of recent pain killer misuse and smoking. Each increase of one day of pain killer misuse raises the odds of smoking prescription drugs by 1.7 percent. Finally, those who participated in the EDM scene had 2.1 times greater odds of smoking relative to those who did not participate in that scene (p < .05). By contrast, those participating in college bar scenes had 59.7 percent lower odds of smoking prescription drugs (p < .01).

Next, we consider smoking prescription drugs as an independent variable in linear regressions of measures of drug problems and mental health net of other covariates, as shown in Table 3. We observe a large and significant effect of smoking prescription drugs on drug problems (p < .01) even net of the effect of the frequency of recent misuse. Those who have smoked prescription drugs scored 40.0 percent higher on the CIDI ([e^0.322 – 1]*100% = 40.0%), controlling for recent misuse and demographics. The independent variables explained 24.4 percent of the variation in the log of CIDI scores. Similarly, those who have smoked scored 34.7 percent higher on the SIP-AD, with this model explaining 29.4 percent of the variation. By contrast, we did not observe significant effects of smoking on mental health, though recent misuse of stimulants and pain killers was consistently significant in those models. The independent variables explain a much smaller proportion of the variance in mental health, ranging from 8.7 to 11.8 percent.

Table 3.

Linear regressions of drug dependence and mental health outcomes

	Model 1: CIDI score		Model 2: SIP-AD score		Model 3: BSI depression		Model 4: BSI anxiety		Model 5: BSI somatization score
	Coef.	(SE)	Coef.	(SE)	Coef.	(SE)	Coef.	(SE)	Coef.	(SE)
Smoked prescription pills	0.322***	(0.089)	0.298**	(0.115)	0.113	(0.119)	0.074	(0.108)	0.068	(0.109)
Age	−0.020	(0.015)	−0.045*	(0.019)	−0.044*	(0.019)	−0.033	(0.018)	−0.064***	(0.018)
Parents' class
Upper middle class	−0.244	(0.158)	−0.181	(0.205)	0.196	(0.212)	0.054	(0.193)	0.084	(0.194)
Middle class	−0.256	(0.155)	−0.163	(0.200)	0.273	(0.207)	0.182	(0.189)	0.250	(0.190)
Working class/poor	−0.173	(0.162)	−0.078	(0.209)	0.386	(0.217)	0.289	(0.197)	0.371	(0.198)
Educational achievement
Some college or Associates degree	0.129	(0.156)	0.232	(0.201)	0.181	(0.208)	0.033	(0.189)	0.142	(0.190)
Currently in college	−0.005	(0.152)	−0.055	(0.196)	0.035	(0.203)	−0.025	(0.184)	−0.076	(0.185)
Bachelor's degree or higher	−0.044	(0.148)	−0.071	(0.191)	0.090	(0.198)	0.017	(0.180)	0.006	(0.181)
Nonwhite	−0.036	(0.072)	0.052	(0.093)	−0.089	(0.096)	−0.199*	(0.087)	−0.016	(0.088)
Male	−0.004	(0.068)	0.093	(0.088)	−0.171	(0.091)	−0.169*	(0.083)	−0.049	(0.083)
Heterosexual	0.027	(0.067)	0.093	(0.086)	−0.121	(0.090)	−0.060	(0.081)	−0.002	(0.082)
Sedatives days of misuse (past 90 days)	0.007***	(0.002)	0.009***	(0.003)	0.003	(0.003)	0.003	(0.002)	0.003	(0.002)
Pain killers days of misuse (past 90 days)	0.007**	(0.002)	0.011***	(0.003)	0.003	(0.003)	0.004	(0.003)	0.005*	(0.003)
Stimulants days of misuse (past 90 days)	0.005***	(0.002)	0.011***	(0.002)	0.006**	(0.002)	0.007***	(0.002)	0.005*	(0.002)
Constant	1.362***	(0.391)	2.140***	(0.505)	2.39***	(0.523)	2.343***	(0.475)	2.507***	(0.477)
F	8.84***		11.43***		2.61**		3.15***		3.67***
R2	0.244		0.294		0.087		0.103		0.118

^*p < .05;

^**p < .01;

^***p < .001

Note: All outcomes are logged.

Qualitative Findings

As indicated by the survey data, almost one out of five young prescription drug misusers reported smoking prescription pills. The interviews indicated that misusers typically progressed from oral consumption to smoking, rather than the reverse. Interviews with the participants who reported smoking pills indicate that, while some youth smoke or ‘vaporize’ prescription opioids, other youth described crushing up sedatives, such as Xanax, or stimulant drugs, such as Adderall, and mixing them with marijuana before smoking them. As an example of the former, one youth said, “You know how Oxycontin was referred to as ‘green beans’? There’s that green outer layer that you have to get off, that’s kind of like the time release part, and then you just basically ‘freebase’ it off of aluminum foil” (respondent 071). Some youth view smoking pills as commonplace. As described by another young man in our study, “Oxycontin just became, out of nowhere, this huge thing. And the thing to do if you were gonna do it back home when I was in high school, was smoke it. We call it ‘smoking beans.’ But it’s like heroin; these kids I used to know they just disappear, and that becomes their life” (001). Thus, some youth recognize that smoking indicates an escalation of drug use, often becoming an option as they increase the frequency of their prescription drug misuse.

Peer networks are the common context in which this occurs. As one young woman described, “I had a friend who had knee surgery or something like that in high school. I was like 15. So she had a bunch of Percocet and one of my friends had the idea to crush it up and roll it up and smoke it” (237). Another young man indicated, “The first time I remember doing oxy and being like ‘I like this where do I get this’ is when I was 22 or 21. I was at a house party. I was going to use the bathroom and two people I casually knew were in the bathroom. They were like—they put stuff away really quickly. And I thought they were doing coke. And I said, ‘Do you guys do coke? Can you share any of it?’ ‘Uh no, you probably don’t wanna do what we are doing.’ And I’m like, ‘I probably do.’ and they had oxy and they were smoking it” (094). This young man eventually transitioned to using heroin.

Some youth report smoking prescription drugs to mitigate the effects of other drug use. Polydrug use is particularly common in the Electronic Dance Music (EDM) scenes, providing context to our EDM finding in the survey results. For example, one young woman stated, “Crushing up Xanax and smoking it in a joint like for me personally I would just do that if I was using either a lot of Adderall or a lot of coke, yeah to curb the effects of the comedown” (002). Prescription drug smoking may therefore be incorporated into wider polydrug use routines among youth.

Our survey findings also indicated associations of prescription drug smoking with symptoms of dependence and drug problems. This concern even resonated among some prescription drug misusers. One youth described his concern as such: “When you smoke it’s just like instantaneous but it freaks me out because my friends compared it to like an opiate. So I was thinking about opium and like Chinese drug lords with their opium pipes and so I didn’t want to be that” (268). Thus, even for youth who smoke prescription drugs, there are concerns about what this practice says about their habit.

DISCUSSION

In this article, we considered escalations in mode of administration for prescription drug misuse, specifically from orally ingesting pills to smoking pills. While studies have examined the escalation in mode of administration for illicit drugs, researchers have provided few focused assessments of escalation for prescription drugs. We examined factors associated with this escalation and consequences of such escalation for dependence and mental health, while contextualizing these survey findings with excerpts from our interviewees. Importantly, our paper identifies that the transition to non-normative routes of administration represents an escalation – indicated by problems and dependence – even after accounting for frequency of use.

In terms of escalation to smoking prescription drugs, our results demonstrate that this mode of administration was more likely among males. Although studies have suggested that the lifetime prevalence of illicit drug use is converging between males and females,³⁰ studies have shown that males often use a wider range of drugs³¹ and have heavier patterns of substance use.³² These findings on escalation to smoking pills cohere with this broader literature. Recent pain killer misusers also reported higher odds of smoking prescription pills.

In terms of subcultural participation, those involved in Electronic Dance Music scenes were much more likely to report smoking pills. This may relate to the misuse of prescription drugs to moderate the effects of “club drugs” popular in such scenes.³³ Further examinations of the role of prescription drug smoking in polydrug combinations are necessary. By contrast, those involved in the college bar scene were less likely to have smoked prescription drugs. This is somewhat surprising given that studies have regularly highlighted the high prevalence of prescription drug misuse among college students,⁴ yet college students may misuse prescription drugs primarily for functional reasons and thus may be less inclined to smoke pills.

Regarding consequences, we found no effect for smoking prescription drugs on mental health outcomes, but strong effects for the CIDI and SIP-AD measures of problem drug use. Thus, our findings provide further support to previous research identifying that transitions to non-oral modes of consumption (mostly sniffing) increase prescription drug abuse related problems.¹⁵ Attesting to the strength of the effect, we observed increased symptoms of dependence associated with smoking prescription drugs even net of the effect of frequency of prescription drug misuse. Thus, smoking is associated with higher dependence and problems above and beyond greater recent misuse. While the effect of frequency of misuse on drug problems is perhaps unsurprising, the effect of smoking provides clinicians and practitioners with another indicator of increased likelihood of dependence and problems associated with prescription drug misuse.

Though the results provide insight into the practice of prescription drug smoking, some limitations should be considered. First, this project was designed to study a sample of youth involved in nightlife scenes. This population is an important one to study due to the salient role that substances often play in nightlife venues, yet these findings may not generalize to all young adults. Yet, although these individuals were recruited in nightlife venues, the locations for pill smoking identified were primarily outside of nightlife venues, thus indicating that this practice extends beyond nightlife scenes. Second, as we used a time-space sampling method, we may have oversampled people who are more frequent nightlife participants. Third, these participants were recruited in a single U.S. city, however, many originated elsewhere and often indicated they initiated into prescription pill smoking prior to moving to New York. Thus, this suggests that the prescription pill smoking phenomenon is not isolated to this region. Additionally, although boroughs with high concentrations of nightlife venues also have relatively high LGBQ populations and young adults are more likely to identify as LGBQ than older adults, our sample contains an oversample of sexual minority young adults. Our mental health outcomes do not account for all mental health conditions. Finally, as subjects were asked to self-report behaviors, there may be a social desirability bias or recall bias in the reporting of drug use behaviors, as is common in such studies. However, studies have shown that computer-assisted surveys improve self-report measures of sensitive topics,^{34, 35} which improves our confidence in these responses.

CONCLUSION

Prescription drug smoking has emerged as a significant drug trend among socially active youth who misuse prescription drugs. It is associated with drug problems and symptoms of dependence. Young men are more likely to smoke prescription drugs, and heterosexual youth appear more likely to smoke prescription drugs than their sexual minority peers. Young adults involved in electronic dance music nightlife may particularly benefit from prevention and intervention efforts targeting prescription drug smoking. Future studies should identify relationships between prescription drug smoking and other risk behaviors. Prevention and intervention efforts for young people who misuse prescription drugs more broadly should be informed by the issue of prescription drug smoking, and this may be an area for clinicians to address with their adolescent patients.

Implications and Contribution.

Little research has been published on prescription drug smoking. This paper highlights that prescription drug smoking, as an escalation of drug use, is associated with drug problems and symptoms of dependence among youth beyond the frequency of misuse. It also identifies characteristics of youth at risk for prescription drug smoking.