The Utility of THC Cutoff Levels in Blood and Saliva for Detection of Impaired Driving

Abstract

Background:

Δ9-Tetrahydrocannabinol (THC) is the psychoactive component in cannabis and a relationship of THC to driving impairment is expected. Despite this, there are discrepant findings with respect to the relationship of blood THC to driving. This study investigated the relationship of blood, urine, and saliva THC/THC-COOH levels to “weaving,” as measured by a driving simulator.

Methods:

Participants smoked cannabis alone or with alcohol. THC/THC-COOH levels in blood, urine, and saliva were correlated with standard deviation of lateral position (SDLP), measuring “weaving.” In addition, SDLP after cannabis and/or alcohol were compared with SDLP after placebo when THC/THC-COOH levels were above or below specified thresholds in blood (5 ng/mL), urine (50 ng/mL), or saliva (25 ng/mL).

Results:

A clear linear relationship between blood THC concentration and SDLP was not observed based on calculation of Spearman coefficients. When compared with placebo, SDLP was significantly increased after cannabis and cannabis combined with alcohol when THC in the blood was above the legal limit. SDLP was increased in drug conditions when saliva cutoffs were above the legal limit.

Conclusions:

The findings of this study suggest that specified thresholds for THC in blood and saliva may be able to detect driving impairment, but future studies are needed. ClinicalTrials.gov ID: NCT03106363.

Introduction

Epidemiological studies have found that cannabis can increase the risk of a motor vehicle collision.^1,2 Indeed, a meta-analysis found that cannabis increases measures of impaired lane control and reaction time, as well as compensatory decreases in speed and increases in following distance.³ Δ9-Tetrahydrocannabinol (THC) is the psychoactive component in cannabis and the impact of cannabis on driving is underscored by evidence for a dose–response relationship between THC and driving impairment,^4–8 as well as changes in driving associated with specific levels of THC in blood.^9,10 By contrast, it has also been reported that there is no correlation between blood and oral fluid (saliva) levels of THC and impaired driving.^11–13 Indeed, there may be a great deal of variability in THC levels in blood¹⁴ and understanding the relationship of THC levels to driving is important because deterrence of driving under the influence of cannabis centers on levels of THC in blood or saliva above certain thresholds.¹⁵

Based on the success of roadside detection of alcohol on breath, many jurisdictions have attempted to set per se THC limits in blood and saliva. In Canada, per se thresholds of 5 ng/mL when measured in blood and 25 ng/mL when measured in saliva have been adopted. The purpose of this study was to determine the relationship between THC-COOH in blood, urine, and saliva and driving, using a driving simulator. Standard deviation of lateral position (SDLP) was chosen as the outcome measure as it is most consistently reported in driving simulator studies evaluating the effects of cannabis.^3,16

Methods

This study received approval by the Centre for Addiction and Mental Health REB (No. 123/2015) and the Health Canada REB (2015-018H). The present study was part of a larger investigation into the effects of alcohol, cannabis, or alcohol + cannabis on driving.¹⁶ This was a within-subjects, double-blind, double-dummy, placebo-controlled, randomized clinical trial. The study included the following four conditions: (1) placebo alcohol and placebo cannabis; (2) alcohol and placebo cannabis; (3) placebo alcohol and active cannabis, and; (4) alcohol and active cannabis.

Alcohol was administered with a target breath alcohol (BrAC) of 0.08%. The cannabis in the active condition contained 12.5%±2% THC in a 750 mg cigarette (∼94 mg THC) and was obtained from Aurora Cannabis Enterprises, Inc., a licensed producer approved by Health Canada. Placebo cannabis (<0.1% THC) was provided by the National Institute of Drug Abuse (NIDA) Drug Supply Program. Participants were instructed to smoke the cannabis as they normally do for a maximum of 10 min and told to stop smoking if they felt unwell or until their usual level of intoxication.

The alcohol beverage was equivalent to ∼3–5 standard drinks in an average woman weighing 66 kg or an average man weighing 83 kg. The beverage was mixed in a 1:3 ratio of vodka to tonic water. The placebo contained the same volume but was only tonic water. All beverages were topped with a few milliliters of lemon-lime juice and the placebo beverage was capped with a minimal amount of vodka to mask the smell. The drinks were divided into three equal portions and each portion was consumed over 5 min, for a total of 15 min.

Following a driving simulator session to ensure that participants were familiar with the simulator, all four of the drug administration sessions were identical with the only difference being the alcohol and/or cannabis exposure conditions. Details of the simulator and scenarios are provided in Fares et al (2022).¹⁶ Each drug administration session was separated by at least 72 h. Participants completed baseline driving simulation scenarios. Also at baseline, a whole blood sample (for quantitative determination of THC) was collected.

A baseline test of saliva was also conducted (DrugWipe 3S) and urine was collected for measurement of 11-Nor-9-carboxy-Δ⁹-tetrahydrocannabinol (THC-COOH). Once all baseline measures were completed, participants were administered alcohol or placebo. Completion of this task was considered time 0. Fifteen minutes after completion of the alcohol/placebo administration, the cannabis or placebo cannabis was administered. A second blood sample and saliva test was collected 45 min after time 0, followed by the driving trial. A second urine sample was collected ∼5 h after time 0.

For analysis of the correlation between blood THC and driving, THC values in blood collected at 45 min (∼20 min after cannabis smoking) were correlated with SDLP measures at the same time point using Spearman's correlation coefficient. For the urine measures, THC-COOH values in urine collected 5 h were correlated with SDLP at 45 min. For the saliva THC measures, outcomes with DrugWipe 3S at 45 min were correlated with SDLP at the same time point. It was of interest to determine whether SDLP was increased in those participants over the legal limit for THC. For analysis of SDLP in those over the legal cutoff, SDLP data were divided into two groups. For one, SDLP for those >5 ng/mL of THC in blood was compared with their respective SDLP under placebo with Wilcoxon signed-rank test.

Similarly, those who tested below the legal limit of 5 ng/mL of THC in blood were compared with their respective SDLP under placebo. A similar analysis was conducted for urine with a THC-COOH of 50 ng/mL (the cutoff in urine toxicology cups). For saliva, data from those that achieved a “positive” outcome (over 25 ng/mL) or “negative” outcome (<25 ng/mL) was compared with their respective SDLP under placebo. A limit of 5 ng/mL was selected for blood because it is the upper legal limit in the Province of Ontario, Canada.¹⁷ The cutoff of 25 ng/mL was chosen for saliva because it is the legal limit in Canada.¹⁸

In Canada, there is no legal cutoff for urine but the cutoff in urine toxicology cups was adopted for COOH-THC. For each of the cutoffs for saliva, urine, and blood, the difference in amount smoked between those who were above or below the cutoff was compared with the Mann–Whitney U-test, separately for the cannabis alone and cannabis + alcohol conditions. The amount of cannabis smoked was calculated as the change in weight in the cigarette after smoking (post–pre). To determine whether driving was different when participants were either above or below the threshold, the difference between drug and placebo (drug-placebo) was compared for the above and below threshold groups with Mann–Whitney U-tests, separately for each biological matrix and drug condition.

Data from the control condition in which participants drank alcohol only (target BrAC 0.08%) were subject to similar correlational and signed-rank analyses. For the analyses of driving under or over the legal alcohol limit, a BrAC of 0.05% was chosen as the legal limit because this is the accepted limit in many jurisdictions and is the legal limit as recommended by the WHO.¹⁹ Owing to the fact that SDLP is a measure of variability (standard deviation), we used nonparametric tests in all analyses.

Results

One outlier was removed from the analyses owing to high levels of THC-COOH throughout the sessions, for a total sample size of 27. For analysis of the saliva data, saliva was not collected from one participant, resulting in a total sample size of 26. The results of correlational analyses are provided in Table 1. Results of Wilcoxon signed-rank test and Mann–Whitney tests are also provided in Tables 2 and 3. Visual inspection of scatter plots of the difference scores between placebo and drug conditions revealed that SDLP was higher during the cannabis session as compared with the placebo session when THC in blood and saliva were above the cutoffs; however, there was some degree of overlap (Fig. 1). For the alcohol and cannabis condition combined, SDLP was more pronounced in the drug session, regardless of whether THC levels were above the legal threshold. There were no differences in amount smoked for any comparisons, with the exception of the comparison of those above and below the cutoff in saliva after smoking cannabis (U=44.00, p=0.04).

FIG. 1.

Difference in SDLP between drug and placebo sessions when blood THC cutoff, saliva cutoff, and urine cutoffs are above (filled symbols) or below (open symbols) thresholds. Circles represent difference in SDLP between placebo and cannabis sessions and squares represent this difference for the alcohol + cannabis combined session. Horizontal lines represent the mean.

Table 1.

Correlation Between SDLP and Values in Biological Matrices

Measure	SDLP
Measure	Alc	AlcCann	Can
BrAC	r_s=−0.123	r_s=−0.360
Saliva THC		r_s=−0.067	r_s=−0.165
Blood THC		r_s=−0.012	r_s=0.201
Urine THC-COOH		r_s=−0.018	r_s=0.062

Spearman's correlation between SDLP and BrAC, blood THC, saliva THC (positive, negative) and urine THC-COOH after participants had either alcohol alone (Alc), cannabis alone (Can), or alcohol combined with cannabis (AlcCann). No correlations were significant.

SDLP, standard deviation of lateral position; BrAC, breath alcohol; THC, Δ9-tetrahydrocannabinol; THC-COOH, 11-nor-9-carboxy-Δ⁹-tetrahydrocannabinol.

Table 2.

SDLP Values Above and Below Thresholds

Matrix	Drug	Threshold	Placebo	Drug	n	Z	p
Blood	Cannabis	Above	28.54±8.12	33.77±8.10	13	−2.949	0.003^a
	Cannabis	Below	30.50±4.90	32.21±7.04	14	−0.596	0.551
	Alcohol–cannabis	Above	27.50±6.81	36.17±10.46	12	−2.945	0.003^a
	Alcohol–cannabis	Below	31.20±6.13	38.27±13.42	15	−2.906	0.004^a
Saliva	Cannabis	Above	29.25±6.33	34.25±7.19	12	−2.803	0.005^a
	Cannabis	Below	30.64±6.43	32.21±7.94	14	−1.087	0.277
	Alcohol–cannabis	Above	29.89±6.39	38.11±13.45	18	−3.451	0.001^a
	Alcohol–cannabis	Below	30.25±6.50	36.13±9.57	8	−2.214	0.027^a
Urine	Cannabis	Above	29.50±7.33	31.67±6.26	12	−1.432	0.152
	Cannabis	Below	29.60±6.20	34.0±8.34	15	−2.166	0.03^a
	Alcohol–cannabis	Above	26.33±5.48	34.67±5.63	9	−2.552	0.011^a
	Alcohol–cannabis	Below	31.17±6.64	38.67±14.15	18	−3.238	0.001^a
BrAC	Alcohol	Above	28.89±6.04	32.40±7.07	25	−2.246	0.025^a
BrAC	Alcohol–cannabis	Above	29.56±6.59	37.33±12.02	27	−4.149	<0.001^a

Mean±SD SDLP (in cm) after placebo or drug (cannabis, alcohol or alcohol combined with cannabis) when blood, saliva and urine THC and BrAC were below the threshold limit (5 ng/mL THC for blood, 25 ng/mL THC for saliva, 50 ng/mL COOH-THC for urine, and BrAC 0.05% for alcohol) or above the threshold limit. Wilcoxon signed-rank tests are presented.

Significant difference.

Table 3.

Comparison of Above and Below Threshold Values for Different Matrices

Matrix	Drug	Below	n (Below)	Above	n (Above)	Z (Mann–Whitney)	p
Blood	Cannabis	1.71±6.47	14	5.23±3.98	13	−2.14	0.033^a
Blood	Alcohol–cannabis	7.07±9.32	15	8.67±6.96	12	−0.954	0.347
Saliva	Cannabis	1.57±5.26	14	5.00±5.54	12	−1.443	0.16
Saliva	Alcohol–cannabis	5.88±7.26	8	8.22±8.79	18	−0.78	0.461
Urine	Cannabis	4.40±6.50	15	2.17±4.20	12	−0.978	0.347
Urine	Alcohol–cannabis	7.50±9.43	18	8.33±5.61	9	−0.954	0.348

Mean+SD difference in SDLP between the drug and placebo conditions when blood, urine, or saliva were above or below the threshold. Mann–Whitney U-test presented to compare the difference between groups.

Significant difference.

Discussion

No significant correlations between SDLP and THC-COOH following cannabis or combinations of alcohol and cannabis were found, regardless of biological matrix. Nonetheless, SDLP was significantly greater in the cannabis and combined alcohol and cannabis conditions relative to placebo when blood and saliva THC levels were above the legal cutoff, suggesting that there may be a threshold above which driving is impaired. Likewise, in the cannabis condition, SDLP was significantly greater among those participants with blood THC levels above the legal cutoff compared with participants with blood THC levels below that level. This speaks of the importance of examining data on the group level, and highlights the degree of variability seen with individual measures. Indeed, inspection of individual data revealed that there was some overlap in these values.

In previous studies we found evidence for dose-related effects of cannabis, using a median split approach to separate those with lower and higher levels of blood THC. We observed greater change in heart rate and self-reported subjective effects of cannabis²⁰ and greater change in memory impairment and mood effects of cannabis²¹ when a median split was performed on blood THC data at 7.3 ng/mL 30 min after smoking, separating participants into low and high blood THC groups. These findings of dose-related effects are in line with animal and human studies examining performance on simple tasks.²²

The fact that SDLP was also increased when blood and saliva THC was below the legal cutoff after the combined use of alcohol and cannabis likely reflects the cumulative effects of alcohol and cannabis on driving²³ and speaks to the importance of roadside deterrence of driving after use of cannabis combined with other drugs. Similarly, SDLP was increased following cannabis alone when saliva levels of THC were above the legal cutoff, but not when they were below this threshold. Statistical comparisons of driving performance when cutoffs were below versus above the threshold revealed a significant difference for the cannabis alone condition but not when cannabis and alcohol were used together; this absence of a statistically significant difference for the alcohol and cannabis combined condition likely reflects that SDLP was altered both below and above the cutoff.

THC-COOH in urine was not correlated with SDLP nor related to driving impairment above or below the legal threshold, a finding that is consistent with THC-COOH being a nonpsychoactive metabolite of THC. It should be noted that, in this study, urine samples were collected 5 h after smoking cannabis and therefore the lack of association between changes in driving and urine levels may reflect not only the fact that urine is not a reliable indicator of impairment, but may also be owing to the lack of temporal relationship between our sample and the time of the drive.

Several limitations of this study are important to consider. First, this study only used one route of administration, the smoked route. Although this is a common route of administration²⁴ and the one that is frequently studied in the literature,²⁵ it may be the case that other routes of administration may not yield similar results. Further, this study used a relatively small sample size and further controlled investigations are warranted to support the present findings. Finally, this study relied on only SDLP as a measure of driver performance, whereas speed, following distance and reaction time are also common outcomes.^3,25 Future investigations will need to include more batteries of tests to evaluate the impact of cutoffs on driving.

Our findings provide some interesting perspectives on the current controversy over the value of per se levels for detecting impairment by cannabis. For example, two recent reviews suggest that current evidence does not support the use of legal limits for cannabis, because these effects may not be dose related.^26,27 Our results suggest that support for the ability to detect cannabis impairment may differ depending on the research approach that is taken. Individual values may be highly variable and conservative group analyses may be warranted. Studies that find little evidence for dose-related effects appear to be those that are based on smoking cannabis using ad libitum smoking procedures, and involve driving simulator measures of performance.^26,27 Interpretation of the results of these studies is complicated by the wide variability seen in blood THC levels after smoking, presumably affected by such factors as depth and length of inhalation, and the complex nature of measures derived from the driving simulator task.

Under these circumstances, the ability to detect dose-related effects using regression approaches may be restricted. In our results here, we find no significant relationship between blood and saliva THC levels and SDLP when analyzed with correlational methods. However, when SDLP after active drug was compared with placebo, we found differences in people who tested above legal limits in blood and saliva. Although there were no differences in driving between the below or above cutoffs after combinations of alcohol and cannabis, this may be related to the greater changes in driving observed in this condition. Future studies are warranted to extend the observations in this study.

Footnotes

Acknowledgments

This manuscript is dedicated to our friend, colleague, and mentor Dr. Robert E. Mann. Placebo cannabis was supplied by the NIDA Drug Supply Program. Active cannabis was provided by Aurora Enterprises.

Author Disclosure Statement

B.L.F. has obtained funding from Pfizer, Inc. (GRAND Awards, including salary support) for investigator-initiated projects. He has obtained funding from Indivior for a clinical trial sponsored by Indivior. He has in-kind donations of cannabis products from Aurora Cannabis Enterprises, Inc., and study medication donations from Pfizer, Inc. (varenicline for smoking cessation) and Bioprojet Pharma. He was also provided a coil for a Transcranial magnetic stimulation study from Brainsway. He has obtained industry funding from Canopy Growth Corporation (through research grants handled by the Centre for Addiction and Mental Health and the University of Toronto), Bioprojet Pharma, Alcohol Countermeasure Systems, Alkermes and Universal Ibogaine. He has received in kind donations of nabiximols from GW Pharmaceuticals for past studies funded by CIHR and NIH. He has participated in a session of a National Advisory Board Meeting (Emerging Trends BUP-XR) for Indivior Canada and has been consultant for Shinogi. He is supported by CAMH, a clinician-scientist award from the department of Family and Community Medicine of the University of Toronto and a Chair in Addiction Psychiatry from the department of Psychiatry of University of Toronto. He is supported by Waypoint Centre for Mental Health Care. No other author has any conflicts to declare.

Funding Information

This research was supported by a project grant from the Canadian Institutes of Health Research (No. MOP-142196) and the Ministry of Transportation Ontario Road Safety Research Partnership Program.

Abbreviations Used

References

Asbridge

, Hayden

, Cartwright

. Acute cannabis consumption and motor vehicle collision risk: systematic review of observational studies and meta-analysis. BMJ. 2012; 344:e536.

, Brady

, DiMaggio

, et al. Marijuana use and motor vehicle crashes. Epidemiol Rev. 2012; 34:65–72.

Alvarez

, Colonna

, Kim

, et al. Young and under the influence: a systematic literature review of the impact of cannabis on the driving performance of youth. Accid Anal Prev. 2021; 151:105961.

Robbe

Marijuana's impairing effects on driving are moderate when taken alone but severe when combined with alcohol. Hum Psychopharmacol. 1998; 13:S70–S78.

Ramaekers

, Robbe

, O'Hanlon

. Marijuana, alcohol and actual driving performance. Hum Psychopharmacol. 2000; 15:551–558.

Lenne

, Dietze

, Triggs

, et al. The effects of cannabis and alcohol on simulated arterial driving: influences of driving experience and task demand. Accid Anal Prev. 2010; 42:859–866.

Ronen

, Gershon

, Drobiner

, et al. Effects of THC on driving performance, physiological state and subjective feelings relative to alcohol. Accid Anal Prev. 2008; 40:926–934.

Downey

, King

, Papafotiou

, et al. The effects of cannabis and alcohol on simulated driving: influences of dose and experience. Accid Anal Prev. 2013; 50:879–886.

Hartman

, Brown

, Milavetz

, et al. Cannabis effects on driving longitudinal control with and without alcohol. J Appl Toxicol. 2016; 36:1418–1429.

10.

Hartman

, Brown

, Milavetz

, et al. Cannabis effects on driving lateral control with and without alcohol. Drug Alcohol Depend. 2015; 154:25–37.

11.

Arkell

, Spindle

, Kevin

, et al. The failings of per se limits to detect cannabis-induced driving impairment: results from a simulated driving study. Traffic Inj Prev. 2021:1–6.

12.

Marcotte

, Umlauf

, Grelotti

, et al. Driving performance and cannabis users' perception of safety: a randomized clinical trial. JAMA Psychiatry. 2022; 79:201–209.

13.

Ramaekers

, Moeller

, van Ruitenbeek

, et al. Cognition and motor control as a function of Delta9-THC concentration in serum and oral fluid: limits of impairment. Drug Alcohol Depend. 2006; 85:114–122.

14.

Ramaekers

, Mason

, Kloft

, et al. The why behind the high: determinants of neurocognition during acute cannabis exposure. Nat Rev Neurosci. 2021; 22:439–454.

15.

Grotenhermen

, Leson

, Berghaus

, et al. Developing limits for driving under cannabis. Addiction. 2007; 102:1910–1917.

16.

Fares

, Wickens

, Mann

, et al. Combined effect of alcohol and cannabis on simulated driving. Psychopharmacology (Berl). 2022; 239:1263–1277.

17.

Canada Go. Frequently Asked Questions—Drug-Impaired Driving Laws. Available from: https://www.justice.gc.ca/eng/cj-jp/sidl-rlcfa/qa2-qr2.html [Last accessed: December 6, 2022].

18.

Use

RC.

Navigating Canada's Cannabis Laws: the Driver's Manual. Available from: https://thercu.org/blogs/blog/navigating-canada-cannabis-laws-the-driver-s-manual [Last accessed: December 6, 2022].

19.

Fell

JC.

Approaches for reducing alcohol-impaired driving: evidence-based legislation, law enforcement strategies, sanctions, and alcohol-control policies. Forensic Sci Rev. 2019; 31:161–184.

20.

Brands

, Mann

, Wickens

, et al. Acute and residual effects of smoked cannabis: impact on driving speed and lateral control, heart rate, and self-reported drug effects. Drug Alcohol Depend. 2019; 205:107641.

21.

Matheson

, Mann

, Sproule

, et al. Acute and residual mood and cognitive performance of young adults following smoked cannabis. Pharmacol Biochem Behav. 2020; 194:172937.

22.

Berghaus

, Scheer

, Schmidt

Effects of Cannabis on Psychomotor Skills and Driving Performance: A Metaanalysis of Experimental Studies. In: Proceedings of the 13th International Conference on Alcohol, Drugs and Traffic Safety. University of Adelaide, 1995; pp. 403–409.

23.

Fares

, Wright

, Matheson

, et al. Effects of combining alcohol and cannabis on driving, breath alcohol level, blood THC, cognition, and subjective effects: a narrative review. Exp Clin Psychopharmacol. 2022; doi: 10.1037/pha0000533

24.

Canada

Canadian Cannabis Survey. Available from: https://www.canada.ca/en/health-canada/services/drugs-medication/cannabis/research-data/canadian-cannabis-survey-2021-summary.html [Last accessed: December 6, 2022].

25.

Brands

, Di Ciano

, Mann

. Cannabis, impaired driving, and road safety: an overview of key questions and issues. Front Psychiatry. 2021; 12:641549.

26.

McCartney

, Arkell

, Irwin

, et al. Are blood and oral fluid Delta(9)-tetrahydrocannabinol (THC) and metabolite concentrations related to impairment? A meta-regression analysis. Neurosci Biobehav Rev. 2022; 134:104433.

27.

White

, Burns

. The risk of being culpable for or involved in a road crash after using cannabis: a systematic review and meta-analyses. Drug Sci Policy Law. 2021; 7:1–20.