Management of substance abuse

The health and social effects of nonmedical cannabis use

New WHO publication on cannabis

Chapter 5. Short-term effects of cannabis

5.1 What do we know?

5.1.1 Cognition and coordination

Crean, Crane & Mason (2011) reviewed a broad spectrum of cognitive functions, designated as executive functions, and identified studies that reported that attention, concentration, decision-making, impulsivity, inhibition (self-control of responses), reaction time, risk taking, verbal fluency and working memory were impaired acutely in a dose-dependent manner, although these effects were not consistently observed.

Cannabis acutely impairs several components of cognitive function, with the most robust effects on short-term episodic and working memory, planning and decision-making, response speed, accuracy and latency (Ranganathan & D’Souza, 2006). Some studies also report increased risk-taking and impulsivity (Crean, Crane & Mason, 2011). Less experienced cannabis users undergo stronger intoxicating effects on attention and concentration than those with established drug tolerance. Cannabis also acutely impairs motor coordination, interferes with driving skills and increases the risk of injuries. Evidence suggests that recent cannabis smoking is associated with substantial driving impairment, particularly in occasional smokers, with implications for work in safety-sensitive positions or when operating a means of transportation, including aircraft (Hartman & Huestis, 2013). Complex human/machine performance can be impaired as long as 24 hours after smoking a moderate dose of cannabis and the user may be unaware of the drug's influence (Leirer, Yesavage & Morrow, 1991).

5.1.2 Anxiety and psychotic symptoms

A minority of first-time cannabis users become very anxious, have panic attacks, experience hallucinations and vomit. These symptoms may be sufficiently distressing to prompt affected users to seek medical care (Smith, 1968; Thomas, 1993; Weil, 1970). Experienced users may also have negative experiences if they use more potent cannabis products than usual, or use cannabis by an unfamiliar route (e.g. oral ingestion) that does not permit them to achieve their usual dose of THC. Hallucinations may occur after using very high doses of THC, and may occur at lower doses in individuals with a pre-existing vulnerability to psychosis (e.g. having experienced psychotic symptoms or having a first-degree relative with a psychotic disorder). These distressing experiences are often time-limited and can usually be managed by reassurance and mild sedation in a safe environment (Dines et al., 2015).

5.1.3 Acute toxicity

The risk of a fatal cannabis overdose is extremely small compared to the risks of opioid and stimulant drug overdoses (Gable, 2004). The dose of THC that reliably kills rodents is extremely high and the equivalent fatal dose in humans extrapolated from animal studies is between 15 g (Gable, 2004) and 70 g (Iversen, 2007; Lachenmeier & Rehm, 2015). This is much greater than the amount of cannabis that a very heavy user would consume in a day (Gable, 2004). There are no reports of fatal overdoses in the epidemiological literature (Calabria et al., 2010b). The lack of respiratory overdoses is consistent with the absence of cannabinoid receptors in brain stem areas that control respiration (Iversen, 2012).

5.1.4 Acute cardiovascular effects

Acute exposure to cannabis increases heart rate and blood pressure and can in some cases cause orthostatic hypotension (Pacher & Kunos, 2013; Schmid et al., 2010). There have been case reports of serious cardiovascular complications, including acute coronary syndromes and strokes, in cannabis users (Jouanjus, 2014). Mittleman and colleagues found that the risk of myocardial infarction was four times higher in patients with a recent myocardial infarction in the hour after smoking cannabis compared to cannabis users without a history of myocardial infarction (Mittleman et al., 2001). The risk then declined rapidly. Many of these more serious events have been reported in heavy daily cannabis smokers and are discussed in more detail in section 7.1.2.

5.1.5 Acute effects on lungs and airways

The acute bronchial effects of smoking tobacco and smoking cannabis differ; tobacco smoking produces acute bronchial constriction, while cannabis smoking causes acute bronchial dilation in proportion to the dose of THC (Tashkin, 2015).This effect has been reported in cannabis users in the United States where cannabis used to be smoked alone Users in many parts of the world frequently smoke cannabis and tobacco together (especially when cannabis resin is used), and this combination is likely to produce different acute bronchial effects. The effects of long- term cannabis smoking on lung function are considered in more detail in section 7.1.1.

5.1.6 Traffic injuries and fatalities

At the time of the last WHO report on cannabis (WHO, 1997), laboratory studies showed that cannabis and THC produced dose-related impairments in reaction time, information processing, perceptual-motor coordination, motor performance, attention and tracking behaviour (Moskowitz, 1985; Robbe & O’Hanlon, 1993). These findings suggested that cannabis could potentially cause car crashes if users drove while intoxicated.

It was unclear in 1997, however, if cannabis use increased traffic accidents. Studies in driving simulators indicated that cannabis-impaired drivers were aware of their impairment and compensated by slowing down and taking fewer risks. There were similar findings in the small number of studies on the effects of cannabis use on driving on the road (Smiley, 1999). In some of these studies, however, cannabis-impaired drivers responded less effectively to simulated emergencies than control drivers did (Robbe, 1994).

Most epidemiological studies of traffic fatalities in the 1990s are reported only on the presence of cannabis metabolites. These indicated only that cannabis had been used hours to days before the accident; they did not establish that the drivers were impaired by cannabis at the time of the accident. Moreover, a substantial proportion of drivers with cannabis in their blood also had high blood alcohol concentrations (BACs), which made it difficult to distinguish the effects of cannabis and alcohol on accident risk (Hall, Solowij & Lemon, 1994).

In the past two decades, better designed epidemiological studies have found that cannabis users who drive while intoxicated double their risk of a car crash (Asbridge, Hayden & Cartwright, 2012). Evidence suggests that recent cannabis smoking is associated with substantial driving impairment, particularly in occasional smokers. The increased risk of motor vehicle accidents in these studies has persisted after statistical adjustment for confounding. For example, Mura et al. (2003) found an increased risk of accidents in a case-control study of 900 persons hospitalized in France with motor vehicle injuries and 900 age- and sex-matched controls admitted to the same hospitals for reasons other than trauma. Laumon and colleagues (2005) compared blood THC levels in 6766 culpable and 3006 non-culpable drivers in France between October 2001 and September 2003. Culpability was higher in drivers with THC levels greater than 1 ng/mL (OR = 2.87) and there was a dose-response relationship between blood THC and culpability that persisted after controlling for BAC, age and time of accident.

A meta-analysis of nine case-control and culpability studies (Asbridge, Hayden & Cartwright, 2012) found that recent cannabis use (indicated by either THC in blood or self-reported cannabis use) doubled the risk of a car crash (OR = 1.92, 95% CI: 1.35, 2.73). The risk was higher in better-designed studies (2.21 vs. 1.78), case-control rather than culpability studies (2.79 vs. 1.65) and studies of fatalities rather than injuries (2.10 vs. 1.74). Very similar results were reported in a meta-analysis by Li et al. (2012) (who reported a pooled risk estimate of 2.66) and in a systematic review of laboratory and epidemiological studies (Hartman & Huestis, 2013). The risk of an accident increases substantially if cannabis users also have elevated blood alcohol levels, as many do (Hartman & Huestis, 2013).

Finally, a meta-analysis of 72 estimates of the risk of injury from cannabis was obtained from 46 studies, including some of the studies referred to above but also several others (Elvik, 2015). A random-effects model of analysis produced estimates of the risk of injury associated with the use of cannabis (95% confidence intervals in parentheses) and after adjustment for publication bias (Table 5.1).

Table 5.1. Estimates of the risk of injury associated with the use of cannabis



Adjusted for publication bias

Fatal injury

1.37 (1.24; 1.52)

1.37 (1.24, 1.51)

Serious injury

1.96 (1.27; 3.02)

1.84 (1.19, 2.85)

Other injury (severity not specified)

1.41 (0.97; 2.05)

1.12 (0.78, 1.62)

Property damage only

1.43 (1.26; 1.63)

1.11 (0.93, 1.32)


A test for publication bias suggested bias at all levels of injury severity, but not severe enough to influence the summary estimates of risk very much.

The analysis also found a relationship between the prevalence of cannabis use in drivers and the risk of injury associated with using cannabis. The fewer drivers that used cannabis, the higher the risk associated with its use. This pattern probably reflected selective recruitment of risky drivers to use cannabis.

The Driving Under the Influence of Drugs, Alcohol, and Medicines (DRUID) study was a population-based study of accident risks related to the use of cannabis and other drugs in nine EU countries (Hels et al., 2012). A pooled analysis found that drivers who tested positive for THC were 1-3 times more likely to be in an accident than sober drivers. This is comparable to a BAC level of 0.05 g/dl to <0.10 g/dl but the confidence intervals around these estimates were wide. A Department of Transportation case-control study in the USA found that drivers who tested positive for THC had 1.25 times higher risk of collision than a sober driver, but the association disappeared when age, gender, ethnicity and BAC levels were taken into account (Berning, Compton & Wochinger, 2015).

The existing evidence points to a small causal impact of cannabis on traffic injury. There are plausible biological pathways, and the pooling of studies found significant effects for cannabis. Overall, even though the effect is small compared to the effects of alcohol, traffic injury may be the most important adverse public health outcome for cannabis in terms of mortality in high-income countries (Fischer et al., 2015).

5.1.7 Other injury (not related to driving)

Some recent epidemiological studies of cannabis use and general injury risk have produced mixed findings. Gerberich and colleagues (2003) found that, among 64 657 patients in a health maintenance organization, cannabis users had higher rates of hospitalization for injury from all causes than former cannabis users or non-users. A meta-analysis of injury studies related to cocaine and cannabis users found that cannabis use was related to intentional injuries, as well as injuries in general, in cannabis-using clients of addiction treatment services (Macdonald et al., 2003). However, the authors argued that the evidence was not conclusive on the risk of injury among cannabis users. A Canadian survey study of 1999 adults who reported a history of traumatic brain injury had higher odds of reported past-year daily smoking (adjusted odds ratio [AOR] = 2.15), use of cannabis (AOR = 2.80) and use of nonmedical opioids (AOR = 2.90) (Ilie et al., 2015).

A case-crossover study among a sample of injured male and female patients in the emergency department in Lausanne, Switzerland, found that acute cannabis use (within a six-hour window) was associated with a reduced risk of injury (Gmel et al., 2009). The combined use of cannabis and alcohol was also not associated with an increased injury risk (Gmel et al., 2009). The authors suggested that the inconsistency between their findings and other studies could be explained by the fact that cannabis users in their study used cannabis at home whereas drinkers of alcohol usually consumed alcohol in bars where smoking cannabis did not frequently occur (Gmel et al., 2009). Another recent study among injured patients in the emergency department in Vancouver, Canada, also did not find an increased risk of injury associated with cannabis use. It did find, however, that the combined use of alcohol and drugs (with cannabis the most frequently reported drug) increased an individual’s risk of being injured compared to non-drug-using controls (Cherpitel et al., 2012). Both studies used self-reported data on cannabis use prior to the injury and at the control time period.

5.1.8 Cannabis and the workplace

The effects of cannabis use on cognition in the context of work and everyday life, and whether off-site cannabis use endangers a worker or his colleagues while at work, are of concern (Phillips et al., 2015, Goldsmith et al., 2015). This topic has not been systematically investigated in recent years.

5.1.9 Areas that require more research

A. The epidemiological evidence on the effects of cannabis on driving is increasing but it is still small compared to evidence on the effects of alcohol.

·       Larger and better-controlled studies are needed:

·       to clarify the magnitude of the risk of traffic injuries and to resolve inconsistent findings from recent studies (Berning, Compton & Wochinger, 2015);

·       on how tolerance may affect accident risk among regular cannabis users. Chronic heavy drinkers develop tolerance to alcohol and show fewer obvious signs of intoxication even at extremely high BAC levels. They can in many cases drive a car with BAC levels at which others with a lower tolerance would not be able to drive (Chesher, Greeley, & Saunders, 1989).

·       Differences in impairments for the same dose of THC also need to be investigated in naïve and experienced users (Berning, Compton & Wochinger, 2015).

·       Studies are needed to investigate the effects of high THC levels on driving.

·      Studies are required to compare the effects of smoking and of ingested cannabis on driving.

B. Some studies in the literature on driving use self-reported cannabis use as a marker.

  • Future research should rely only on biological specimens which are more reliable markers of cannabis use. At least one report found inconsistencies between self-reported cannabis use and biological specimens collected from crash victims (Asbridge et al., 2014), though all measures showed elevated risk.

C. A number of developed countries have introduced roadside drug testing to deter drivers from driving while impaired by cannabis.

  • Evaluations of the effectiveness of these countermeasures would provide some indication of the magnitude of the effect that cannabis use has on road crash risk (Hall, 2012).

D. Although a recent study found no increased risk of injury associated with cannabis use, which suggests that the setting in which cannabis is used may affect the risk (Gmel et al., 2009), other studies show the use of cannabis to be associated with increased injuries in adolescents and increased burns.

  • Research is needed to understand the effect on injury risk of the social environment in which cannabis is typically consumed.