Factors associated with health-related cannabis use intentions among a community sample of people who inject drugs in Los Angeles and San Francisco, CA 2016 to 2018

Abstract

Objective:

Cannabis motivations have been studied extensively among patients of medicinal cannabis dispensaries, but less is known about motivations in community samples of opioid-using people who inject drugs. Our objective is to describe cannabis use motivations associated with self-treatment of physical pain, emotional issues, and as an opioid substitute.

Methods:

Data come from 6-month follow-up interviews with people who inject drugs who participated in a study on the efficacy of an injection initiation prevention intervention in Los Angeles and San Francisco, California from 2016-18. The analytic sample consists of 387 people who inject drugs who reported past-month cannabis use. We developed multivariable logistic regression models by reported cannabis use motivations: physical pain relief, emotional problems, and opioid substitute.

Results:

The most common cannabis use motivations reported by people who inject drugs was to “get high,” relieve physical pain and emotional problems, and reduce opioid use. In separate multivariate models, using cannabis for physical pain relief was associated with higher odds of using cannabis as a substitute for opioids; cannabis for emotional problems was associated with being diagnosed with depression; and cannabis as a substitute for opioids was associated with non-prescribed, non-injection methadone use.

Conclusion:

People who inject drugs reported using cannabis for health-related motivations. This motivation aligns with health needs and suggests the acceptability of cannabis use for health reasons in this population. Studies to determine the medical effectiveness of cannabis products for these common health and mental health needs among people who inject drugs are needed.

1.0. Introduction

1.1. United States trends in cannabis

Cannabis is the most widely used substance globally, with an estimated 188 million people having used it in 2017 (UNODC, 2019). Public acceptance of cannabis use in the United States (U.S.) is growing, with 20% of the U.S. population now living in states that have passed ballot initiatives to allow cannabis to be sold for recreational use (Kilmer, 2017; National Conference of State Legislatures, 2019; Marijuana Policy Project, 2020). Changing acceptance of cannabis is reflected in studies showing lower perceived risk of cannabis (Carliner et al., 2017; Pacek et al., 2015) and increased use among adults (Compton et al., 2019). With growing use, legalization, and social acceptance of cannabis, understanding cannabis use practices and motivations is warranted.

1.2. Cannabis use, opioid related health outcomes, and people who inject drugs.

While more recent analyses have called into question the expanded legalization and proposed benefits of cannabis (Chihuri and Li, 2019; Powell et al., 2018), clinical, experimental, and ethnographic research has explored cannabis as an alternative strategy to alleviate pain and manage non-medical opioid use (Cooke et al., 2019; Elikkottil et al., 2009; Hall et al., 2018; Hurd et al., 2019; Wenger et al., 2014). For instance, state-level ecological studies have suggested that broader access to medical and/or recreational cannabis (Livingston et al., 2017) as a result of state legalization might be associated with decreased prescription opioid use (Bradford et al., 2018; Wen and Hockenberry, 2018), hospitalization (Shi, 2017), traffic fatalities (Kim et al., 2016), opioid prescribing (Bradford et al., 2018; Lopez et al., 9000), and overdose mortality rates (Bachhuber et al., 2015; Bachhuber et al., 2014; Finney et al., 2015; Powell et al., 2018).

Studies of community recruited people who inject drugs (PWID) provide some evidence of an inverse relationship between cannabis and opioid use. For instance, Kral et al (2015), in a cross-sectional study, found that any cannabis use was less likely among PWID with high frequency opioid use (Kral et al., 2015) More promisingly, Lake et al. (2019), found that in cohorts of people who use drugs, some of whom also injected drugs, daily cannabis use was associated with lower odds of daily opioid use (Lake et al., 2019).

Implications of this association for people who inject drugs (PWID) could be significant. Many of the health ailments caused by opioid injection are related to the frequency and mode of use (Beyrer et al., 2009; Bluthenthal et al., 2020; Caudarella et al., 2016). Declines in opioid use due to cannabis use could therefore reduce fatal and non-fatal overdoses, Human Immunodeficiency Virus and Hepatitis C Virus transmissions, and bacterial infections and sequalae like injection-related infective endocarditis in this vulnerable population is suggested by others (Lake et al., 2020; Lucas, 2017; Wiese and Wilson-Poe, 2018). In this context, it is useful to know how PWID characterize their own cannabis use and whether cannabis use is associated with health-related intentions.

1.3. Cannabis use intentions among opioid-using populations

An important contributor to the effort to legalize cannabis has been its potential medicinal benefits (Bostwick, 2012; Takakuwa, 2020). General population surveys have found that health-related motivations for cannabis use are common and some report using cannabis to reduce opioid use (Azcarate et al., 2020; Ishida et al., 2019; Keyhani et al., 2018). Further, medicinal cannabis patients report using cannabis for a wide range of emotional issues, physical pain, and even as a substitute for chronic opioid use, defined as at least 90 days of opioid use with a gap of no greater than 30 days between consecutive use within a 180-day period (Baron et al., 2018; Lucas et al., 2013; Lucas and Walsh, 2017; Shah et al., 2019). However, motivations for cannabis use among chronic opioid-using PWID have not been closely examined. In the following, we conducted analyses to determine factors correlated with cannabis use motivations related to physical pain relief, emotional regulation, and as a substitute for opioids among a community-recruited sample of chronic opioid-using PWID.

2.0. Methods

2.1. Study procedures, eligibility, and recruitment

Data for these analyses were drawn from a randomized controlled trial that assessed the efficacy of a behavioral intervention to reduce injection initiation among PWID in Los Angeles and San Francisco, California from 2016 to 2018 (Strike et al., 2014). Participants in the parent study were recruited using targeted sampling methods (Bluthenthal and Watters, 1995; Kral et al., 2010; Watters and Biernacki, 1989). Targeted sampling is a method for recruiting “hidden” or other populations who are not easily identified due to the stigma and illegality associated with injection drug use. This approach to identifying and recruiting PWID has been found to be equivalent to respondent-driven sampling (Kral et al., 2010; Robinson et al., 2006) and has been used by this investigative team since 1991. Targeted sampling methods include: 1) assessment of administrative dataset about local patterns of substance use, 2) ethnographic mapping of substance buying scenes and service utilization locations (Koester, 1994), and 3) ongoing assessment of recruitment to ensure that known demographic and substance-using populations are included in the sample.

Eligible study participants self-reported injection drug use in the past 30 days (i.e., verified visible signs of venipuncture), being 18 years of age or older, and were able to provide informed consent. A total of 978 participants were enrolled. This analysis was restricted to individuals who completed their 6-month follow-up interview, and who self-reported using cannabis within the prior 30 days (N=387). Due to concerns about respondent burden, we did not ask items specifically related to cannabis motivations on the baseline survey. These items were added to the 6-month interview, which was shorter than the baseline interview and thus had sufficient space and time to ask for these additional items.

All data were obtained using a ~45 minute quantitative survey administered by trained research assistants (including co-author K.S.) in a private setting using computer-assisted personal interviewing software (Nova Research, Bethesda, MD). Participants received $20 U.S. Dollars for completing the quantitative survey. The institutional review board at the University of Southern California approved the study procedures.

2.2. Key study domains and variables

Our key variables were collected in the following manner. To determine cannabis use in the last 30 days we asked, “Have you used marijuana or cannabis in the last 30 days?” Participants who responded affirmatively were asked questions assessing frequency of use in the last 30 days. Frequency of cannabis use was recoded into “less than daily,” “once or twice a day,” or “three or more times a day.”

Motivation for cannabis use was collected using the following items: “When you used marijuana or cannabis in the last 30 days, how often did you use it to relieve physical problems such as pain, withdrawal symptoms, or discomfort from coming down from other drugs?”; “When you used marijuana or cannabis in the last 30 days, how often did you use it to relieve emotional problems such as depression, anxiety, or feelings of sadness?”; “When you used marijuana or cannabis in the last 30 days, how often did you use it to reduce your heroin or prescription opiate use?”; and “When you used marijuana or cannabis in the last 30 days, how often did you use it to get high?” Response options for both questions were “None,” “Some,” “About half,” “Most,” “All,” “Don’t know,” “Refuse to answer.” and “Not applicable.” To capture regular use of cannabis by motivation, we recoded each item such that those who reported using “about half or more” were classified as “mostly used for [motivation].” Those responding “none” or “some” were classified as “less than half” on these items.

We were particularly interested in patterns of cannabis motivation by age, gender, race/ethnicity (Asian/Pacific Islander, Native American, Latinx, Black, and White), sexual orientation (lesbian, gay, bisexual, or heterosexual), income level and sources of income, diagnosis for a mental health problem (yes/no), and specific diagnoses for depression, anxiety, post-traumatic stress disorder (PTSD), schizophrenia, and bipolar disorder (yes/no) as predictors of cannabis use motivations in this analysis. We also examined drug use patterns, including 30-day use of heroin, methamphetamine, cocaine, crack cocaine, and non-prescribed use of opioids, tranquilizers, stimulants, sedatives, methadone, and buprenorphine. In addition, we considered the total usage of opioids (heroin, speedballs [heroin/cocaine admixture], prescription opioids), cocaine (and crack cocaine), and methamphetamine (goofballs [heroin/methamphetamine admixture]) using summed variables.

2.3. Statistical analysis

After descriptive analyses, we conducted bivariate analyses to determine factors associated at the p < 0.05 level with cannabis use motivations using chi-square test for categorical variables and t-test for continuous variables. We considered variables in the domains of demographics, socioeconomic characteristics, mental health, and drug use patterns as potential confounders. Variables significant in bivariate analyses were then assessed for collinearity within domains using Pearson correlation coefficient. Collinear variables (Pearson correlation coefficient >0.300) were considered in the final analyses based on the strength of association with the dependent variable. Multivariable logistic regression models (using listwise deletion) were then conducted with cannabis use motivations as the outcome variable in two independent models. Variables not significant in multivariate analyses (p > 0.05) were removed from models. For both models, Odds Ratios (OR) and 95% Confidence Intervals (CI) were estimated. All analyses were performed using SPSS, Version 25.

We considered all variables in Table 1 for bivariate analysis. Only variables significant in bivariate and multivariate analyses were presented. We have included variables found to be significant in studies on cannabis motivations among medicinal cannabis patients and related studies, such as motivations to use cannabis to manage a mental health diagnosis (Lev-Ran et al., 2014 Gobbi et al., 2019; Rhew et al., 2017; Metrik et al., 2018; Wilkinson et al., 2015; Kansagara et al., 2017). As published studies on cannabis use among medicinal cannabis patients rarely include PWID, we have drawn upon additional studies on PWID and cannabis use to examine cannabis motivations in our study population of PWID (Boehnke et al., 2016; Cooper et al., 2018; Haroutounian et al., 2016; Kral et al., 2015; Lake et al., 2019; Slawson et al., 2015; Takakuwa, 2020; Wenger et al., 2014).

Table 1:

Selected demographic, socioeconomic, and cannabis use characteristics of opioid-using people who use inject drugs in Los Angeles and San Francisco, CA 2016-2018 (N=387)

Characteristic	N (%)
Study site
Los Angeles	160 (41%)
San Francisco	227 (59%)
Gender
Transgender	3 (1%)
Other	4 (1%)
Heterosexual female	77 (20%)
Heterosexual male	303 (78%)
Age
<30 years old	70 (18%)
30 to 39 years old	93 (24%)
40 to 49 years old	108 (28%)
≥50 years old	116 (30%)
Race / ethnicity
Asian/Pacific Islander	3 (1%)
Native American	25 (6%)
Mixed race	28 (7%)
African American	87 (23%)
Latinx	73 (19%)
White	171 (44%)
High school education or more
Yes	289 (75%)
No	98 (25%)
Gay, lesbian, or bisexual
Yes	78 (20%)
No	309 (80%)
HIV positive
Yes	27 (7%)
No/Unknown	360 (93%)
Monthly income
<$1,401	300 (78%)
$1,400 plus	87 (22%)
Currently homeless
Yes	290 (75%)
No	97 (25%)
Income source, last 6 months
Welfare, general relief	228 (59%)
Illegal or possible illegal source	188 (49%)
Panhandling	99 (26%)
Recycling	89 (23%)
Job	58 (15%)
Veteran’s benefits	10 (3%)
Years of drug injection ≥ (Don’t know=1)
<10 years	128 (33%)
10-19 years	84 (22%)
≥20 years	174 (45%)
Injection drug use, last 30 days
Heroin	275 (71%)
Methamphetamine	203 (53%)
Goofball (admixture of heroin and methamphetamine)	175 (45%)
Speedball (admixture of heroin and cocaine)	121 (31%)
Non-prescribed Opioid medication	30 (8%)
Crack cocaine	33 (9%)
Powder cocaine	33 (9%)
Non-injection drug use in the last 30 days
Cannabis	387 (100%)
Methamphetamine	196 (51%)
Crack cocaine	145 (38%)
Non-prescribed Opioid medication	74 (19%)
Heroin	97 (25%)
Non-prescribed Tranquilizer medication	93 (24%)
Powder cocaine	56 (15%)
Non-prescribed Methadone	34 (9%)
Non-prescribed Buprenorphine	26 (7%)
Non-prescribed Stimulant medication	22 (6%)
Cannabis use frequency, last 30 days
Less than once a day (<30 times)	231 (60%)
Once or twice a day (30 to 89 times)	92 (24%)
Three or more times a day (90 or more)	64 (16%)
Injection frequency, last 30 days
None	30 (8%)
Less than once a day (<30 times)	103 (26%)
Once or twice a day (30-89 injections)	99 (26%)
Three times or more a day (90 or more injections)	155 (40%)
Opioid use frequency, last 30 days
None	67 (18%)
Less than once a day (<30 times)	90 (23%)
Once or twice a day (30-89 times)	86 (22%)
Three times or more a day (90 or more times)	144 (37%)
Mental health conditions
Any diagnosis	235 (52%)
Depression	178 (46%)
Anxiety	138 (36%)
Bipolar disorder	106 (27%)
Post-traumatic stress disorder (PTSD)	104 (27%)
Schizophrenia	51 (13%)
Drug treatment in the last 6 months
Methadone maintenance	81 (21%)
Outpatient	34 (9%)
Buprenorphine	18 (5%)
Detoxification	15 (4%)
Residential	15 (4%)
Inpatient	13 (3%)
Cannabis use intentions
Half or more times to get high (Don’t know=1)	249 (65%)
Less than half	137 (35%)
Half or more times to relief physical pain (Don’t know=2)	195 (51%)
Less than half	190 (41%)
Half or more times to relieve emotional problems (Don’t know=2)	162 (42%)
Less than half	223 (58%)
Have a medicinal cannabis card	83 (21%)
Do not have a medicinal cannabis card	304 (79%)
Half or more times to reduce your use of opioids (Don’t know=13)	75 (20%)
Less than half	299 (80%)
Half or more times for at least one health-related motivation	260 (67%)
Less than half	129 (33%)

3.0. Results

3.1. Sample characteristics

Our sample of 387 total PWID, who reported past-month cannabis use, was socio-demographically diverse with the majority of participants being younger than 50 years of age, low-income, and unstably housed (U.S. Department of Housing and Urban Development 2020; Los Angeles Almanac 2018)(see Table 1). Mean age was 42 years (SD=11.80; Median=44; IQR=32, 51). Average number of drug injections in the last 30 days was 83.6 (SD=127.2; Median=60; IQR=15, 115), and mean frequency of opioid use in the last month was 70.8 (SD=90.4; Median=50; IQR=5, 103).

Among opioid-using PWID, cannabis use of less than once a day in the last 30 days was reported by 59% of the sample (231/387). Looking at cannabis use intentions, 65% of the sample (249/387) reported using cannabis to get high 50% or more of the time, 51% to relieve physical pain, 42% to relieve emotional problems, and 20% to reduce their opioid use. Possession of a medical cannabis card, which is an identification card issued in the state of California that allows a person to obtain medical cannabis from dispensaries, was reported by 21% of participants.

3.2. Factors associated with mostly using cannabis for physical pain relief

In a multivariable analysis model (Table 2), we found that using cannabis for physical pain relief was associated with higher odds of using cannabis as a substitute for opioids (AOR=4.46, 95% 95% CI=2.30, 8.64), while controlling for confounders (e.g., demographics [sex, race], socioeconomic status [income, homelessness], mental health, and drug use patterns). We also found that using cannabis for physical pain relief was significantly associated with increased odds of non-prescribed, non-injection prescription opioid us. (AOR=2.18, 95% CI=1.18, 4.03).

Table 2:

Factors associated with mostly using cannabis for physical pain relief among opioid-using people who inject drugs in Los Angeles and San Francisco, CA 2016-2018 (N = 386)

Variables	Unadjusted odds ratio (95% confidence interval)	P value	Adjusted Odds Ratio (95% Confidence interval)	P value
Opioid use frequency, last 30 days
None	Referent		Referent
Less than once a day	1.08 (0.57, 2.05)	0.82	0.84 (0.40, 1.79)	0.65
Once or twice a day	0.65 (0.34, 1.25)	0.20	0.39 (0.18, 0.85)	0.02
Three times or more a day	0.54 (0.30, 0.97)	0.04	0.44 (0.22, 0.90)	0.02
Cannabis use frequency, last 30 days
Less than once a day	Referent		Referent
Once or twice a day	1.93 (1.18, 3.16)	0.009	1.96 (1.12, 3.43)	0.02
Three or more times a day	2.26 (1.27, 4.02)	0.005	2.16 (1.10, 4.23)	0.03
Cannabis use intention	Referent		Referent
To get high <50%	0.43 (0.28, 0.66)	<0.001	0.37 (0.23, 0.61)	<0.001
To get high 50 or +	Referent		Referent
For emotional problems <50%	3.08 (2.01, 4.70)	<0.001	2.28 (1.41, 3.69)	<0.001
For emotional problems 50% or +	Referent		Referent
As opioid substitute <50% As opioid substitute 50% or +	5.18 (2.81, 9.55)	<0.001	4.46 (2.30, 8.64)	<0.001
Non-prescribed, non-injection prescription opioid use
No	Referent		Referent
Yes	1.95 (1.15, 3.31)	0.01	2.18 (1.18, 4.03)	0.01

3.3. Factors associated with mostly using cannabis for emotional problems

Results from a multivariable analysis model show that mostly using cannabis for emotional problems was associated with increased odds of being diagnosed with a depressive disorder (Table 3)(AOR=2.01, 95% CI=1.28, 3.15), while controlling for panhandling (i.e., begging or soliciting people for donations) as an income source. We controlled for socioeconomic variables, including income and income source (e.g., panhandling) because economic resources may relate to cannabis access and use patterns. We also found that using cannabis for emotional problems was significantly associated with increased odds of using cannabis as a substitute for opioids (AOR=2.58, 95% CI=1.45, 4.58) and for physical pain relief (AOR=2.30, 95% CI= 1.45, 3.67).

Table 3:

Factors associated with mostly using cannabis for emotional problems among opioid-using people who inject drugs in Los Angeles and San Francisco, CA 2016-2018 (N = 385)

Variables	Unadjusted odds ratio (95% confidence interval)	P value	Adjusted Odds Ratio (95% Confidence interval)	P value
Panhandling income source
No	Referent		Referent
Yes	1.77 (1.12, 2.81)	0.02	1.80 (1.09, 2.97)	0.02
Cannabis use frequency, last 30 days
Less than once a day	Referent		Referent
Once or twice a day	1.68 (1.03, 2.75)	0.04	1.72 (1.00, 2.95)	0.05
Three or more times a day	1.99 (1.14, 3.49)	0.02	1.77 (0.96, 3.26)	0.07
Diagnosed with Depression
No	Referent
Yes	2.00 (1.33, 3.02)	0.001	2.01 (1.28, 3.15)	0.002
Cannabis use intention for
Physical pain relief <50%	Referent		Referent
Physical pain relief 50% or +	3.08 (2.01, 4.70)	<0.001	2.30 (1.45, 3.67)	<0.001
Opioid substitute <50%	Referent		Referent
Opioid substitute 50% or +	3.45 (2.02, 5.90)	<0.001	2.58 (1.45, 4.58)	0.001

3.4. Factors associated with mostly using cannabis as a substitute for opioids

Results from a multivariable analysis model show that mostly using cannabis as a substitute for opioids was associated with significantly increased odds of non-prescribed, non-injection methadone use (AOR=3.60, 95% CI=1.53, 5.06), while controlling for sexual orientation and veterans benefits income source, which provides income to individuals who have served in the U.S. military, navy, or air service (Table 4). We also found that using cannabis as a substitute for opioids was associated with higher odds of non-prescribed, non-injection stimulant prescription medication use (AOR=3.00, 95% CI=1.03, 8.77).

Table 4:

Factors associated with mostly using cannabis as a substitute for opioids (N = 374)

Variables	Unadjusted odds ratio (95% confidence interval)	P value	Adjusted Odds Ratio (95% Confidence interval)	P value
Gay, lesbian, bisexual
No	Referent		Referent
Yes	0.24 (0.09, 0.61)	0.003	0.21 (0.08, 0.57)	0.002
Veterans benefits income source
No	Referent		Referent
Yes	4.20 (1.18, 14.91)	0.03	5.98 (1.37, 26.02)	0.02
Non-prescribed, non-injection Stimulant prescription medication use
No	Referent		Referent
Yes	3.26 (1.32, 8.06)	0.01	3.00 (1.03, 8.77)	0.04
Non-prescribed, non-injection Methadone use
No	Referent		Referent
Yes	2.93 (1.38, 6.20)	0.005	3.60 (1.53, 5.06)	0.003
Cannabis use intention for
Physical pain relief <50%	Referent		Referent
Physical pain relief 50% or more	5.18 (2.81, 9.55)	<0.001	4.46 (2.29, 8.69)	<0.001
Emotional problems <50%	Referent		Referent
Emotional problems 50% or more	3.45 (2.07, 5.90)	<0.001	2.75 (1.50, 5.06)	0.001

4.0. Discussion

Motivations for using cannabis have been studied extensively among patients of medicinal cannabis dispensaries, but less is known about cannabis use motivations in community recruited samples, particularly for those who inject drugs (Valleriani et al., 2020). To the best of our knowledge, this study is the first quantitative assessment of motivations for cannabis use among PWID. We found that health-related motivations for cannabis use ranged from 51% for pain relief to 20% to reduce opioid use, and that 67% reported mostly using cannabis for at least one health-related motivation. In this population, cannabis use for these health-related reasons could be important, particularly if cannabis use is effective at reducing pain, emotional problems, or reducing opioid use for a population that often experiences social hardships that contribute to poor health. While opioids are among the most effective pain relievers, prescription access to opioids (including highly potent fentanyl-containing medicines and its analogues) for PWID is not recommended, as it could lead to greater dependence or overdose (Fairbairn et al., 2017; Green and Gilbert, 2016; Volkow et al., 2019). Similarly, common pharmacological treatments for emotional problems like anxiety and depression are benzodiazepines, a known risk factor for overdose among people with opioid use disorder (Cho et al., 2020; Kerr et al., 2007). Lastly, in the context of the opioid overdose death crisis, use of substances with lower risk profiles has obvious benefits. Research to determine then the effectiveness of cannabis use for these common needs among PWID is warranted.

4.1. Use of cannabis to self-manage physical pain and reduce non-injection opioid use

Chronic pain is a growing public health concern, particularly among PWID (Heimer et al., 2015) who often engage in self-management of physical pain through nonmedical prescription opioid use (Dahlman et al., 2017). Using cannabis to reduce physical pain was common among our study participants and was positively associated with more frequent cannabis use in this sample. The use of cannabis to address PWID pain control needs may serve as a low-risk and economical strategy to prevent higher risk methods of self-treatment, such as buying unprescribed pain medication from unknown sources or using heroin to alleviate pain (Voon et al., 2015).

While previous studies often position cannabis use as a high-risk behavior that increases the risk of using illicit substances (i.e., the “gateway drug” theory, see National Institute of Drug Abuse, 2020; Institute of Medicine, 2001), our research shows that cannabis use among PWID seeking relief from physical pain was inversely related to non-injection opioid use. Our findings align closely with those of recent studies that indicate cannabis as having potential harm reduction benefits and functioning as a “gateway” out of high-risk substance use for populations like PWID (Reddon et al., 2018). The potential implications of cannabis use to manage physical pain for this population is important because clinicians are often reluctant to prescribe PWID opioid-based analgesia because of their history of substance use (Voon et al., 2015), resulting in untreated or undertreated pain that may promote higher-risk substance use to self-treat pain (Dahlman et al., 2017). In addition to PWlD’s lack of access to adequate pain management through the healthcare system, our sample of PWID experienced a lack of stable employment (only 15% had a job), income (78% earned less than $1400/month), and housing (75% currently homeless), which often contributes to poor health and may increase the risk of physical pain (Dahlman et al., 2017).

We also found that PWID who reported using cannabis for physical pain relief were four times as likely to use cannabis as a substitute for non-injection opioids. Given previous research on the potential of cannabis to manage pain (Bostwick, 2014; Degenhardt et al., 2015; Kral et al., 2015; Lucas, 2012), our findings suggest that PWID experiencing pain may be using cannabis as a self-medicated strategy to reduce the frequency of non-injection opioid use in order to achieve pain relief. These findings are encouraging and may suggest that PWID who use cannabis to relieve pain are motivated to do so for the “opioid-sparing effect” of cannabis, where a smaller dose of opioids provides equivalent pain relief to a larger dose of opioids when it is paired with cannabis (Cooper et al., 2018; Nielsen et al., 2017). Our findings on cannabis use motivations among PWID build on current studies indicating that cannabis may be a strategy to alleviate pain and/or reduce opioid use (Lake et al., 2019). We also build upon our previous research with PWID in California to show that cannabis use correlates with a lower frequency of illicit opioid use (Kral et al., 2015). Notably, our findings suggest that PWID may be intentionally using cannabis as a pain management tool in order to reduce the frequency of opioid injection because of its perceived effectiveness in managing both pain and opioid use.

4.2. Use of cannabis to manage emotional health and non-injection opioid use

Nearly 1 in 3 PWID meet a clinical diagnosis of a depressive disorder (Colledge et al., 2020). PWID are often exposed to medical, social, and legal harms (Milloy et al., 2009), as well as traumatic events (Darke and Torok, 2013; Havens et al., 2009) that may increase the risk of mental disorders among this population (Colledge et al., 2020). Our findings show that PWID who reported using cannabis for mostly emotional issues were using cannabis frequently (3 or more times a day) and were twice as likely to self-report having been diagnosed with depression. Our findings align with previous studies which show that people who use substances and cannabis are likely to self-report mental health as a reason for frequent cannabis use relative to occasional cannabis use (Lake et al., 2019).

Given the high rates of attempted suicide, PTSD, depression, and child abuse and historical emotional abuse among PWID (Colledge et al., 2020; Darke and Torok, 2013; Lake et al., 2015; Plotzker BA et al., 2007), our findings suggest that PWID may be intentionally using cannabis to manage their depression. Existing research has shown that depression is consistently associated with a higher prevalence of overdose among PWID (Lemstra et al., 2011), with major depression being a significant predictor of continued heroin use (Teesson et al., 2015). Research shows that cannabis use may provide temporary relief from symptoms of depression and that people may be using cannabis to cope with higher levels of depression (Glodosky and Cuttler, 2020). While some research has shown that cannabis may negatively contribute to the development of depressive disorders (Artenie et al., 2015; Lev-Ran et al., 2014) in adolescents (Gobbi et al., 2019; Rhew et al., 2017) and veterans (Metrik et al., 2018; Wilkinson et al., 2015), there is little research about the potential benefits of cannabis in PWID populations. For PWID that often experience a high burden of emotional distress with significant unmet need of accessible mental health services and/or treatment, our research suggests that cannabis for this population may be intentionally used to self-treat depression. Our findings suggesting the use of cannabis for self-treating depression align with previous research on PWID that found that their use of cannabis was negatively associated with accessing treatment for depression (Genberg et al., 2019).

We also found that for PWID who used cannabis for emotional issues were also twice as likely to use cannabis for physical pain relief and as a substitute for opioids. Some research shows the co-use of cannabis and opioids to be associated with elevated anxiety and depression symptoms among adults with chronic pain (Rogers et al., 2019). Yet for PWID, cannabis may function as a tool to regulate depression as well as non-injection opioid use. These findings contribute to a harm reduction strategy (i.e., discouraging abstinence-only conditions) of supporting a shift toward a less harmful substance, such as cannabis for PWID seeking to reduce their opioid use.

4.3. Use of cannabis and non-prescribed methadone to reduce opioid use and pain

Non-prescribed use of opioid substitution medication, such as methadone, among PWID is often associated with opioid dependence and frequency of injection drug use (Genberg et al., 2015). Yet research has shown that use of cannabis during methadone maintenance treatment has no impact on a person’s treatment outcomes (e.g., dose titration, induction time, attendance, or early discharge)(Scavone et al., 2013). In our current study, we found that PWID who mostly used cannabis as a substitute for opioids were four times as likely to use cannabis for physical pain, and three times as likely to report non-prescribed, non-injection methadone use. Our findings build upon previous research to suggest that PWID may be intentionally using cannabis, along with non-medical methadone, to self-treat and reduce their opioid use in the context of managing pain.

While non-prescription methadone may indicate a relevant source of opioids for PWID and thus raise concerns about potentially fatal overdoses and related high-risk behaviors (Reddon et al., 2018), methadone is often not used as a main substance of choice. Furthermore, non-prescription methadone is found to be more prevalent among people who use opioids and are not engaged in opioid treatment (Schulte et al., 2016), suggesting the use of non-prescription methadone as a strategy to reduce opioid consumption and self-treat opioid withdrawal symptoms, such as withdrawal pain (Tucker et al., 2015). Our previous research in California (Zhao et al., 2020) showed that PWID who used non-medical methadone were more likely to have participated in methadone treatment, suggesting that PWID may be self-managing their own opioid agonist therapy (Johnson and Richert, 2019) in order to prevent withdrawal symptoms while also avoiding illicit opioids in response to inadequate methadone dosing from clinics. Together, our previous work and current findings suggest that PWID are engaging in a harm reduction strategy where they are intentionally using cannabis and non-prescribed methadone to reduce their opioid use and manage pain that may be related to withdrawal.

4.4. Limitations

Interpretations of these results should be considered in light of a number of potential limitations. We have used a cross-sectional data analysis approach, therefore causal inferences cannot be made. Further, all data are self-report. For example, self-report of diagnoses for depression in this analysis may be affected because PWID in our sample experienced a lack of stable employment, income, and housing that would make it difficult for them to access services where they could receive formal mental health diagnosis for a depressive disorder. However, a recent systematic review of the prevalence of mental health indicators among PWID show that depression, PTSD, suicidality, and self-harm are high among PWID (Colledge et al., 2020). Furthermore, reliability and validity of self-reported responses to drug use and risk variables have generally been found to be good in studies of PWID (Dowling-Guyer et al., 1994; Weatherby et al., 1994). While our findings suggest that PWID may be intentionally using cannabis for health-related motivations as a result of having limited access to health services to treat their health issues, we did not examine PWID access to health services and unmet health needs. Future research on cannabis use motivations among PWID would benefit from further study on how social and economic hardships, such as access to mental health services and unmet needs around chronic opioid use, impact self-treatment of health-related issues.

Changes in cannabis use motivation are likely as recreational legalization of this product in California has resulted in product differentiation and increased potency and advertising. How these developments influence cannabis motivation and use patterns over time is an important research question to address that we are addressing in an ongoing study.

5.0. Conclusions

We found that health-related motivations for cannabis use were common among PWID. The acceptability of cannabis for addressing health-related problems suggests that cannabis might be a reasonable “first step” for helping some PWID address physical pain and emotional health. Nonetheless, much more research is needed to determine if consistent cannabis use is associated with reductions in opioids, and improvement in physical pain and emotional health. This research will occur in the context of changing cannabis product properties. The increasing potency of cannabis appears to be a trend and its impacts on motivations and drug use patterns are just being studied (Chandra et al., 2019). Further, carefully controlled clinical studies are needed to evaluate the effectiveness of various forms and potencies of cannabis on physical pain and emotional distress. Based on our findings, we recommend for future research to focus on chronic opioid users and those who inject drugs. Should further research find cannabis to reduce opioid use and injection, the potential benefits of cannabis for PWID are substantial and research focused on reducing harms in this population could have enormous benefits to individuals, families, and communities.

Highlights.

• People who inject drugs reported cannabis use for common health-related motivations

• Cannabis motivations included reducing pain, emotional issues, or opioid use

• Motivations suggest the acceptability of cannabis use for health reasons

• Cannabis use may have implications for pain management among this population