HHS Public Access Author manuscript Author Manuscript
Addict Behav. Author manuscript; available in PMC 2017 September 01. Published in final edited form as: Addict Behav. 2016 September ; 60: 8–12. doi:10.1016/j.addbeh.2016.03.029.
No association between HIV status and risk of non-fatal overdose among people who inject drugs in Vancouver, Canada Daniel J Escudero1, Brandon DL Marshall1, Thomas Kerr2,3, Kanna Hayashi2,3, Cindy Feng2,4, Silvia A Guillemi2, Robert S Hogg2, Julio Montaner2,3, Evan Wood2,3, and M-J Milloy2,3 of Epidemiology, Brown University School of Public Health, 2nd Floor, 121 S. Main St., Providence, RI, 02906, United States 1Department
Author Manuscript
2British
Columbia Centre for Excellence in HIV/AIDS, 608 – 1081 Burrard Street, Vancouver, B.C., Canada V6Z 1Y6
3Division
of AIDS, Department of Medicine, University of British Columbia, 667-1081 Burrard Street, Vancouver, BC, Canada V6Z1Y6
Abstract
Author Manuscript
Background—The evidence to date on whether HIV infection increases the risk of accidental drug overdose among people who inject drugs (PWID) is equivocal. Thus, we sought to estimate the effect of HIV infection on risk of non-fatal overdose among two parallel cohorts of HIVpositive and –negative PWID. Methods—Data were collected from a prospective cohort of PWID in Vancouver, Canada between 2006 and 2013. During biannual follow-up assessments, non-fatal overdose within the previous 6 months was assessed. Bivariable and multivariable generalized mixed-effects regression models were used to determine the unadjusted and adjusted associations between HIV status, plasma HIV-1 RNA viral load, and likelihood of non-fatal overdose.
Author Manuscript
Results—A total of 1,760 eligible participants (67% male, median age = 42, and 42% HIVpositive at baseline) were included. Among 15,070 unique observations, 649 (4.3%) included a report of a non-fatal overdose within the previous 6 months (4.4% among seropositive and 4.3% among seronegative individuals). We did not observe a difference in the likelihood of overdose by HIV serostatus in crude (odds ratio [OR]: 1.05, p=0.853) analyses or analyses adjusted for known overdose risk factors (adjusted OR [AOR]: 1.19, p=0.474). In a secondary analysis, among HIVpositive PWID, we did not observe an association between having a detectable viral load and overdose (AOR: 1.03, p=0.862).
Send correspondence to: M-J Milloy, PhD, Research Scientist, Urban Health Research Initiative, British Columbia Centre for Excellence in HIV/AIDS, St., Paul's Hospital, 608-1081 Burrard Street, Vancouver, B.C., V6Z 1Y6 CANADA, Tel: (604) 682-2344 X 68473, Fax: (604) 806-8464, ; Email: [email protected] 4Present Address: School of Public Health, University of Saskatchewan, 104 Clinic Place Saskatoon SK, Canada S7N 5E5 Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Escudero et al.
Page 2
Author Manuscript
Conclusions—Despite the evidence that HIV infection is a risk factor for fatal overdose, we found no evidence for a relationship between HIV disease and non-fatal overdose. However, overdose remains high among PWID, indicating the need for ongoing policy addressing this problem, and research into understanding modifiable risk factors that predict non-fatal overdose. Keywords Overdose; HIV; mortality; injection drug users
1. Introduction
Author Manuscript Author Manuscript
There is contradictory evidence as to whether HIV infection confers additional risk for accidental overdose among people who inject drugs (PWID). Although some studies have found no association between HIV status and either fatal or non-fatal overdose (Ferreros, Lumbreras, Hurtado, Perez-Hoyos, & Hernandez-Aguado, 2008; Ochoa, et al., 2005), others have suggested HIV infection may be associated with higher risk of either type of overdose among both non-injecting and injecting drug users (Eskild, Magnus, Samuelsen, Sohlberg, & Kittelsen, 1993; Green, McGowan, Yokell, Pouget, & Rich, 2012; Solomon, et al., 2009; van Ameijden, Langendam, & Coutinho, 1999b; van Haastrecht, et al., 1996; Wang, et al., 2005; Zaccarelli, et al., 1994). Several biological mechanisms have been proposed as potential underlying reasons for increased risk, including immunosuppression (Lyles, et al., 1997; van Ameijden, Krol, et al., 1999a), liver or pulmonary disease (Solomon, et al., 2009; Wang, et al., 2005), and higher drug use-associated risk behaviors among those with HIV infection (Goedert, et al., 1995; van Ameijden, Langendam, et al., 1999b). Further explanations for potential causal relationships are summarized in a 2012 systematic review and meta-analysis (Green, et al., 2012). Pooling results from 27 studies published between 1988-2010, Green et al. estimated that HIV seropositivity was associated with a 74% increase in fatal overdose among samples consisting primarily of PWID. However, it remains uncertain whether HIV status is causally related to increased overdose risk, or if past observational studies have been subject to biases, including unmeasured confounding.
Author Manuscript
The effect of HIV status on overdose may also differ for fatal versus non-fatal overdose events: specifically, there has been more compelling evidence for an increased risk of fatal overdose among HIV-infected populations (Green, et al., 2012). For example, a study of heroin overdose indicated that systemic disease affecting the pulmonary system (i.e., sequelae from immunosuppression caused by HIV disease) may have distinct effects on the risk of non-fatal versus fatal overdoses, although such disease was expected to increase susceptibility to both events (Warner-Smith, Darke, Lynskey, & Hall, 2001). Further work is therefore needed to examine the relationship between HIV status and non-fatal overdose experiences, which until now has not garnered the same attention, and may have been presumed to have the same positive relationship as fatal overdose. In addition, increased treatment coverage of highly active antiretroviral therapy (HAART) among drug users in some settings warrants a re-examination of the relationship between HIV status and non-fatal overdose, based on more recent data. We are not aware of any currently available evidence that indicates whether those who are virally suppressed are at
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 3
Author Manuscript
the same risk of non-fatal overdose as those with detectable levels. With this analysis we hypothesized that, among HIV positive individuals, those with undetectable viral loads would have a reduced risk of non-fatal overdose; thus, the overall relationship between positive HIV status and risk of non-fatal overdose may have diminished in the modern treatment era. To test our hypotheses, we used 8 years of data from two community-recruited cohorts of HIV-positive and –negative PWID in Vancouver, Canada.
2. Material and Methods 2.1 Study eligibility and recruitment
Author Manuscript
We analyzed data from two parallel open prospective cohort studies in Vancouver, Canada: the AIDS Care Cohort to evaluate Exposure to Survival Services (ACCESS), consisting of HIV-positive PWID, and the Vancouver Injection Drug Users Study (VIDUS), consisting of HIV-negative PWID. The recruitment, assessment, and follow-up procedures for each of these cohorts have been described elsewhere in detail (Strathdee, et al., 1998; Tyndall, et al., 2003). For both cohorts, community-based recruitment strategies were used, including wordof-mouth, posters, and snowball sampling to recruit PWID in Vancouver's Downtown Eastside neighborhood, an area with high levels of illicit drug use, poverty, and homelessness.
Author Manuscript
Individuals were eligible for inclusion if they were aged ≥ 18 years and, in the case of VIDUS, had used illicit drugs via injection in the 30 days prior to the baseline interview. Although ACCESS includes those who do not inject drugs, we restricted this analysis to participants from both cohorts who had injected illicit drugs within the previous 180 days. HIV status for enrollment in VIDUS (HIV-negative) and ACCESS (HIV-positive) is determined by serologic assay at baseline. Participants who seroconvert during follow-up are automatically transitioned from VIDUS to ACCESS. All participants complete intervieweradministered surveys that elicit information on lifetime and recent characteristics, behaviors, and exposures. Participants also undergo examination by a study nurse, including blood sampling for serological analysis. The ACCESS and VIDUS studies have been approved by the University of British Columbia/Providence Healthcare Research Ethics Board. All participants provided written informed consent and received 20 CAD for each study visit. We included in this analysis all participants (n = 1,760) who completed at least one baseline or follow-up questionnaire during the study period (2006-2013). During each assessment, participants were asked whether they had experienced a non-fatal overdose (i.e., “a negative reaction from using too much drugs”) within the previous 6 months.
Author Manuscript
2.2 Statistical analyses and models First, we plotted the proportion of study participants reporting at least one overdose during each assessment, stratified by HIV status. Next, bivariable and multivariable generalized mixed-effects regression models were used to determine the longitudinal unadjusted and adjusted association between HIV status (time-updated to account for participants who seroconverted during follow-up), and non-fatal overdose (Model 1). In a secondary analysis, we examined the relationship between having a detectable viral load (≥50 copies/mL), and
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 4
Author Manuscript
non-fatal overdose among HIV-positive individuals (Model 2). For this analysis, we accessed the results of viral load tests held by the Drug Treatment Programme of the BC Centre for Excellence in HIV/AIDS, which dispenses all antiretroviral therapy within British Columbia and maintains a complete clinical monitoring database, including all viral load tests conducted through ACCESS or by each individual's physician. Of note, all HIV/AIDS treatment and care is delivered free of charge in this setting through the universal no-cost medical insurance plan.
Author Manuscript
In both models, we assessed for inclusion other established risk factors for overdose among PWID, including age; homelessness; heroin injection; cocaine injection; heavy alcohol use (classified here by self-reported binging or no binging); recent incarceration; and methadone treatment (Fischer, et al., 2004; Kerr, et al., 2007; Sergeev, Karpets, Sarang, & Tikhonov, 2003). We also assessed the following potential risk factors: gender (female vs. male); years of since injection initiation; ethnicity (Caucasian vs. non-Caucasian); education level (≥ high school diploma vs. < high school diploma); crack cocaine use; sex work; exposure to violence; hepatitis C virus (HCV) antibody status; the year of study enrollment; and the year of observation. Consistent with previous analyses, all of these variables, with the exception of gender and ethnicity, were time-updated and referred to the 180-day period prior to the study interview. To determine inclusion of covariates for each model, we used a stepwise selection procedure. This backwards selection protocol, as performed previously (Milloy, et al., 2011), and suggested by Greenland et al. (Maldonado & Greenland, 1993; Rothman & Greenland, 1998), sequentially removes from the full set of potential confounders any that did not change the estimate of the primary association by >5%. All p-values were two-sided.
3. Results Author Manuscript
Between December, 2006 and June, 2013, 1,760 participants completed at least one study interview, producing 15,070 unique assessments, or a median of 8 (inter-quartile range [IQR]: 4-12) per participant. In total, 1279 (67%) were male, 1153 (60%) self-reported Caucasian ancestry, and the median age at baseline was 42 (IQR: 36 – 48) years. Individuals with >1 interview did not differ from those with only one study interview (205, 11%) by gender or self-reported ethnicity (both p > 0.5). Individuals who never returned for followup were younger (40 vs. 43 years; p < 0.001) and more likely to report an overdose in the 6month period prior to the baseline interview (11% vs. 6%, p = 0.01) but not more likely to be HIV-positive (47% vs. 41%, p = 0.051).
Author Manuscript
Among these 15,070 observations, there were 649 (4.3%) reports of overdose within the previous 6 months (4.4% among seropositive and 4.3% among seronegative individuals). Figure 1 presents the proportion of participants reporting at least one overdose during each assessment. As shown, trends in the point prevalence of non-fatal overdose between HIV-positive and negative participants were similar. Results of the unadjusted analysis of HIV serostatus and longitudinal risk of overdose were null (odds ratio [OR]: 1.05, 95% confidence interval [CI]: 0.62–1.78, p=0.853). Full unadjusted results are presented in Table 1. Among those HIV-
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 5
Author Manuscript
positive, no association was observed between having a detectable viral load and risk of nonfatal overdose (OR: 0.89, CI: 0.56–1.43, p=0.630). Multivariate analysis also failed to identify a statistically significant association (adjusted OR [AOR]: 0.91, p=0.548) between serostatus and non-fatal overdose (Model 1). In Model 2, restricting to HIV-positive participants (n=815), the association between having detectable viral load and non-fatal overdose also yielded a null result (AOR: 1.03, p=0.862).
4. Discussion
Author Manuscript
In contrast to several earlier reports, we did not observe a significant association between non-fatal drug overdose and HIV serostatus. Within this large sample of HIV-positive and – negative PWID recruited from community settings, we observed 649 overdose events over an 8-year study period, indicating that both HIV-negative and -positive PWID remain at high risk for overdose.
Author Manuscript
These results contrast with several analyses that have detected a relationship between positive HIV serostatus and non-fatal overdose or overall overdose (Green, et al., 2012; Solomon, et al., 2009; van Haastrecht, et al., 1996; Wang, et al., 2005). The majority of studies included in the systematic review by Green et al. did not include non-fatal overdose as an endpoint, but rather only fatal overdose (Green, et al., 2012). Although it seems likely that potential causal mechanisms between HIV serostatus and fatal overdose would similarly be present for non-fatal overdose, HIV infection may confer specific risk for mortality among those who experience an overdose event. In assessing victims of fatal overdose in Baltimore, Wang et al. found that autopsy reports showed lower plasma levels of morphine and multidrug use among HIV-positive drug users, compared to HIV–negative drug users, suggesting that HIV positivity (along with other associated factors) may have increased the risk of mortality in the presence of lower drug doses (Wang, et al., 2005). Wang et al. also proposed that temporary decreases in drug-using behavior among those recently diagnosed with HIV may have resulted in future relapses in which users are particularly susceptible to fatal overdose (Wang, et al., 2005). If HIV infection does in fact confer increased risk of specifically fatal overdose, then this would highlight the need and potential benefits of increased access to naloxone for HIV-positive PWID.
Author Manuscript
Second, previous studies may have been confounded by risk factors for overdose, which are also known to be associated with HIV infection among PWID. For instance, chronic HCV infection, commonly observed in PWID, may have inflated previous estimates for the specific effect of HIV disease on overdose risk, as mediated by liver disease. HCV antibody status, however, was not associated with non-fatal overdose in our study (p = 0.782), so its potential role as a confounder elsewhere is uncertain. Similarly, long-term drug use, polysubstance use, and even withdrawal have been associated with immunosuppression, making it difficult to isolate the independent effect of HIV status on overdose risk apart from drug-related factors (Dinda, Gitman, & Singhal, 2005; Friedman & Eisenstein, 2004; Roy, et al., 2011).
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 6
Author Manuscript
The wide array of harm reduction, HIV treatment programs, and addiction services in the study setting (Kerr, et al., 2010; Marshall, Milloy, Wood, Montaner, & Kerr, 2011; Montaner, et al., 2010), may also contribute to the lack of an observed association. Specifically, coverage of HAART and subsequent levels of viral suppression were relatively high, with the prevalence of undetectable plasma viral load among HIV positive participants at 63% by 2013. It is possible that, as hypothesized, the effect of HIV infection on non-fatal overdose risk was attenuated compared to populations with lesser HAART coverage; however, we did not observe a significant association when comparing non-fatal overdose among HIV positive individuals with and without detectable plasma viral load.
Author Manuscript Author Manuscript
This study has some notable limitations. First, our primary endpoint is self-reported, and thus subject to potential biases and misclassification. Second, the relatively high proportion of HIV-positive PWID who have achieved viral suppression in the ACCESS cohort may have diminished our ability to observe the effects of uncontrolled HIV viral replication or its downstream effects (i.e., inflammation, immunosuppression) on non-fatal overdose. Additionally, although we adjusted for many known and suspected risk factors for non-fatal overdose, it remains possible that the presence of unknown effect measure modifiers in this setting limits the generalizability of our findings. For instance, the high access to harm reduction, addiction, and HIV treatment services within our study setting may make generalizability of our results outside of North America or other highly-resourced settings limited. Also, since we did not have the ability to distinguish between one or multiple reported non-fatal overdose events within the same 6-month period, we were not able to estimate the incidence of multiple events, or perform subgroup analysis. Finally, there remains some potential for selection bias due to loss to follow-up. Individuals who did not return for a second study assessment were more likely than those who did to report an overdose in the 6 months before the baseline interview (p=0.01); in addition, the baseline prevalence of HIV among individuals who did not return was 47% compared to 41% among those who did return (p=0.051). The results for HIV prevalence indicate that although selection bias was possible, it is unlikely it would be strong enough to result in the null associations observed in our study.
5. Conclusions
Author Manuscript
In this prospective cohort study of PWID in Vancouver, Canada, we did not observe an association between HIV serostatus or viral load and the likelihood of non-fatal overdose. However, questions remain as to whether there exist biological, behavioral, or structural mechanisms that increase the risk of non-fatal overdose among some sub-groups of PWID. The evidence for HIV playing a significant role in non-fatal overdose is scarce, suggesting that HIV infection may specifically increase the risk of mortality given the occurrence of an overdose, rather than induce the overdose itself. Although our findings are specific to a setting with substantial harm reduction, addiction, and HIV treatment services, we were able to account for a substantial amount of known and suspected risk factors for non-fatal overdose, enhancing our ability to find any true association. Our results do indicate, however, that overdose remains a substantial risk for all PWID, regardless of HIV status.
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 7
Author Manuscript
Acknowledgments The authors thank the study participants for their contributions to the research, as well as current and past researchers and staff. We would specifically like to thank: Kristie Starr, Deborah Graham, Tricia Collingham, Carmen Rock, Jennifer Matthews, Steve Kain, Benita Yip and Guillaume Colley for their research and administrative assistance.
Author Manuscript
Funding: The authors thank the study participants for their contribution to the research, as well as current and past researchers and staff. The study was supported by the US National Institutes of Health (VIDUS: R01DA011591, U01DA038886; ACCESS: R01DA021525). This research was undertaken, in part, thanks to funding from the Canada Research Chairs program through a Tier 1 Canada Research Chair in Inner City Medicine which supports Dr. Evan Wood. Dr. Milloy is supported in part by the United States National Institutes of Health (R01-DA021525). Dr. Montaner is supported by the British Columbia Ministry of Health and through an Avant-Garde Award (DP1DA026182) from the National Institute of Drug Abuse (NIDA), at the US National Institutes of Health (NIH). He has also received financial support from the International AIDS Society, United Nations AIDS Program, World Health Organization, National Institutes of Health Research-Office of AIDS Research, National Institute of Allergy & Infectious Diseases, The United States President's Emergency Plan for AIDS Relief (PEPfAR), UNICEF, the University of British Columbia, Simon Fraser University, Providence Health Care and Vancouver Coastal Health Authority. Daniel Escudero is supported by the National Institute on Drug Abuse (F31-DA037808-01). Dr. Hayashi is supported by the Canadian Institutes of Health Research.
References
Author Manuscript Author Manuscript
Dinda A, Gitman M, Singhal PC. Immunomodulatory effect of morphine: therapeutic implications. Expert Opin Drug Saf. 2005; 4:669–675. [PubMed: 16011446] Eskild A, Magnus P, Samuelsen SO, Sohlberg C, Kittelsen P. Differences in mortality rates and causes of death between HIV positive and HIV negative intravenous drug users. Int J Epidemiol. 1993; 22:315–320. [PubMed: 8505190] Ferreros I, Lumbreras B, Hurtado I, Perez-Hoyos S, Hernandez-Aguado I. The shifting pattern of cause-specific mortality in a cohort of human immunodeficiency virus-infected and non-infected injecting drug users. Addiction. 2008; 103:651–659. [PubMed: 18339110] Fischer B, Brissette S, Brochu S, Bruneau J, el-Guebaly N, Noel L, Rehm J, Tyndall M, Wild C, Mun P, Haydon E, Baliunas D. Determinants of overdose incidents among illicit opioid users in 5 Canadian cities. CMAJ. 2004; 171:235–239. [PubMed: 15289420] Friedman H, Eisenstein TK. Neurological basis of drug dependence and its effects on the immune system. J Neuroimmunol. 2004; 147:106–108. [PubMed: 14741438] Goedert JJ, Pizza G, Gritti FM, Costigliola P, Boschini A, Bini A, Lazzari C, Palareti A. Mortality among drug users in the AIDS era. Int J Epidemiol. 1995; 24:1204–1210. [PubMed: 8824864] Green TC, McGowan SK, Yokell MA, Pouget ER, Rich JD. HIV infection and risk of overdose: a systematic review and meta-analysis. AIDS. 2012; 26:403–417. [PubMed: 22112599] Kerr T, Fairbairn N, Tyndall M, Marsh D, Li K, Montaner J, Wood E. Predictors of non-fatal overdose among a cohort of polysubstance-using injection drug users. Drug Alcohol Depend. 2007; 87:39– 45. [PubMed: 16959438] Kerr T, Small W, Buchner C, Zhang R, Li K, Montaner J, Wood E. Syringe sharing and HIV incidence among injection drug users and increased access to sterile syringes. Am J Public Health. 2010; 100:1449–1453. [PubMed: 20558797] Lyles CM, Margolick JB, Astemborski J, Graham NM, Anthony JC, Hoover DR, Vlahov D. The influence of drug use patterns on the rate of CD4+ lymphocyte decline among HIV-1-infected injecting drug users. AIDS. 1997; 11:1255–1262. [PubMed: 9256944] Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol. 1993; 138:923–936. [PubMed: 8256780] Marshall BD, Milloy MJ, Wood E, Montaner JS, Kerr T. Reduction in overdose mortality after the opening of North America's first medically supervised safer injecting facility: a retrospective population-based study. Lancet. 2011; 377:1429–1437. [PubMed: 21497898] Milloy MJ, Kerr T, Buxton J, Rhodes T, Guillemi S, Hogg R, Montaner J, Wood E. Dose-response effect of incarceration events on nonadherence to HIV antiretroviral therapy among injection drug users. J Infect Dis. 2011; 203:1215–1221. [PubMed: 21459814]
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 8
Author Manuscript Author Manuscript Author Manuscript
Montaner JS, Lima VD, Barrios R, Yip B, Wood E, Kerr T, Shannon K, Harrigan PR, Hogg RS, Daly P, Kendall P. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a population-based study. Lancet. 2010; 376:532–539. [PubMed: 20638713] Ochoa KC, Davidson PJ, Evans JL, Hahn JA, Page-Shafer K, Moss AR. Heroin overdose among young injection drug users in San Francisco. Drug Alcohol Depend. 2005; 80:297–302. [PubMed: 15961257] Rothman, KJ.; Greenland, S. Modern epidemiology. 2nd. Philadelphia, PA: Lippincott-Raven; 1998. Roy S, Ninkovic J, Banerjee S, Charboneau RG, Das S, Dutta R, Kirchner VA, Koodie L, Ma J, Meng J, Barke RA. Opioid drug abuse and modulation of immune function: consequences in the susceptibility to opportunistic infections. J Neuroimmune Pharmacol. 2011; 6:442–465. [PubMed: 21789507] Sergeev B, Karpets A, Sarang A, Tikhonov M. Prevalence and circumstances of opiate overdose among injection drug users in the Russian Federation. J Urban Health. 2003; 80:212–219. [PubMed: 12791797] Solomon SS, Celentano DD, Srikrishnan AK, Vasudevan CK, Anand S, Kumar MS, Solomon S, Lucas GM, Mehta SH. Mortality among injection drug users in Chennai, India (2005-2008). AIDS. 2009; 23:997–1004. [PubMed: 19367155] Strathdee SA, Palepu A, Cornelisse PG, Yip B, O'Shaughnessy MV, Montaner JS, Schechter MT, Hogg RS. Barriers to use of free antiretroviral therapy in injection drug users. JAMA. 1998; 280:547–549. [PubMed: 9707146] Tyndall MW, Currie S, Spittal P, Li K, Wood E, O'Shaughnessy MV, Schechter MT. Intensive injection cocaine use as the primary risk factor in the Vancouver HIV-1 epidemic. AIDS. 2003; 17:887–893. [PubMed: 12660536] van Ameijden EJ, Krol A, Vlahov D, Flynn C, van Haastrecht HJ, Coutinho RA. Pre-AIDS mortality and morbidity among injection drug users in Amsterdam and Baltimore: an ecological comparison. Subst Use Misuse. 1999a; 34:845–865. [PubMed: 10227114] van Ameijden EJ, Langendam MW, Coutinho RA. Dose-effect relationship between overdose mortality and prescribed methadone dosage in low-threshold maintenance programs. Addict Behav. 1999b; 24:559–563. [PubMed: 10466851] van Haastrecht HJ, van Ameijden EJ, van den Hoek JA, Mientjes GH, Bax JS, Coutinho RA. Predictors of mortality in the Amsterdam cohort of human immunodeficiency virus (HIV)-positive and HIV-negative drug users. Am J Epidemiol. 1996; 143:380–391. [PubMed: 8633622] Wang C, Vlahov D, Galai N, Cole SR, Bareta J, Pollini R, Mehta SH, Nelson KE, Galea S. The effect of HIV infection on overdose mortality. AIDS. 2005; 19:935–942. [PubMed: 15905674] Warner-Smith M, Darke S, Lynskey M, Hall W. Heroin overdose: causes and consequences. Addiction. 2001; 96:1113–1125. [PubMed: 11487418] Zaccarelli M, Gattari P, Rezza G, Conti S, Spizzichino L, Vlahov D, Ippolito G, Lelli V, Valenzi C. Impact of HIV infection on non-AIDS mortality among Italian injecting drug users. AIDS. 1994; 8:345–350. [PubMed: 8031512]
Author Manuscript Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 9
Author Manuscript
Highlights •
It is unclear if HIV status is a risk-factor for non-fatal overdose among PWID
•
We examined this longitudinal association among 1,760 prospectively-enrolled PWID
•
We found no association between HIV status/viral load and overdose risk
•
These results contrast the predominant findings from earlier research
Author Manuscript Author Manuscript Author Manuscript Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 10
Author Manuscript Author Manuscript
Figure 1.
Proportion of people who inject drugs who reported a recent non-fatal overdose, stratified by HIV status (n = 1,760).
Author Manuscript Author Manuscript Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 11
Table 1
Author Manuscript
Unadjusted risk factors for non-fatal overdose among people who inject drugs in Vancouver, Canada (n = 1,760) Characteristic
Author Manuscript
OR (95% CI)
Model 1: AOR (95% CI)
Model 2: AOR (95% CI)
HIV serostatus (positive vs. negative)
1.05 (0.62-1.78)
0.91 (0.68-1.23)
-
Detectable VL (>50 copies/mL)*
0.89 (0.56-1.43)
-
1.03 (0.70-1.54)
Age (per year older)
0.98 (0.95-1.01)
0.99 (0.97-1.00)
-
Gender (female vs. male)
1.00 (0.57-1.72)
-
-
Years injecting (per year longer)
1.00 (0.97-1.02)
-
1.01 (0.99-1.04)
Caucasian (yes vs. no)
1.50 (0.88-2.56)
1.33 (0.98-1.80)
1.46 (0.94-2.27)
Education (≥HS vs.
Addict Behav. Author manuscript; available in PMC 2017 September 01. Published in final edited form as: Addict Behav. 2016 September ; 60: 8–12. doi:10.1016/j.addbeh.2016.03.029.
No association between HIV status and risk of non-fatal overdose among people who inject drugs in Vancouver, Canada Daniel J Escudero1, Brandon DL Marshall1, Thomas Kerr2,3, Kanna Hayashi2,3, Cindy Feng2,4, Silvia A Guillemi2, Robert S Hogg2, Julio Montaner2,3, Evan Wood2,3, and M-J Milloy2,3 of Epidemiology, Brown University School of Public Health, 2nd Floor, 121 S. Main St., Providence, RI, 02906, United States 1Department
Author Manuscript
2British
Columbia Centre for Excellence in HIV/AIDS, 608 – 1081 Burrard Street, Vancouver, B.C., Canada V6Z 1Y6
3Division
of AIDS, Department of Medicine, University of British Columbia, 667-1081 Burrard Street, Vancouver, BC, Canada V6Z1Y6
Abstract
Author Manuscript
Background—The evidence to date on whether HIV infection increases the risk of accidental drug overdose among people who inject drugs (PWID) is equivocal. Thus, we sought to estimate the effect of HIV infection on risk of non-fatal overdose among two parallel cohorts of HIVpositive and –negative PWID. Methods—Data were collected from a prospective cohort of PWID in Vancouver, Canada between 2006 and 2013. During biannual follow-up assessments, non-fatal overdose within the previous 6 months was assessed. Bivariable and multivariable generalized mixed-effects regression models were used to determine the unadjusted and adjusted associations between HIV status, plasma HIV-1 RNA viral load, and likelihood of non-fatal overdose.
Author Manuscript
Results—A total of 1,760 eligible participants (67% male, median age = 42, and 42% HIVpositive at baseline) were included. Among 15,070 unique observations, 649 (4.3%) included a report of a non-fatal overdose within the previous 6 months (4.4% among seropositive and 4.3% among seronegative individuals). We did not observe a difference in the likelihood of overdose by HIV serostatus in crude (odds ratio [OR]: 1.05, p=0.853) analyses or analyses adjusted for known overdose risk factors (adjusted OR [AOR]: 1.19, p=0.474). In a secondary analysis, among HIVpositive PWID, we did not observe an association between having a detectable viral load and overdose (AOR: 1.03, p=0.862).
Send correspondence to: M-J Milloy, PhD, Research Scientist, Urban Health Research Initiative, British Columbia Centre for Excellence in HIV/AIDS, St., Paul's Hospital, 608-1081 Burrard Street, Vancouver, B.C., V6Z 1Y6 CANADA, Tel: (604) 682-2344 X 68473, Fax: (604) 806-8464, ; Email: [email protected] 4Present Address: School of Public Health, University of Saskatchewan, 104 Clinic Place Saskatoon SK, Canada S7N 5E5 Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Escudero et al.
Page 2
Author Manuscript
Conclusions—Despite the evidence that HIV infection is a risk factor for fatal overdose, we found no evidence for a relationship between HIV disease and non-fatal overdose. However, overdose remains high among PWID, indicating the need for ongoing policy addressing this problem, and research into understanding modifiable risk factors that predict non-fatal overdose. Keywords Overdose; HIV; mortality; injection drug users
1. Introduction
Author Manuscript Author Manuscript
There is contradictory evidence as to whether HIV infection confers additional risk for accidental overdose among people who inject drugs (PWID). Although some studies have found no association between HIV status and either fatal or non-fatal overdose (Ferreros, Lumbreras, Hurtado, Perez-Hoyos, & Hernandez-Aguado, 2008; Ochoa, et al., 2005), others have suggested HIV infection may be associated with higher risk of either type of overdose among both non-injecting and injecting drug users (Eskild, Magnus, Samuelsen, Sohlberg, & Kittelsen, 1993; Green, McGowan, Yokell, Pouget, & Rich, 2012; Solomon, et al., 2009; van Ameijden, Langendam, & Coutinho, 1999b; van Haastrecht, et al., 1996; Wang, et al., 2005; Zaccarelli, et al., 1994). Several biological mechanisms have been proposed as potential underlying reasons for increased risk, including immunosuppression (Lyles, et al., 1997; van Ameijden, Krol, et al., 1999a), liver or pulmonary disease (Solomon, et al., 2009; Wang, et al., 2005), and higher drug use-associated risk behaviors among those with HIV infection (Goedert, et al., 1995; van Ameijden, Langendam, et al., 1999b). Further explanations for potential causal relationships are summarized in a 2012 systematic review and meta-analysis (Green, et al., 2012). Pooling results from 27 studies published between 1988-2010, Green et al. estimated that HIV seropositivity was associated with a 74% increase in fatal overdose among samples consisting primarily of PWID. However, it remains uncertain whether HIV status is causally related to increased overdose risk, or if past observational studies have been subject to biases, including unmeasured confounding.
Author Manuscript
The effect of HIV status on overdose may also differ for fatal versus non-fatal overdose events: specifically, there has been more compelling evidence for an increased risk of fatal overdose among HIV-infected populations (Green, et al., 2012). For example, a study of heroin overdose indicated that systemic disease affecting the pulmonary system (i.e., sequelae from immunosuppression caused by HIV disease) may have distinct effects on the risk of non-fatal versus fatal overdoses, although such disease was expected to increase susceptibility to both events (Warner-Smith, Darke, Lynskey, & Hall, 2001). Further work is therefore needed to examine the relationship between HIV status and non-fatal overdose experiences, which until now has not garnered the same attention, and may have been presumed to have the same positive relationship as fatal overdose. In addition, increased treatment coverage of highly active antiretroviral therapy (HAART) among drug users in some settings warrants a re-examination of the relationship between HIV status and non-fatal overdose, based on more recent data. We are not aware of any currently available evidence that indicates whether those who are virally suppressed are at
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 3
Author Manuscript
the same risk of non-fatal overdose as those with detectable levels. With this analysis we hypothesized that, among HIV positive individuals, those with undetectable viral loads would have a reduced risk of non-fatal overdose; thus, the overall relationship between positive HIV status and risk of non-fatal overdose may have diminished in the modern treatment era. To test our hypotheses, we used 8 years of data from two community-recruited cohorts of HIV-positive and –negative PWID in Vancouver, Canada.
2. Material and Methods 2.1 Study eligibility and recruitment
Author Manuscript
We analyzed data from two parallel open prospective cohort studies in Vancouver, Canada: the AIDS Care Cohort to evaluate Exposure to Survival Services (ACCESS), consisting of HIV-positive PWID, and the Vancouver Injection Drug Users Study (VIDUS), consisting of HIV-negative PWID. The recruitment, assessment, and follow-up procedures for each of these cohorts have been described elsewhere in detail (Strathdee, et al., 1998; Tyndall, et al., 2003). For both cohorts, community-based recruitment strategies were used, including wordof-mouth, posters, and snowball sampling to recruit PWID in Vancouver's Downtown Eastside neighborhood, an area with high levels of illicit drug use, poverty, and homelessness.
Author Manuscript
Individuals were eligible for inclusion if they were aged ≥ 18 years and, in the case of VIDUS, had used illicit drugs via injection in the 30 days prior to the baseline interview. Although ACCESS includes those who do not inject drugs, we restricted this analysis to participants from both cohorts who had injected illicit drugs within the previous 180 days. HIV status for enrollment in VIDUS (HIV-negative) and ACCESS (HIV-positive) is determined by serologic assay at baseline. Participants who seroconvert during follow-up are automatically transitioned from VIDUS to ACCESS. All participants complete intervieweradministered surveys that elicit information on lifetime and recent characteristics, behaviors, and exposures. Participants also undergo examination by a study nurse, including blood sampling for serological analysis. The ACCESS and VIDUS studies have been approved by the University of British Columbia/Providence Healthcare Research Ethics Board. All participants provided written informed consent and received 20 CAD for each study visit. We included in this analysis all participants (n = 1,760) who completed at least one baseline or follow-up questionnaire during the study period (2006-2013). During each assessment, participants were asked whether they had experienced a non-fatal overdose (i.e., “a negative reaction from using too much drugs”) within the previous 6 months.
Author Manuscript
2.2 Statistical analyses and models First, we plotted the proportion of study participants reporting at least one overdose during each assessment, stratified by HIV status. Next, bivariable and multivariable generalized mixed-effects regression models were used to determine the longitudinal unadjusted and adjusted association between HIV status (time-updated to account for participants who seroconverted during follow-up), and non-fatal overdose (Model 1). In a secondary analysis, we examined the relationship between having a detectable viral load (≥50 copies/mL), and
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 4
Author Manuscript
non-fatal overdose among HIV-positive individuals (Model 2). For this analysis, we accessed the results of viral load tests held by the Drug Treatment Programme of the BC Centre for Excellence in HIV/AIDS, which dispenses all antiretroviral therapy within British Columbia and maintains a complete clinical monitoring database, including all viral load tests conducted through ACCESS or by each individual's physician. Of note, all HIV/AIDS treatment and care is delivered free of charge in this setting through the universal no-cost medical insurance plan.
Author Manuscript
In both models, we assessed for inclusion other established risk factors for overdose among PWID, including age; homelessness; heroin injection; cocaine injection; heavy alcohol use (classified here by self-reported binging or no binging); recent incarceration; and methadone treatment (Fischer, et al., 2004; Kerr, et al., 2007; Sergeev, Karpets, Sarang, & Tikhonov, 2003). We also assessed the following potential risk factors: gender (female vs. male); years of since injection initiation; ethnicity (Caucasian vs. non-Caucasian); education level (≥ high school diploma vs. < high school diploma); crack cocaine use; sex work; exposure to violence; hepatitis C virus (HCV) antibody status; the year of study enrollment; and the year of observation. Consistent with previous analyses, all of these variables, with the exception of gender and ethnicity, were time-updated and referred to the 180-day period prior to the study interview. To determine inclusion of covariates for each model, we used a stepwise selection procedure. This backwards selection protocol, as performed previously (Milloy, et al., 2011), and suggested by Greenland et al. (Maldonado & Greenland, 1993; Rothman & Greenland, 1998), sequentially removes from the full set of potential confounders any that did not change the estimate of the primary association by >5%. All p-values were two-sided.
3. Results Author Manuscript
Between December, 2006 and June, 2013, 1,760 participants completed at least one study interview, producing 15,070 unique assessments, or a median of 8 (inter-quartile range [IQR]: 4-12) per participant. In total, 1279 (67%) were male, 1153 (60%) self-reported Caucasian ancestry, and the median age at baseline was 42 (IQR: 36 – 48) years. Individuals with >1 interview did not differ from those with only one study interview (205, 11%) by gender or self-reported ethnicity (both p > 0.5). Individuals who never returned for followup were younger (40 vs. 43 years; p < 0.001) and more likely to report an overdose in the 6month period prior to the baseline interview (11% vs. 6%, p = 0.01) but not more likely to be HIV-positive (47% vs. 41%, p = 0.051).
Author Manuscript
Among these 15,070 observations, there were 649 (4.3%) reports of overdose within the previous 6 months (4.4% among seropositive and 4.3% among seronegative individuals). Figure 1 presents the proportion of participants reporting at least one overdose during each assessment. As shown, trends in the point prevalence of non-fatal overdose between HIV-positive and negative participants were similar. Results of the unadjusted analysis of HIV serostatus and longitudinal risk of overdose were null (odds ratio [OR]: 1.05, 95% confidence interval [CI]: 0.62–1.78, p=0.853). Full unadjusted results are presented in Table 1. Among those HIV-
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 5
Author Manuscript
positive, no association was observed between having a detectable viral load and risk of nonfatal overdose (OR: 0.89, CI: 0.56–1.43, p=0.630). Multivariate analysis also failed to identify a statistically significant association (adjusted OR [AOR]: 0.91, p=0.548) between serostatus and non-fatal overdose (Model 1). In Model 2, restricting to HIV-positive participants (n=815), the association between having detectable viral load and non-fatal overdose also yielded a null result (AOR: 1.03, p=0.862).
4. Discussion
Author Manuscript
In contrast to several earlier reports, we did not observe a significant association between non-fatal drug overdose and HIV serostatus. Within this large sample of HIV-positive and – negative PWID recruited from community settings, we observed 649 overdose events over an 8-year study period, indicating that both HIV-negative and -positive PWID remain at high risk for overdose.
Author Manuscript
These results contrast with several analyses that have detected a relationship between positive HIV serostatus and non-fatal overdose or overall overdose (Green, et al., 2012; Solomon, et al., 2009; van Haastrecht, et al., 1996; Wang, et al., 2005). The majority of studies included in the systematic review by Green et al. did not include non-fatal overdose as an endpoint, but rather only fatal overdose (Green, et al., 2012). Although it seems likely that potential causal mechanisms between HIV serostatus and fatal overdose would similarly be present for non-fatal overdose, HIV infection may confer specific risk for mortality among those who experience an overdose event. In assessing victims of fatal overdose in Baltimore, Wang et al. found that autopsy reports showed lower plasma levels of morphine and multidrug use among HIV-positive drug users, compared to HIV–negative drug users, suggesting that HIV positivity (along with other associated factors) may have increased the risk of mortality in the presence of lower drug doses (Wang, et al., 2005). Wang et al. also proposed that temporary decreases in drug-using behavior among those recently diagnosed with HIV may have resulted in future relapses in which users are particularly susceptible to fatal overdose (Wang, et al., 2005). If HIV infection does in fact confer increased risk of specifically fatal overdose, then this would highlight the need and potential benefits of increased access to naloxone for HIV-positive PWID.
Author Manuscript
Second, previous studies may have been confounded by risk factors for overdose, which are also known to be associated with HIV infection among PWID. For instance, chronic HCV infection, commonly observed in PWID, may have inflated previous estimates for the specific effect of HIV disease on overdose risk, as mediated by liver disease. HCV antibody status, however, was not associated with non-fatal overdose in our study (p = 0.782), so its potential role as a confounder elsewhere is uncertain. Similarly, long-term drug use, polysubstance use, and even withdrawal have been associated with immunosuppression, making it difficult to isolate the independent effect of HIV status on overdose risk apart from drug-related factors (Dinda, Gitman, & Singhal, 2005; Friedman & Eisenstein, 2004; Roy, et al., 2011).
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 6
Author Manuscript
The wide array of harm reduction, HIV treatment programs, and addiction services in the study setting (Kerr, et al., 2010; Marshall, Milloy, Wood, Montaner, & Kerr, 2011; Montaner, et al., 2010), may also contribute to the lack of an observed association. Specifically, coverage of HAART and subsequent levels of viral suppression were relatively high, with the prevalence of undetectable plasma viral load among HIV positive participants at 63% by 2013. It is possible that, as hypothesized, the effect of HIV infection on non-fatal overdose risk was attenuated compared to populations with lesser HAART coverage; however, we did not observe a significant association when comparing non-fatal overdose among HIV positive individuals with and without detectable plasma viral load.
Author Manuscript Author Manuscript
This study has some notable limitations. First, our primary endpoint is self-reported, and thus subject to potential biases and misclassification. Second, the relatively high proportion of HIV-positive PWID who have achieved viral suppression in the ACCESS cohort may have diminished our ability to observe the effects of uncontrolled HIV viral replication or its downstream effects (i.e., inflammation, immunosuppression) on non-fatal overdose. Additionally, although we adjusted for many known and suspected risk factors for non-fatal overdose, it remains possible that the presence of unknown effect measure modifiers in this setting limits the generalizability of our findings. For instance, the high access to harm reduction, addiction, and HIV treatment services within our study setting may make generalizability of our results outside of North America or other highly-resourced settings limited. Also, since we did not have the ability to distinguish between one or multiple reported non-fatal overdose events within the same 6-month period, we were not able to estimate the incidence of multiple events, or perform subgroup analysis. Finally, there remains some potential for selection bias due to loss to follow-up. Individuals who did not return for a second study assessment were more likely than those who did to report an overdose in the 6 months before the baseline interview (p=0.01); in addition, the baseline prevalence of HIV among individuals who did not return was 47% compared to 41% among those who did return (p=0.051). The results for HIV prevalence indicate that although selection bias was possible, it is unlikely it would be strong enough to result in the null associations observed in our study.
5. Conclusions
Author Manuscript
In this prospective cohort study of PWID in Vancouver, Canada, we did not observe an association between HIV serostatus or viral load and the likelihood of non-fatal overdose. However, questions remain as to whether there exist biological, behavioral, or structural mechanisms that increase the risk of non-fatal overdose among some sub-groups of PWID. The evidence for HIV playing a significant role in non-fatal overdose is scarce, suggesting that HIV infection may specifically increase the risk of mortality given the occurrence of an overdose, rather than induce the overdose itself. Although our findings are specific to a setting with substantial harm reduction, addiction, and HIV treatment services, we were able to account for a substantial amount of known and suspected risk factors for non-fatal overdose, enhancing our ability to find any true association. Our results do indicate, however, that overdose remains a substantial risk for all PWID, regardless of HIV status.
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 7
Author Manuscript
Acknowledgments The authors thank the study participants for their contributions to the research, as well as current and past researchers and staff. We would specifically like to thank: Kristie Starr, Deborah Graham, Tricia Collingham, Carmen Rock, Jennifer Matthews, Steve Kain, Benita Yip and Guillaume Colley for their research and administrative assistance.
Author Manuscript
Funding: The authors thank the study participants for their contribution to the research, as well as current and past researchers and staff. The study was supported by the US National Institutes of Health (VIDUS: R01DA011591, U01DA038886; ACCESS: R01DA021525). This research was undertaken, in part, thanks to funding from the Canada Research Chairs program through a Tier 1 Canada Research Chair in Inner City Medicine which supports Dr. Evan Wood. Dr. Milloy is supported in part by the United States National Institutes of Health (R01-DA021525). Dr. Montaner is supported by the British Columbia Ministry of Health and through an Avant-Garde Award (DP1DA026182) from the National Institute of Drug Abuse (NIDA), at the US National Institutes of Health (NIH). He has also received financial support from the International AIDS Society, United Nations AIDS Program, World Health Organization, National Institutes of Health Research-Office of AIDS Research, National Institute of Allergy & Infectious Diseases, The United States President's Emergency Plan for AIDS Relief (PEPfAR), UNICEF, the University of British Columbia, Simon Fraser University, Providence Health Care and Vancouver Coastal Health Authority. Daniel Escudero is supported by the National Institute on Drug Abuse (F31-DA037808-01). Dr. Hayashi is supported by the Canadian Institutes of Health Research.
References
Author Manuscript Author Manuscript
Dinda A, Gitman M, Singhal PC. Immunomodulatory effect of morphine: therapeutic implications. Expert Opin Drug Saf. 2005; 4:669–675. [PubMed: 16011446] Eskild A, Magnus P, Samuelsen SO, Sohlberg C, Kittelsen P. Differences in mortality rates and causes of death between HIV positive and HIV negative intravenous drug users. Int J Epidemiol. 1993; 22:315–320. [PubMed: 8505190] Ferreros I, Lumbreras B, Hurtado I, Perez-Hoyos S, Hernandez-Aguado I. The shifting pattern of cause-specific mortality in a cohort of human immunodeficiency virus-infected and non-infected injecting drug users. Addiction. 2008; 103:651–659. [PubMed: 18339110] Fischer B, Brissette S, Brochu S, Bruneau J, el-Guebaly N, Noel L, Rehm J, Tyndall M, Wild C, Mun P, Haydon E, Baliunas D. Determinants of overdose incidents among illicit opioid users in 5 Canadian cities. CMAJ. 2004; 171:235–239. [PubMed: 15289420] Friedman H, Eisenstein TK. Neurological basis of drug dependence and its effects on the immune system. J Neuroimmunol. 2004; 147:106–108. [PubMed: 14741438] Goedert JJ, Pizza G, Gritti FM, Costigliola P, Boschini A, Bini A, Lazzari C, Palareti A. Mortality among drug users in the AIDS era. Int J Epidemiol. 1995; 24:1204–1210. [PubMed: 8824864] Green TC, McGowan SK, Yokell MA, Pouget ER, Rich JD. HIV infection and risk of overdose: a systematic review and meta-analysis. AIDS. 2012; 26:403–417. [PubMed: 22112599] Kerr T, Fairbairn N, Tyndall M, Marsh D, Li K, Montaner J, Wood E. Predictors of non-fatal overdose among a cohort of polysubstance-using injection drug users. Drug Alcohol Depend. 2007; 87:39– 45. [PubMed: 16959438] Kerr T, Small W, Buchner C, Zhang R, Li K, Montaner J, Wood E. Syringe sharing and HIV incidence among injection drug users and increased access to sterile syringes. Am J Public Health. 2010; 100:1449–1453. [PubMed: 20558797] Lyles CM, Margolick JB, Astemborski J, Graham NM, Anthony JC, Hoover DR, Vlahov D. The influence of drug use patterns on the rate of CD4+ lymphocyte decline among HIV-1-infected injecting drug users. AIDS. 1997; 11:1255–1262. [PubMed: 9256944] Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol. 1993; 138:923–936. [PubMed: 8256780] Marshall BD, Milloy MJ, Wood E, Montaner JS, Kerr T. Reduction in overdose mortality after the opening of North America's first medically supervised safer injecting facility: a retrospective population-based study. Lancet. 2011; 377:1429–1437. [PubMed: 21497898] Milloy MJ, Kerr T, Buxton J, Rhodes T, Guillemi S, Hogg R, Montaner J, Wood E. Dose-response effect of incarceration events on nonadherence to HIV antiretroviral therapy among injection drug users. J Infect Dis. 2011; 203:1215–1221. [PubMed: 21459814]
Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 8
Author Manuscript Author Manuscript Author Manuscript
Montaner JS, Lima VD, Barrios R, Yip B, Wood E, Kerr T, Shannon K, Harrigan PR, Hogg RS, Daly P, Kendall P. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a population-based study. Lancet. 2010; 376:532–539. [PubMed: 20638713] Ochoa KC, Davidson PJ, Evans JL, Hahn JA, Page-Shafer K, Moss AR. Heroin overdose among young injection drug users in San Francisco. Drug Alcohol Depend. 2005; 80:297–302. [PubMed: 15961257] Rothman, KJ.; Greenland, S. Modern epidemiology. 2nd. Philadelphia, PA: Lippincott-Raven; 1998. Roy S, Ninkovic J, Banerjee S, Charboneau RG, Das S, Dutta R, Kirchner VA, Koodie L, Ma J, Meng J, Barke RA. Opioid drug abuse and modulation of immune function: consequences in the susceptibility to opportunistic infections. J Neuroimmune Pharmacol. 2011; 6:442–465. [PubMed: 21789507] Sergeev B, Karpets A, Sarang A, Tikhonov M. Prevalence and circumstances of opiate overdose among injection drug users in the Russian Federation. J Urban Health. 2003; 80:212–219. [PubMed: 12791797] Solomon SS, Celentano DD, Srikrishnan AK, Vasudevan CK, Anand S, Kumar MS, Solomon S, Lucas GM, Mehta SH. Mortality among injection drug users in Chennai, India (2005-2008). AIDS. 2009; 23:997–1004. [PubMed: 19367155] Strathdee SA, Palepu A, Cornelisse PG, Yip B, O'Shaughnessy MV, Montaner JS, Schechter MT, Hogg RS. Barriers to use of free antiretroviral therapy in injection drug users. JAMA. 1998; 280:547–549. [PubMed: 9707146] Tyndall MW, Currie S, Spittal P, Li K, Wood E, O'Shaughnessy MV, Schechter MT. Intensive injection cocaine use as the primary risk factor in the Vancouver HIV-1 epidemic. AIDS. 2003; 17:887–893. [PubMed: 12660536] van Ameijden EJ, Krol A, Vlahov D, Flynn C, van Haastrecht HJ, Coutinho RA. Pre-AIDS mortality and morbidity among injection drug users in Amsterdam and Baltimore: an ecological comparison. Subst Use Misuse. 1999a; 34:845–865. [PubMed: 10227114] van Ameijden EJ, Langendam MW, Coutinho RA. Dose-effect relationship between overdose mortality and prescribed methadone dosage in low-threshold maintenance programs. Addict Behav. 1999b; 24:559–563. [PubMed: 10466851] van Haastrecht HJ, van Ameijden EJ, van den Hoek JA, Mientjes GH, Bax JS, Coutinho RA. Predictors of mortality in the Amsterdam cohort of human immunodeficiency virus (HIV)-positive and HIV-negative drug users. Am J Epidemiol. 1996; 143:380–391. [PubMed: 8633622] Wang C, Vlahov D, Galai N, Cole SR, Bareta J, Pollini R, Mehta SH, Nelson KE, Galea S. The effect of HIV infection on overdose mortality. AIDS. 2005; 19:935–942. [PubMed: 15905674] Warner-Smith M, Darke S, Lynskey M, Hall W. Heroin overdose: causes and consequences. Addiction. 2001; 96:1113–1125. [PubMed: 11487418] Zaccarelli M, Gattari P, Rezza G, Conti S, Spizzichino L, Vlahov D, Ippolito G, Lelli V, Valenzi C. Impact of HIV infection on non-AIDS mortality among Italian injecting drug users. AIDS. 1994; 8:345–350. [PubMed: 8031512]
Author Manuscript Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 9
Author Manuscript
Highlights •
It is unclear if HIV status is a risk-factor for non-fatal overdose among PWID
•
We examined this longitudinal association among 1,760 prospectively-enrolled PWID
•
We found no association between HIV status/viral load and overdose risk
•
These results contrast the predominant findings from earlier research
Author Manuscript Author Manuscript Author Manuscript Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 10
Author Manuscript Author Manuscript
Figure 1.
Proportion of people who inject drugs who reported a recent non-fatal overdose, stratified by HIV status (n = 1,760).
Author Manuscript Author Manuscript Addict Behav. Author manuscript; available in PMC 2017 September 01.
Escudero et al.
Page 11
Table 1
Author Manuscript
Unadjusted risk factors for non-fatal overdose among people who inject drugs in Vancouver, Canada (n = 1,760) Characteristic
Author Manuscript
OR (95% CI)
Model 1: AOR (95% CI)
Model 2: AOR (95% CI)
HIV serostatus (positive vs. negative)
1.05 (0.62-1.78)
0.91 (0.68-1.23)
-
Detectable VL (>50 copies/mL)*
0.89 (0.56-1.43)
-
1.03 (0.70-1.54)
Age (per year older)
0.98 (0.95-1.01)
0.99 (0.97-1.00)
-
Gender (female vs. male)
1.00 (0.57-1.72)
-
-
Years injecting (per year longer)
1.00 (0.97-1.02)
-
1.01 (0.99-1.04)
Caucasian (yes vs. no)
1.50 (0.88-2.56)
1.33 (0.98-1.80)
1.46 (0.94-2.27)
Education (≥HS vs.
