Drug use generations and patterns of injection drug use: Birth cohort differences among people who inject drugs in Los Angeles and San Francisco, California

Highlights

• National drug trends are associated with injection drug use patterns.

• Recent birth cohorts of people who inject drugs (PWID) are initiating injection more rapidly than older birth cohorts.

• Efforts to prevent injection initiation among susceptible drug users are urgently needed.

Abstract

Objectives

A robust literature documents generational trends in drug use. We examined the implications of changing national drug use patterns on drug injection histories of diverse people who inject drugs (PWID).

Methods

Drug use histories were collected from 776 active PWID in 2011-13. Using descriptive statistics, we examine drug use initiation by year and birth cohort (BC) differences in drug first injected. A multivariate linear regression model of time to injection initiation ([TTII] (year of first injection minus year of first illicit drug use) was developed to explore BC differences.

Results

The first drug injected by BC changed in tandem with national drug use trends with heroin declining from 77% for the pre-1960’s BC to 58% for the 1960’s BC before increasing to 71% for the 1990’s BC. Multivariate linear regression modeling found that shorter TTII was associated with the 1980’s/1990’s BC (−3.50 years; 95% Confidence Interval [CI] = −0.79, −6.21) as compared to the 1970’s BC. Longer TTII was associated with being female (1.65 years; 95% CI = 0.40, 2.90), African American (1.69 years; 95% CI = 0.43, 2.95), any substance use treatment prior to injection (4.22 years; 95% CI = 2.65, 5.79), and prior non-injection use of drug that was first injected (3.29 years; 95% CI = 2.19, 4.40).

Conclusion

National drug trends appear to influence injection drug use patterns. The prescription opiate drug era is associated with shorter TTII. Culturally competent, demographically and generationally-targeted prevention strategies to combat transitions to drug injection are needed to prevent or shorten upstream increases in risky drug use practices on a national level.

Introduction

For much of the previous twenty years, people who inject drugs (PWID) were regarded as an aging population (Armstrong, 2007) and rates of drug injection were understood to be declining (Brady et al., 2008). Recent reports indicate that drug injection may be increasing among young persons in the North America. For instance, Klevens and colleagues report that despite stable prevalence of injection drug use among high school seniors from 1995 to 2013, increases in some subgroups (non-Hispanic Blacks) and regions (Arkansas, Hawaii, Maine, Maryland, and New York) have been detected (Klevens et al., 2016). Using a modelling approach that relies on multiple local and national datasets, Tempalski and colleagues estimated that per population portion of PWID increased in the 15–29 years of age subgroup in 2006–07 as compared to this subgroup in 1996-97 (Tempalski et al., 2013). The causes of these emerging trends are worthy of study. In the following, we use a drug use generations framework to explore how patterns of injection drug use are changing among birth cohorts of PWID.

North America has experienced national and regional drug epidemics of heroin, powder cocaine, crack cocaine, methamphetamine, and prescription drugs (especially pain relievers) over the last 40 years (Bourgois, 2003a, Bourgois and Schonberg, 2000, Compton et al., 2005, Golub et al., 2005, Gruenewald et al., 2013, Maxwell and Rutkowski, 2008, Roy et al., 2012). One method for understanding the trajectory and implications of these trends is to use the “drug generation framework” (Golub and Johnson, 1994a, Golub and Johnson, 1994b, Golub et al., 2005, Johnson and Golub, 2002). Key to this approach is tracking birth cohort changes in drug use preferences as a way of identifying transitions between drug “generations” or eras. This approach has been successfully used to examine transitions from heroin injection in the 1970’s to crack cocaine smoking in the 1980’s (Golub and Johnson, 1999), from crack cocaine in the 1980’s to marijuana/blunt smoking in the 1990’s (Golub and Brownstein, 2013), and the emergence of nonmedical use of opiate prescription drugs in the 2000’s (Golub et al., 2015, Golub et al., 2013).

The origins of the current nonmedical use of opiate prescription drugs lies in the development of long-acting opioid pain medications and the new focus on pain management, both of which occurred in the early-1990’s (Kolodny et al., 2015, Manchikanti et al., 2012, Maxwell, 2011). These long-acting opioid-based pain medications have become widely available and while the vast majority of patients prescribed these medications use them as indicated, documented increases in nonmedical use are indisputable (Cheatle, 2015, Roland et al., 2016).

The current nonmedical use of opiate prescription drugs has been characterized by significant increases in overdose deaths (Unick et al., 2013), increased mortality (Cottler et al., 2016), at least one significant HIV/HCV outbreak (CDC, 2015, Peters et al., 2016), and a substantial rise in the number of heroin users (Jones et al., 2015b, Maxwell, 2015). Another potential consequence of this trend is changes in injection drug use patterns.

A key question is whether the most recent drug epidemic of nonmedical use of prescription opiate drugs is associated with changes in drug use administration routes (Firestone and Fischer, 2008, Fischer et al., 2006, Maxwell, 2011, Maxwell, 2015, Roy et al., 2011, Roy et al., 2012). Based on published studies, it appears that nonmedical use of prescription opiates can lead to injection drug use through a 2-step process.

First, nonmedical use of prescription opiate drugs can lead to heroin use. Using national surveys, Jones found that past year use of heroin increased significantly among nonmedical users of prescription opioid drugs as compared to people who did not use prescription opioid drugs during the decade of the 2000’s (Jones, 2013). In a more recent national study, Cerda and colleagues found that nonmedical prescription opioid drug use among youth was a significant predictor of subsequent heroin use (Cerda et al., 2015). An analysis of data from the Veterans Aging Cohort Study, found that nonmedical use of prescription opioid drugs resulted in a 5-fold higher hazard ratio of initiating heroin as compared to participants who without nonmedical prescription opioid drug use (Banerjee et al., 2016). Other studies have documented the connection between nonmedical prescription opiate drug use at the local and regional level (Cicero and Kuehn, 2014), including an observational cohort of young adult nonmedical prescription opioid drug users in Ohio that reported an 2.8% annual rate of heroin uptake (Carlson et al., 2016). Several studies among PWID have now documented that nonmedical use of prescription opioid drugs preceded heroin use and injection for most (Novak et al., 2016, Peavy et al., 2012, Pollini et al., 2011). Qualitative accounts in diverse North American regions have described patterns of injection drug use among younger birth cohorts that differ under the influence of nonmedical prescription opioid drug use and heroin injection as compared to those patterns observed in earlier birth cohorts (Firestone and Fischer, 2008, Mars et al., 2014, Siegal et al., 2003).

Second, heroin use is strongly associated with injection modes of use. For instance, Novak and Kral found that 50% of people who use heroin reported injection as compared to users of methamphetamines and cocaine where injections rates of 13% and 3%, respectively, were reported (Novak and Kral, 2011). It is possible therefore for the non-medical use of prescription opiates to influence patterns of injection drug use.

One way to explore drug use epidemics and their impact on injection drug use is to compare drug use histories among birth cohorts of PWID through retrospective accounts. Key questions include whether patterns of drug use among PWID conform to changing national trends and whether these national trends are associated with changes in drug injection initiation patterns (i.e., first drug injected) by birth cohort and time to injection initiation (TTII).

TTII is measured here as the number of years between first illicit drug use (including nonmedical use of prescription drugs) and first injection drug use. TTII has been studied in a variety of locales (Clatts et al., 2011, DeBeck et al., 2016b, Malekinejad and Vazirian, 2012, Mehta et al., 2012, Ross et al., 2008, Vorobjov et al., 2013, Young and Havens, 2011) but to date, no reports are available on US samples that include a wide range of substance use patterns. TTII is one way to examine how injection initiation patterns change in relationship to different “drug generations” (Golub and Johnson, 1994b, Golub et al., 2005, Johnson and Golub, 2002).

In the following, we examine whether national drug use patterns are associated with drug use patterns and injection patterns among birth cohorts of PWID. We also examine whether drug use trends are associated differences in birth cohort TTII among a large, diverse sample of PWID in California.