New Study: Cannabis and Driving in Older Adults January 18, 2024

Key PointsQuestion What is the association between retail cannabis available to the consumer, driving, and associated blood tetrahydrocannabinol (THC) levels in people over 65 years of age?

Findings In this cohort study, 31 regular users of cannabis aged 65 to 79 years chose on average high potency (18.74% THC) THC-dominant cannabis. Weaving was increased and speed was decreased at 30 minutes after smoking, which was not correlated with blood THC concentrations; subjective experience and self-reports of impaired driving persisted for 3 hours.

Meaning These findings suggest that older drivers, even if they regularly use cannabis, show evidence of impaired driving performance after smoking cannabis.


Importance Epidemiological studies have found that cannabis increases the risk of a motor vehicle collision. Cannabis use is increasing in older adults, but laboratory studies of the association between cannabis and driving in people aged older than 65 years are lacking.

Objective To investigate the association between cannabis, simulated driving, and concurrent blood tetrahydrocannabinol (THC) levels in older adults.

Design, Setting, and Participants Using an ecologically valid counterbalanced design, in this cohort study, regular cannabis users operated a driving simulator before, 30 minutes after, and 180 minutes after smoking their preferred legal cannabis or after resting. This study was conducted in Toronto, Canada, between March and November 2022 with no follow-up period. Data were analyzed from December 2022 to February 2023.

Exposures Most participants chose THC-dominant cannabis with a mean (SD) content of 18.74% (6.12%) THC and 1.46% (3.37%) cannabidiol (CBD).

Main outcomes and measures The primary end point was SD of lateral position (SDLP, or weaving). Secondary outcomes were mean speed (MS), maximum speed, SD of speed, and reaction time. Driving was assessed under both single-task and dual-task (distracted) conditions. Blood THC and metabolites of THC and CBD were also measured at the time of the drives.

Results A total of 31 participants (21 male [68%]; 29 White [94%], 1 Latin American [3%], and 1 mixed race [3%]; mean [SD] age, 68.7 [3.5] years), completed all study procedures. SDLP was increased and MS was decreased at 30 but not 180 minutes after smoking cannabis compared with the control condition in both the single-task (SDLP effect size [ES], 0.30;b = 1.65; 95% CI, 0.37 to 2.93; MS ES, −0.58;b = −2.46; 95% CI, −3.56 to −1.36) and dual-task (SDLP ES, 0.27;b = 1.75; 95% CI, 0.21 to 3.28; MS ES, −0.47;b = −3.15; 95% CI, −5.05 to −1.24) conditions. Blood THC levels were significantly increased at 30 minutes but not 180 minutes. Blood THC was not correlated with SDLP or MS at 30 minutes, and SDLP was not correlated with MS. Subjective ratings remained elevated for 5 hours and participants reported that they were less willing to drive at 3 hours after smoking.

Conclusions and relevance In this cohort study, the findings suggested that older drivers should exercise caution after smoking cannabis.


Epidemiological studies have established that cannabis increases the risk of a motor vehicle collision.15Laboratory studies have demonstrated that this impairment results in increased weaving, slowed reaction time, and compensatory changes in speed and following distance.68At present, controlled investigations of the effects of cannabis on driving have enrolled younger participants.68Cannabis use is on the rise in older adults,911yet the effects of cannabis on driving remain unknown in this population. Older adults may be particularly affected by cannabis, given age-related changes in cognition,1220metabolic changes that may prolong or enhance the effects of cannabis,2124and the concomitant use of medications. Conversely, older users of cannabis may have been using cannabis for many years and cannabis may have a diminished impact in this population due to development of tolerance.2528

Delta-9-tetrahydrocannabinol (THC), the main psychoactive component of cannabis, is believed to be responsible for the impact of cannabis on driving.2933Based on the success of the deterrence of alcohol-impaired driving, many jurisdictions have adopted per se limits for blood THC levels. These limits set a cutoff of THC in blood that are permitted while driving; these limits vary by jurisdiction but generally range between 0 and 5 ng/mL. Although dose-dependent increases in driving impairment have been observed after cannabis, the exact relationship between blood THC and driving abilities remains debated.34,35

The purpose of the present study was to investigate the association between cannabis and driving and blood THC levels in older adults. The present study used an ecologically valid approach in which participants were invited to smoke their usual cannabis in the laboratory. This is important because it is known that legally available cannabis is more potent than the cannabis that has typically been studied in the laboratory.36An ecological approach may yield more valid results into the outcomes of cannabis in everyday situations. In the present study, participants drove the simulator before smoking in a dedicated smoking room and then again at 30 and 180 minutes afterwards; blood was collected for measurement of THC and metabolites as well as cannabidiol (CBD) at the time of the drives. In an ecologically valid control condition, participants relaxed in the dedicated smoking room instead of smoking. We hypothesized that SD of lateral position (SDLP; weaving) after smoking cannabis would be increased compared with the no smoking condition, while reaction time would be slowed and speed decreased. We further hypothesized that measures of driving performance would be associated with blood THC levels.


This study was approved by research ethics boards at both the Centre for Addiction and Mental Health and Health Canada. This study was conducted at the Centre for Addiction and Mental Health in Toronto, Canada, between March 2022 and November 2022 with no follow-up period. This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.


Adults aged 65 to 79 years were recruited from advertisements placed on public transit and social media. After provision of written informed consent, participants were evaluated for inclusion criteria (see eAppendix 1 inSupplement 1).

Design and Procedures

This was a within-participants counterbalanced study of the association between cannabis and simulated driving and blood THC. After the telephone interview, potentially eligible participants had an in-person assessment and those eligible participated in a practice session to familiarize them with the study procedures, including an opportunity to drive the simulator. If the participant experienced illness on the simulator, they were given a break. If the illness persisted, they were withdrawn from the study. At this session demographic information was collected, including self-reported race and ethnicity for statistical purposes (participants were provided with the following options: Asian-East [eg, Chinese, Japanese, Korean], Asian-South [eg, Indian, Pakistani, Sri Lankan]. Asian-South East [eg, Malaysian, Filipino, Vietnamese], Black-African [eg, Ghanaian, Kenyan, Somali], Black-Caribbean [eg, Barbadian, Jamaican], Black-North American [eg, Canadian, American] First Nations, Indian-Caribbean [eg, Guyanese with origins in India], Indigenous or Aboriginal not included elsewhere, Inuit, Latin American [eg, Argentinian, Chilean, Salvadoran], Metis, Middle Eastern [eg, Egyptian, Iranian, Lebanese], White-European [eg, English, Italian, Portuguese, Russian], White-North American [eg, Canadian, American], mixed heritage [eg, Black-African and White-North American], other[s], prefer not to answer, do not know). In total, the practice session was about 2 to 3 hours in duration.

This was followed by 2 test sessions of about 7 hours each (5 hours after smoking and 2 hours of baseline), separated by at least 72 hours. Participants were asked to abstain from cannabis, alcohol, and other recreational drugs for 12 hours and received the following 2 conditions in counterbalanced order: (1) cannabis, in which they smoked cannabis in a dedicated negative pressure room; and (2) an ecologically valid control, with no placebo or cannabis, in which they relaxed in the smoking room for approximately 10 minutes.

Before each test session, breathalyzer (Alert J5 model) and saliva sampling (DrugWipe, 5 ng/ml cutoff) were used to confirm self-reported abstinence from alcohol and cannabis in the past 12 hours. In addition, a 14-panel urine screen was used to determine that there was no recent use of other recreational or psychoactive drugs. At the start of each test session, participants were asked about their degree of withdrawal from cannabis, as assessed by the Marijuana Withdrawal Checklist (MWC).37,38

Participants then drove the simulator and provided blood for measurement of THC and metabolites before smoking cannabis or relaxing in the smoking room and then again at 30 and 180 minutes after; the timing of blood draws for measurement of THC corresponded to the time of the drives. Cognitive and subjective assessments followed each drive (to be published in a separate report). The visual analog scales (VAS; see eAppendix 2 inSupplement 1for the definitions) were administered at baseline and then again at 30 minutes, 60 minutes, and hourly until 5 hours after smoking. Before the drive at baseline and at 180 minutes, participants were asked (1) how willing they would be to drive a real vehicle (5-point Likert scale); and (2) how impaired they were at the time to drive35(VAS from 0 to 100). Participants were compensated for their participation in the study. For a schematic of the test day and details of the blood analysis, see eFigure 1 for the test session and eAppendix 3 for analysis of blood inSupplement 1.


Participants were asked to bring their own legally purchased cannabis to the laboratory. They were asked to smoke the cannabis as a joint with no tobacco. They were given the choice of rolling their own joint or bringing a preroll that was purchased from a retail outlet. Participants were allowed to smokead libitum. Since this was an ecologically valid design, participants were provided with only a few instructions: they were told to smoke their usual amount to achieve their desired effect and to stop if they felt strange or unwell. They were told that they were not required to smoke the entire cannabis cigarette. The cannabis cigarette was weighed before and after smoking to estimate the amount consumed, with an electronic balance (model VWR-123P) that was calibrated at least weekly in accordance with the manufacturer instructions. The amount of THC and CBD present in the cannabis was determined from the packaging of the cannabis. To estimate the amount of THC smoked, the potency of THC in the cigarette (expressed as a percentage) was multiplied by the change in weight of the cigarette (in mg), and divided by 100.


The primary outcome was SDLP, the measure most consistently found to be associated with cannabis.27,33,3945SDLP (measured in centimeters) is a sensitive measure of the effects of psychoactive drugs on driving and measures the amount of weaving. Secondary outcome measures were mean speed (MS, km/h), brake latency (reaction time; seconds), SD of speed (SDSP), and maximal speed (MAX, km).30,32,33,46MS is the mean speed (km/h) during the drive when asked to maintain a speed of 80 km/h. SDSP represents the variability of speed during a drive. Larger numbers mean that the driver was not able to maintain a consistent speed. MAX is the maximal speed obtained during the drive. Reaction time, or brake latency, is the time for a participant to move their foot from the gas to the brake pedal after a stop sign appears on the road facing them; participants were instructed not to brake when a stop sign appeared that was not facing them. For details of the driving simulator and simulations, see eAppendix 4 inSupplement 1.

All driving outcomes were assessed under both single- and dual-task conditions. In the dual-task condition, participants were required to count backward by threes aloud while driving, starting at a number between 700 and 1000. This condition was included to mimic the situation of driving while distracted or under extra cognitive load and has been shown to be a good measure of driving while distracted.47,48

Primary Sponsor


Karma Koala Podcast

Top Marijuana Blog