Medical, psychosocial, and treatment predictors of opioid overdose among high risk opioid users

Correlates of overdose among 2711 people who use drugs and live in 7 rural US sites

BACKGROUND

Overdose rates in rural areas have been increasing globally, with large increases in the United States. Few studies, however, have identified correlates of non-fatal overdose among rural people who use drugs (PWUD). The present analysis describes correlates of nonfatal overdose among a large multistate sample of rural PWUD.

METHODS

This is a cross-sectional analysis of data gathered via surveys with PWUD recruited through seven Rural Opioid Initiative (ROI) sites. Descriptive analyses were conducted to assess the prevalence of past 30-day overdose. Generalized estimating equations were used to estimate a series of multivariable models quantifying relationships of select factors to past-month overdose; factors were selected using the Risk Environment Framework.

RESULTS

The multisite sample included 2711 PWUD, 6% of whom reported overdosing in the past 30 days. In the fully adjusted model, houselessness (AOR=2.27, 95%CI[1.48, 3.48]), a positive test result for Hepatitis C infection (AOR=1.73 95%CI[1.18, 2.52]) and heroin/fentanyl use (AOR= 8.58 95%CI [3.01, 24.50]) were associated with an increased risk of reporting past 30-day overdose, while having a high-school education or less was associated with reduced odds of overdose (AOR=0.52, 95% CI[0.37, 0.74]).

CONCLUSION

As in urban areas, houselessness, Hepatitis C infection, and the use of heroin and fentanyl were significant correlates of overdose. Widespread access to overdose prevention interventions - including fentanyl test strips and naloxone - is critical in this rural context, with particular outreach needed to unhoused populations, people living with Hepatitis C, and people using opioids.

Psychosocial factors associated with overdose subsequent to Illicit Drug use: a systematic review and narrative synthesis

BACKGROUND AND AIMS

Psychological and social status, and environmental context, may mediate the likelihood of experiencing overdose subsequent to illicit drug use. The aim of this systematic review was to identify and synthesise psychosocial factors associated with overdose among people who use drugs.

METHODS

This review was registered on Prospero (CRD42021242495). Systematic record searches were undertaken in databases of peer-reviewed literature (Medline, Embase, PsycINFO, and Cinahl) and grey literature sources (Google Scholar) for work published up to and including 14 February 2023. Reference lists of selected full-text papers were searched for additional records. Studies were eligible if they included people who use drugs with a focus on relationships between psychosocial factors and overdose subsequent to illicit drug use. Results were tabulated and narratively synthesised.

RESULTS

Twenty-six studies were included in the review, with 150,625 participants: of those 3,383-4072 (3%) experienced overdose. Twenty-one (81%) studies were conducted in North America and 23 (89%) reported polydrug use. Psychosocial factors associated with risk of overdose (n = 103) were identified and thematically organised into ten groups. These were: income; housing instability; incarceration; traumatic experiences; overdose risk perception and past experience; healthcare experiences; perception of own drug use and injecting skills; injecting setting; conditions with physical environment; and social network traits.

CONCLUSIONS

Global rates of overdose continue to increase, and many guidelines recommend psychosocial interventions for dependent drug use. The factors identified here provide useful targets for practitioners to focus on at the individual level, but many identified will require wider policy changes to affect positive change. Future research should seek to develop and trial interventions targeting factors identified, whilst advocacy for key policy reforms to reduce harm must continue.

Predictors of problematic adult alcohol, cannabis, and other substance use: A longitudinal study of two samples

This study examined whether a key set of adolescent and early adulthood risk factors predicts problematic alcohol, cannabis, and other substance use in established adulthood. Two independent samples from the Child Development Project (CDP; n = 585; 48% girls; 81% White, 17% Black, 2% other race/ethnicity) and Fast Track (FT; n = 463; 45% girls; 52% White, 43% Black, 5% other race/ethnicity) were recruited in childhood and followed through age 34 (CDP) or 32 (FT). Predictors of substance use were assessed in adolescence based on adolescent and parent reports and in early adulthood based on adult self-reports. Adults reported their own problematic substance use in established adulthood. In both samples, more risk factors from adolescence and early adulthood predicted problematic alcohol use in established adulthood (compared to problematic cannabis use and other substance use). Externalizing behaviors and prior substance use in early adulthood were consistent predictors of problematic alcohol and cannabis misuse in established adulthood across samples; other predictors were specific to the sample and type of substance misuse. Prevention efforts might benefit from tailoring to address risk factors for specific substances, but prioritizing prevention of externalizing behaviors holds promise for preventing both alcohol and cannabis misuse in established adulthood.

Detoxification, 12-step meeting attendance, and non-fatal opioid overdoses among a suburban/exurban population with opioid use disorder

BACKGROUND

Drug overdoses are the leading cause of injury death in the United States with an estimated 105,752 individuals dying from an overdose in the United States in a 12-month period ending October 2021. Given that people who have opioid use disorder (OUD) are at an increased risk of death, it is crucial to assess risk factors associated with opioid overdose to improve interventions.

OBJECTIVES

We examine factors associated with non-fatal overdose among a suburban/exurban population with OUD in Southern California.

METHODS

Participants were recruited by convenience sampling (n = 355) and were interviewed between November 2017 to August 2018. Participants were eligible for the study if they had a history of pharmaceutical opioid use.

RESULTS

A total of 198 (55.8%) participants reported at least one overdose in their lifetime. A total of 229 participants identified as male, 124 identified as female, and 2 identified as non-binary. When controlling for demographic factors, non-oral opioid administration at first opioid use (AOR 2.82, 95% CI 1.52-5.22), having a history of methadone detoxification, (AOR 2.23, 95% CI 1.27-3.91), history of buprenorphine detoxification (AOR 1.77, 95% CI 1.02-3.07), and history of 12 step attendance (AOR 1.89, 95% CI 1.12-3.20) were found to be independently and positively associated with lifetime opioid overdose.

CONCLUSIONS

Detoxification with buprenorphine and methadone was found to be associated with having a non-fatal opioid overdose. Buprenorphine and methadone should not be prescribed as a detoxification medication as long-term use of medication for OUD results in better outcomes than medication that is used short-term.

Syndemic factors associated with non-fatal overdose among young opioid users in New York City

Introduction

Rates of illicit opioid use are particularly high among young adults, yet research on overdose experience and factors associated with overdose in this population remains limited. This study examines the experiences and correlates of non-fatal overdose among young adults using illicit opioids in New York City (NYC).

Methods

539 participants were recruited via Respondent-Driven Sampling in 2014-2016. Eligibility criteria included: aged 18-29 years old; current residence in NYC; and nonmedical prescription opioid (PO) use and/or heroin use in the past 30 days. Participants completed structured interviews to assess their socio-demographics, drug use trajectories, current substance use and lifetime and most recent overdose experiences, and were tested on-site for hepatitis C virus (HCV) antibodies.

Results

43.9% of participants reported lifetime overdose experience; of these, 58.8% had experienced two or more overdose events. The majority of participants' most recent overdoses (63.5%) were due to polysubstance use. In bivariable analyses, after RDS adjustment, having ever overdosed was correlated with: household income of >$100,00 growing up (vs. $51,000-100,000); lifetime homelessness; HCV antibody-positive status; lifetime engagement in regular nonmedical benzodiazepine use, regular heroin injection and regular PO injection; and using a non-sterile syringe in the past 12 months. Multivariable logistic regression identified childhood household income >$100,00 (AOR=1.88), HCV-positive status (AOR=2.64), benzodiazepine use (AOR=2.15), PO injection (AOR=1.96) and non-sterile syringe use (AOR=1.70) as significant independent correlates of lifetime overdose. A multivariable model with multiple overdoses (vs. one) found only lifetime regular heroin use and PO injection to be strong correlates.

Discussion

Results indicate a high prevalence of lifetime and repeated overdose among opioid-using young adults in NYC, highlighting a need for intensified overdose prevention efforts for this population. The strong associations of HCV and indices of polydrug use with overdose suggest that prevention efforts should address the complex risk environment in which overdose occurs, attending to the overlapping nature of disease-related risk behavior and overdose risk behavior among young people who inject opioids. Overdose prevention efforts tailored for this group may find it useful to adopt a syndemic conception of overdose that understands such events as resulting from multiple, and often interrelated, risk factors.

Clinical Pharmacology, Toxicity, and Abuse Potential of Opioids

Opioids were the most common drug class resulting in overdose deaths in the United States in 2019. Widespread clinical use of prescription opioids for moderate to severe pain contributed to the ongoing opioid epidemic with the subsequent emergence of fentanyl-laced heroin. More potent analogues of fentanyl and structurally diverse opioid receptor agonists such as AH-7921 and MT-45 are fueling an increasingly diverse illicit opioid supply. Overdose from synthetic opioids with high binding affinities may not respond to a typical naloxone dose, thereby rendering autoinjectors less effective, requiring higher antagonist doses or resulting in a confusing clinical picture for health care providers. Nonscheduled opioid drugs such as loperamide and dextromethorphan are associated with dependence and risk of overdose as easier access makes them attractive to opioid users. Despite a common opioid-mediated pathway, several opioids present with unique pharmacodynamic properties leading to acute toxicity and dependence development. Pharmacokinetic considerations involve half-life of the parent opioid and its metabolites as well as resulting toxicity, as is established for tramadol, codeine, and oxycodone. Pharmacokinetic considerations, toxicities, and treatment approaches for notable opioids are reviewed.

Drug type and risk behaviors associated with non-fatal overdose among people who use drugs: a systematic review and meta-analysis

The present study aimed to determine the association between drug type, risk behaviors and non-fatal overdose among people who use drugs (PWUD). We searched for studies in English published before February 1, 2021, on PubMed, Scopus, Cochrane, and Web of Science to identify primary studies on the factors associated with non-fatal overdose among PWUD. After reviewing for study duplicates, the full-text of selected articles were assessed for eligibility using Population, Intervention, Comparator, Outcomes (PICO) criteria. After a detailed assessment of over 13,845 articles, a total of 49 studies met the eligibility criteria. We found that non-injection opioid use, heroin injection, cocaine use, concurrent use of buprenorphine and benzodiazepines, benzodiazepine use, incarceration, injecting drugs, and duration of injecting were associated with greater odds of non-fatal overdose among PWUD. The findings of the current meta-analysis support the requirement to improve suitable harm reduction strategies for drug users, such as peer-based overdose management, and further focusing on the need to balance the current emphasis on enforcement-based responses to illegal drug use with health-related interventions.

Mortality in an Opioid Treatment Program

This retrospective study examined mortality in an Opioid Treatment Program (OTP) with policies designed to admit and retain patients with as few barriers as possible. Methadone provided in the context of an OTP is known to decrease mortality and morbidity. Historically, patients have been discharged or turned away because of continued substance use. We examined patient deaths over three years in an OTP serving approximately 2400 patients daily. Demographics and causes of death were gathered from electronic health records and medical examiner reports. Pairwise comparisons were used to compare drug poisoning versus non-drug poisoning deaths. There were 155 deaths during the study period. The average age was 54, and half of the participants had positive results on their most recent drug screen. Forty one (26%) died from "drug poisoning." Drug poisoning deaths were more common among patients who: 1) had recent positive drug test results; 2) had documented alprazolam use; 3) were younger; 4) had shorter treatment durations; 5) were female. The majority of deaths were among long-term patients over 50 and were not caused by drug poisoning. These results support keeping patients in treatment despite ongoing drug use, and linking them to appropriate primary care.

Engagement in Harm Reduction Strategies After Suspected Fentanyl Contamination Among Opioid-Dependent Individuals

The evolving opioid epidemic in the United States has increased drug-related overdose rates exponentially (Centers for Disease Control and Prevention in Opioid overdose, 2020c, https://www.cdc.gov/drugoverdose/data/otherdrugs.html#:~:text=Polysubstance%20drug%20use%20occurs%20with,or%20other%20non%2Dopioid%20substances ). Fentanyl, a synthetic opioid, has recently fueled the epidemic, increasing overdose death rates (Centers for Disease Control and Prevention in Drug overdose deaths involving fentanyl, 2011-2016, 2019a, https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_03-508.pdf ). Harm reduction strategies (drug checking, naloxone administration, etc.) are at the forefront of preventing opioid-related overdoses in high-risk populations (Kennedy et al. in Drug Alcohol Depend 185:248-252, 2018, https://doi.org/10.1016/j.drugalcdep.2017.12.026 ; Laing et al. in Int J Drug Policy 62:59-66, 2018, https://doi.org/10.1016/j.drugpo.2018.10.001 ). Little is known, however, about how people who inject drugs (PWID) may modify their drug use behaviors after suspected fentanyl contamination in their drugs. We conducted a cross-sectional survey among 105 opioid-dependent PWID enrolled in a methadone maintenance program. We assessed their willingness to engage in various harm reduction methods (i.e., slowing down drug use, not using drugs, carrying naloxone, using with someone who has naloxone) after suspected fentanyl contamination of their drugs. In a multivariable analysis, participants who were white, low-income, polysubstance users, and had previously experienced an overdose or had previously administered naloxone were more likely to report a willingness to engage in harm reduction measures. These findings provide an evidence-based understanding of PWID's engagement in harm reduction behaviors after suspecting potential fentanyl exposure as well as a basis for tailoring intervention strategies in the context of fentanyl-adulterated markets.

Prevalence of Mental Health Disorders among Individuals Enrolled in an Emergency Response Program for Treatment of Opioid Use Disorder

Background

Psychiatric co-occurring disorders among individuals with opioid use disorder has primarily focused on epidemiological surveys of patients in continuous or long-term treatment, yet little is known about the socio-economically vulnerable who are non-treatment seeking prior to an emergency response.

Methods

We retrospectively analyzed data from patients who had enrolled in a study involving home outreach to patients referred from police, emergency medical services (EMS), or hospital emergency departments following an emergency response. The sample is largely socio-economically vulnerable with high rates of unemployment and uninsured. Eligible consenting patients received an intervention consisting of medication (buprenorphine), behavioral counseling, and peer support. Participants completed semi-structured, psychological instruments to assess current and prior history for both substance use and mental health disorders. We used descriptive statistics to evaluate rates of co-occurring mental health comorbidity.

Results

Among 102 patients (average age = 36.5 years old), approximately 61 (59.8%) reported a prior mental health diagnosis, with 31 (30.3%) currently on medications for their diagnoses. Mood and anxiety disorders were most frequently recorded. Just over half (51%) had received any prior treatment for their substance use. Of those with dual disorders, 67.2% had experienced prior suicidal thoughts, and 63.7% reported polysubstance use of 5 or more substances.

Conclusion

Co-occurring psychiatric disorders, and specifically mood and anxiety disorders, appear to be prevalent in vulnerable populations at an increased rate. Mental health assessments should routinely be performed in the emergency setting and in early stages of treatment.

The Role of Cannabis Use in Suicidal Ideation Among Patients With Opioid Use Disorder

OBJECTIVES

Cannabis use is associated with suicide risk in the general population; however, it is unknown if this association is also present in patients with opioid use disorder (OUD). The purpose of this study is to investigate the association between cannabis use and suicidal ideation in patients with OUD.

METHODS

We conducted a multivariable logistic regression analysis to assess the association between cannabis use and suicidal ideation, amongst a large cohort of patients with OUD. Current cannabis use and suicidal ideation over the past 30 days were obtained by self-report.

RESULTS

Cross-sectional data from 2335 participants with OUD were included in the analysis, of whom 51% report current cannabis use. We found a positive association between cannabis use and suicidal ideation (OR = 1.41, 95% CI 1.11, 1.80, P = 0.005). We found that men (OR = 1.84, 95% CI 1.44, 2.35, P < 0.001), younger individuals (OR = 1.02, 95% CI 1.01, 1.03), P = 0.004), and that those with more symptoms of anxiety or depression (OR = 1.16, 95% CI 1.15, 1.18, P < 0.001) were more likely to report suicidal ideation.

CONCLUSIONS

Cannabis use is associated with a heightened propensity for suicidal ideation amongst patients with OUD, who are already a high-risk population. Further research into the potential harms of cannabis use in this population is required given the prevalence of its use and potential benefits in mitigating opioid withdrawal.

Opioid overdose in the age of fentanyl: Risk factor differences among subpopulations of overdose survivors

BACKGROUND

Although fentanyl is the drug most frequently implicated in overdose deaths, the association between overdose risk and attitudes and behaviors surrounding fentanyl in opioid-using communities has remained understudied. Possible subpopulation differences in fentanyl-related overdose risk remain equally unexamined. This paper addresses these gaps by exploring the association between overdose and fentanyl-related attitudes/behaviors in three subpopulations of overdose survivors.

METHODS

In this cross-sectional study, we sampled 432 individuals who currently or recently used opioids from New Jersey methadone and acute residential detoxification programs. Using multinomial regression analysis, we compared overdose risk factors, including fentanyl-related attitudes/behaviors, of those who never overdosed with three subgroups of overdose survivors who experienced: 1. recent overdoses occurring after, but not before, fentanyl expansion; 2. past overdoses occurring before, but not after, fentanyl expansion; 3. persistent overdoses occurring before and after fentanyl expansion.

RESULTS

Forty percent of respondents had knowingly used fentanyl and 38% deliberately sought overdose-implicated drugs. Respondents with persistent overdoses represented under 10% of the sample but accounted for 44% of all lifetime overdoses (x̅ =8.03 vs. 1.71 for the full sample). This was also the only subgroup for whom PTSD (AOR=3.84; 95%CI=1.45-10.16; p=.01) and fentanyl-seeking (AOR=1.50; 95% CI=1.16-1.94; p=.01) were significant overdose risk factors. Those with recent overdoses engaged in frequent drug combining (AOR=2.28; 95% CI=1.19-6.98; p=.05), which could have led to inadvertent fentanyl use. Those with past overdoses were not at overdose risk from fentanyl-seeking or drug combining and had rates of methadone treatment comparable to rates of those with no overdoses.

CONCLUSION

Harm reduction strategies will need to address consumers' evolving drug preferences as fentanyl continues to saturate local drug markets. Targeting comprehensive interventions, including mental health treatment, to the small group of opioid users with longstanding overdose histories may reduce the burden of overdose in opioid-using communities.

Evaluation of Opioid Overdose Reports in Patients Treated with Extended-Release Naltrexone: Postmarketing Data from 2006 to 2018

INTRODUCTION

After treatment with naltrexone extended-release injectable suspension (XR-NTX), a µ-opioid receptor antagonist, opioid tolerance is reduced from pretreatment baseline. Patients may be vulnerable to opioid overdose if they attempt to override the blockade during treatment, at the end of a dosing interval, after missing a dose, or after discontinuing treatment.

OBJECTIVE

We analyzed postmarketing data to characterize reporting rates of opioid overdose during treatment with and after discontinuation of XR-NTX.

METHODS

Postmarketing adverse event reports within the XR-NTX safety database, received 2006-2018, for patients treated with XR-NTX for any indication were reviewed for opioid overdose cases. Assessable cases were categorized by timing of the event from the last dose of XR-NTX (latency): ≤28 days (on treatment), 29-56 days, and >56 days from last dose of XR-NTX. Within each latency group, cases were further classified as serious and, of those, cases that had a fatal outcome.

RESULTS

During the 12-year period, an estimated 495,602 patients received XR-NTX. Opioid overdose was reported in 161 cases; of these, 66 contained sufficient information to determine latency. Reporting rates of opioid overdose per 10,000 patients treated were similar among latency groups: 0.54 for ≤28 days (0.24 fatal), 0.34 for 29-56 days (0.16 fatal), and 0.44 for >56 days (0.40 fatal) from the last dose of XR-NTX.

CONCLUSIONS

Over the 12-year period, the reporting rates of opioid overdose were similar during treatment with or after discontinuation of XR-NTX and <10/10,000 patients exposed. Our findings are limited by the nature of spontaneously reported safety data.

Endogenous Opiates and Behavior: 2018

This paper is the forty-first consecutive installment of the annual anthological review of research concerning the endogenous opioid system, summarizing articles published during 2018 that studied the behavioral effects of molecular, pharmacological and genetic manipulation of opioid peptides and receptors as well as effects of opioid/opiate agonists and antagonists. The review is subdivided into the following specific topics: molecular-biochemical effects and neurochemical localization studies of endogenous opioids and their receptors (2), the roles of these opioid peptides and receptors in pain and analgesia in animals (3) and humans (4), opioid-sensitive and opioid-insensitive effects of nonopioid analgesics (5), opioid peptide and receptor involvement in tolerance and dependence (6), stress and social status (7), learning and memory (8), eating and drinking (9), drug abuse and alcohol (10), sexual activity and hormones, pregnancy, development and endocrinology (11), mental illness and mood (12), seizures and neurologic disorders (13), electrical-related activity and neurophysiology (14), general activity and locomotion (15), gastrointestinal, renal and hepatic functions (16), cardiovascular responses (17), respiration and thermoregulation (18), and immunological responses (19).

Prevalence of and Risk Factors Associated With Nonfatal Overdose Among Veterans Who Have Experienced Homelessness

IMPORTANCE

Individuals with a history of homelessness are at increased risk for drug or alcohol overdose, although the proportion who have had recent nonfatal overdose is unknown. Understanding risk factors associated with nonfatal overdose could guide efforts to prevent fatal overdose.

OBJECTIVES

To determine the prevalence of recent overdose and the individual contributions of drugs and alcohol to overdose and to identify characteristics associated with overdose among veterans who have experienced homelessness.

DESIGN, SETTING, AND PARTICIPANTS

This survey study was conducted from November 15, 2017, to October 1, 2018, via mailed surveys with telephone follow-up for nonrespondents. Eligible participants were selected from the records of 26 US Department of Veterans Affairs medical centers and included veterans who had received primary care at 1 of these Veterans Affairs medical centers and had a history of experiencing homelessness according to administrative data. Preliminary analyses were conducted in October 2018, and final analyses were conducted in January 2020.

MAIN OUTCOMES AND MEASURES

Self-report of overdose (such that emergent medical care was obtained) in the previous 3 years and substances used during the most recent overdose. All percentages are weighted according to propensity to respond to the survey, modeled from clinical characteristics obtained in electronic health records.

RESULTS

A total of 5766 veterans completed the survey (completion rate, 40.2%), and data on overdose were available for 5694 veterans. After adjusting for the propensity to respond to the survey, the mean (SD) age was 56.4 (18.3) years; 5100 veterans (91.6%) were men, 2225 veterans (38.1%) were black, and 2345 veterans (40.7%) were white. A total of 379 veterans (7.4%) reported any overdose during the past 3 years; 228 veterans (4.6%) reported overdose involving drugs, including 83 veterans (1.7%) who reported overdose involving opioids. Overdose involving alcohol was reported by 192 veterans (3.7%). In multivariable analyses, white race (odds ratio, 2.44 [95% CI, 2.00-2.98]), self-reporting a drug problem (odds ratio, 1.66 [95% CI, 1.39-1.98]) or alcohol problem (odds ratio, 2.54 [95% CI, 2.16-2.99]), and having witnessed someone else overdose (odds ratio, 2.34 [95% CI, 1.98-2.76]) were associated with increased risk of overdose.

CONCLUSIONS AND RELEVANCE

These findings suggest that nonfatal overdose is relatively common among veterans who have experienced homelessness. While overdose involving alcohol was more common than any specific drug, 1.7% of veterans reported overdose involving opioids. Improving access to addiction treatment for veterans who are experiencing homelessness or who are recently housed, especially for those who have experienced or witnessed overdose, could help to protect this population.

Association of Nonmedical Prescription Opioid Use With Subsequent Heroin Use Initiation in Adolescents

IMPORTANCE

There is concern that nonmedical prescription opioid use is associated with an increased risk of later heroin use initiation in adolescents, but to our knowledge, longitudinal data addressing this topic are lacking.

OBJECTIVE

To determine whether nonmedical prescription opioid use is associated with subsequent initiation of heroin use in adolescents.

DESIGN, SETTING, AND PARTICIPANTS

This prospective longitudinal cohort study conducted in 10 high schools in Los Angeles, California, administered 8 semiannual surveys from 9th through 12th grade that assessed nonmedical prescription opioid use, heroin use, and other factors from October 2013 to July 2017. Students were baseline never users of heroin recruited through convenience sampling. Cox regression models tested nonmedical prescription opioid use statuses at survey waves 1 through 7 as a time-varying and time-lagged regressor and subsequent heroin use initiation across waves 2 to 8 as the outcome.

EXPOSURES

Self-reported nonmedical prescription opioid use (past 30-day [current] use vs past 6-month [prior] use without past 30-day use vs no past 6-month use) at each wave from 1 to 7.

MAIN OUTCOMES AND MEASURES

Self-reported heroin use initiation (yes/no) during waves 2 to 8.

RESULTS

Of 3298 participants, 1775 (53.9%) were adolescent girls, 1563 (48.3%) were Hispanic, 548 (17.0%) were Asian, 155 (4.8%) were African American, 529 (16.4%) were non-Hispanic white, and 220 (6.8%) were multiracial. Among baseline never users of heroin in ninth grade with valid data (3298 [97% of cohort enrollees]; mean [SD] age, 14.6 [0.4] years), the number of individuals with outcome data available at each follow-up ranged from 2987 (90.6%) to 3200 (97.0%). The mean per-wave prevalence of prior and current nonmedical prescription opioid use from waves 1 to 7 was 1.9% and 2.7%, respectively. Seventy students (2.1%) initiated heroin use during waves 2 to 8. Prior vs no (hazard ratio, 3.59; 95% CI, 2.14-6.01; P < .001) and current vs no (hazard ratio, 4.37; 95% CI, 2.80-6.81; P < .001) nonmedical prescription opioid use were positively associated with subsequent heroin use initiation. For no, prior, and current nonmedical prescription opioid use statuses at waves 1 to 7, the estimated cumulative probabilities of subsequent heroin use initiation by wave 8 (42-month follow-up) were 1.7%, 10.7%, and 13.1%, respectively. In covariate-adjusted models, associations were attenuated but remained statistically significant and current nonmedical prescription opioid use risk estimates were stronger than corresponding associations of nonopioid substance use with subsequent heroin use initiation.

CONCLUSIONS AND RELEVANCE

Nonmedical prescription opioid use was prospectively associated with subsequent heroin use initiation during 4 years of adolescence among Los Angeles youth. Further research is needed to understand whether this association is causal.