Association of Cannabis Use with Intimate Partner Violence among Couples with Substance Misuse

Abstract

Background and Objectives:

There is a well-established causal link between substance use and intimate partner violence (IPV) perpetration and victimization. However, little is known about the complex emerging relationship between cannabis use and IPV. Because cannabis is the most commonly used drug in the United States and is associated with numerous IPV risk factors such as alcohol use, it is important to examine this relationship in greater detail.

Method:

The current exploratory study examined the association between 1) self-report cannabis use during the past 90 days and 2) Δ9-Tetrahydrocannabinol (THC) urine drug screens and IPV perpetration and victimization in a sample of 30 alcohol or drug-misusing community couples (N=60 individual participants).

Results:

The majority of participants (n=50 individuals, 83.3%) had concordant cannabis self-report and urine drug screen results. After accounting for demographic variables and quantity and frequency of alcohol and stimulant use, greater quantity and frequency of cannabis use as well as positive THC urine drug screen results were associated with greater physical IPV victimization, and greater quantity and frequency of cannabis was associated with greater IPV psychological victimization and perpetration, and physical IPV victimization.

Conclusions and Scientific Significance:

Findings emphasize the unique and important role that cannabis plays in the occurrence of IPV among intact couples. Findings also underscore the feasibility and utility of integrating confirmatory biological samples into future studies on this topic in order to advance the science in this area.

Introduction

Cannabis use and intimate partner violence (IPV) are both highly prevalent and often co-occur. IPV is a pervasive public health problem, with approximately one third of women and of men reporting IPV in their lifetime¹. Cannabis use is one of the most commonly used drugs in the United States², and particularly prevalent in IPV populations. Approximately one third to one half of individuals in court-mandated batterer intervention programs reporting past year cannabis use, and cannabis use reported among up to 70% of individuals experiencing IPV victimization^3,4.

Emerging evidence suggests that cannabis use specifically may increase risk for IPV perpetration and victimization^4–6. However, the limited amount of data available to date suggests that this association is complex and more research is needed to elucidate the association between cannabis use and IPV. The importance of examining the link between cannabis use and IPV in greater detail is further underscored by the continuously changing landscape surrounding cannabis use and related cultural norms. For example, numerous policy changes including decriminalization of cannabis possession in over 20 states to date and legalization of medical cannabis in 35 states have occurred during the past several years. Relatedly, factors such as potency, novel formulations and delivery methods, and cannabis use attitudes and expectancies are changing^7,8. Thus, the goal of the current study is to examine the relation between both self-report cannabis use and THC-positive urine drug screens and IPV perpetration and victimization in a sample of substance-misusing couples.

Cannabis Use and IPV

The limited data linking cannabis use with IPV is complex and there is little theoretical or empirical research on the mechanisms underlying this association. Although widespread belief regarding cannabis use suggests that it may produce anxiolytic and positive mood and behavioral effects, the affective and behavioral effects of cannabis varies based on pharmacological, individual, and contextual differences^9,10. Among some individuals, cannabis use may be associated with increased alcohol consumption, aggressive behavior, physiological arousal, anxiety, and cognitive impairment, all of which may increase the likelihood of IPV occurring within a relationship. Furthermore, individuals who are prone to IPV risk factors such as attentional bias to threat sensitivity, and those with other substance use and psychiatric disorders incur the greatest risk for cannabis use and use disorder¹¹.

Emerging research suggests for some individuals, cannabis use is associated with both perpetration and victimization of IPV perpetration^5,12. Specifically, the most comprehensive reviews conducted to date suggest that there may be an association between cannabis use and IPV perpetration^5,12 but one of these reviews found that there was no significant association between cannabis use and sexual IPV¹². Conversely, recent work found that cannabis use was associated with sexual IPV perpetration³. Few studies have examined temporal associations between cannabis use and IPV. Using retrospective interview data and ecological momentary assessment, three prior studies among individuals receiving substance abuse treatment, university students, and women in court-mandated batterer intervention programs showed that cannabis use was not temporally associated with IPV^13–15. One study conducted among cannabis-using community couples found that cannabis use was associated with verbal but not physical IPV two hours after use¹⁶. Conversely, another study conducted among university students found that physical or sexual IPV victimization predicted next-day cannabis use¹⁷. Collectively, the existing research suggests that there is a complex bidirectional association between cannabis use and IPV; however, inconsistent findings indicate that more work is needed to understand this complex relationship.

Current Study

It is critical to employ state of the science methods to add to the scant existing data and disentangle the complex association between cannabis use and IPV. The current study makes two contributions to the existing literature. First, this study is the first to our knowledge to examine the association between both self-reported cannabis use and urine drug screens positive for the THC metabolite THC-COOH and IPV perpetration and victimization among a community sample of substance misusing couples. This specific methodological advancement is an important addition to the existing substance use-IPV literature for several reasons. First, reliability of self-reported alcohol and drug use remains a persistent challenge in the substance use field¹⁸. Second, agreement between self-report and biological measures of cannabis use is also a topic of continued investigation in the cannabis literature given the lack of standard units of use for measurement in research and treatment settings¹⁹. Indeed, emerging literature suggests that just as estimation of standard drink units has become essential to understanding alcohol use behavior and pharmacokinetics, related risk behaviors, and neurobiological markers, estimating cannabis in grams is essential to clearly understand individuals’ cannabis use patterns¹⁹. One recent study demonstrated that measuring cannabis in grams was a superior predictor of THC biomarkers compared to assessments of frequency or number of joints per day²⁰.

Additionally, previous studies linking cannabis use with IPV have utilized either single-item assessment of frequency of cannabis use^3,21, a yes/no single item assessing daily cannabis use²², or the daily number of joints and/or number of cannabis use episodes as the primary outcome variable¹⁶. The present study is the first in the IPV literature to examine cannabis use quantity in grams, which might reflect superior validity compared to alternative measurement approaches^20,23. Further, the current study examines both quantity and frequency to allow for a more comprehensive understanding of cannabis use than quantity alone. The current study examines the association between cannabis use and IPV among a community sample of substance misusing couples. We hypothesized that both greater self-report cannabis use and positive urine drug screen would be associated with greater IPV victimization and perpetration.

Method

Participants

Participants were 33 couples (66 individual participants). Two same-sex female couples (n=4) enrolled in the study. The same sex couples are not included in the current analyses due to the lack of adequate power necessary to test effects of sex constellations within couples. Another couple was excluded due to questionable reliability of data, thus the final sample was comprised of 30 opposite-sex couples (N = 60). Participants were enrolled in a larger randomized controlled laboratory study examining the effects of oxytocin on cortisol reactivity and conflict behaviors among couples. The procedures have been reported previously²⁴. This study was IRB approved and all participants completed written informed consent prior to taking part in any study procedures. Participants responded to advertisements of a study examining an investigational drug on couples conflict resolution behaviors. Advertisements were placed in the community, local treatment clinics, and on the internet. Participants were required to be 18-65 years of age, and one or both partners in each dyad must have engaged in at least one episode of hazardous drinking (i.e., 4 or more standard drinks for women, 6 or more for men on one occasion^25,26) or illicit drug use during the past 60 days. Couples were not required to be married or cohabitating. Exclusion criteria for the larger study included 1) pregnancy or breastfeeding; 2) history of or current physical or psychiatric diagnosis known to impact HPA axis function; 3) a BMI ≥ 39; 4) use of prescription medications that interfere with HPA axis activity; 4) active suicidal or homicidal ideation and intent. Participants who had reported severe and unilateral IPV in the past year (assessed by the Revised Conflict Tactics Scale [CTS-2]) were also excluded in order to maximize participant safety and sample generalizability. Both partners’ CTS-2 reports were examined to establish eligibility.

Demographic sample characteristics are presented in Table 1. The sample was primarily comprised of opposite sex couples, with two same-sex female couples. The same sex couples are not included in the current analyses due to the lack of adequate power necessary to test effects of sex constellations within couples. On average, participants were 32.18 years old (SD = 9.93 years) and were cohabitating with their current partner (80%; n = 48 individuals). Thirteen couples were discordant on self-report of cannabis use, meaning that one partner reported use while the other reported zero cannabis use days. Over half (n=32 individuals) of participants tested positive for THC in this sample. Six participants who reported zero cannabis use on the Time Line Follow Back (TLFB) tested positive for THC in urine drug screens. Four participants who reported using cannabis on the TLFB (three participants reported two days of use; one participant reported three days of use) tested negative for THC urine drug screens. Positive results for substances other than cannabis in urine drug screens were low in this sample (amphetamines: n=3; methamphetamine: n=2; benzodiazepines: n=2; cocaine: n=8; opiates: n=1).

Table 1.

Demographic and Cannabis Use Characteristics

	N	%	M	SD
Total Sample	60
Age			32.18	9.93
Gender
Female	30	50%
Male	30	50%
Racial Identity
African American	32	53.3%
Asian/Pacific Islander	1	1.7%
Native American/Alaskan Native	3	5%
More than one race	1	1.7%
White	23	38.3%
Marital Status
Single (never married)	8	13.3%
Married	3	5%
Cohabitating	48	80%
Relationship Length (in months)			52.19	60.26
Household Income			29,904	20,621
Cannabis Use Characteristics
Days of Cannabis Use in Past 90 days			17.97	18.69
Positive THC UDS	32	53.3%
Total Grams Used in Past 90 Days			33.49	84.60

Note. TLFB=Time Line Follow Back. UDS=Urine Drug Screen.

Measures

Intimate partner violence.

IPV victimization and perpetration were measured using the 78-item Revised Conflict Tactics Scales (CTS-2)²⁷. Participants were asked to indicate the frequency with which various behaviors had occurred. Response categories that comprised a range of values were recoded [i.e., 3-5 times (recoded to 4), 6-10 times (recoded to 8), 11-20 times (recoded to 15), and more than 20 times (recoded to 25)]. Consistent with published scoring procedures, total scores for each participant were calculated for victimization and perpetration by summing the respective responses for the psychological and physical IPV subscales.

Substance Use.

The Time Line Followback (TLFB)²⁸ is a calendar-assisted, semi-structured interview which was used to assess quantity and frequency of alcohol and illicit and prescription drugs (e.g., cannabis, prescription opioids, benzodiazepines, and psychostimulants). Participants report the total number of days substances are used and the amount of substance used (e.g., standard drink units for alcohol, grams of marijuana). Grams of cannabis use were estimated consistent with the .33g/joint value described by Ridgeway & Kilmer²⁹ during the 90 days prior to study participation. In this study, we examined the total quantity of each substance used during the assessment period divided by the total number of days each substance was used.

Given that the quantity (i.e., total grams of cannabis) and frequency (i.e., total number of days of cannabis use) were highly and significantly correlated (r = .69) we examined the total number of grams used/total number of days used, providing a measure of average number of grams of cannabis used on days of cannabis use during the past 90 days. THC urine drug screen results were examined dichotomously (i.e., positive=1; negative=0).

Urine Drug Screens.

All participants completed a six-panel urine drug screen testing for the primary metabolite of Δ9-Tetrahydrocannabinol (THC), amphetamines, methamphetamines, benzodiazepines, cocaine, and opiates. The threshold for THC-positive samples in urine drug screens is consistent with the standard 50 ng/ml. UDS were conducted on the same day as all other study procedures.

Statistical Analysis

Demographic characteristics of the sample were examined using descriptive statistics. Correlation analyses were conducted to examine univariate associations between variables (Table 2). In order to examine the association between cannabis use and IPV victimization and perpetration in this sample while accounting for the interdependence of the data, we estimated parallel multilevel models in which individuals at Level 1 were nested under couples at Level 2 ³⁰. With highly skewed and overdispersed physical IPV victimization (skewness = 2.53, kurtosis = 5.32) and perpetration (skewness = 4.05, kurtosis = 17.56) outcome data, models were estimated using a negative binomial distribution with a logit link function. Psychological victimization and perpetration variables fell within acceptable ranges of skewness (psychological perpetration skewness=.85; psychological victimization skewness=.60) and kurtosis (psychological perpetration kurtosis =.29; psychological victimization kurtosis =−.61) and were estimated using ordinary least squares (OLS) regression. All models included age, gender, race, and quantity and frequency of alcohol and stimulant use as covariates. Only alcohol and stimulant use were controlled because only one participant reported one instance of opioid use in the sample. No other drug use was self-reported. Quantity/frequency of cannabis use was the independent variable in one set of analyses while THC urine drug screen was the independent variable of interest on a second set of analyses. Results are reported as incidence response ratios (IRR; i.e., exponentiated coefficients) for negative binomial models and may be interpreted as the factor change in intimate partner violence for each unit increase in the predictor variable.

Table 2.

Descriptive Statistics and Intercorrelations

	M	SD	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.
1. Gender
2. Racial Identity			0.16	0.14	−.54*	0.06	−0.04	0.35	0.15	0.14	0.18	0.27	0.07
3. Age	32.18	9.93	0.3	0.76**	−0.09	0.1	0.14	−0.04	0.21	−0.02	0.07	0.05	−0.09
4. Income	$24452	$13929	0.22	−0.05	0.12	0.02	0.07	−0.06	0.28	0.04	−0.03	−0.06	0.11
5. Alcohol Use	3.31	2.81	0.21	0.06	0.2	−0.001	.42*	−0.03	0.07	−0.22	−0.17	−0.13	−0.11
6. Stimulant Use	6.59	15.72	0.28	.49**	−0.02	0.08	−0.001	0.33	−0.12	0.13	0.21	−0.04	0.04
7. Cannabis Use	.46	.75	0.03	0.03	0.01	−0.12	−0.01	−0.001	.42*	.46*	0.34	.41*	.42*
8. Positive THC UDS	32	53.3%	−0.24	0.01	0.06	−0.11	−0.08	−0.02	−0.03	0.16	0.17	0.12	0.25
9. Psych IPV Perpetration	41.56	31.57	0.11	−0.13	−0.18	−0.14	−0.03	0.2	−0.08	---	0.16	0.17	0.12
10. Psych IPV Victimization	43.51	31.74	0.14	−0.1	−0.1	−0.12	−0.07	0.25	−0.02	.95**	---	.47*	.42*
11. Physical IPV Perpetration	6.85	15.24	0.23	−0.11	0.17	0.2	−0.06	0.02	−0.07	.50**	.46*	---	.52**
12. Physical IPV Victimization	1.83	4.81	0.03	−0.29	0.21	0.16	−0.04	0.04	−0.07	.42*	.42*	.52**	---

Note. Correlations for men are presented above the diagonal, women are presented below the diagonal, and the correlation between partners is bolded and presented on the diagonal. Correlations between partners is not used for IPV because max scores between partners were used.

^*p < .05,

^**p < .001.

M=Mean. SD=Standard deviation. THC= Δ9-Tetrahydrocannabinol. UDS=Urine drug screen. Psych= Psychological. Alcohol, stimulant, and cannabis use reflect average daily use on days using from the Time Line Follow Back (i.e., total quantity of use divided by number of days using).

^†The number and proportion of THC-positive urine drug screens are presented in place of Mean and SD.

Results

Findings from multilevel models examining cannabis use from the TLFB are presented in Table 3. Greater self-reported cannabis use was significantly associated with greater psychological IPV victimization (β=15.24, p=.001) and perpetration (β=12.64, p=.007), as well as physical IPV victimization (IRR = 6.01, β=1.79, p=.017), but not physical IPV perpetration. Alcohol use was also positively associated with physical IPV victimization (β=1.10, p=.032). Findings from multilevel models examining cannabis use measured by urine drug screens are presented in Table 4. Consistent with self-report cannabis use, positive THC urine drug screens were associated with greater physical IPV victimization (IRR = 8.22, β=2.11, p<.001). Alcohol use was negatively associated with psychological IPV perpetration (β=−2.27, p=.023). No other associations were detected.

Table 3.

Multilevel models examining associations between self-report cannabis use and IPV

	IRR	B	Std. Error	p
	Psychological Victimization
Gender		2.47	3.44	0.472
Race		10.78	10.07	0.284
Age		−0.06	0.45	0.886
Alcohol Use		−1.23	1.17	0.293
Stimulant Use		−0.22	0.36	0.546
Cannabis Use		15.24	4.66	0.001
	Psychological Perpetration
Gender		1.91	2.91	0.512
Race		8.82	10.65	0.408
Age		−0.27	0.35	0.434
Alcohol Use		−1.71	1.01	0.092
Stimulant Use		−0.11	0.33	0.743
Cannabis Use		12.64	4.73	0.007
	Physical Victimization
Gender	1.69	0.53	0.54	0.327
Race	0.65	−0.43	0.86	0.616
Age	0.96	−0.04	0.03	0.262
Alcohol Use	1.10	0.10	0.05	0.032
Stimulant Use	0.93	−0.07	0.04	0.066
Cannabis Use	6.01	1.79	0.75	0.017
	Physical Perpetration
Gender	1.61	0.48	0.60	0.427
Race	6.88	1.93	0.93	0.037
Age	1.00	0.00	0.04	0.960
Alcohol Use	1.07	0.07	0.06	0.241
Stimulant Use	0.95	−0.06	0.04	0.136
Cannabis Use	2.70	1.00	0.78	0.201

Note: Psychological IPV was estimated using OLS regression, physical IPV was estimated using negative binomial regression. Significant effects are bolded. IRR = Incidence rate ratio. Gender (female =1, male =2). Racial identity (1=White, 2=Non-White).

Table 4.

Multilevel models examining associations between THC urine drug screen and IPV

	IRR	β	Std. Error	p
	Psychological Victimization
Gender		6.11	3.29	0.063
Race		16.02	11.00	0.145
Age		−0.13	0.54	0.808
Alcohol Use		−1.88	1.22	0.125
Stimulant Use		0.02	0.43	0.967
THC UDS		16.00	9.71	0.100
	Psychological Perpetration
Gender		5.42	3.43	0.114
Race		12.67	11.54	0.272
Age		−0.32	0.40	0.427
Alcohol Use		−2.27	1.00	0.023
Stimulant Use		0.08	0.34	0.808
THC UDS		13.56	10.05	0.177
	Physical Victimization
Gender	1.19	0.17	0.29	0.559
Race	1.58	0.46	0.69	0.505
Age	0.95	−0.05	0.02	0.022
Alcohol Use	1.07	0.07	0.06	0.255
Stimulant Use	1.00	0.00	0.02	0.940
THC UDS	8.22	2.11	0.56	>.001
	Physical Perpetration
Gender	1.20	0.19	0.38	0.622
Race	10.83	2.38	0.82	0.004
Age	0.99	−0.02	0.03	0.608
Alcohol Use	1.05	0.05	0.07	0.479
Stimulant Use	0.98	−0.02	0.02	0.319
THC UDS	2.54	0.93	0.75	0.213

Note: Psychological IPV was estimated using OLS regression, physical IPV was estimated using negative binomial regression. Significant effects are bolded. IRR = Incidence rate ratio. Gender (female =1, male =2). Racial identity (1=White, 2=Non-White).

Discussion

The aim of the present study was to examine the association between cannabis use and IPV victimization and perpetration among community couples who misused substances. Findings demonstrated that greater quantity and frequency of cannabis use was significantly associated with greater physical IPV perpetration and victimization, after controlling for age, gender, race, and quantity and frequency of alcohol and stimulant use. The self-report cannabis use findings were confirmed using corroborating THC urine drug screen outcomes for physical IPV victimization only. These findings extend the limited existing evidence on the relation between cannabis use and IPV in several ways, and warrant replication in larger samples in order to make causal attributions regarding why different associations emerged for self-report versus UDS variables. One possible explanation for these findings is that UDS is one of several existing and emerging biological measurement approaches, and it is possible that in this study, self-report TLFB data captured a more nuanced assessment of quantity/frequency of cannabis use compared a dichotomous UDS measurement.

This is the first study of cannabis-related IPV to corroborate cannabis use through urine drug screens. Overall, findings from the analyses examining self-report of cannabis use and urine drug screen outcomes demonstrated agreement for physical IPV victimization, however there were inconsistent findings for psychological IPV victimization and perpetration. It is possible that the four participants who endorsed cannabis use on the TLFB but had negative urine drug screens might have used cannabis outside the window of detection (3-30 days), and that the reported psychological IPV occurred outside this window. It is also important to note that six participants who had positive urine drug screens declined having used THC in the past 90 days. These data underscore the importance of utilizing a multimethod assessment which might include urine drug screens in IPV research related to substance use, an approach which is rarely incorporated into study designs, in order to maximize the validity of the data in the field. It is also possible that because individuals who use cannabis more frequently or in greater quantities are more likely to have THC-positive UDS outcomes, the more robust associations found between UDS outcomes and IPV in this sample suggests that future studies should examine dose-response associations between cannabis use and IPV. Given that these findings are correlational in nature, it may also be worthwhile to incorporate IPV screening procedures into clinical care if an individual has a THC-positive urine drug screen.

Prior research on this topic has exclusively relied on retrospective self-report and self-report via ecological momentary assessment, which is susceptible to recall-bias and has sometimes been limited to extracting single items from self-report surveys³. Further, self-report regarding use of a drug that is illegal in many states may pose additional concerns for accuracy. While the present study also utilized TLFB to assess cannabis use, previous studies have assessed the number of joints or cannabis use episodes. While some research suggests that examining cannabis use in joints may be sufficient, other recent literature suggests that measurement in grams is a more accurate method^20,23. Because research examining the association between cannabis use and IPV is still in its infancy⁵, utilizing state of the science methods in collaboration with experts in the cannabis field is a critical consideration for future research in this area. For example, given the current lack of standardized assessment of cannabis use in research settings¹⁹, future studies of cannabis use and IPV can improve on current designs by assessing samples from varying biological sources such as hair, saliva, or plasma; including assessments of the potency and strain of cannabis used; and using weighted substitute measurements^20,23. Indeed, confirming associations that emerge using self-report and interview data with additional biological evidence in ongoing and future projects is an important step in this literature.

The association between cannabis use and IPV remains an important yet emerging area of investigation. Individual differences and variability in proximal subjective responses to cannabis use, as well as reasons for cannabis use, are likely to impact the relationship with IPV. For example, among some individuals, acute cannabis administration has resulted in aversive subjective effects³¹. These acute effects of administration also appear to change as use becomes more severe and cannabis use disorders develop and as individuals become more susceptible to cannabis withdrawal³². Prior research suggests that cannabis withdrawal commonly occurs following 1-3 days with varying levels of severity that depend on numerous characteristics of use and individual differences³³. While it is unlikely that participants in this study might have been influenced by withdrawal due to the brief nature of study participation, it is important to note that some individuals experience substantial irritability in response to cannabis use and cannabis withdrawal, which might in part explain the emerging association between cannabis use, use disorders, and IPV found in the current literature. This is a critical variable to assess in future studies examining the association between cannabis use and IPV, regardless of whether self-report, ecological momentary assessment, or laboratory paradigms are employed for assessment. Several variables contribute to these individual differences, which can also be more accurately measured in future studies. For example, one recent study found that age and medical versus recreational cannabis use influenced both subjective effects and cannabis withdrawal³⁴. Notable sex differences in the neurobiological mechanisms of cannabis effects have also been reported³⁵, which are particularly relevant when examining motivations for use of IPV, IPV self-report, and viable mechanisms for prevention and intervention, which might be influenced by gender. Overall, the literature linking cannabis with IPV has not yet been well integrated with the state of the science cannabis literature. Further work should examine the reasons for cannabis use, taking into consideration its use to manage medical and/or psychiatric symptomology, as well as severity of use, and how individual, neurobiological, relational, and/or pharmacological factors predispose individuals for increased risk of violence versus mitigated risk.

An important consideration for future research is that the scant literature in this area remains limited in the scope of populations examined. For example, this study examined IPV in a community sample of couples wherein at least one partner must have reported substance use in the past 90 days. Previous research has focused on cannabis-using community couples¹⁶, men in court mandated treatment³, or college samples²². It is possible that these associations might not generalize to a study designed to assess individual participants or without having nested data within couples. Employing larger and more representative samples, including same sex couples and gender minority populations, in future studies can contribute to the generalizability of findings. Behavioral and neurobiological mechanisms that might explain risk and protective factors for IPV in the context of cannabis use should also be examined.

Limitations and Future Directions

The present investigation utilized a cross-sectional design, thus, temporal or causal conclusions about the variables under consideration cannot be confirmed and should be considered tentative. Expanding the use of longitudinal methodologies such as ecological momentary assessment in combination with laboratory methods would be fruitful in examining the temporal associations between cannabis use and IPV¹⁶. It is particularly crucial to leverage such designs to examine the complex mutually causal association between cannabis use and both IPV perpetration and victimization, and whether compounding effects emerge over time between cannabis use and occurrence of IPV. Further, novel laboratory studies and randomized controlled trials are needed to examine the acute pharmacological effects and behavioral, relational, and stable personality characteristics that might explain the effects of cannabis on IPV. The pharmacologic, behavioral, environmental, and relational aspects of cannabis use (including long-term use, withdrawal, and acute intoxication) in precipitating or possibly inhibiting couple conflict and IPV remains unclear. Laboratory research examining the pharmacological effects of alcohol on IPV using self-administration methods have been employed with valuable benefit to our understanding of the proximal effects of alcohol intoxication on IPV, however this work has yet to be extended to cannabis. Some prior studies examining general aggression rather than IPV have been conducted. One study completed in 1976 examined male undergraduate students (N=40) and compared the effects of high and low doses of alcohol, one dose of THC, and placebo on general aggression. High alcohol precipitated aggression more than low dose of alcohol while THC was associated with a weak dampening effect. Another study completed in 1985 employed a small sample of male undergraduate students (N=30) and examined three doses of THC without a placebo condition. Participants in the low THC group demonstrated greater aggression compared to the moderate or high dose groups³⁶. A more recent study was conducted among heavy alcohol or cannabis users and healthy controls (N=60) who were randomized to receive either alcohol (.08g/L) or vaporized THC (300 μg THC/kg bodyweight). While alcohol increased subjective aggression, THC administration was associated with lower subjective aggression³⁷. Neither study examined simultaneous alcohol and THC intoxication, and both were limited in directly testing putative mechanisms of action underlying aggression.

Smoked cannabis has long been employed in laboratory settings and in randomized controlled trials in fields outside of IPV, and the proliferation of studies examining various orally-administered synthetic cannabinoids, and cannabidiol (which lacks the psychoactive component THC), all demonstrate feasibility and acceptability of these methods that can be readily applied to IPV populations. While the current study cannot establish causal associations, future studies can leverage the use of advanced methods to disentangle causal associations and identify individual differences that might contribute to those relations.

Consistent with prior research, couples in the sample were characterized by bidirectional IPV; however, it is unclear how acute cannabis use may impact this association among individuals or couples experiencing greater frequency or severity of IPV or unilateral IPV. Our measure of IPV does not take into account the context in which IPV occurs, which should be considered in future research. Further, due to our small sample size, we were unable to examine partner cannabis effects that should be explored in future research. The time frame of TLFB measurement overlapped with the CTS-2 in this study, but it is possible that some IPV incidents as well as cannabis use might have occurred outside the window of TLFB measurement. Future studies can improve on this assessment by more closely aligning measurement windows. Moreover, while we found that twelve couples were cannabis-use discordant on self-report measurement using the TLFB, the sample is too limited to examine whether discordant cannabis use might result in greater IPV frequency or severity. Recent evidence from another study suggests that discordant cannabis use is associated with negative behaviors (e.g., demand/withdraw) during conflict³⁸, however this work has yet to be extended to IPV.

Next, the present study did not account for concurrent use of more than one substance, nor did it include a healthy control group or other comparator groups to better contextualize findings. Comparator groups including non-distressed couples, couples who use substances but do not report problematic use or meet diagnostic criteria for substance use disorder, or couples with more specific patterns of cannabis use (i.e., daily use) might contribute to further disentangling this association. Although some participants self-reported using multiple substances, and this was controlled for in the current analyses, we did not examine whether drugs were used simultaneously at the event level. Prior research has yielded inconsistent findings when examining polysubstance use disorders and IPV. One study demonstrated that a diagnosis of alcohol use and cannabis use disorder decreases the likelihood of IPV perpetration compared to a single diagnosis³⁹. Conversely, another study found that for men, but not women, an alcohol and cannabis use disorder significantly predicted IPV perpetration and victimization, as well as the severity of IPV⁴⁰. However, research has yet to examine how acute simultaneous verses single use is associated with IPV. Finally, it is also possible that variables beyond the scope of the current study, such as stable personality characteristics, might influence the association between cannabis use and IPV perpetration and victimization. A well-designed sampling plan consisting of relevant comparator groups in which substance use and IPV are clearly defined a priori is also warranted for future studies to clarify the association between cannabis use and IPV.

Conclusions

This study found that in a sample of substance misusing community couples, both quantity and frequency of cannabis use in the past 90 days was associated with increased IPV victimization and perpetration while THC-positive urine drug screens were associated with IPV victimization only. These findings are consistent with those of previous studies indicating that cannabis use is related to IPV and suggest that more research using state of the science methodologies are required to examine this association in greater detail, and to determine the pharmacological, individual, contextual, and relational factors that might influence how cannabis, and the interaction of cannabis with alcohol and other drugs, influences IPV.