Violence among men and women in substance use disorder treatment: A multi-level event-based analysis

Objective: Numerous studies have documented an association between violence and substance use, particularly alcohol and cocaine. However, controversy remains regarding the influence of substances in specific violent incidents and other factors underlying the observed associations. Moreover, few studies have obtained detailed measures of acute alcohol and drug use associated with violent incidents. Understanding the relationship between substance use and violence can play a key role in developing effective treatments. This study uses an event-based approach to examine the influence of different types of substances types on violence severity.

Method: Participants were recruited within 30 days of enrolling in substance use disorder treatment who screened positive for past-year violence (N = 180; 77% men, 23% women). A semi-structured interview using the Conflict Tactics Scale and timeline follow-back measures were used to obtain details about substance use associated with violence (e.g., use of alcohol-, cocaine-, marijuana- and heroin-use associated with conflict incidents) and violence severity (no violence, violence without injury toward other, and violence with injury toward other). Multilevel multinomial regression analysis was used to examine the associations between types of substances used and episodes of violence, while adjusting sociodemographics and accounting for the clustering of events within individuals.

Results: Approximately 50% of the sample reported no physical violence, 31% physical violence without injury (toward other), and 19% physical violence with injury (toward other). Descriptive analyses showed that alcohol and other drug use appeared to be at higher levels on days with violence (with and without injury). The multilevel multinomial regression model showed that heroin increased the odds of violence with and without injury; cocaine and alcohol increased the odds of violence with injury; and marijuana was not associated with any type of violence.

Conclusions: Overall, consistent with other studies, alcohol and cocaine use were related to violence severity (e.g., resulting in injury). The findings regarding heroin and marijuana use are novel given that few studies have examined acute heroin or marijuana use associated with conflict incidents. A notable finding of this study is that usage of marijuana was associated with a decreased likelihood of violence with injury. These findings underscore the importance of event-based analysis in social work research, particularly as it relates to factors that are likely to vary within an individual over time. When treating substance use disorder, social workers should also be aware of how different types of drugs increase the risk for violence.