Alcohol and cannabis are among the most frequently used drugs worldwide (EMCDDA 2012). The elicitation of aggressive behavior following alcohol consumption, also called “intoxicated aggression,” has been frequently reported on a global scale (Murdoch et al. 1990). Cannabis intoxication, however, does not typically lead to aggression in most individuals (Hoaken and Stewart 2003), but it might increase or facilitate aggression in certain subgroups (i.e., violent offenders, clinical population) (Cherek et al. 1993). However, not everybody who uses alcohol or cannabis engages in aggressive behaviors (Heinz et al. 2001; Kopak et al. 2014; Lammers et al. 2014). A clear relationship between alcohol, drugs, and intoxicated aggression is neither linear nor invariant. Some drugs can facilitate aggressive behavior through their direct pharmacological effects during intoxication, through neurotoxic effects caused by chronic drug use over time or through withdrawal effects during abstinence (Hoaken and Stewart 2003).

The relation between alcohol consumption and aggression has been well established. Experimental studies on aggression have demonstrated that acute doses of alcohol facilitate aggressive behavior in a dose-related manner as assessed by vocal recordings and questionnaires (Bushman and Cooper 1990; Ito et al. 1996). Studies using laboratory-based measures of aggression have generally found that aggression was higher in participants who were intoxicated compared to those who received no alcohol (for a review see Giancola and Chermack, 1998). Longitudinal and observational studies suggest that acute episodes of heavy alcohol consumption are more strongly related to aggressive behavior than chronic alcohol consumption (Chermack and Blow 2002; Fals-Stewart 2003). This indicates that alcohol-induced aggression is more likely to occur in users who are consuming excessively within a given drinking episode (Heinz et al. 2011), although it is only a minority of people who become aggressive when under the influence of alcohol (Beck and Heinz 2013).

The relation between cannabis use and aggression has also been investigated in studies with animals and with humans. In animals, studies have shown a decrease in aggressive behavior of rodents and primates following cannabis administration (Miczek and Barry 1977; Miczek 1978). In humans, experimental findings on acute effects of cannabis on aggression are mixed (Taylor et al. 1976; Moore and Stuart 2005). Some studies indicate that cannabis intoxication is associated with the elicitation of aggression (Cherek et al. 1993; Howard and Menkes 2007). However, interpretation of these findings is difficult as they are based on relatively small sample sizes (Cherek et al. 1993; Howard and Menkes 2007) or only included male participants with self-reported anti-social tendencies (Cherek et al. 1993). Moreover, dose and route of administration differed considerably between studies. Cannabis effects on aggression are exerted in a dose-dependent manner. Low doses (0.1 mg/kg) of tetrahydrocannabinol (THC) slightly increased the willingness of participants to increase shock intensity given to opponents, were moderate to high doses (0.25–0.4 mg/kg) decreased aggressive response during a laboratory-based aggression study (Myerscough and Taylor 1985). In the latter study, however, participants were randomly assigned to one of the three dose conditions without a placebo condition or control group, making it difficult to assess whether the effect was pharmacological, contextual, or due to individual differences. One study monitored aggression in long-term heavy cannabis users (Kouri et al. 1999) and reported increased aggressive responses relative to controls when tested 3 and 7 days into abstinence.

Aggressive behavior is modulated by neuroendocrine mechanisms, and it is suggested that changes in cortisol and testosterone are predictive of changes in aggression (Brown and Dobs 2002; Brown et al. 2008; Terburg et al. 2009). Fluctuations in cortisol levels can affect the relationship between testosterone and the expression of aggression (Popma et al. 2007). It is unclear whether hormones could play a mediating role in the relationship between drugs and aggression. Previous studies report significant changes in testosterone and cortisol levels following acute alcohol and cannabis administration (Ylikahri et al. 1974; Mendelson et al. 1977; Välimäki et al. 1990; Lovallo et al. 2000; Sarkola et al. 2001; Brown and Dobs 2002; Frias et al. 2002; Thayer et al. 2006). Suppression of male testosterone levels has been reported after short-term heavy drinking (Sarkola and Eriksson 2003), and a reduction in cortisol reactivity was found in heavy drinkers compared to light drinkers after a high (0.8 g/kg) alcohol dose (King et al. 2006). Studies on the effects of cannabis show decreased male testosterone levels after both acute and chronic cannabis use (Kolodny et al. 1974, 1976) and elevated cortisol levels in occasional smokers (Cone et al. 1986); these findings were not corroborated by subsequent studies however (Mendelson et al. 1974; Schaefer et al. 1975; Block et al. 1991).

While aggression is defined objectively as any type of behavior aimed at harming another living being who is motivated to avoid such a behavioral act (Baron 1977), aggression in humans could also be operationalized on a subjective level as the increase in aggressive inclination that is triggered by an aversive/aggressive stimulus or event underlying an emotional-cognitive state. The present study investigated the acute effects of alcohol and cannabis on subjective aggression following aggression exposure in heavy alcohol and regular cannabis users, respectively. Subjective aggression was directly measured by means of a visual analogue scale (VAS) that allowed participants to rate their feeling of aggression on a linear scale ranging from “not aggressive at all” to “very aggressive.” Previous studies (Bond and Lader 1974; Cleare and Bond 1995) that have used rating scales of subjective aggression in human drug studies demonstrated that subjective feelings of aggression and hostility are positively correlated to behavioral acts or measures of aggression. The relevance of subjective aggression therefore lies in the notion that it may predict or coincide with behavioral acts of aggression. Aggression exposure consisted of two tasks developed to evoke and measure aggressive responses: i.e. the single category implicit association test (SC-IAT) and the point-subtraction aggression paradigm (PSAP). Subjective aggression occurring in response to some perceived provocation can be categorized as subjective affective/reactive aggression (Anderson and Bushman 2002). A control group served as between group reference in order to compare aggressive responses of alcohol and cannabis users with non-drug users. Subjective aggression in alcohol and cannabis users was compared with controls while sober and compared to placebo while intoxicated. It was expected that aggression exposure would increase subjective aggression in alcohol users during alcohol intoxication and decrease subjective aggression in cannabis users during cannabis intoxication. It was further expected that subjective aggression would increase after aggression exposure in all groups when sober. Neuroendocrine measures of cortisol and testosterone in response to alcohol/cannabis treatment and after aggression exposure were recorded as additional, secondary outcome measures.